Hyaluronic Acid (HA): Medical Device Material Safety Summary

MEDICAL DEVICE MATERIAL PERFORMANCE STUDY

Hyaluronic Acid (HA)

Safety Profile

Report Details

Date of Submission

December 5, 2021

Prepared For

U.S. FDA Center for Devices and Radiological Health

Submitted to

Ed Margerrison, PhD

Director, Office of Science and Engineering Laboratories (OSEL)

Center for Devices and Radiological Health

U.S. Food and Drug Administration

ECRI Corporate Governance

Michael Argentieri, MS

Vice President, Technology and Safety

Phone: (610) 316-2766

Email: [email protected]

Dheerendra Kommala, MD

Chief Medical Officer, Executive Committee

Phone: (610) 825-6000 ext. 5090

Email: dkommala@ECRI.org

Project Manager

Christopher N. Schabowsky, PhD, CCE

Director, Accident and Forensic Investigation

Phone: (540) 272-2740

Email: cschabowsky@ECRI.org

If you have any questions or require additional information, please do not hesitate to call the Project Manager.

Material Performance Study - Hyaluronic Acid

Table of Contents

Introduction ..................................................................................................................................................................... 4

Executive Summary – Muscle/Skeletal Applications ............................................................................................................. 5

Executive Summary – Dermal, Facial and Eye Applications ................................................................................................ 10

Executive Summary – Adhesion Barrier and Bulking Agent Applications .............................................................................. 13

Project Overview ............................................................................................................................................................. 16

Literature Search and Systematic Review Framework ........................................................................................................ 17

ECRI Surveillance Search Strategy ................................................................................................................................... 18

Safety Profile – Muscle/Skeletal Applications ..................................................................................................................... 20

Safety Brief - Systematic Review Results ........................................................................................................................ 20

ECRI Surveillance Data .................................................................................................................................................. 29

Potential Gaps .............................................................................................................................................................. 31

Safety Profile – Dermal, Facial and Eye Applications .......................................................................................................... 33

Safety Brief - Systematic Review Results ........................................................................................................................ 33

ECRI Surveillance Data .................................................................................................................................................. 40

Potential Gaps .............................................................................................................................................................. 41

Safety Profile – Adhesion Barrier and Bulking Agent Applications........................................................................................ 43

Safety Brief - Systematic Review Results ........................................................................................................................ 43

ECRI Surveillance Data .................................................................................................................................................. 52

Potential Gaps .............................................................................................................................................................. 53

Appendix A. Inclusion/Exclusion Criteria and Quality of Evidence Criteria ............................................................................ 54

Appendix B1. Search Summary – Muscle/Skeletal Applications ........................................................................................... 55

Appendix B2. Search Summary – Dermal, Facial and Eye Applications ................................................................................ 58

Appendix B3. Search Summary – Adhesion Barrier and Bulking Agent Applications .............................................................. 61

Appendix C1. Study Flow Diagram – Muscle/Skeletal Applications ...................................................................................... 68

Appendix C2. Study Flow Diagram – Dermal, Facial and Eye Applications ........................................................................... 69

Appendix C3. Study Flow Diagram – Adhesion Barrier and Bulking Agent Applications ......................................................... 69

Appendix D1. Evidence Tables – Muscle/Skeletal Applications ............................................................................................ 71

Appendix D2. Evidence Tables – Dermal, Facial and Eye Applications .................................... Error! Bookmark not defined.

Appendix D3. Evidence Tables – Adhesion Barrier and Bulking Agent Applications .............................................................. 130

Appendix E. References .................................................................................................................................................. 166

Appendix F. Surveillance Event Reports - PSO and Accident Investigation .......................................................................... 176

Material Performance Study - Hyaluronic Acid

Appendix G. Regulatory and Manufacturer Safety Alerts ................................................................................................... 177

Table of Tables

Table 1: Medical Devices Containing HA for Muscle/Skeletal Applications provided by FDA to Guide ECRI Searches.............. 20

Table 2: Summary of Primary Findings from Systematic Review – Muscle/Skeletal Applications ........................................... 20

Table 3" Complications in HA-related PSO Event Reports – Muscle/Skeletal Applications ..................................................... 30

Table 4: Harm Score Associated with HA-related PSO Event Reports – Muscle/Skeletal Applications .................................... 30

Table 5: Summary of Regulatory and Manufacturer Alerts - Muscle/Skeletal Applications .................................................... 31

Table 6: Medical Devices Containing HA for Dermal, Facial and Eye Applications provide by FDA to Guide ECRI Searches .... 33

Table 7: Summary of Primary Findings from Systematic Review - Dermal, Facial and Eye Applications ................................ 34

Table 8: Complications in HA-related PSO Event Reports - Dermal, Facial and Eye Applications .......................................... 40

Table 9: Harm Score Associated with HA-related PSO Event Reports - Dermal, Facial, and Eye Applications ........................ 40

Table 10: Summary of Regulatory and Manufacturer Alerts - Dermal, Facial and Eye Applications ....................................... 41

Table 11: Medical Devices Containing HA for Adhesion Barrier and Bulking Agent Applications provided by FDA to Guide ECRI

Searches ........................................................................................................................................................................ 43

Table 12: Summary of Primary Findings from Systematic Review - Adhesion Barrier and Bulking Agent Applications ............ 44

Table 13: Summary of Regulatory and Manufacturer Alerts - Adhesion Barrier and Bulking Agent Applications .................... 53

Table 14: Viscosupplementation – knee: Health Effect (In Vivo) Human Studies ................................................................ 71

Table 15: Viscosupplementation – hip: Health Effect (In Vivo) Human Studies ................................................................... 88

Table 16: Viscosupplementation – hand and ankle: Health Effect (In Vivo) Human Studies ................................................. 90

Table 17: Viscosupplementation – shoulder: Health Effect (In Vivo) Human Studies ........................................................... 92

Table 18: Viscosupplementation – temporomandibular joint disorders: Health Effect (In Vivo) Human Studies ..................... 93

Table 19: Scaffold - Health Effect (In Vivo) Human Studies .............................................................................................. 95

Table 20: Dermal Fillers - Health Effect (In Vivo) Human Studies ...................................................................................... 97

Table 21: Dermal Fillers - Health Effects (In Vivo) Animal Studies .................................................................................... 116

Table 22: Intraocular/Ophthalmic Viscoelastic Solution/Fluid - Health Effect (In Vivo) Human Studies ................................ 117

Table 23: Eye Drops - Health Effect (In Vivo) Human Studies .......................................................................................... 123

Table 24: Hyaluronic acid as a Material – Health Effect (In Vivo) Human Studies .............................................................. 130

Table 25: Adhesion Barrier - Health Effects (In Vivo) Human Studies ............................................................................... 132

Table 26: Anti-adhesion: Nasal packing - Health Effect (In Vivo) Human Studies .............................................................. 151

Table 27: Barrier gel for oral lesions - Health Effect (In Vivo) Human Studies ................................................................... 154

Table 28: Bulking Agents - Health Effect (In Vivo) Human Studies ................................................................................... 156

Table 29: : Intravesical agents for bladder pain syndrome/interstitial cystitis therapy - Health Effect (In Vivo) Human Studies

..................................................................................................................................................................................... 159

Table 30: Organ spacer - Health Effect (In Vivo) Human Studies ..................................................................................... 161

Table 31: Protective topical agent for esophageal and gastric lesions in GERD - Health Effect (In Vivo) Human Studies ...... 163

Table 32: Vocal cord/fold medialization - Health Effect (In Vivo) Human Studies ............................................................... 165

Material Performance Study - Hyaluronic Acid

Introduction

This Medical Device Materials Performance Report divides hyaluronic acid into three major categories:

1. Muscle/skeletal applications

2. Dermal, facial and eye applications

3. Adhesion barrier and bulking agent applications

Accordingly, the report has dedicated sections and discussions for each application. These include separate executive

summaries, safety briefs, surveillance data, gap analyses, and appendices. The goal of this report structure is to allow the

reader to focus on the current state of knowledge with regard to medical device material biocompatibility for each application

in an organized and discrete manner.

Material Performance Study - Hyaluronic Acid

Executive Summary – Muscle/Skeletal Applications

Key Points

1. Searches identified 1795 citations; 39 articles were selected for inclusion.

2. Thirty-six (92%) studies in the evidence base focused on viscosupplementation, the intra-articular (IA) injection of

hyaluronic acid (HA). Of these 36 studies, 26 studies addressed knees, 4 studies addressed hips, while 2 studies each

addressed hands or ankles, shoulders, and temporomandibular joint (TMJ) disorders. Three remaining studies

examined HA as cartilage scaffolds.

3. The most commonly reported local responses included swelling, pain at injection site, arthralgia/joint pain, and

effusion which were associated with moderate to low quality of evidence. Additionally, joint stiffness, musculoskeletal

pain, post-injection pain, pain flare ups and edema were frequently reported, and they were associated with low

quality of evidence. Outcomes infrequently reported (e.g., hematoma) or device categories with no evidence (HA as

a material, and bone putty/filler) were rated very low quality of evidence.

4. Evidence for systemic responses was reported for viscosupplementation in knees, shoulders, and TMJ disorders. One

systematic review (SR) investigating fatal adverse events (AEs) from intraarticular injections of HA listed in the

Manufacturer and User Facility Device Experience (MAUDE) and Alternative Summary Reporting databases between

2014 and 2019 identified 63 unique fatalities from knee viscosupplementation. Eight (12%) fatalities were

categorized as possibly related to the intraarticular injection of HA (IAHA), but available information was “insufficient

to make a firm determination of causality.” Two studies reporting skin reactions (1 rash, 1 peeling of skin on hands

and toes) directly attributed the response to knee viscosupplementation.

5. Healthcare Technology Alerts identified 5 manufacturer issued alerts describing problems with adverse events,

contamination, packaging issues, and impartially filled products.

6. Patient Safety Organization identified 7 complications associated with HA including wrong side procedure - 2

(28.5%), 2) wrong product - 2 (28.5%), 3) product expired - 2 (28.5%), and 4) wrong time – 1 (14.2%).

7. Evidence gaps:

a. Additional research is needed in viscosupplementation in non-knee indications. Small evidence bases were

identified for hip, hand, ankle, shoulder and TMJ disorders. While studies for other indications (e.g., plantar

fasciitis, trigger finger, and rotator cuff) were identified, these studies were excluded at final prioritization

level due to being low quality (e.g., single arm design) or having small enrollment (<100 patients).

b. Evidence for viscosupplementation was mostly focused on 2 HA products (Hyalgan and Synvisc). Additional

evidence is needed for Monovisc, HYADD, Sinovial, Gel-One, Supartz, Go-On, Hanox M-XL, Ostenil, and

Hyalubrix which were investigated in ≤3 studies. Additionally, evidence is lacking for numerous HA products

that were not evaluated for viscosupplementation (e.g., Synojoynt, Triluron, Trivisc), bone putty/filler (e.g.,

DBX Putty, Kinex Gel), cartilage scaffold or (e.g., Agili-C).

c. While many studies were high-quality (SRs or randomized controlled trials (RCTs)), several studies did not

report important study characteristics (e.g., HA products investigated, HA dose), or details on clinical results

Material Performance Study - Hyaluronic Acid

(e.g., type of treatment-related AE (TRAE), number of events, number of patients experiencing events,

timing of events).

d. Long-term human RCTs for local responses to HA as a material and for all device categories to better

ascertain associations with these responses to HA.

e. Additional research on systemic responses, including those on patient or material factors, for all HA device

categories. Systemic responses were only investigated in 13 (33%) studies with no studies investigating HA

for viscosupplementation (in hips, hands, and ankles), bone putty/fillers, or cartilage scaffolds.

Overview - Muscle/Skeletal Applications

FDA engaged ECRI to perform a comprehensive literature search and systematic review to identify the current state of

knowledge with regard to medical device material biocompatibility. Additionally, data derived from ECRI’s Patient Safety

Organization (PSO), accident investigations, Problem Reporting Network (PRN), and healthcare technology alerts were

analyzed. This report focuses on answering five key questions provided by FDA and summarized below, regarding a host’s

local and systemic response to Hyaluronic Acid in muscle/skeletal applications. If data did not exist to sufficiently address

these questions, a gap was noted in this report. These gaps could represent areas of further research.

1. What is the typical/expected local host response to these materials?

Local responses/device events varied somewhat across different device categories and between human studies (see

specific responses/events under 1a. below). The majority of ECRI surveillance data were related to wrong side

application, packaging issues, and contamination, which were not related to material response due to insufficient

biocompatibility or mechanical integrity and use of the device.

Can that response vary by location or type of tissue the device is implanted in or near?

i. Swelling occurred with viscosupplementation in knees and ankles, and scaffold placement in knees.

This local response was the most commonly reported after viscosupplementation in knees. One study

reported mild-to-moderate swelling of the knee joint occurred similarly with HA vs. platelet-rich plasma

(PRP) with HA rates ranging from 87% (6 months) to 100% (6 weeks). Other studies reported higher

but low incidence of knee swelling with IAHA vs. saline (3.7% Synvisc, 0.9% saline; 1.1% Monovisc,

0.5% saline). One RCT reported that swelling after injection of Synvisc was device-related, while 1 SR

of 12 RCTs reported significantly more TRAEs (including joint swelling) occurred with HA vs

corticosteroid (CS; 746 events vs. 606 events). Mild swelling was reported after scaffold placement in

knees (n=3), and also after viscosupplementation in ankles.

ii. Pain at injection site occurred after viscosupplementation in knees, shoulders, ankles, and TMJ

disorders. One study reported knee pain and local swelling at injection site in 21% of patients

administered HA injections of different molecular weights and concentrations (DMW), high molecular

weights (HMWs), and low molecular weights (LMWs). Studies also reported knee pain at injection site

in 8 (13.4%) patients with IAHA, and a 0.5% incidence rate with Synvisc. This outcome was commonly

reported in 1 SR of 15 studies addressing viscosupplementation in shoulders, and in 1 SR of 18 studies

addressing viscosupplementation in ankles. Lastly, 2 SRs examining TMJ disorders reported a lower

incidence of pain at injection site with HA vs. CS (37% vs. 70%) and HA vs. PRP or platelet-rich growth

factor (PRGF; 60% vs. 76%).

iii. Arthralgia/joint pain was reported in 7 studies focused on viscosupplementation in knees. Studies

reported a significantly higher occurrence of arthralgia with HA vs. nonsteroidal anti-inflammatory drugs

(NSAIDs; 8.1% vs. 2.9%), but similar rates vs. saline (<4% in 2 studies). 1 SR reported severe joint

pain in 28 (100%) patients with pseudoseptic arthritis, a rare complication after IAHA injection. 2 RCTs

reported that arthralgia with Synvisc or Monovisc was device-related.

iv. Effusion was reported in 5 studies focused on viscosupplementation in knees. One SR (n=5,354)

reported no significant difference in effusion after injection based on molecular weights (1.9% HMW,

Material Performance Study - Hyaluronic Acid

1.7% medium molecular weight (MMW), 1.8% LMW), but a significantly higher incidence of effusion

with products processed through extraction of avian-derived molecules (AD-HA) vs. bacterial processes

of biological fermentation (Bio-HA; 3.4% vs. 0.5%). Another SR reported that 28 (100%) patients with

pseudoseptic arthritis presented with effusion.

v. Joint stiffness only occurred with knee viscosupplementation. Studies reported no significant differences

in stiffness with HA vs. PRGF, lower rates with Monovisc vs. saline (0.5% vs. 1.1%), and incidence

rates of 7% with varying weights of IAHA.

vi. Musculoskeletal pain was reported after viscosupplementation in knees and shoulders. For knee

viscosupplementation, 1 study reported more frequent musculoskeletal pain with HA vs. PRP (5% vs.

1.8%). For shoulder viscosupplementation, musculoskeletal pain was commonly reported in 13 (87%)

studies in 1 SR; severe musculoskeletal, a serious adverse event (SAE), was reported in some cases.

vii. Post-injection pain was reported after viscosupplementation in knees and hips. For knee

viscosupplementation, studies reported post-injection pain in 3 patients with IAHA, significantly more

post-injection pain with HA vs. PRP, and lower incidence with Synvisc-One vs. cooled radiofrequency

ablation (CRFA; 3% vs. 8%). For hip viscosupplementation, 1 study reported significantly higher post-

injective pain with PRP vs. HA.

viii. Pain flare ups were reported with viscosupplementation in knees and hips. One SR (n=5,354) reported

significantly more knee flare ups with HA products with HMWs vs. lower molecular weights (13.7%

HMW, 3.3% MMW, 10.7% LMW), and a significantly lower incidence of acute flare ups with AD-HA vs.

Bio-HA (3.0% vs. 13.2%). One RCT reported knee flare up rates of 6% to 36% with Synvisc. Lastly, a

higher occurrence of hip pain flare ups were reported with HA vs. placebo (7% vs. 2.4%).

ix. Edema was reported after knee viscosupplementation and scaffold placement. Injection-site edema and

peripheral edema were only reported with HA vs. placebo (0.5 % vs. 0%) in 1 RCT. Edema or cyst

formation occurred similarly with a Hyalofast-based scaffold vs. a chitosan-glycerol phosphate/blood

implant (BST-CarGel®) (8 (38%) Hyalofast, 10 (40%) BST-CarGel) in 1 nonrandomized comparative

study.

x. Less commonly reported local responses included 1) device-related cutaneous vasculitis, Baker’s cyst

(n=2) with Synvisc-One, and elevated inflammatory cell count (n=16) after knee

viscosupplementation; 2) hematoma after hip viscosupplementation and HA scaffold placement in the

knee; 3) pseudogout after ankle viscosupplementation; 4) abscess after shoulder viscosupplementation;

5) postoperative discomfort and ear pressure after viscosupplementation for TMJ disorders; and 6)

persistent pain after an HA scaffold placement in knees.

xi. The overall quality of evidence related to local host responses was moderate to very low, with variation

across different device categories.

xii. Evidence was limited for HA as a scaffold, however no evidence was included for HA as a material, and

bone putty/filler.

Over what time course does this local host response appear?

Swelling was reported after knee and ankle viscosupplementation, from 2 weeks to 6 weeks with an HA

scaffold; and at 6 weeks, 3 months and 6 months with knee viscosupplementation. Pain at injection site was

reported up to 3 days with viscosupplementation of knees, ankles, shoulders, and TMJ disorders. After knee

viscosupplementation: 1) arthralgia was reported in 7 studies as immediately after up to 9 days; 2) effusion

was reported immediately after up to 3 days; and 3) joint stiffness only occurred immediately after. Pain

flare ups were described at weeks 6, 13, 26, 39, and 52 with Synvisc in knee viscosupplementation, and

during or immediately after with hip viscosupplementation. Timing for musculoskeletal pain

(viscosupplementation in knees, shoulders), post-injection pain (viscosupplementation in knees, hips), and

edema was not reported.

2. Does the material elicit a persistent or exaggerated response that may lead to systemic signs or

symptoms – beyond known direct toxicity problems?

What evidence exists to suggest or support this?

Material Performance Study - Hyaluronic Acid

Overall, 13 studies investigated systemic responses; studies addressed viscosupplementation in knees (11),

shoulders (1), and TMJ disorders (1). Of studies investigating, 11 (85%) studies identified persistent or

exaggerated immune responses, while 26 (67%) did not investigate systemic responses.

What are the likely systemic manifestations?

For knee viscosupplementation, evidence for systemic responses was provided in studies examining IAHA

alone, and studies grouped in the following categories (vs. PRP, anti-inflammatories, saline/placebo,

miscellaneous treatments).

IAHA

: 1 SR indicated that 8 (12%) of 63 fatalities were possibly IAHA-related but available

information was “insufficient to make a firm determination of causality.” Authors noted the

following responses occurred prior to death: septic shock followed by paralysis, necrotizing fasciitis,

and septicemia. Another SR investigating acute pseudoseptic arthritis reported elevated C reactive

protein (CRP), elevated erythrocyte sedimentation rates (ESR), elevated polymorphonuclear

leukocytes (PMNs) in 10 (37%) patients, fever in 6 (22%) patients, and leukocytosis above 10,000

in 4 (15%) patients. Systemic responses in a related case report included a cell count of 123,260

and white blood cell count (WBC) of 15.5 in 1 patient.

IAHA vs. PRP

: 1 of 2 studies investigating identified nasopharyngitis and backache in 2 patients

each, and headache in 1 patient.

IAHA vs. anti-inflammatories

: Of 3 SRs investigating, 1 SR reported significantly more headache vs.

oral NSAIDs (8.4% vs. 4.4%), and cardiac-related complications (n=1). Data for other responses

(nausea and vomiting) was not provided in another SR. Lastly, 1 SR (n=1794) did not identify any

treatment-related systemic responses.

IAHA vs. saline/placebo

: 1 SR of 9 RCTs (n=1967), focused on TRAEs, reported no significant

difference in system-organ class(SOC)-related AEs with IAHA vs. placebo. Cardiac disorders,

vascular disorders, and renal/urinary disorders occurred in <10 patients each with IAHA.

Skin/subcutaneous tissue disorders and hypersensitivity reactions occurred in 22 and 23 patients,

respectively. Gastrointestinal (GI) disorders, respiratory/thoracic/mediastinal disorders, and

nervous system disorders occurred in 66 to 71 patients each. Musculoskeletal and connective tissue

disorders occurred in 145 patients with HA, and 149 patients with placebo. 1 RCT reported a

Monovisc-related rash in 1 patient. Lastly, 1 SR of 10 RCTs attributed 1 serious AE (skin reaction

characterized by peeling of skin on hands and toes) to IAHA.

IAHA vs. miscellaneous treatments

: GI complaints were reported in 3% to 7% of patients.

For shoulder viscosupplementation: 1 SR with followup up to 3 years, commonly reported systemic

responses including diarrhea, flu symptoms, and headache. Serious AEs such as chest pain and cancer were

also reported (N not reported). Responses were not deemed as device-related.

For viscosupplementation in TMJ disorders: 1 individual RCT in a SR reported slight chills in 5.7% of patients

with HA, vs. 0% with CS.

What is the observed timeline(s) for the systemic manifestations?

Evidence for timing was limited to the following:

For knee viscosupplementation:

IAHA:

death in 1 of 8 patients was reported 4 months post-IAHA injection. 1 SR addressing pseudoseptic

arthritis indicated that most cases presented within 24 hours of injection, however time from injection to

presentation ranged from 1 hour to 9 days. In the related case report, arthrocentesis was undertaken at 20

hours post-injection.

IAHA vs.

saline/placebo:

Skin reaction in 1 patient occurred 8 days post-IAHA injection.

IAHA vs. miscellaneous treatments

: GI complaints were tracked at weeks 6, 13, 26, 39, and 52.

Material Performance Study - Hyaluronic Acid

For viscosupplementation in TMJ disorders: Slight chills were reported at 1 week.

Have particular cellular/molecular mechanisms been identified for such manifestations?

No studies investigated cellular/molecular mechanisms for systemic responses.

3. Are there any patient-related factors that may predict, increase, or decrease the likelihood and/or

severity of an exaggerated, sustained immunological/systemic response?

No studies investigated patient-related factors that may predict, increase, or decrease the likelihood of an

exaggerated, sustained immunological/systemic response.

4. Are there any material-related factors that may predict, increase, or decrease the likelihood and/or

severity of an exaggerated, sustained immunological/systemic response?

1 SR indicated that the pathophysiological response for pseudoseptic arthritis (identified in 28 patients) may be due

to the accumulation of viscous material (e.g., phagocytised HA). This same review noted that significantly more

reactions occurred in patients receiving more than a single administration. Results indicated that 7 cases (25%)

occurred after 2 injections, 5 cases (17.9%) after 3 injections, and 13 cases (46.4%) after ≥4 injections.

5. What critical information gaps exist and what research is needed to better understand this issue?

All gaps listed here could benefit from future research.

a. Additional research is needed in viscosupplementation in non-knee indications. Small evidence bases were

identified for hip, hand, ankle, shoulder and TMJ disorders. While we did identify studies for other

indications (e.g., plantar fasciitis, trigger finger, and rotator cuff), these studies were excluded at final

prioritization level due to being low quality (e.g., single arm design) or having small enrollment (<100

patients).

b. Evidence for viscosupplementation was mostly focused on 2 HA products (Hyalgan and Synvisc). Additional

evidence is needed for Monovisc, HYADD, Sinovial, Gel-One, Supartz, Go-On, Hanox M-XL, Ostenil, and

Hyalubrix which were investigated in ≤3 studies. Additionally, evidence is lacking for numerous HA products

that were not evaluated for viscosupplementation (e.g., Synojoynt, Triluron, Trivisc), bone putty/filler (e.g.,

DBX Putty, Kinex Gel), and cartilage scaffold (e.g., Agili-C).

c. While many studies were high-quality (SRs or RCTs), several studies did not report important study

characteristics (e.g., HA products investigated, HA dose), or details on clinical results (e.g., type of TRAE,

number of events, number of patients experiencing events, timing of events).

d. Long-term human and animal RCTs for local responses to HA as a material and for all device categories to

better ascertain associations with these responses to HA.

e. Additional research on systemic responses, including those on patient or material factors, for all HA device

categories. Systemic responses were only investigated in 13 (33%) studies with no studies investigating HA

for viscosupplementation (in hips, hands, and ankles), bone putty/fillers, or cartilage scaffolds.

Material Performance Study - Hyaluronic Acid

Executive Summary – Dermal, Facial and Eye Applications

Key Points

1. Searches identified 1401 citations; 58 articles were selected for inclusion

2. For HA dermal fillers, the most common complications of facial injections included lumpiness, tenderness, swelling,

and bruising. Delayed inflammatory reactions/events occurred in about 1% of patients; in rare instances granulomas

formed that required surgical removal. Most complication rates were similar between different HA fillers, but delayed-

onset nodules had a higher occurrence rate for Juvederm Volbella. Rare but serious events included partial or

complete vision loss following upper face injections, as well as brain infarcts or hemorrhage. The overall quality of

evidence is moderate.

3. For intraocular/ophthalmic viscoelastic HA solutions, intraoperative pressure (IOP) elevation was a common

complication following eye surgery and injection of HA solutions. IOP usually decreased over time, but Healon GV

and Healon5 were associated with lesser elevations and faster decreases in IOP over 1 week compared to other HA

solutions. Corneal and/or macular edema were relatively rare events occurring at similar rates for different HA

solutions. The overall strength of evidence is moderate for IOP elevation and corneal/macular edema, low for other

adverse events.

4. For eye drops, reported adverse events included itching/stinging/irritation/erythema/keratitis, eye pain, eye

disorders/dry eye, hyperemia/hemorrhage, and infection/viral conjunctivitis. Overall, most studies reported no

difference in adverse event rates between HA eye drops and other treatments or controls. The overall quality of

evidence is moderate.

5. Systemic responses. A small number of patients developed late-onset, inflammatory, non-infectious adverse reactions

related to dermal fillers/implants (including HA fillers) that could be totally or partially considered as autoimmune/

inflammatory syndrome induced by adjuvants (ASIA)-related disorders. Reported symptoms of HA cases included

myalgia, arthralgia/arthritis, fatigue, neurologic complaints, cognitive features, fever, Sicca syndrome, skin

manifestations (including facial nodules), and autoimmune disease. The overall strength of evidence is low for

systemic responses.

6. ECRI PSO identified one incident of hemorrhage/hematoma associated with a dermal implant composed of HA. There

were no relevant reports found in ECRI’s accident investigation, or PRN databases related to devices composed of

HA.

7. Evidence gaps:

a. Across all device categories, the overall quality of evidence was low to very low for potential systemic

responses.

Material Performance Study - Hyaluronic Acid

b. For intraocular/ophthalmic/viscoelastic solution/fluids, other than IOP elevation and corneal/macular edema

(moderate quality of evidence), all identified local host responses (e.g., ocular tension, inflammation) were

of low quality of evidence.

c. There were no studies that addressed any particular cellular or molecular mechanisms for systemic

manifestations.

d. Additionally, no studies addressed patient-related or material-related factors that could predict the likelihood

and/or severity of immunological/systemic responses.

Overview - Dermal, Facial and Eye Applications

FDA engaged ECRI to perform a comprehensive literature search and systematic review to identify the current state of

knowledge with regard to medical device material biocompatibility. Additionally, data derived from ECRI’s Patient Safety

Organization (PSO), accident investigations, Problem Reporting Network (PRN), and healthcare technology alerts were

analyzed. This report focuses on answering five key questions provided by FDA and summarized below, regarding a host’s

local and systemic response to Hyaluronic Acid in muscle/skeletal applications. If data did not exist to sufficiently address

these questions, a gap was noted in this report. These gaps could represent areas of further research.

1. What is the typical/expected local host response to these materials?

Local responses/device events varied somewhat across different device categories (see specific responses/events under 1a.

below).

Can that response vary by location or type of tissue the device is implanted in or near?

i. For HA dermal fillers, the most common complications of facial injections included lumpiness,

tenderness, swelling, and bruising. Most complication rates were similar between different HA fillers,

but delayed-onset nodules had a higher occurrence rate for Juvederm Volbella. Delayed inflammatory

reactions/events occurred in about 1% of patients; in rare instances granulomas formed that required

surgical removal. Rare but serious events included partial or complete vision loss following upper face

injections, as well as brain infarcts or hemorrhage. HA (Macrolane) injection of breasts was associated

with cysts, capsular contraction, early absorption and sometimes removal of product was required.

ii. For intraocular/ophthalmic viscoelastic HA solutions, intraoperative pressure (IOP) elevation was a

common complication following eye surgery and injection of HA solutions. IOP usually decreased over

time, but Healon GV and Healon5 were associated with lesser elevations and faster decreases in IOP

over 1 week compared to other HA solutions. Corneal and/or macular edema were relatively rare events

occurring at similar rates for different HA solutions. One study reported Descemet membrane

detachment (DMD) in 7% of patients following canaloplasty and Healon GV injection (most cases

resolved without further surgery). One study reported 34 cases of toxic anterior segment syndrome

(TASS) after cataract surgery; the source was identified as HA solution derived from rooster comb.

iii. For eye drops, reported adverse events included itching/stinging/irritation/erythema/keratitis, eye pain,

eye disorders/dry eye, hyperemia/hemorrhage, and infection/viral conjunctivitis. Overall, most studies

reported no difference in adverse event rates between HA eye drops and other treatments or controls.

Over what time course does this local host response appear?

i. For dermal fillers, most events occurred within the first 2 weeks; a smaller percentage of patients

experienced delayed-onset inflammatory reactions/events that occurred months or even years after

injection (3 to 10 years for granulomas).

ii. For intraocular HA solutions, IOP tended to peak at 1 day post-surgery and decrease afterward. Other

complications occurred within 6 hours to 3 months post-surgery, most within the first week.

Material Performance Study - Hyaluronic Acid

iii. For eye drops, adverse responses occurred within the first 3 months, most within the first month.

2. Does the material elicit a persistent or exaggerated response that may lead to systemic signs or

symptoms – beyond known direct toxicity problems?

What evidence exists to suggest or support this?

Two retrospective single-arm case series reported on cases of patients suffering from late-onset,

inflammatory, non-infectious adverse reactions related to dermal fillers/implants that could be totally or

partially considered as autoimmune/ inflammatory syndrome induced by adjuvants (ASIA)-related disorders.

What are the likely systemic manifestations?

Reported symptoms of HA cases in the larger study included 4 myalgia, 6 arthralgia/arthritis, 6 fatigue, 2

neurologic complaints, 2 cognitive features, 1 fever, 1 Sicca syndrome, 7 skin manifestations (3 facial

nodules), 3 evolvements into autoimmune disease.

What is the observed timeline(s) for the systemic manifestations?

The symptoms occurred between 6- and 317-months following exposure to dermal filler.

Have particular cellular/molecular mechanisms been identified for such manifestations?

The search did not identify any studies that addressed this question.

3. Are there any patient-related factors that may predict, increase, or decrease the likelihood and/or

severity of an exaggerated, sustained immunological/systemic response?

The number of cases related to HA was low and analyzed with cases who had received other dermal fillers, so the

evidence was unclear concerning patient-related factors that predict sustained systemic response.

4. Are there any material-related factors that may predict, increase, or decrease the likelihood and/or

severity of an exaggerated, sustained immunological/systemic response?

The number of cases related to HA was low and analyzed with cases who had received other dermal fillers, so the

evidence was unclear concerning material-related factors that predict sustained systemic response.

5. What critical information gaps exist and what research is needed to better understand this issue?

All gaps listed here could benefit from future research.

a. Across all device categories, the overall quality of evidence was low to very low for potential systemic responses.

b. For intraocular/ophthalmic/viscoelastic solution/fluids, other than IOP elevation and corneal/macular edema

(moderate quality of evidence), all identified local host responses (e.g., ocular tension, inflammation) were of

low quality of evidence.

c. There were no studies that addressed any particular cellular or molecular mechanisms for systemic

manifestations.

d. Additionally, no studies adequately addressed patient-related or material-related factors that could predict the

likelihood and/or severity of immunological/systemic responses.

Material Performance Study - Hyaluronic Acid

Executive Summary – Adhesion Barrier and Bulking Agent

Applications

Key Points

1. Searches identified 1217 citations; 52 articles were selected for inclusion

2. Adhesion barriers was the most studied application of hyaluronic acid (34 human studies). There are few adverse

events or complications caused by HA-containing products, and the quality of evidence is High (for both local and

systemic responses). One trial found statistically significantly higher adverse event rates with HA adhesion barriers in

the context of colorectal or small bowel surgery while another trial involving gastrointestinal surgery also found

higher adverse event rates with HA adhesion barriers. It is possible that the powder formulation of HA adhesion

barriers was a contributing factor to increased frequency of adverse events due to the potential for greater diffusion

and migration away from the application site to anastomoses

3. For nasal packaging as well as barrier gel for oral lesions, there are few adverse events or complications caused by

HA-containing products. For both of these categories, the evidence was Moderate for local responses and Very Low

for systemic responses. .

4. In two trials of bulking agents, statistically significantly higher adverse event rates with HA were found in the context

of treatment for fecal incontinence or urinary incontinence which occurred within 12 months of surgery.

5. There were no relevant reports found in ECRI’s PSO, accident investigation, or PRN databases related to adhesion

barriers composed of HA. Healthcare Technology Alerts search returned 4 manufacturer issued alerts describing

problems with compromised sterility and off label use.

6. Evidence gaps:

a. For 7 of 9 device categories, no local adverse events were statistically significantly more likely in patients

receiving HA devices compared to patients receiving non-HA devices. This is predominantly because of a

lack of studies with strong design that investigated these events. The exceptions were adhesion barrier and

bulking agents.

Material Performance Study - Hyaluronic Acid

b. For 7 of 9 device categories, no systemic adverse events were statistically significantly more likely in

patients receiving HA device compared to patients received non-HA devices. This is predominantly because

of a lack of studies with strong design that investigated these events. The exceptions were adhesion barrier

and bulking agents.

c. Only one study related to HA as a material was identified. However, because treated patients received HA

as well as chondroitin sulfate, one cannot determine whether the reported adverse events were due to HA

or chondroitin sulfate or other factors

d. For 8 of 9 device categories, only study involving adhesion barriers investigated the cellular or molecular

mechanisms for systemic manifestations.

Overview - Adhesion Barrier and Bulking Agent Applications

FDA engaged ECRI to perform a comprehensive literature search and systematic review to identify the current state of

knowledge with regard to medical device material biocompatibility. Additionally, data derived from ECRI’s Patient Safety

Organization (PSO), accident investigations, Problem Reporting Network (PRN), and healthcare technology alerts were

analyzed. This report focuses on answering five key questions provided by FDA and summarized below, regarding a host’s

local and systemic response to Hyaluronic Acid in Adhesion Barrier and Bulking Agent applications. If data did not exist to

sufficiently address these questions, a gap was noted in this report. These gaps could represent areas of further research.

1. What is the typical/expected local host response to these materials?

For 7 of 9 device categories, no local adverse events were statistically significantly more likely in patients receiving HA

devices compared to patients receiving non-HA devices. The exceptions were adhesion barrier and bulking agents.

Can that response vary by location or type of tissue the device is implanted in or near?

i. In two trials of adhesion barriers, statistically significantly higher adverse event rates with HA were

found in the context of colorectal or small bowel surgery (one trial), and “gastrointestinal surgery” (the

other trial. In two trials of bulking agents, statistically significantly higher adverse event rates with HA

were found in the context of treatment for fecal incontinence or urinary incontinence.

Over what time course does this local host response appear?

i. In the two aforementioned trials of adhesion barriers, adverse events occurred within 30 days of

surgery. In the two aforementioned trials of bulking agents, adverse events occurred within 12 months

of surgery.

2. Does the material elicit a persistent or exaggerated response that may lead to systemic signs or

symptoms – beyond known direct toxicity problems?

What evidence exists to suggest or support this?

For 7 of 9 device categories, no systemic adverse events were statistically significantly more likely in

patients receiving HA device compared to patients received non-HA devices. The exceptions were adhesion

barrier and bulking agents.

What are the likely systemic manifestations?

For adhesion barriers, creatinine was significantly higher in HA patients than in non-HA patients. For bulking

agents, fever was significantly higher in HA patients than in non-HA patients.

c. What is the observed timeline(s) for the systemic manifestations?

Material Performance Study - Hyaluronic Acid

Creatinine was significantly higher in HA patients on day 5 but not day 7 after surgery. Fever occurred

within the first 6 months after injection of the bulking agent.

Have particular cellular/molecular mechanisms been identified for such manifestations?

The study authors suggested the HA barrier may cause a temporary increase in serum creatinine in the

early stage after surgery, but the effect is transitory. In the other study, authors did not speculate as to why

fever was more likely in HA-treated patients.

3. Are there any patient-related factors that may predict, increase, or decrease the likelihood and/or

severity of an exaggerated, sustained immunological/systemic response?

For adhesion barriers, one systematic review of laparoscopic gynecologic surgery

stated “Filmy mobile adhesions

resulted in the highest verbal pain scores followed by dense mobile adhesions, filmy fixed adhesions, and dense fixed

adhesions. Pain scores were highest when adhesions were between a segment of bowel and an ovary.

4. Are there any material-related factors that may predict, increase, or decrease the likelihood and/or

severity of an exaggerated, sustained immunological/systemic response?

For adhesion barriers, an RCT of laparoscopic colorectal and/or small bowel surgery stated, “The powder formulation

of HA/CMC used in this study is likely to be an important contributing factor for the increased frequency of adverse

events and serious adverse events.” “Due to the potential for greater diffusion of the powder formulation, the

authors speculate that migration away from the application site to anastomoses could have occurred in some cases.

Furthermore, over-hydration of the HA/CMC powder might have resulted in pooling of the resulting gel away from

the application site, raising the possibility of migration onto an anastomosis or provision of a nidus for abscess; such

migration to anastomoses might increase the rate of SSIs.”

5. What critical information gaps exist and what research is needed to better understand this issue?

All gaps listed here could benefit from future research.

a. For 7 of 9 device categories, no local adverse events were statistically significantly more likely in patients

receiving HA devices compared to patients receiving non-HA devices. This is predominantly because of a

lack of studies with strong design that investigated these events. The exceptions were adhesion barrier and

bulking agents.

b. For 7 of 9 device categories, no systemic adverse events were statistically significantly more likely in

patients receiving HA device compared to patients received non-HA devices. This is predominantly because

of a lack of studies with strong design that investigated these events. The exceptions were adhesion barrier

and bulking agents.

c. Only one study related to HA as a material was identified. However, because treated patients received HA

as well as chondroitin sulfate, one cannot determine whether the reported adverse events were due to HA

or chondroitin sulfate or other factors

d. For 8 of 9 device categories, only study involving adhesion barriers investigated the cellular or molecular

mechanisms for systemic manifestations.

Material Performance Study - Hyaluronic Acid

Project Overview

FDA engaged ECRI to perform a comprehensive literature search and systematic review to identify the current state of

knowledge with regard to medical device material biocompatibility. Specific materials or topics were selected by FDA based on

current priority. For 2021, the following 18 topics have been chosen.

1. Magnesium (Mg)

2. Complications associated with Polypropylene Mesh in Pre-, Peri-, and Post-Menopausal Women

3. Polytetrafluoroethylene (PTFE)

4. Acrylics 1: PMMA

5. Acrylics 2: pHEMA

6. Acrylics 3: Cyanoacrylates (PET)

7. Correlations between complications with polypropylene mesh and surgical procedure/anatomical location and

chemical/mechanical device properties

8. Dimethacrylates, Trimethacrylates (EDMA, EGDMA, TEGDMA, PEGDMA), and glycerol methacrylate (bis-GMA)

9. Polyethylene glycol (PEG)

10. Other Fluoropolymers (PFPE, PVDF, PVDF-HFP, PCTFE)

11. Silver

12. Small Molecule Per- and polyfluoroalkyl substances (SM-PFAS)

13. Hyaluronic Acid (HA) - Muscle/Skeletal Applications

14. Hyaluronic Acid (HA) - Dermal, Facial, and Eye Applications

15. Hyaluronic Acid (HA) – Adhesion Barriers

16. Polycaprolactone (PCL)

17. Zirconia

18. Nitinol

The systematic review was guided by key questions mutually agreed upon by FDA and ECRI. Data were extracted from

literature articles and ECRI surveillance databases accordingly.

Key Questions

1. What is the typical/expected local host response to Hyaluronic Acid?

Can that response vary by location or type of tissue the device is implanted in or near?

Over what time course does this local host response appear?

2. Does the material elicit a persistent or exaggerated response that may lead to systemic signs or symptoms – beyond

known direct toxicity problems?

What evidence exists to suggest or support this?

What are the likely systemic manifestations?

What is the observed timeline(s) for the systemic manifestations?

Have particular cellular/molecular mechanisms been identified for such manifestations?

3. Are there any patient-related factors that may predict, increase, or decrease the likelihood and/or severity of an

exaggerated, sustained immunological/systemic response?

4. Are there any material-related factors that may predict, increase, or decrease the likelihood and/or severity of an

exaggerated, sustained immunological/systemic response?

5. What critical information gaps exist and what research is needed to better understand this issue?

Material Performance Study - Hyaluronic Acid

If data did not exist to sufficiently address these questions, a gap was noted in this report. These gaps could represent areas

of further research.

Safety Profiles were written for the six materials listed above to include the summary of key findings from the systematic

review and surveillance search and are included in this report.

Literature Search and Systematic Review Framework

The ECRI-Penn Evidence-based Practice Center (EPC) conducts research reviews for the Agency for Healthcare Research and

Quality (AHRQ) Effective Health Care (EHC) Program. ECRI’s scientific staff within our Center for Clinical Excellence has

authored hundreds of systematic reviews and health technology assessments on 3,500+ technologies/interventions for ECRI’s

public- and private-sector clients. In addition to this work, ECRI staff have coauthored several methods papers on evidence

synthesis published on the AHRQ Effective Health Care website and in peer-reviewed journals.

For this project, the clinical and engineering literature was searched for evidence related to biocompatibility of each material.

Searches of PubMed/Medline and Embase were conducted using the Embase.com platform. Scopus was used initially to search

nonclinical literature; however, it was determined that the retrieved citations did not meet inclusion criteria and that database

was subsequently dropped from the search protocol. Search limits included publication dates between 2011 and 2021 and

English as the publication language. ECRI and FDA agreed on appropriate host and material response search concepts as

follows:

• Material Response

o Strength

o Embrittlement

o Degradation

o Migration

o Delamination

o Leaching

• Host Response

o Local

 Inflammation

 Sensitization

 Irritation

 Scarring/fibrosis

•

Keloid formation

•

Contracture

 Ingrowth

 Erosion

o Systemic

 Cancer

 Inflammation

 Immune Response

 Fatigue

 Memory Loss

 Rash

 Joint Pain

 Brain Fog

Search strategies were developed for each concept and combined using Boolean logic. Several search approaches were used

for comprehensiveness. Strategies were developed for devices of interest as indicated by FDA as well as the material-related

strategies. Each of these sets were combined with the material and host response strategies. Detailed search strategies and

contextual information are presented in Appendix B. Resulting literature was screened by title review, then abstract review,

and finally full article review. Data were extracted from the articles meeting our inclusion criteria to address the key questions

for each material.

Material Performance Study - Hyaluronic Acid

ECRI Surveillance Search Strategy

There are four key ECRI sources for medical device hazards and patient incidents. These databases were searched by key

terms and device models. Relevant data were extracted to address the key questions agreed upon by FDA and ECRI. Patient

demographics were extracted when available. All data presented were redacted and contain no protected health information

(PHI).

ECRI surveillance data comprise ECRI Patient Safety Organization (PSO) event reports, accident investigations, problem

reporting network (PRN) reports, and alerts. The PSO, investigations, and PRN reports included in this report include mostly

acute patient events. We rarely find chronic conditions or patient follow-up reports, which are more prevalent in the clinical

literature. Complications are reported directly by clinical staff; thus, reports vary greatly in the level of detail provided.

ECRI Patient Safety Organization (PSO)

ECRI is designated a Patient Safety Organization by the U.S. Department of Health and Human Services and has collected

more than 3.5 million serious patient safety events and near-miss reports from over 1,800 healthcare provider organizations

around the country. Approximately 4% of these reports pertain to medical devices. Most of these reports are acute (single

event) reports and do not include patient follow-up. These data were filtered by complication, and relevant reports were

included in the analysis. “Harm Score” refers to the National Coordinating Council Medication Error Reporting and Prevention

(NCC MERP) taxonomy of harm, ranging from A to I with increasing severity (see

Figure 1). The entire PSO database was

included in the search, with reports ranging from year 2004 through August 2021, unless otherwise noted.

Figure 1. NCC MERP “harm score,” which is now regularly used by patient safety organizations.

Category A (No Error)

Circumstances or events that have the capacity to cause error.

Category B (Error, no harm)

An error occurred, but the error did not reach the patient (an “error of omission” does reach the patient).

Category C (Error, no harm)

An error occurred that reached the patient but did not cause patient harm.

Category D (Error, no harm)

An error occurred that reached the patient and required monitoring to confirm that it resulted in no harm to the patient and/or

required intervention to preclude harm.

Category E (Error, harm)

An error occurred that may have contributed to or resulted in temporary harm to the patient and required intervention.

Category F (Error, harm)

An error occurred that may have contributed to or resulted in temporary harm to the patient and required initial or prolonged

hospitalization.

Category G (Error, harm)

An error occurred that may have contributed to or resulted in permanent patient harm.

Material Performance Study - Hyaluronic Acid

Category H (Error, harm)

An error occurred that required intervention necessary to sustain life.

Category I (Error, death)

An error occurred that may have contributed to or resulted in patient’s death.

Definitions

Harm: Impairment of the physical, emotional, or psychological function or structure of the body and/or pain resulting

therefrom.

Monitoring: To observe or record relevant physiological or psychological signs.

Intervention: may include change in therapy or active medical/ surgical treatment.

Intervention necessary to sustain life: includes cardiovascular and respiratory support (e.g., CPR, defibrillation, intubation).

Accident Investigation

ECRI has performed thousands of independent medical-device accident investigations over more than 50 years, including on-

site and in-laboratory investigations, technical consultation, device testing and failure analysis, accident simulation, sentinel

event and root-cause analyses, policy and procedure development, and expert consultation in the event of litigation. Our

investigation files were searched by keywords, and the search was limited to the past 10 years unless we found landmark

investigations that are particularly relevant to biocompatibility.

Problem Reporting Network (PRN)

For more than 50 years, ECRI’s Problem Reporting Network (PRN) has gathered information on postmarket problems and

hazards and has been offered as a free service for the healthcare community to submit reports of medical device problems or

concerns. Each investigation includes a search and analysis of the FDA MAUDE database for device-specific reports. Based on

our search findings, we may extend our analysis to all devices within that device’s FDA-assigned product code. The PRN

database was searched by keywords, and the search was limited to the past 10 years.

Healthcare Technology Alerts

We regularly analyze investigation and PRN data to identify trends in use or design problems. When we determine that a

device hazard may exist, we inform the manufacturers and encourage them to correct the problem. ECRI publishes the

resulting safety information about the problem and our recommendations to remediate the problem in a recall-tracking

management service for our members. The Alerts database contains recalls, ECRI exclusive hazard reports, and other safety

notices related to Medical Devices, Pharmaceuticals, Blood Products, and Food Products. This database was searched by

keywords and specific make and model, and the search was limited to the past 10 years.

Material Performance Study - Hyaluronic Acid

Safety Profile – Muscle/Skeletal Applications

Full Name: Hyaluronic Acid

CAS Registry Number: 9067-32-7

Safety Brief - Systematic Review Results

The systematic review included clinical and engineering literature on biocompatibility (i.e., host response and material

response) of Hyaluronic Acid used in medical devices. In addition to fundamental material biocompatibility, we focused on

specific devices known to be made of Hyaluronic Acid. The devices in Table 1 were recommended by FDA CDRH to guide ECRI

in searching this literature and ECRI’s surveillance data.

Table 1: Medical Devices Containing HA for Muscle/Skeletal Applications provided by FDA to Guide ECRI Searches

Regulatory Description

Product Code

Class

Acid, hyaluronic, intraarticular (viscosupplement)

MOZ

III

Filler, Bone Void, Calcium Compound

MQV

Filler, Bone Void, Osteoinduction (W/O Human Growth Factor)

MBP

Prosthesis, Toe, Hemi-, Phalangeal

KWD

The Safety Brief summarizes the findings of the literature search on toxicity/biocompatibility of Hyaluronic Acid.

Inclusion/exclusion criteria and quality of evidence criteria appear in Appendix A in the Appendices below. Quality of evidence

ratings reflected a combination of the quality of comparative data (study designs), quantity of evidence (number of relevant

studies), consistency of evidence, magnitude of effect, directness of evidence, and evidence for a dose response or response

over time. The search strategy appears in Appendix B1, and a flow diagram documenting inclusion/exclusion of studies

appears in Appendix C1. Summary evidence tables with individual study data appear in Appendix D1, and a reference list of

studies cited in the Safety Brief appears in Appendix E.

A summary of our primary findings is shown in Table 2. We then turn to a detailed discussion of research on Hyaluronic Acid

as a material as well as research on the various device categories.

In the discussion section, please note that a statement of “no difference” or “no significant difference” between

devices/materials does not imply equivalence between devices/materials, as studies with low numbers of patients or events

often lack sufficient statistical power to detect a difference between comparators. In addition, when we cite odds ratio(s), an

odds ratio >1 means that the rate was higher in the HA group than in the non-HA group.

Table 2: Summary of Primary Findings from Systematic Review – Muscle/Skeletal Applications

Application

Local Host

Responses/Device Events

Quality of Evidence

(local responses)

Systemic

Responses

Quality of Evidence

(systemic responses)

Hyaluronic Acid as a

material

No studies

Very low (no evidence)

No studies

Very low (no

evidence)

Material Performance Study - Hyaluronic Acid

Application

Local Host

Responses/Device Events

Quality of Evidence

(local responses)

Systemic

Responses

Quality of Evidence

(systemic responses)

Viscosupplementation –

knee

(26 human studies)

Arthralgia/joint pain,

Baker’s cyst, bleeding,

cutaneous vasculitis,

edema (injection-site,

peripheral), effusion,

erythema, heaviness of

injection, joint stiffness,

musculoskeletal pain, pain

at injection site, pain flare

ups, pain post-injection,

purulent aspirate, skin

reaction, swelling,

tenderness of knee joint

Moderate for swelling,

pain at injection site,

arthralgia/joint pain,

and effusion

Low for all other local

responses/device

events

Backache,

cardiac

disorders,

cellulitis,

death

possibly

device-

related, GI

disorders,

headache,

hypersensiti

vity

reaction,

musculoskel

etal and

connective

tissue

disorders,

nausea,

nervous

system

disorders,

paralysis,

rash, renal

and urinary

disorders,

respiratory/

thoracic/

mediastinal

disorders,

septic

shock,

septicemia,

skin

reaction,

skin and

subcutaneo

us tissue

disorders,

pseudosepti

c arthritis

(presenting

with

elevated

CRP, ESR,

PMN;

leukocytosis

and fever),

nasopharyn

gitis,

Low

Material Performance Study - Hyaluronic Acid

Application

Local Host

Responses/Device Events

Quality of Evidence

(local responses)

Systemic

Responses

Quality of Evidence

(systemic responses)

vascular

disorders,

vomiting

Viscosupplementation – hip

(4 human studies)

Hematoma, nausea, pain

flare ups, pain post-

injection, pruritus

Low

No studies

investigated

Very low

Viscosupplementation –

hand and ankle

(2 human studies)

Enlarged lymph node in

ipsilateral groin, pain at

injection site, pseudogout,

swelling

Low for swelling and

pain at injection site

Very low for all other

local responses/device

events

No studies

investigated

Very low

Viscosupplementation –

shoulder

(2 human studies)

Abscess, musculoskeletal

pain, pain at injection site

Low for

musculoskeletal pain

and pain at injection

site

Very low for all other

local responses/device

events

Cancer,

chest pain,

diarrhea, flu

symptoms,

headache

Very low

Viscosupplementation – TMJ

disorders

(2 human studies)

Discomfort post-injection,

ear pressure, pain at

injection site

Low for pain at

injection site

Very low for all other

local responses/device

events

Chills

Very low

Cartilage scaffold

(3 human studies)

Cyst formation, edema,

hematoma, persistent

pain, swelling

Low for swelling

Very low for all other

local responses/device

events

No studies

investigated

Very low

Bone putty/filler

No studies

Very low (no evidence)

No studies

Very low (no

evidence)

GI: gastrointestinal; TMJ: temporomandibular joint

Hyaluronic Acid as a Material

Our literature searches did not identify any studies of these devices that met inclusion criteria.

Viscosupplementation

Due to the primary focus on viscosupplementation in this evidence base, we provide results separately for five indications for

this device category (knee, hip, hand and ankle, shoulder, and TMJ disorders).

Viscosupplementation – knee

26 human studies (16 SRs

2-17

and 10 randomized controlled trials (RCTs)

18-27

). For further information see Table 1 Appendix D.

Material Performance Study - Hyaluronic Acid

Local Responses/Device Events (human studies)

IAHA:

2 SRs addressed the use of IAHA injections. The first SR

investigated fatal adverse events (AEs) from IAHA injections

listed in the MAUDE and Alternative Summary Reporting databases between 2014 and 2019. Brand names searched for in this

review included Durolane, Euflexxa, Gel-One/Gel One/GelOne, Gelsyn, Genvisc/Gen Visc, Hyalgan/Hyalgan LL, Hymovis,

Monovisc, Orthovisc, Sinovial, Supartz/Supartz FX, Suprahyal, Synojoynt, Synvisc/Synvisc-One, Triluron, Trivisc, and Visco-3.

Of 63 unique fatalities identified, 8 (12%) fatalities were possibly IAHA-related. Local responses affiliated with these cases

included pain after injection (3 patients), purulent aspirate (1 patient), and swelling of the knee (2 patients).

The second SR

investigated reporting of pseudoseptic arthritis, a rare complication after IAHA injection. 27 patients (28

knees) with acute pseudoseptic arthritis after HA injection (mostly Synvisc) were identified in 11 studies (5 single arm studies,

6 case reports). Results indicated that 22 cases (78.6%) of pseudoseptic arthritis presented within 24 hours of injection; 15 of

the 22 cases (68.2%) presented within the first 12 hours. Time from injection to presentation ranged from 1 hour to 9 days. 7

cases (25%) occurred after 2 injections, 5 cases (17.9%) after 3 injections, and 13 cases (46.4%) after ≥4 injections. All

patients presented with severe joint pain and effusion. 3 patients had synovial leucocyte elevation in the range typically

concerning for septic arthritis (above 50,000), and 16 patients had elevations in an inflammatory cell count range of 5,000 to

50,000. Results from a separate case report included progressive and increasing pain with obvious suprapatellar effusion at

12 hours post-injection; inflammatory symptoms resolving by 4 weeks.

IAHA versus saline/placebo:

8 studies (4 SRs

12-15

and 4 RCTs

20-23

investigated the safety of IAHA vs. saline/placebo in mostly

middle-aged women with mild-to-moderate OA. Overall, approximately 6,000 individuals enrolled in over 50 RCTs received

IAHA. The SRs and 1 RCT

investigated the safety of several IAHAs, while the remaining RCTs focused on one IAHA

(Synvisc,

Monovisc

21,22

). Dose was reported in 6 (75%) studies; and administrations ranged from 1 to 11 injections. Overall

mean followup was 6 months, and ranged up to 1 year

13,14

and 2 years.

12,23

Studies reported on overall rates for TRAEs,

13,14

AEs,

12,20

device-related AEs,

serious AEs (SAEs),

12,15,20,22,23

and non-serious

AEs.

• TRAEs: 2 SRs focused on TRAEs (not specified) up to 52 weeks.

13,14

The first SR

of 9 RCTs indicated significantly

increased odds of TRAEs with IAHA vs. saline (Odds ratio (OR) 1.78, 95% CI: 1.21 to 2.63), but no significant

differences in any severe AEs (OR 1.08, 95% CI: 0.50 to 2.31). The second SR

of 30 RCTs reported significantly

more TRAEs with IAHA in studies using ≥5 injections vs. saline (RR 1.70, 95% CI: 1.12 to 2.59), but no significant

difference for studies examining 1 injection or 2 to 4 injections vs. saline. This same SR reported no significant

difference in number of patients experiencing a TRAE (RR 1.13, 95% CI: 0.95 to 1.35).

• AEs: 2 studies reported similar overall AEs with IAHA vs. saline.

12,20

• Device-related AEs: 1 RCT reported device-related AEs occurred more frequently with Monovisc vs. saline (13 (7.1%)

vs. 10 (5.4%)).

• SAEs and non-serious AEs: 3 RCTs reported no SAEs

20,22,23

up to 104 weeks. 1 SR

reported no difference in risk of

SAEs (1.8% vs. 1.2%, RR 1.44, 95% CI: 0.91 to 2.26, p=0.12)

12,15

or non-serious AEs (415 HA, 375 placebo; RR

1.03, 95% CI: 0.93 to 1.15) up to 2 years.

Arthralgia and joint swelling were the most common AEs reported in 6 studies.

12,15,20-23

Arthralgia was reported in 5 (83%)

studies.

12,15,20-22

Rates were reported as similar versus saline and ranged from 1.3% to 3.8% up to 26 weeks.

20-22

2 RCTs

reported arthralgia with Synvisc and Monovisc was device-related.

20,22

Timing reported in 2 SRs (rates NR) was immediately

post-injection

and up to 3 days post-injection.

Joint swelling was reported in 4 (66%) studies;

12,15,20,21

one study indicated response was device-related.

Rates ranged from

1.1% to 3.7% with IAHA. 1 RCT reported a higher incidence of joint swelling with Synvisc (3.7% vs. 0.9%),

while another

RCT reported no significant difference in incidence of joint swelling (1.1% Monovisc, 0.5% saline).

Timing reported in 2 SRs

(rates NR) was immediately post-injection

and up to 3 days post-injection.

Less frequently reported AEs were injection site pain,

12,20,23

injection site swelling,

20,23

and joint stiffness.

21,23

Incidence of

injection site pain

was 0.5% with Synvisc,

and 21% with IAHA.

Timing was immediately after

and up to 3 days post

injection (rate NR).

Incidence of

injection site swelling

was 0.5% with Synvisc

and 21% with varying weights of IAHA.

Incidence of

joint stiffness

in the index knee ranged from 0.5% (with Monovisc)

to 7% (with varying weights of IAHA).

Rates were lower vs. saline (1.1%)

and timing was immediately post-injection.

Additional AEs reported by 1 study each included the following:

Material Performance Study - Hyaluronic Acid

• joint pain immediately post-injection

• injection site joint pain (0.5% with Synvisc, 0% placebo)

• device-related non-serious cutaneous vasculitis during 1st week post-injection

• peripheral edema with Synvisc (0.5% HA, 0% placebo)

• effusion immediately post-injection

)

• joint effusion up to 3 days post-injection

• injection site edema with Synvisc(0.5% HA, 0% placebo)

• injection site erythema (12% overall rate for all IAHA weights)

• joint tenderness immediately post-injection

• transient increase in pain immediately post-injection

IAHA versus platelet-rich plasma (PRP):

3 studies (1 SR

and 2 RCTs

18,19

) addressed this topic in 1,137 individuals with mild-

to-moderate knee osteoarthritis (OA). A majority of patients were females with mean age of 60 years. The sample size was

greater than 100 patients for the RCTs,

18,19

and ranged from 55 to 183 in the SR (mean 128 patients).

Dose was not

reported. Administrations were reported in the 2 RCTs as single injections of HA,

and 3 weekly injections of Hyalgan

(FidiaFarmaceutici S.p.A.).

Followup was 6 months

and 12 months.

4,19

Local responses reported in the SR

included pain and swelling immediately after injection. Of the 7 included RCTs (n= 908), 3

RCTs reported no significant differences in AEs, 2 studies reported significantly more post-injection pain and/or swelling with

PRP, while 2 studies reported no AEs.

The first individual RCT

reported no significant differences for mild-to-moderate AEs (musculoskeletal pain, swelling and

tenderness of the knee joint) at 6 months in 110 individuals (55 each arm) receiving single injections of HA or PRP.

Musculoskeletal pain occurred in 4 patients (3 (5%) HA, 1 (1.8%) PRP; timing NR). Mild-to-moderate swelling of the knee

joint occurred similarly at 6 weeks (100%), and a lower incidence with HA at 3 months (85% vs. 96%), and 6 months (87%

vs. 91%). Mild tenderness was reported in 3 patients (2 with HA at baseline and 6 weeks, 1 with PRP at baseline).

The second individual RCT

reported no significant differences in minor complications (stiffness and heaviness of injection

site) immediately post-injection in 13 of 119 patients (3 (6%) HA, 10 (20%) PRP-derived growth factor (PRGF)).

IAHA versus intra-articular oxygen-ozone:

One SR

of 4 RCTs addressed this topic in 298 individuals with end-staged knee OA.

Sample size ranged from 42 to 141 patients. The population was 75% female aged 60 to 71 years. Dose was reported as 10

mg/ml x 2.0 ml (3 studies), and 10 mg/ml x 2.5 ml (1 study). Frequency of injection was not reported. Local responses

included bleeding and skin reaction at 6 to 12 months. No significant differences were reported in incidence rate of AEs (risk

difference 0.006, 95% CI: -0.047 to 0.058).

IAHA versus IAHA (different weights):

2 SRs

6,7

including 40 RCTs compared the safety of IAHA of varying molecular weights in

individuals with knee OA. 1 SR reported the average molecular weight ranged from 500 to 3600 kDa with IAHA mostly being

administered in 3 injections.

Mean followup was 6 months in 1 SR reporting.

One SR

of 30 RCTs examined IAHA injections of high molecular weight (HMW), medium MW (MMW), and low MW (LMW),

and whether the products were processed through extraction of avian-derived molecules (AD-HA) or through bacterial

processes of biological fermentation (Bio-HA). 11 RCTs (n=2094) addressed HMW, 4 RCTs (n=621) addressed MMW, and 15

RCTs (n=2639) addressed LMW. Bio-HA and AD-HA treatments were administered to 1,776 and 3,070 patients, respectively.

Local responses from products such as Euflexxa, Synvisc, Supartz, Orthovisc, and Durolane included effusion after injection

and flare ups. Followup was NR.

Results based on weight: HMW products had the highest rate of injection site flare-ups. Significantly more injection

flare ups with HMW vs. MMW (13.73% vs. 3.31%) and HMW vs. LMW (13.73% vs. 10.73%). Significantly more flare

ups with LMW vs. MMW (10.73% vs. 3.31%). No significant difference was reported in effusion after injection based

on molecular weights (1.9% HMW, 1.7% MMW, 1.8% LMW).

Results based on process: Use of AD-HA products resulted in a significantly lower incidence of acute flare ups (3.0%

vs. 13.2%), but a significantly higher incidence of effusion (3.4% vs. 0.5%) versus Bio-HA.

Another SR

included 10 RCTs in a safety analysis comparing Synvisc of a relatively HMW versus IAHA with LMW (Artzal,

Orthovisc, Bio-HA, Hyalgan, Sinovial, Variofil, Hydros) in 2,616 patients. Authors reported no significant difference in TRAEs at

mean 26.8 weeks (14.7% Synvisc vs. 11.6% LMWHA; risk difference 0.02, 95% CI: -0.01 to 0.05; p=0.13) or number of

TRAEs (23.2% Synvisc vs. 16.3% LMWHA; risk difference 0.03, 95% CI: -0.01 to 0.07; p=0.14).

Material Performance Study - Hyaluronic Acid

IAHA versus anti-inflammatories

: Of 4 SRs addressing this comparison,

8-11

3 SRs examined local responses with IAHA,

8,10,11

while 1 SR only examined systemic responses (see below).

2 SRs examined IAHA versus NSAIDs.

8,11

The first SR of 6 RCTs

included 831 middle-aged patients with mild-to-moderate

knee OA. Sample size was 60 to 327 patients. Administrations were one injection (Durolane), 3 weekly injections (Synvisc, and

Suplasyn), and 5 weekly injections (Hyalgan and Suvenyl). Responses from 5 weeks to 26 weeks included arthralgia, injection

site pain, and lower leg effusion (1 (0.3%) HA, 2 (0.5%) NSAID). Results from a meta-analysis of 5 RCTs indicated:

• no significant difference in risk of serious AEs (1.2% vs. 0.9%, risk ratio 1.37, 95% CI: 0.26 to 7.14),

• a significantly lower risk of AEs with HA (19.8% vs. 29%; risk ratio 0.74, 95% CI: 0.59 to 0.94),

• a significantly higher risk of arthralgia with HA (8.1% vs. 2.9%).

The second SR of 5 RCTs (n=712) reported no IAHA-related AEs up to 12 weeks.

TRAEs were reported in one SR of 12 RCTs (n=1794)

comparing IAHA with intraarticular corticosteroids (CS; triamcinolone

hexacetonide, methylprednisolone acetate, 6-methylprednisolone acetate, triamcinolone acetonide). Sample size was 41 to

391 patients. IAHA doses ranged from 16 to 60 mg, while administrations ranged from 1 to 5 weekly injections. Knee pain,

joint swelling, and joint stiffness were reported from 1 to 6 months. Results indicated significantly more TRAEs with HA (746

events vs. 606 events; risk ratio 1.66, 95% CI: 1.34 to 2.06) which may have been caused by the higher injection frequency

IAHA versus miscellaneous treatments:

6 studies (2 SRs,

16,17

and 4 RCTs

24-27

addressed this comparison. One study each

compared IAHA with usual care,

debridement,

cooled radiofrequency ablation (CRFA),

acupotomy,

and

polydeoxyribonucleotide (PDRN).

One study compared IAHA with physical therapy (PT), dextrose prolotherapy, and

botulinum neurotoxin.

Over 100 patients were examined in each study; individuals with mild-to-moderate knee OA,

26,27

or at

least moderate knee OA

24,25

in studies reporting. Dose was reported in all studies, and administrations ranged from 1 to 5

injections. Followup was mostly 6 months and 12 months.

One study noted no device-related responses with Hyalgan vs. PT, dextrose prolotherapy, and botulinum neurotoxin.

2 RCTs

reported fewer TRAEs with IAHA. One RCT

noted a TRAE rate of 14% with Synvisc-One versus 19% TRAE rate with CRFA up

to 6 months. While procedural pain (3 (3%) in both arms), and post-procedural pain (3 (3%) with Synvisc-One, 7 (8%) with

CRFA) occurred in both arms, Baker’s cyst (a fluid-filled cyst that causes a bulging and feeling of stiffness behind the knee)

only occurred in 2 (2%) patients with Synvisc-One. The other RCT (n=120)

reported pain at injection site in 8 (13.4%)

patients after IAHA vs. pain and mild effusion in 13 (26%) patients after debridement up to 6 months.

The remaining 3 studies reported a similar incidence of TRAE/AEs. One SR of 5 RCTs

reported a similarly low incidence of

TRAEs (unspecified) with HA and PDRN up to 12 months. One SR of 4 RCTs (n=309) reported AE incidence rates of 2.2% with

HA, and 4.3% with acupotomy up to 6 months. AEs included redness and swelling post-injection in 3 patients with each

treatment.

1 RCT comparing 3 weekly injections of Synvisc (HMW-HA) vs. usual care

in 156 patients reported similar

overall TRAEs (77 Synvisc, 79 UC). Events, described as flare-ups or “other”, were reported at 6 weeks, 13 weeks, 26 weeks,

39 weeks, and 52 weeks. Flare-up rates with Synvisc were highest at 6 weeks (36%) and lowest at 52 weeks (6%).

Systemic Responses

IAHA:

Both SRs addressing this topic reported systemic responses. The first SR

identified 63 unique fatalities, 8 fatalities were

possibly IAHA-related. Of the 8 fatalities, 4 patient’s experienced local responses such as pain and swelling (see above) prior

to death. 1 death involved a suicide due to worsening pain. The remaining 4 patients experienced the following systemic

responses prior to death:

• septic shock, followed by paralysis, and death 4 months post-injection.

• necrotizing fasciitis after starting chemotherapy and receiving an HA injection.

• septicemia, cause of death unknown.

• cellulitis.

Authors noted that the available information was “insufficient to make a firm determination of causality.”

The second SR

reported results from 27 patients (11 studies) with acute pseudoseptic arthritis after mostly Synvisc

injections. Systemic responses included elevated CRP above 50, elevated ESR, elevated PMNs above 75% in 10 (37%)

patients, fever in 6 (22%) patients, and leukocytosis above 10,000 in 4 (15%) cases. Results from a separate case report

included a cell count of 123,260 and white blood cell count of 15.5.

Material Performance Study - Hyaluronic Acid

IAHA versus saline/placebo:

Of the 8 studies addressing this comparison, 3 studies investigated systemic responses.

13,15,22

SR, focused on TRAEs,

reported no significant differences in any system-organ class(SOC)-related AEs (GI, cardiac, vascular,

respiratory, nervous system, skin and subcutaneous tissue disorders, musculoskeletal, renal and urinary disorders, and

hypersensitivity reaction).

• GI disorders: 69 HA, 83 placebo; OR 0.81, 95% CI: 0.52 to 1.27

• cardiac disorders: 5 HA, 4 placebo; OR 1.25, 95% CI: 0.36 to 4.41

• vascular disorders: 5 HA, 3 placebo; OR 1.70, 95% CI: 0.39 to 7.29

• respiratory, thoracic and mediastinal disorders: 71 HA, 52 placebo; OR 1.21, 95% CI: 0.82 to 1.78

• nervous system disorders: 66 HA, 55 placebo; OR 1.15, 95% CI: 0.77 to 1.70

• skin and subcutaneous tissue disorders: 22 HA, 10 placebo; OR 1.71, 95% CI: 0.52 to 5.63

• musculoskeletal and connective tissue disorders: 145 HA, 149 placebo; OR 0.99, 95% CI: 0.71 to 1.39

• renal and urinary disorders: 6 HA, 12 placebo; OR 0.54, 95% CI: 0.21 to 1.41

• hypersensitivity reaction: 23 HA, 19 placebo; OR 0.64, 95% CI: 0.05 to 7.94

IAHA versus PRP:

One RCT reported nasopharyngitis in both arms (2 HA, 1 PRP); and backache (2 HA) and headache (1) only

with HA.

One SR investigated but did not identify any systemic responses.

IAHA versus anti-inflammatories

: 3 of 4 SRs addressing this comparison investigated systemic responses. One SR

comparing

IAHA with intraarticular methylprednisolone only reported on systemic responses. Results from 4 of 5 included RCTs indicated

headache, nausea, and vomiting with no significant difference in AE incidence up to 52 weeks (risk difference -0.042, 95% CI:

-0.092 to 0.009). One SR

(n=831) reported a significantly higher risk of headache (8.4% vs. 4.4%) with HA and a rare

occurrence of cardiac-related complications (1 (0.3%) HA, 0 NSAID). Lastly, one SR (n=1794) did not identify any treatment-

related systemic responses.

1 RCT reported rash in 1 patient with Monovisc, and indicated this response was device-related.

Lastly, 1 SR of 10 RCTs

indicated 1 serious AE attributed to HA was skin reaction characterized by peeling of skin on hands and toes in 1 patient. This

reaction occurred 8 days post-injection.

IAHA versus miscellaneous treatments:

1 of 6 studies addressing this comparison reported systemic TRAEs included GI

complaints from 6 weeks to 52 weeks. Rates ranged from 3% (at 26 weeks) to 7% (at 6 weeks).

Overall Quality of Evidence

The evidence for swelling, pain at injection site, arthralgia/joint pain, and effusion was consistently reported across high

quality studies, in agreement with reporting with other HA categories (e.g., viscosupplementation-hand and ankle,

viscosupplementation-shoulder, viscosupplementation-TMJ; and cartilage scaffold), so the quality of evidence is moderate. For

other local responses/events and systemic responses, the quality of evidence is low due to the large evidence base of high

quality studies.

Viscosupplementation - hip

4 human studies (4 SRs

28-31

). For further information see Table 2 Appendix D.

Local Host Responses (human studies)

One SR

included 5 RCTs (n = 591) comparing HA (n=298) with placebo injection (n=293) for hip OA. The HA products used

in the studies were Durolane 3 ml, hylan G-F 20 6 ml, Hyalubrix 8 ml, Hyalgan 6 ml, and Adant 2.5 ml, and follow-up ranged

from 56 to 182 days. No meta-analysis of AEs was performed. The SR lists numbers of patients in each study who

experienced slight or moderate pain flare during or after injection, permitting calculation of the following pooled AE rates: 21

of 298 patients in the HA groups (7%) and 7 of 293 patients in the placebo groups (2.4%). Pruritus and hematoma at the

injection site was reported in 1 patient each in the HA group.

Another SR

included 9 RCTs (n=1,164) comparing HA (n=558), methylprednisolone (n=201), PRP (n=115), mepivacaine

(n=20), and placebo (n=270) for hip OA. The HA products used in the studies were Durolane 3 ml, Hyalubrix 5 ml (2 studies),

Synvisc 6ml, Hyalubrix 8 ml, Hyalgan 6 ml, Adant 2.5 ml, and hylan G-F 20 2 ml; follow-up ranged from 2 to 12 months. Meta-

analysis of overall AEs was performed for comparisons of HA with placebo and methylprednisolone; otherwise, rates of overall

AEs were reported by study, permitting calculation of pooled rates. A 29.0% overall AE rate was reported for patients in the

HA groups across 9 studies (range: 0% to 49.7%). The rate was 23.7% for patients in the placebo groups across 4 studies

Material Performance Study - Hyaluronic Acid

(range: 0% to 34.9%), 8.7% for patients in the PRP groups across 3 studies (range: 0% to 20%), 38.3% for patients in the

methylprednisolone groups across 3 studies (range: 0% to 48.7%), and 5% for patients in the mepivacaine group in 1 study.

In the meta-analyses, no statistically significant differences in rates of overall AEs were found for HA versus placebo or for HA

versus methylprednisolone.

Another SR

included 4 RCTs (n=303) comparing HA (n=148) with PRP (n=155) for hip OA. The SR did not specify the

products used in the studies. Follow-up time was 12 months. Meta-analysis revealed a relative risk of nausea of 1.15 (95% CI:

0.34 to 3.93) for PRP relative to HA. One included study reported significantly higher post-injective pain for PRP than for HA.

Another SR

included 6 RCTs (n=630) comparing HA (n=291) with placebo (n=114) or methylprednisolone (n=207) for hip

OA. The HA products used in the studies were hylan G-F 20 2 ml, Hyalubrix 8 ml, Ostenil 2 ml, Durolane 3 ml, Adant 2.5 ml,

and Hyalgan 6 ml. (One included study compared Ostenil with hylan G-F 20; the SR appears to have treated hylan G-F 20 as a

control treatment in this study, although it treated the same product as HA in another included study.) Follow-up ranged from

56 to 180 days. Four of the six included studies reported AEs by treatment group; meta-analysis compared overall AEs in the

HA groups versus the control groups (apparently both placebo and control treatments) and obtained a relative risk of 0.94

(95% CI 0.41 to 2.20).

Systemic Responses

We did not identify any studies reporting systemic responses to HA as viscosupplementation in hips.

Overall Quality of Evidence

The 4 systematic reviews included high quality studies with a high number of patients, and most outcomes were in agreement

with other device categories. Since reporting of specific responses was limited to one systematic review each, the overall

quality of evidence was rated low. Since systemic responses were not investigated in these studies, the evidence is very low.

Viscosupplementation – hand and ankle

2 human studies (2 SRs

32,33

). For further information see Table 3 Appendix D.

Local Responses/Device Events (human studies)

One SR

included 24 studies (n = 844) of intra-articular injective treatments for ankle lesions, 18 of which examined HA. Of

the 24 studies, 8 were RCTs, 1 was a nonrandomized comparative study, and 15 were single-arm studies. The HA products

used in the studies were Euflexxa 6 ml (2 studies), Durolane 1ml, Suplasyn 6 ml, Hanox M-XL 2.2 ml, Hyalgan 2 ml to 10 ml

(4 studies), Synvisc 2 ml to 6 ml (3 studies), Supartz 2.5 ml to 12.5 ml (2 studies), Orthovisc 1 ml to 3 ml, Adant 7.5 mg, and

unspecified HA products 6 ml to 12.5 ml (2 studies). 581 patients received HA; other treatments were PRP (n = 83),

methylprednisolone (n = 48), botulinum toxin type A (n = 38), prolotherapy (n = 27), exercise therapy (n = 15),

mesenchymal stem cells (n = 6), and placebo/saline (n = 46). Follow-up ranged from 1 to 45.5 months. AEs with HA in 1

patient each included enlarged lymph node in ipsilateral groin, osteochondritis dissecans (not considered treatment-related),

and pseudogout. The SR also noted that “in some cases, pain and swelling were reported at the injection site soon after the

injection.”

The other SR

included 13 studies (n = 809) of intra-articular therapies for hand OA; 9 studies examined HA. Of the 13

studies, 12 were RCTs and one was a nonrandomized comparative study. The HA products used in the studies included hylan

G-F 20 1 ml to 2 ml (3 studies) and unspecified sodium hyaluronate products 1.5 ml to 3 ml (6 studies). 360 patients received

HA; other treatments were corticosteroids (n = 280), dextrose (n = 30), infliximab (n = 10), and placebo (n = 172). (These

numbers total more than 809 because two studies were within-subject and each patient in those studies [33 HA/placebo, 10

infliximab/placebo] is counted as having received both treatments.) Follow-up time was 24 weeks. Of the nine studies

including HA, one did not report numbers of patients experiencing AEs. In the 8 remaining studies, AE rates were as follows:

HA, 11.5%; corticosteroid, 9.0%; placebo, 2.7%. Numbers of patients with specific types of AE were not reported, either

overall or by group; AEs were described as minor, and included surgeries unrelated to study medication, pain and local

swelling, skin and nail abnormalities, heat and/or redness, and unspecified “local AEs.”

Systemic Responses

We did not identify any studies reporting systemic responses to HA as viscosupplementation in hands and ankles.

Material Performance Study - Hyaluronic Acid

Overall Quality of Evidence

The evidence for swelling and pain at injection site was reported in 2 SRs that included numerous high-quality studies. These

outcomes were also in agreement with other indications for viscosupplementation (e.g., knee, shoulder, TMJ), so we rated the

quality of evidence as low. For other local responses and systemic responses, the quality of evidence is very low.

Viscosupplementation – shoulder

2 human studies (2 SRs

34,35

). For further information see Table 4 Appendix D.

Local Host Responses (human studies)

One SR evaluated the safety of HA in patients 18 years of age and older with glenohumeral OA. Follow-up time ranged from

12 weeks to 3 years and sample size for the HA arms of the studies varied from 15 to 265, with 50.5% to 76% male patients.

Common AEs from 13 of the 15 included studies (5 RCTs, 10 single arm) were musculoskeletal pain, and pain at injection with

HA.

Serious AEs such as severe musculoskeletal pain and abscess were also reported. Similar AEs were reported in

individuals receiving intra-articular injections of corticosteroids or saline. The overall pooled AE rate (local and systemic) was

33.92% (406 of 1197 patients) and the serious AE rate was 5.35% (64 of 1197). Most events were deemed by study

investigators to not be product-related.

Another SR of 4 RCTs investigated the safety of HA administration versus conventional adhesive capsulitis (AC) therapies

(e.g., intra-articular corticosteroid injection and physical therapy) versus HA administration plus conventional therapy

regimens in patients with AC.

Follow-up ranged from 2 weeks to 6 months. Doses ranged from 20 to 30 mg, while

administrations ranged from 2 to 8 total injections. Mean patient age ranged from 54.5 to 64.2 years, and 69% were female.

2 studies reported no major AEs, and 1 study reported intra-procedural pain in 12 out of 45 participants undergoing capsular

distension combined with intra-articular HA injection. It was unclear if the pain was a result of the capsular distension

procedure, the intra-articular HA injection, or both.

Systemic Responses

Diarrhea, flu symptoms, and headache were commonly reported (data not provided) in 1 SR with up to 3 years followup.

Serious AEs such as chest pain and cancer were also reported (N not reported). Most responses were deemed by study

investigators to not be product related.

Overall Quality of Evidence

Musculoskeletal pain and pain at injection site were both commonly reported in 2 high-quality studies and in agreement with

other indications for viscosupplementation (e.g., knee, hand and ankle, TMJ), so we rated the quality of evidence as low. For

other local responses and systemic responses, the quality of evidence is very low.

Viscosupplementation – TMJ disorders

2 human studies (2 SRs

36,37

). For further information see Table 5 Appendix D.

Local Responses/Device Events (human studies)

Studies analyzed HA injections as viscosupplementation versus other types of intra-articular injections (corticosteroid [CS],

PRP, or platelet-rich growth factor [PRGF]) in the treatment of TMJ disorders. Dose ranged from 0.5 to 5 ml with single

administration in both studies. Follow-up was 1 week to 8 years

and 12 months.

Sample size was 13 to 102 patients with a

mean age of 25 to 50 years.

Pain at injection site: Three RCTs included in a SR

reported TMJ pain post-injection with HA compared to CS. Zero percent to

37% of patients receiving HA reported pain post-injection versus 0 to 70% patients reporting pain after CS injection (OR 2.98,

95% CI: 0.08 to 111.17). Overall, the review found no relevant difference between the HA and CS treatment groups in terms

of adverse events. The SR comparing HA to PRP or PRGF

included one RCT reporting AEs, assessing that 60% percent of

patients receiving HA reported pain during injection versus 88% of patients receiving PRP or PRGF.

Ear pressure: One RCT included in a SR

reported 5% of HA patients with ear pressure compared to 10% of CS patients (OR

= 2.11, 95% CI: 0.18 to 25.35) at two-week follow-up.

Material Performance Study - Hyaluronic Acid

Post-operative discomfort: The SR comparing HA to PRP or PRGF

included one RCT reporting AEs, assessing that 8% of

patients receiving HA reported post-operative discomfort versus 76% of patients receiving PRP or PRGF.

Systemic Responses

An RCT included in a SR

reported slight chills at 1 week follow-up for HA compared to CS. Zero percent of CS patients

reported chills versus 5.7% of patients receiving HA (OR=0.24, 95% CI: 0.01 to 5.90; p=0.36). Overall, the review found no

relevant difference between the HA and CS treatment groups in terms of AEs.

Overall Quality of Evidence

Pain at injection site was reported by both SRs and in agreement with other indications for viscosupplementation (e.g., knee,

shoulder, hand and ankle), so we rated the quality of evidence as low. For other local responses and systemic responses, the

quality of evidence is very low.

Cartilage scaffold

3 human studies (1 RCT,

and 2 nonrandomized comparative studies

39,40

). For further information see Table 6 Appendix D.

Local Host Responses (human studies)

The RCT compared polyglycolic acid and hyaluronan matrix-augmented bone marrow stimulation (m-BMS) with

Chondrotissue (BioTissue AG) after microfracture (MF) to MF alone in 24 patients with grade III/IV International Cartilage

Repair Society cartilage degradation.

One patient in the m-BMS group reported an infected hematoma up to 108 weeks.

Three patients in the m-BMS group reported mild swelling after 2 weeks, which resolved by 6 weeks; swelling did not occur in

the MF group. There was 1 severe effusion after 6 weeks in the MF group, and none in the m-BMS group. Neither group had

allergic reactions (moderate or severe).There were no events that were more likely in the group receiving HA (the m-BMS

group).

Two nonrandomized comparative studies compared HA-based cell-free scaffold with Hyalofast (Anika Therapeutics) with a

chitosan-glycerol phosphate/blood implant (BST-CarGel®, Piramal Life Sciences)

or MF alone

in patients with focal

osteochondral lesions of the knee joint (Outerbridge grade III or IV).

The first study enrolled 21 patients in the Hyalofast scaffold group and 25 in the BST-CarGel implant group.

At mean 24

months follow-up, edema or cyst formation in the subchondral bone was detected in 8 (38%) knees with Hyalofast, and 10

(40%) knees with BST-CarGel). One patient from each group had persistent pain as well as early degenerative changes of the

knee joint and planned to undergo replacement surgery at the latest follow-up. Deep venous thrombosis, septic arthritis,

neurovascular complication, or intra-articular adhesion were not detected in any patients.

The second study enrolled 24 patients in the MF group (mean age 43±6.8 years) and 19 in the Hyalofast group (mean age

40±9.8 years), 62.8% female overall.

1 patient from the HA scaffold group and 2 patients from the MF group had persistent

pain as well as early degenerative changes of the knee joint up to mean 25 months.

Systemic Responses

We did not identify any studies investigating systemic responses to HA as cartilage scaffolds.

Overall Quality of Evidence

None of the three studies found any higher rates in groups receiving HA, but they may not have been large enough to detect

effects. The evidence for swelling was reported in 1 high-quality study and in agreement with reporting with other HA

categories (e.g., viscosupplementation-knee, and viscosupplementation-hand and ankle), so the quality of evidence is low. For

other local responses/device events and systemic responses (no studies investigated), the quality of evidence is very low.

Bone putty/filler

Our literature searches did not identify any studies of these devices that met inclusion criteria.

ECRI Surveillance Data

Refer to Appendix F for a list of devices that guided our searches of ECRI Surveillance Data.

Material Performance Study - Hyaluronic Acid

Patient Safety Organization

Search Results: ECRI PSO identified 115 reports that involved HA related to muscle/skeletal applications that occurred

between April 2016 and November 2021. Seven of these involved pertinent events. The events included: 1) Wrong side

procedure - 2 (28.5%), 2) Wrong product - 2 (28.5%), 3) Product expired - 2 (28.5%), 4) Wrong time – 1 (14.2%). Table 3

and Table 4 outline the complications and harm scores of these events, respectively.

All individual PSO event reports are redacted and included in Appendix F.

Table 3" Complications in HA-related PSO Event Reports – Muscle/Skeletal Applications

Complication

Viscosupplement

Bone Filler

Total

Wrong side procedure

Wrong product

Product expired

Wrong time

Total

Table 4: Harm Score Associated with HA-related PSO Event Reports – Muscle/Skeletal Applications

Harm Scores

(NCC-MERP)

Harm Scores (NCC-MERP)

Viscosupplement

Bone Filler

Total

No Error

Error, No Harm

Error, Harm

Error, Death

Material Performance Study - Hyaluronic Acid

Harm Scores

(NCC-MERP)

Harm Scores (NCC-MERP)

Viscosupplement

Bone Filler

Total

NULL*

Total

*Harm score was not reported

Accident Investigations

Search Results: Zero investigations were recovered from the accident investigations database.

ECRI Problem Reports

Search Results: The search returned zero reports submitted by ECRI members.

Healthcare Technology Alerts

Search Results: The search returned 5 manufacturer issued alerts describing problems with adverse events, contamination,

packaging issues, and impartially filled products, summarized in Table 5.

Table 5: Summary of Regulatory and Manufacturer Alerts - Muscle/Skeletal Applications

Device Type

# Alerts

Reported Problem

Viscosupplemenation:

MOZ (Acid, Hyaluronic,

Intraarticular)

2 manufacturer issued

• Reports of post-injection pain and swelling

• Contaminated product may lead to infection

Bone Filler:

MQV (Filler, Bone Void,

Calcium Compound)

2 manufacturer issued

• Impartially filled tubes may delay surgery

• Unsealed Tyvek packaging

Bone Filler:

MQV (Filler, Bone Void,

Calcium Compound);

MBP (Filler, Bone Void,

Osteoinduction (w/o

Human Growth Factor)

1 manufacturer issued

• Incomplete outer packaging seal

Potential Gaps

ECRI surveillance searches reflect mostly acute patient incidents that involved medical devices made of Hyaluronic Acid. Areas

of particular concern involve incidents that result in direct tissue exposure to the material if there is moderate to high-quality

evidence of acute or systemic reaction to this exposure, as determined by the systematic review. Topics with very low or low

quality of evidence represent areas of potential gaps in the literature. If the literature revealed areas of new concern (e.g.,

systemic response to long-duration contact) and there is little supporting evidence, these are considered gaps.

Material Performance Study - Hyaluronic Acid

HA as a Material

There were no studies that met inclusion criteria for HA as a material indicating an area of future research.

Viscosupplementation

14 HA products were investigated in 36 studies addressing five various indications (knee, hip, hand/ankle, shoulder, and TMJ

disorders). Hyalgan and Synvisc were investigated in 15 studies and 16 studies, respectively. 9 products (Monovisc, HYADD,

Sinovial, Gel-One, Supartz, Go-On, Hanox M-XL, Ostenil, and Hyalubrix) were investigated in ≤3 studies. Most studies were

high-quality and enrolled a large number of patients. In general, studies enrolled middle-aged women with mild-to-moderate

OA. Several studies did not report important study characteristics (e.g., HA products investigated, HA dose), or details on

clinical results (e.g., type of TRAE, number of events, number of patients experiencing events, timing of events).

Viscosupplementation – knee

26 studies (16 SRs, 10 RCTs) focused on knee viscosupplementation. Several outcomes such as swelling and pain at injection

site were consistently reported across studies and in agreement with other HA categories (viscosupplementation in

hands/ankles, shoulders, TMJ disorders; and scaffold), so we rated the quality of evidence as moderate. Other local responses

(e.g., flare ups, edema) and all systemic responses (e.g., cellulitis, nervous system disorders) were rated low quality due to

infrequent or rare reporting. Of 11 (42%) studies investigating systemic responses, 2 systemic skin responses (1 rash, 1

peeling of skin on hands and toes) were reported as device-related. Only 1 systemic response (GI disorders/complaints) was

reported in more than 1 study, but this outcome occurred similarly versus placebo or usual care so in both instances, the

association with HA is unclear.

Viscosupplementation – hip

4 SRs (n=2688) investigated HA products such as Durolane, Synvisc, Hyalbrix, Hyalgan, Adant. 3 SRs included placebo

injections as comparators. Quality of evidence for local responses was rated low due to the high number of patients enrolled

and similar reporting of outcomes (e.g., pain flare up, pain post-injection, hematoma) versus other categories. Systemic

responses were not investigated resulting in a rating of very low quality.

Viscosupplementation – hand and ankle

2 SRs (n=1653) investigated HA products such as Euflexxa, Durolane, and Suplasyn. 1 study each examined

viscosupplementation in hands and ankles, comparators were varied (e.g., PRP, methylprednisolone, botulinum toxin A,

prolotherapy, corticosteroids, and placebo). 1 SR of 24 studies only included 9 (37%) studies with control groups and followup

was as short as 1 month. Evidence for local responses such as swelling and pain at injection were rated low quality, while

other local responses and systemic responses (not investigated) were rated very low quality.

Viscosupplementation – shoulder

2 SRs addressed this indication in patients with glenohumeral OA and adhesive capsulitis. 1 SR only included 5 (33%) studies

with control groups. This SR pooled the AE rate for local and systemic events making it difficult to determine the AE rates

separately. This same SR did not report data for several systemic responses including serious AEs such as cancer and chest

pain, however study investigators deemed “most events” to not be product-related. The other SR reported intra-procedural

pain in 12 patients however it was unclear whether the pain was a result of the capsular distension procedure, the IAHA

injection, or both. Two local responses (musculoskeletal pain, pain at injection) were rated low quality. Other local responses

and systemic responses were rated very low quality.

Viscosupplementation – TMJ disorders

2 SRs addressing this indication reported 3 local responses including pain at injection site, ear pressure, and post-operative

discomfort. Studies reported events as occurring within 2 weeks of IAHA injections, however these responses were also

reported after other types of IA injections (CS, PRP, and PRGF) making the association with HA unclear. Pain at injection site

was reported by both SRs and in agreement with other indications for viscosupplementation (e.g., knee, shoulder, hand and

ankle), so we rated the quality of evidence as low. For other local responses and systemic responses, the quality of evidence

is very low.

Material Performance Study - Hyaluronic Acid

Cartilage scaffold

Evidence for HA scaffolds was limited to 3 studies enrolling 75 patients in 2 (66%) scaffolds of interest. Evidence for Agili-C

was lacking, and dose was NR for any study. 1 RCT reported the placement of an HA scaffold Chondrotissue in only 12

patients, while 2 nonrandomized comparative studies (led by the same investigator) reported on Hyalofast in 42 patients

(possibly including duplicate patients). These latter 2 studies both reported persistent pain, which is inconsistent with

reporting in other studies. The evidence for swelling however was reported in the RCT and was in agreement with reporting

with other HA categories (e.g., viscosupplementation-knee, and viscosupplementation-hand and ankle), so the quality of

evidence is low. For other local responses and systemic responses (no study investigating), the quality of evidence is very low.

Bone putty/filler

There were no studies that met inclusion criteria for HA bone putty/filler devices indicating an area of future research.

Safety Profile – Dermal, Facial and Eye Applications

Full Name: Hyaluronic Acid

CAS Registry Number: 9067-32-7

Safety Brief - Systematic Review Results

The systematic review included clinical and engineering literature on biocompatibility (i.e., host response and material

response) of Hyaluronic Acid used in medical devices. In addition to fundamental material biocompatibility, we focused on

specific devices known to be made of Hyaluronic Acid. The devices in Table 6 were recommended by FDA CDRH to guide ECRI

in searching this literature and ECRI’s surveillance data.

Table 6: Medical Devices Containing HA for Dermal, Facial and Eye Applications provide by FDA to Guide ECRI Searches

Regulatory Description

Product Code

Class

Aid, Surgical, Viscoelastic. (Intraocular fluid)

LZP

Implant, Dermal, For Aesthetic Use

LMH

Implant, Dermal, For Aesthetic Use In the Hands

PKY

Products, Contact Lens Care, Rigid Gas Permeable

MRC

Accessories, Soft Lens Products

LPN

The Safety Brief summarizes the findings of the literature search on toxicity/biocompatibility of HA for dermal, facial and eye

applications. Inclusion/exclusion criteria and quality of evidence criteria appear in Appendix A in the Appendices document.

Quality of evidence ratings reflected a combination of the quality of comparative data (study designs), quantity of evidence

(number of relevant studies), consistency of evidence, magnitude of effect, directness of evidence, and evidence for a dose

response or response over time. The search strategy appears in Appendix B2, and a flow diagram documenting

Material Performance Study - Hyaluronic Acid

inclusion/exclusion of studies appears in Appendix C2. Summary evidence tables with individual study data appear in Appendix

D2, and a reference list of studies cited in the Safety Brief appears in Appendix E.

A summary of our primary findings is shown in Table 7. We then turn to a detailed discussion of research on HA for dermal

fillers, intraocular/ophthalmic viscoelastic solution/fluid, and eye drops.

In the discussion section, please note that a statement of “no difference” or “no significant difference” between

devices/materials does not imply equivalence between devices/materials, as studies with low numbers of patients or events

often lack sufficient statistical power to detect a difference between comparators. In addition, when we cite odds ratio(s), an

odds ratio >1 means that the rate was higher in the HA group than in the non-HA group.

Table 7: Summary of Primary Findings from Systematic Review - Dermal, Facial and Eye Applications

Application

Local Host

Responses/Device Events

Quality of Evidence

(local responses)

Systemic

Responses

Quality of Evidence

(systemic responses)

Dermal Fillers

Lumpiness, tenderness,

pain, swelling, bruising,

hematoma, edema,

numbness, irregular

surface, disfiguring

surface, nodules,

granuloma, necrosis,

migration, vascular

adverse events, infection,

erythema, delayed

inflammatory reactions,

delayed-type

hypersensitivity, partial or

total vision loss, visual

changes, brain infarcts or

hemorrhage, capsular

contraction, early

resorption, embolism

Moderate

ASIA

Low

Intraocular/Ophthalmic

Viscoelastic Solution/Fluid

IOP elevations, corneal or

macular edema, ocular

hypertension,

inflammation, capsule

break, Viscoat bubbles,

layered hyphema,

choroidal effusion,

clogged tube shunt, TASS,

Descemet membrane

detachment

Moderate for IOP

elevation and corneal/

macular edema

Low for all other

responses/events

Not

evaluated

Very low

Eye Drops

Itching/stinging/irritation/

erythema/keratitis, eye

pain, eye disorders/dry

eye,

hyperemia/hemorrhage,

Moderate

Not

evaluated

Very low

Material Performance Study - Hyaluronic Acid

Application

Local Host

Responses/Device Events

Quality of Evidence

(local responses)

Systemic

Responses

Quality of Evidence

(systemic responses)

infection/viral

conjunctivitis

ASIA = autoimmune/inflammatory syndrome induced by adjuvants; IOP = intraoperative pressure; TASS = toxic anterior

segment syndrome

Dermal Fillers

The literature search identified 33 relevant human studies (6 systematic reviews (SRs),

41-46

11 randomized controlled trials

(RCTs),

47-57

4 comparative observational studies,

58-61

12 single-arm

62-73

) and 1 comparative observational animal study.

For

further information see Tables 1 and 2 in Appendix D.

The human studies examined over 70 different HA dermal fillers including Aesthefill, Allaxin, Bellast, Belotero, BioHyalux, a

biphasic HA filler (NASHA), Captique, CPMHA, Dermalax implant plus, Dermicol-P35, DGE, DHF-001, DIVAVIVA, Elravie,

Emervel Classic, HA (Titan) + Infrared, HA-biphasic, HA-G-monophasic, HA-monophasic, HA-P-biphasic, Hyacorp, Hylaform,

hylan B plus, Juvederm, Juvina, Koken, Macrolane, Matridex (HA and dextranomer), Matrifill, Modelis, Neuramis, Perlane, PP-

501-A/B, Prevelle SILK, Princess Volume, Puragen, Redensity II Eyes, Restylane, RHA 4, Surgiderm, Teosyal, Terafill, Uma

Jeunesse, YVOIRE volume plus, and Zyplast. Non-HA fillers examined in these studies included calcium hydroxyapatite (CaHa;

Radiesse), autologous fibroblast and keratin gel (AFKG), collagen-based fillers, Ial System Duo, polycaprolactone (PCL),

autologous fat, lipofilling, polyalkymide, poly-L-lactic acid, and acrylamide.

The 1 animal study compared the HA filler Restylane with polymethyl methacrylate (PMMA).

Local Responses/Device Events (human studies)

Comparisons of HA fillers to non-HA fillers: One SR

included 28 studies of various dermal fillers for treatment of tear trough

deformity and reported adverse events for HA and non-HA fillers. Overall complication rates were statistically significantly

different between HA and non-HA fillers (p <0.0001), but the clinical significance is unclear since the complications are minor:

HA (50.75%); calcium hydroxylapatite (CaHa, 19.95%); autologous fibroblast and keratin gel (AFKG, 11.43%); and collagen-

based filler (90%). Prolonged edema was more likely to occur among patients receiving HA injections (11.2%) than in patients

injected with the other materials (0 to 2.1%, P < 0.00001). The Tyndall effect and lumpiness were only reported in the HA

group. Thirty of 31 patients who were noted to have product migration were in the HA group, an event rate of 1.96%

(30/1585 patients) in the HA group. However, the number of patients who received AFKG or collagen-based filler was very

low, so for these fillers some complications may not have been captured simply due to the low patient numbers. Most of the

event rate comparisons were indirect (not compared within the same studies) and therefore could have been confounded by

differences in patient characteristics across different studies.

Another SR

included 51 studies that evaluated dermal fillers (HA and non-HA) for treatment of nasolabial folds. Overall

complication rates were 0.59 (95% CI: 0.46–0.72) for all HA fillers, 0.4 (95% CI: 0.12–0.7) for all collagen fillers, 0.82 (95%

CI: 0.69–0.93) for Mesoglow, and 0.88 (95% CI: 0.72–0.99) for IAL-systems. However, the number of studies reporting on

non-HA fillers was relatively small compared to the number of HA studies. The most common events associated with HA

injections included lumpiness (43%), tenderness (39%), swelling (32%) and bruising (28%).

One RCT

with 37 patients compared HA (Juvederm Ultra 4) to CaHa for hand augmentation (randomization was by hand; all

37 patients received HA in one hand and CaHa in the other). While the HA hands experienced no adverse events, six patients

experienced 11 adverse events in hands injected with CaHa. These events included edema (n = 4), pain (n = 3), impaired

movement (n = 2) and loss of sensitivity/ touch-sensitivity (number of patients not reported). Most events resolved within 2

weeks, and no serious adverse events occurred.

One prospective controlled study

compared HA (Juvederm Voluma) to autologous fat for treatment of temporal hollowing.

Early complications (within 2 weeks of injection) included bruising, pain and swelling which did not differ between Juvederm

and autologous fat injection sites. Juvederm had more cases of irregular surface (5/22) compared to autologous fat (1/24).

Among late complications (> 2 weeks post-injection), swelling occurred in 3/22 Juvederm sites and no autologous fat sites; it

resolved within 12 months. Subcutaneous scarring occurred in 5/24 autologous fat sites and no Juvederm sites. Abdominal

Material Performance Study - Hyaluronic Acid

pain occurred in 2 patients due to autologous fat harvest. Disfiguring surface occurred in 2/22 Juvederm sites and 1/24

autologous fat sites.

One retrospective controlled study

compared HA (Macrolane) to lipofilling for penile augmentation. Patients treated with

Macrolane had no complications, while 8 patients treated with lipofilling developed granulomas. In 5 patients granulomas

resolved spontaneously within 6 months with the help of massage, while 2 patients reported no relief from massage. After 20

days one patient experienced fat necrosis with progressive skin loss which was treated conservatively with weekly dressing,

and secondary healing occurred after 3 months.

Comparisons of different HA fillers: One SR

reported that overall complication rates did not differ between monophasic

(0.59, 95% CI: 0.36–0.8) and biphasic HA fillers (0.6, 95% CI: 0.43–0.76). Complications included redness, bruising, swelling,

pruritus, skin induration, tenderness, skin discoloration, pain, nodulus, hematoma, infection, vascular adverse events,

migration, numbness, and lumpiness. The most common complications included lumpiness, tenderness, swelling, and bruising.

Individual RCTs

48,50-52,54,56

did not report a difference in adverse event rates between different HA fillers. However, one

retrospective comparative study

reported that delayed-onset nodules occurred at a higher rate among patients receiving

Juvederm Volbella (1%) compared to Juvederm Vollure (0%), Juvederm Voluma (0%) and Restylane Silk (0.25%). The

nodules developed 6 to 37 weeks after the last injection.

Single-arm studies of HA fillers: One SR

evaluated prospective and retrospective studies of HA fillers to estimate the

incidence and prevalence of delayed inflammatory reactions and delayed-type hypersensitivity (DTH). Combined data from 35

prospective studies led to an estimated incidence of delayed inflammatory reactions of 1.1% per year, and that of possible

DTH reaction was 0.06% per year. The majority of 7 retrospective studies estimated a percentage of delayed inflammatory

reactions of less than 1% in 1 to 5.5 years. The incidence of DTH reaction cannot be determined from retrospective studies

but would likely be similar to the rate calculated from the prospective studies. Only about 5% of all reported DTH cases were

proven to be genuine DTH reactions. Two small single-arm case series

63,67

reported on delayed inflammatory reactions at HA

filler injection sites following Covid-19 exposure (3 cases) or exposure to influenza-like illness (14 cases). Most cases occurred

within 3 to 5 days of exposure for influenza, but the Covid-related cases showed a wider range (1 day to 2 weeks following

exposure either to Covid or Covid vaccine).

One large retrospective single-arm cohort study

evaluated the rate of delayed adverse events (DAEs) among 4500 patients

who received 9,324 treatments of HA filler (Juvederm Voluma). Forty-four patients experienced DAEs (incidence rate 0.98%

per patient, 0.47% per treatment). The most common reactions were delayed swelling and nodule formation (29 each), along

with erythema, warmth, and tenderness or pain. The median time to reaction was 4 months after the final HA injection.

Fifteen of 44 cases (34.1%) had an identifiable immunologic stimulus such as flu-like illness, infection, or dental procedure

immediately before DAE onset. Another large retrospective cohort study

reported that delayed-onset nodules occurred in

0.5% of Juvederm injections (23 cases out of 4,702 facial injections in 2,342 patients) with median time of onset 4 months

post-injection.

One small retrospective study

evaluated 11 cases of delayed granuloma formation in the orofacial region 3 to 10 years after

dermal filler orofacial injections (8 had received HA injections). All cases required surgical removal of granulomas.

One SR

included 26 studies reporting on 44 cases of vision loss following HA dermal filler injections. The symptoms included

partial or complete loss of vision and periocular changes including ptosis, ophthalmoplegia and exotropia; 8/44 patients had

CNS involvement in the form of brain infarcts or hemorrhage. Vision loss symptoms were almost always immediate; CNS

involvement was 7 and 9 hours after injection in 2 cases. The most common injection sites associated with visual loss were

the nose, forehead, and glabella, and the prognosis was better for patients with partial vision loss. No cases occurred

following injections in the lower face. A recent case series

reported on 7 cases with similar symptoms due to periorbital

vessel occlusion following HA facial injections.

Two single-arm studies reported complications related to use of Macrolane for breast enhancement. A retrospective case

series

of 20 patients with Macrolane-related complications presented with breast lumpiness and noted rapid, asymmetric

losses in breast volume. Cysts developed within 12 months of injection At least 85% of Macrolane filler was surgically removed

from the breasts of all patients. A prospective study

reported complication rates among 194 patients who received

Macrolane: minor adverse events 12.4% (24 patients), major adverse events 8.7% (17 patients), infection 0.5% (1 patient),

capsular contraction 4.6% (9 patients), early resorption 3.1% (6 patients), and removal of product 0.5% (1 patient). Most

events occurred within the first 6 months.

Material Performance Study - Hyaluronic Acid

A large retrospective case series

of HA fillers used for non-surgical rhinoplasty reported that the only complication was

persistent tip redness in 2% of patients that resolved spontaneously.

One prospective single-arm ultrasound study

of 63 female patients found alterations of the lacrimal, parotid, and

submandibular glands after filler injection (86% of patients had HA injections) that suggest subclinical inflammatory responses

to fillers. Further research would be needed to determine if ultrasound findings are clinically useful for preventing adverse

events related to dermal fillers.

Local Responses/Device Events (animal studies)

One non-randomized animal study

evaluated the risk of embolism and necrosis of HA (Restylane) versus PMMA (Artecoll)

injected intra-arterially in rabbit ears. With 0.1 ml injected volume, 5% of PMMA-injected ears had mild necrosis on day 7,

while HA tended to form obvious transparent emboli and 60% of HA-injected ears showed necrosis on day 7. With 0.2 ml

injected volume, 30% of PMMA-injected ears had necrosis and 100% of 0.2 ml HA-injected ears showed transparent emboli

and necrosis. HA injection led to substantially increased areas of necrosis compared to PMMA group regardless of injection

volume.

Systemic Responses

Two retrospective single-arm studies

72,73

reported on cases of patients suffering from late-onset, inflammatory, non-infectious

adverse reactions related to dermal fillers/implants that could be totally or partially considered as autoimmune/ inflammatory

syndrome induced by adjuvants (ASIA)-related disorders. The largest study (45 cases) included 7 cases who had received HA

as the only filler, and the other study (15 cases) included 2 cases where HA was the only filler (and 1 case where HA was the

second filler, following an earlier injection of silicone). Reported symptoms of HA cases in the larger study included 4 myalgia,

6 arthralgia/arthritis, 6 fatigue, 2 neurologic complaints, 2 cognitive features, 1 fever, 1 Sicca syndrome, 7 skin manifestations

(3 facial nodules), 3 evolvement into autoimmune disease. The symptoms occurred between 6 and 317 months following

exposure to dermal filler. These appear to be rare events but the rate of occurrence could not be determined because the

total number of HA-treated patients was not reported in either study.

Overall Quality of Evidence

Several SRs and RCTs provided evidence concerning local responses/adverse events associated with HA dermal fillers, so the

overall quality of evidence for local responses/events is moderate. Two small retrospective single-arm case series reported

evidence of rare systemic responses in the form of autoimmune/inflammatory syndrome associated with HA fillers; the quality

of evidence for systemic responses is therefore low.

Intraocular/Ophthalmic Viscoelastic Solution/Fluid

The literature search identified 10 human studies (1 SR,

2 RCTs,

76,77

3 observational comparative studies,

78-80

and 4 single-

arm studies

81-84

). For further information see Table 3 in Appendix D.

The human studies examined ophthalmic viscoelastic devices (OVDs) including Healon, Healon5, Healon GV, 2% HPMC,

Viscoat, OcuCoat, Provisc, Restylane-L, SKGEL, Soft Shell, DisCoVisc, Duovisc, and Twinvisc.

Local Host Responses (human studies)

All studies evaluated local responses/device events related to various ophthalmic viscoelastic (HA) solutions in human

subjects.

Four studies compared intraoperative pressure (IOP) elevations following eye surgery and injection of HA solutions to other

HA solutions or no HA solution. One SR

reported that Healon, Viscoat, Provisc, and Soft Shell were associated with

statistically significant increases in IOP at 1 day post-surgery while other HA solutions (Healon GV, Healon5, 2%HPMC,

OcuCoat, and Viscoat + Provisc showed either non-statistically significant increases or decreases in IOP. At 1-week follow-up,

most HA solutions were associated with non-statistically significant reductions in IOP; only Healon GV and Healon5 were

associated with statistically significant decreases in IOP. The remaining 3 studies (2 RCTs

76,77

and 1 retrospective comparative

study

) showed no statistically significant difference in post-operative IOP levels between different HA solutions

76,77

between HA and no HA.

These 3 studies did not evaluate Healon GV or Healon5. Overall, IOP tended to peak at 1 day post-

surgery and decrease afterward.

Material Performance Study - Hyaluronic Acid

Three studies (1 SR

and 2 RCTs

76,77

) reported cases of corneal and/or macular edema. These were rare events (the SR

reported rates of corneal edema ranging from 0.6% to 2.5% across 4 studies, while macular edema rates were 0.6% and

0.9% in 2 studies, respectively) and there was no statistically significant difference in edema rates between groups of patients

receiving different HA solutions.

One RCT

reported ocular hypertension in 12.6% of patients who received Twinvisc and 17.6% of patients who received

Duovisc. This study also reported 1 case each of inflammation (0.9%), capsule break (0.9%) and Viscoat bubbles (0.9%),

respectively.

One retrospective comparative study

reported that 20 of 95 (21.1%) of patients with HA fill (Provisc) had a layered hyphema

compared with 2 of 45 (4.4%) in the control group (no HA fill) at 1 week post-surgery (P=0.01). One patient (1.0%) in the HA

group and 1 patient (2.2%) in the non-HA group required drainage of choroidal effusions within the first postoperative month.

Other complications included a retinal detachment in a patient without HA fill and a clogged tube shunt that occurred in a

patient who received HA fill.

One prospective observational study

compared non-penetrating very deep sclerectomy (NPVDS) with the use of HA implant

(SKGEL) to trabeculectomy (TB) in patients with medically uncontrolled glaucoma. The overall complication rate was higher in

the TB group (79.5% vs 35.9%, p = 0.00011); whether the lower rate in the NPVDS group is related to HA or the sclerectomy

procedure (or both) is unclear.

One retrospective single-arm study

reported on 34 patients who developed toxic anterior segment syndrome (TASS) after

routine cataract surgery. The authors identified the source of TASS as HA (2.0% or 1.4%) derived from rooster comb; when

they substituted with OVDs derived from bacterial fermentation no further cases of TASS were observed.

One retrospective single-arm study

with 115 patients (162 eyes) who received canaloplasty for glaucoma reported that 12

patients (7.4%) developed Descemet membrane detachment (DMD) with or without intracorneal hemorrhage. All cases

resolved in 4 to 10 weeks (3 patients required additional surgery to resolve the DMD). Although the cause of DMD was

unclear, the authors noted that “DMD could be related to excessive amounts of Healon GV injection into the Schlemm canal

during the viscodilation portion of the surgery.”

One retrospective study

compared a viscoelastic device with brilliant blue G (Visco-BBG) versus balanced salt solution with

BBG (BSS-BBG) in patients who were receiving par plana vitrectomy (PPV) combined with internal limiting membrane (ILM)

peeling. The authors reported no complications in the Visco-BBG group and only 1 complication (retinal perforation) in the

BSS-BGG group.

Two single-arm observational studies

81,84

reported only complications that were considered to be unrelated to HA solutions.

Systemic Responses

No studies reported whether any systemic responses occurred.

Overall Quality of Evidence

The evidence regarding IOP came from 1 SR and 3 comparative studies (2 RCTs) and was of moderate quality, supporting an

initial increase followed by decrease in IOP after 1 day post-surgery. Some HA solutions (Healon GV, Healon5) appeared to

increase IOP less and decrease IOP sooner than others. Evidence that corneal and macular edema are rare events and have

similar rates among different HA solutions was also moderate quality. Evidence for other outcomes were reported in fewer

studies (mostly single studies, mostly observational) and the quality of evidence is low. Because no studies reported systemic

responses, the quality of evidence regarding systemic responses is very low.

Eye Drops

The literature search identified 14 human RCTs. For further information see Table 4 in Appendix D. The studies examined HA-

containing eye drops including Hylo Confort Plus/HydoGel, Lacure, Restasis, Tearin, Vislub®, Vismed®, Vismed® Multi, HA-

trehalose (Thealoz Duo®/Thealose®). These products contained 0.1% to 0.2% HA.

Local Host Responses (human studies)

All 14 studies evaluated local responses/device events related to procedures utilizing HA eye drops in human subjects.

Material Performance Study - Hyaluronic Acid

Eight studies

85-92

addressed general adverse events or complications related to HA. Four studies

85,88,90,92

reported no difference

in adverse events between HA and a control or another treatment group. One study

found statistically significantly fewer

adverse events when compared to another treatment group (p=0.0186). Three studies

86,87,91

found less than 10% of patients

experienced adverse events where both groups contained HA.

Seven studies

85,93-98

addressed itching/stinging/irritation/pruritus. Two studies

85,98

showed a decrease in

itching/stinging/irritation/erythema/keratitis when HA was compared to another treatment or a control. Four studies

93,95-97

showed no statistically significant difference in itching/stinging/irritation/erythema/keratitis when HA was compared to another

treatment or a control. One study

showed less than 2% of patients experience itching/stinging/irritation/erythema/keratitis

where both groups contained HA.

Six studies

85,94-98

addressed eye pain. Four studies

85,95,97,98

showed no statistically significant difference in pain when HA was

compared to another treatment or a control. One study

showed less than 2% of patients experience pain where both groups

contained HA. One study

showed an increase in pain when HA was compared to another treatment or a control.

Five studies

89,93,94,97,98

addressed eye disorders/dry eye. Two studies

93,94

showed 12% or fewer patients experienced eye

disorders/dry eye where both groups contained HA. One study

showed no statistically significant difference in eye

disorders/dry eye when HA was compared to another treatment or a control. Two studies

89,98

showed a decrease in eye

disorders/dry eye when HA was compared to another treatment or a control.

Four studies

95-98

addressed hyperemia/hemorrhage. All 4 studies

95-98

showed no statistically significant difference in

hyperemia/hemorrhage when HA was compared to another treatment or a control.

Three studies

94,95,98

addressed infection/viral conjunctivitis. Two studies

95,98

showed no statistically significant difference in

infection/viral conjunctivitis when HA was compared to another treatment or a control. One study

showed less than 2% of

patients experience infection/viral conjunctivitis where both groups contained HA.

Systemic Responses

No studies reported whether any systemic responses occurred.

Overall Quality of Evidence

The evidence base for human studies is large and includes 14 RCTs. The findings were mostly consistent in showing that HA

did not play a large role in local host response or may have reduced the incidence of local host response. Some of the studies

did not address the effects HA may have locally, as both experimental groups contained HA. Therefore, the quality of evidence

for local host response is moderate

Because no studies reported on systemic responses, the quality of evidence for systemic

responses is very low.

Material Performance Study - Hyaluronic Acid

ECRI Surveillance Data

Patient Safety Organization

Search Results: ECRI PSO identified a total of 6 reports that involved HA materials that occurred between July 2016 and

August 2021. One of these reports is pertinent to the topic of this report and it involved a hemorrhage/hematoma associated

with a dermal implant. Table 8 and Table 9 outline the PSO event report complications and harm scores, respectively.

All individual PSO event reports are redacted and included in Appendix F.

Table 8: Complications in HA-related PSO Event Reports - Dermal, Facial and Eye Applications

Complication

Dermal implant

Total

Hemorrhage/Hematoma

Total

Table 9: Harm Score Associated with HA-related PSO Event Reports - Dermal, Facial, and Eye Applications

Harm Scores (NCC-MERP) Harm Scores (NCC-MERP)

Dermal implant

Total

No Error

Error, No Harm

Error, Harm

Material Performance Study - Hyaluronic Acid

Harm Scores (NCC-MERP) Harm Scores (NCC-MERP)

Dermal implant

Total

Error, Death

NULL*

Total

1 1

*Harm score was not reported

Accident Investigations

Search Results: The search returned zero accident investigations associated with HA for dermal, facial and eye applications.

ECRI Problem Reports

Search Results: The search returned zero reports submitted by ECRI members.

Healthcare Technology Alerts

Search Results: The search returned 7 manufacturer issued alerts describing problems with technique, updated IFU,

discontinued use, clogged equipment, difficulty removing from body, contamination, leakage, and compromised sterility,

summarized in Table 10.

Table 10: Summary of Regulatory and Manufacturer Alerts - Dermal, Facial and Eye Applications

Device Type

# Alerts

Reported Problem

LZP (Aid, Surgical

Viscoelastic) –

Intraocular Fluid

5 manufacturer issued

• Product is difficult to remove from the eye, increasing post-op

intraocular pressure

• Clogged phacoemulsification equipment can lead to ocular

injury

• Product may contain microscopic glass particles

• Cannulae may leak viscoelastic material

• Inadequately sealed syringes compromise sterility

• Manufacturer reminds users of proper technique

Dermal Implant

2 manufacturer issued

• Updated IFU

• Discontinued use for breast augmentation

Potential Gaps

ECRI surveillance searches reflect mostly acute patient incidents that involved medical devices made of Hyaluronic Acid. Areas

of particular concern involve incidents that result in direct tissue exposure to the material if there is moderate to high-quality

evidence of acute or systemic reaction to this exposure, as determined by the systematic review. Topics with very low or low

quality of evidence represent areas of potential gaps in the literature. If the literature revealed areas of new concern (e.g.,

systemic response to long-duration contact) and there is little supporting evidence, these are considered gaps.

Across all device categories, the overall quality of evidence was low to very low for potential systemic responses.

Material Performance Study - Hyaluronic Acid

For intraocular/ophthalmic/viscoelastic solution/fluids, other than IOP elevation and corneal/macular edema (moderate quality

of evidence), all identified local host responses (e.g., ocular tension, inflammation) were of low quality of evidence.

There were no studies that addressed any particular cellular or molecular mechanisms for systemic manifestations.

Additionally, no studies addressed patient-related or material-related factors that could predict the likelihood and/or severity of

immunological/systemic responses.

Material Performance Study - Hyaluronic Acid

Safety Profile – Adhesion Barrier and Bulking Agent

Applications

Full Name: Hyaluronic Acid

CAS Registry Number: 924474

Safety Brief - Systematic Review Results

The systematic review included clinical and engineering literature on biocompatibility (i.e., host response and material

response) of Hyaluronic Acid used in medical devices. In addition to fundamental material biocompatibility, we focused on

specific devices known to be made of Hyaluronic Acid. The devices in Table 11 were recommended by FDA CDRH to guide

ECRI in searching this literature and ECRI’s surveillance data.

Table 11: Medical Devices Containing HA for Adhesion Barrier and Bulking Agent Applications provided by FDA to Guide ECRI

Searches

Regulatory Description

Product Code

Class

Balloon, Epistaxis

EMX

Mesh, Surgical, Polymeric

FTL

Mesh, Surgical

FTM

Cuff, Nerve

JXI

Polymer, Ent Synthetic-Polyamide (Mesh or Foil

Material

KHJ

Implant, Dermal, for Aesthetic Use

LMH

Agent, Bulking, Injectable for Gastro-Urology Use

LNM

Splint, Intranasal Septal

LYA

Barrier, Absorbable, Adhesion

MCN

Dressing, Wound and Burn, Hydrogel with Drug and/or

Biologic

MGQ

Unclassified

System, Vocal Cord Medialization

MIX

Polymer, Ear, Nose and Throat, Synthetic, Absorbable

NHB

The Safety Brief summarizes the findings of the literature search on toxicity/biocompatibility of Hyaluronic Acid.

Inclusion/exclusion criteria and quality of evidence criteria appear in Appendix A in the Appendices below. Quality of evidence

ratings reflected a combination of the quality of comparative data (study designs), quantity of evidence (number of relevant

studies), consistency of evidence, magnitude of effect, directness of evidence, and evidence for a dose response or response

over time. The search strategy appears in Appendix B3, and a flow diagram documenting inclusion/exclusion of studies

appears in Appendix C3. Summary evidence tables with individual study data appear in Appendix D3, and a reference list of

studies cited in the Safety Brief appears in Appendix E.

Material Performance Study - Hyaluronic Acid

Hyaluronic acid (HA) is a natural polymer found in the extracellular matrix and in high concentrations in the skin and synovial

fluid. HA suppresses immune responses and inflammation and aids wound healing. Based on these properties HA has been

incorporated into a variety of medical products designed to reduce postsurgical inflammation.

99,100

A summary of our primary findings is shown in Table 12. Even though the table mentions local host “responses” as well as

systemic “responses”, the events listed were not necessarily caused by the material. Only a few of these events were

statistically significantly more likely with HA than without HA. Some listed events were actually statistically significantly less

likely with vs without HA (in which case we write the word “less” next to the event). After the table, we present a detailed

discussion of research on HA as a material as well as research on the various device categories. For three additional device

categories, our searches identified no evidence (anti-adhesion protective oral gels, intravesical agents, and protective topical

agents).

Table 12: Summary of Primary Findings from Systematic Review - Adhesion Barrier and Bulking Agent Applications

Application

Local Host Responses/Device Events

Quality of

Evidence

(local

responses)

Systemic Responses

Quality of

Evidence

(systemic

responses)

Hyaluronic Acid as a

material (1 human

study, no animal

studies)

Hematuria, nausea

Low

Urticaria

Low

Adhesion barrier (34

human studies, no

animal studies)

Abdominal abscess, Acute pyelonephritis,

Anastomotic fistula, Anastomotic leakage,

Blood transfusion, Bowel injury, Cardiovascular

complications, Changes in menstrual pattern,

Hyperthermia, Ileus, Incision site abscess,

Intestinal fistulas , Intra-abdominal abscess,

Less Adhesion bowel obstruction, Less Post-

operative intestinal obstruction , Paralytic

ileus, Parietal abscess, Pelvic abscess,

Peritonitis, Pleural effusion, Pneumonia,

Postoperative biloma, Postoperative right

portal vein embolization, Respiratory

complications, Septic shock, Surgical-site

infection, Suture failure, Transfusion, Urinary

tract infection, Uterine dehiscence, Wound

infection

High

Alanine

aminotransferase,

Albumin, Aspartate

aminotransferase,

Bilirubin, Blood urea

nitrogen, C-reactive

protein, Creatinine,

Fever, Glucose,

Hemoglobin, Liver

enzymes, Liver

function, Renal

function, WBC

count, White blood

cell count

High

Anti-adhesion: Nasal

packaging (5 human

studies, no animal

studies)

No events

Moderate

None investigated

Very Low

Barrier gel for oral

lesions (3 human

studies, no animal

studies)

Less inflammation

Moderate

None investigated

Very Low

Material Performance Study - Hyaluronic Acid

Application

Local Host Responses/Device Events

Quality of

Evidence

(local

responses)

Systemic Responses

Quality of

Evidence

(systemic

responses)

Bulking agents (3

human studies, no

animal studies)

Urinary tract infection proctalgia, rectal

hemorrhage, diarrhea, injection site

complications, injection site sterile abscess,

injection site mass, pseudo-cyst formation

Moderate

Fever

Moderate

Intravesical agents

for bladder pain

syndrome/interstitial

cystitis therapy (2

human studies, no

animal studies)

Less urinary tract infection

Low

None investigated

Very Low

Organ spacer (2

human studies, no

animal studies)

Injection site pain, lower abdomen pain,

hematuria

Low

Asthenia

Low

Protective topical

agent for

esophageal and

gastric lesions in

GERD (1 human

study, no animal

studies)

Cough, rhinitis, throat irritation

Low

Nervous system

event (unspecified),

vertigo,

hypertension

Low

Vocal cord/fold

medialization (1

human study, no

animal studies)

Hematoma, edema of the aryteno-epiglottic

fold and false vocal fold

Low

None investigated

Very Low

Hyaluronic Acid as a Material

One human study (one randomized controlled trial [RCT]), no animal studies. For more information, see Table 1 in Appendix

Local Responses (human studies)

One RCT of men (n = 49) undergoing radiotherapy for prostate cancer assessed using HA and chondroitin sulfate instilled into

the bladder to prevent radiation-induced cystitis.

101

Patients also received oral HA and chondroitin sulfate. Control patients

received no treatment. Mild to moderate drug-related adverse events were reported in 7 HA and chondroitin sulfate treated

patients. Only 2 local adverse events were described, hematuria (blood in the urine) and nausea. Four of the seven adverse

events were not described. There were no serious local drug-related adverse events or treatment-related withdrawals. No

local adverse events were reported in control patients.

Because treated patients received HA as well as chondroitin sulfate, one cannot determine whether the reported adverse

events were due to HA or chondroitin sulfate or other factors.

Systemic Responses

In the same RCT, only 1 systemic adverse event was described: a case of urticaria (skin rash due to an allergic reaction) in

the treated group. There were no serious systemic drug-related adverse events or treatment-related withdrawals. As with

local events, the study design does not permit a determination of the cause of this event.

Material Performance Study - Hyaluronic Acid

Overall Quality of Evidence

The quality of the evidence is low for local responses and for systemic responses since the evidence comes from only one RCT

with confounding factors that included multiple additional compounds administered to the treated patients and an unblinded

control group that did not receive a placebo.

Adhesion barrier

34 human studies (14 SRs, 20 RCTs, no animal studies). For more information, see Table 2 in Appendix D.

Local Responses/Device Events (human studies)

Below, we discuss the evidence in five categories:

• Seprafilm (HA and carboxymethylcellulose)

• Sepraspray (HA and carboxymethylcellulose)

• Guardix-sol (HA + CMC gel)

• MateRegen (HA gel)

• Other HA gels

Seprafilm (HA and carboxymethylcellulose):

One SR examined SSI rate, anastomotic leak, ileus, intra‑abdominal abscess, and small bowel obstruction in RCTs using HA

and carboxymethylcellulose (CMC) barrier (Seprafilm) versus no treatment (control) during abdominal surgery.

102

The SR

included 13 RCTs with 3,665 patients, of which 1,800 were in the Seprafilm group. Use of Seprafilm was associated with a

significantly higher risk of anastomotic leak (3.1%) compared to control (1.6%), but a significantly lower incidence of small

bowel obstruction (3.0%) compared to control (5.9%).

One SR examined complications in clinical studies using HA+CMC barrier (Seprafilm) versus control on rates of postoperative

small bowel obstruction.

103

The SR included 4 RCTs and 5 clinical studies with 4,351 patients, of which 2,123 were in the

Seprafilm group. No complications were related to Seprafilm. The SR reported “the risk of postoperative small bowel

obstruction can be significantly decreased by the application of Seprafilm.”

One SR examined postoperative complications including intra-abdominal infection, wound infection, anastomotic leakage, and

postoperative hospital stay in clinical studies using HA+CMC barrier (Seprafilm) versus control after gastrointestinal neoplasms

surgery.

104

The SR included 5 RCTs and 5 clinical studies with 2,937 patients, of which 1,334 were in the Seprafilm group. The

incidence of postoperative intestinal obstruction was significantly lower in the Seprafilm group than the control (risk ratio,

0.52; 95% CI, 0.38–0.70; p<0.0001). Combined postoperative complications (intra-abdominal infection, wound infection,

anastomotic leakage, and postoperative hospital stay) were also significantly lower in the Seprafilm group than the control

(risk ratio 0.77; 95% CI, 0.61–0.97; p = .03). However, separate analysis of each postoperative complication did not find a

significant difference.

One SR examined complications in RCTs using HA+CMC barrier (Seprafilm) versus control after gastrointestinal surgery.

105

The SR included 5 RCTs with 2,177 patients. The SR reported “As regards adverse effects, the use of Seprafilm has not been

shown to result in a significant increase in morbidity, but…when Seprafilm was placed in direct contact with intestinal

anastomoses, a significant increase in the rate of intestinal fistulas and their secondary morbidity was observed.”

One RCT of 95 metastatic colorectal cancer patients requiring 2-stage hepatectomy compared HA+CMC barrier (Seprafilm)

with no HA control.

106

There was no significant difference in complication rates between Seprafilm and control groups. Major

complications included postoperative biloma that required percutaneous drainage and postoperative right portal vein

embolization. Minor complications included urinary tract infections, minor pleural effusion, blood transfusion and parietal

abscess.

One RCT of 345 patients undergoing elective colectomy compared HA+CMC barrier (Seprafilm) with no HA control.

107

complications were related specifically to Seprafilm. Surgery-related complications seen equally in both groups included

surgical-site infection, suture failure, and respiratory and cardiovascular complications.

One RCT of 753 females undergoing caesarean delivery compared HA+CMC barrier (Seprafilm) with no HA control.

108

complications were related specifically to Seprafilm. Surgery-related complications seen equally in both groups included bowel

injury, intraoperative transfusion, and uterine dehiscence.

Material Performance Study - Hyaluronic Acid

One RCT of 488 patients undergoing colorectal surgery compared HA+CMC barrier (Seprafilm), HA+CMC gel (Guardix), and

no HA control groups.

109

No complications were specifically related to Seprafilm or Guardix. Surgery-related complications

seen equally in each group included pneumonia, surgical-site infection, anastomosis leakage, and intra-abdominal abscess.

Sepraspray (HA + CMC):

One RCT of 209 patients undergoing laparoscopic colorectal and/or small bowel surgery compared HA+CMC powder

(Sepraspray Adhesion Barrier) with no HA control (n=104).

110

The Sepraspray group had significantly higher rates of adverse

events (62.9%) and serious adverse events (27.6%) compared to control. Adverse events included hyperthermia (HA 5.7%,

Control 2.9%), incision site abscess (HA 4.8%, Control 2.9%), urinary tract infection (HA 4.8%, Control 1.0%), and

anastomotic fistula (HA 3.8%, Control 3.8%). Serious adverse events included pelvic abscess (HA 4.8%, Control 1.9%),

abdominal abscess (HA 3.8%, Control 0%), septic shock (HA 1.0%, Control 1.9%), peritonitis (HA 1.9%, Control 2.9%), ileus

(HA 2.9%, Control 0%), and anastomotic fistula (HA 2.9%, Control 3.8%).

One RCT of 41 females undergoing laparoscopic myomectomy compared HA+CMC powder (Sepraspray Adhesion Barrier) with

no HA control.

111

No adverse events were specifically related to Sepraspray. Adverse events occurred in both groups but were

not defined. The authors reported that surgical-site infections, intra-abdominal abscesses, and deep vein thrombosis did not

occur in either group.

Guardix-sol (HA + CMC gel):

One RCT of 76 patients undergoing laparoscopic radical cystectomy compared HA+CMC gel (Guardix-sol) with no HA

control.

112

Complication rates did not differ significantly in the Guardix-sol and control groups, “except for adhesive bowel

obstruction, which occurred in zero and six patients, respectively (p = 0.025).” Reported complications included transfusion

(HA 26.3%, Control 34.2%), paralytic ileus (HA 15.8%, Control 10.5%), wound disruption (HA 2.6%, Control 7.9%), and

acute pyelonephritis (HA 5.3%, Control 7.9%).

One RCT of 43 males undergoing surgery for chronic epididymitis compared HA+CMC gel (Guardix-sol) with no HA control.

113

No complications occurred in either group (specifically looked for wound infection, wound hematoma, wound deadhesion,

hematoma, cord injury, and testicular injury).

MateRegen (HA gel):

One RCT of 306 females with moderate to severe intrauterine adhesions (IUAs) undergoing operative hysteroscopy followed

by embryo transfer compared intrauterine injection of HA gel (MateRegen) with no HA control.

114

No adverse events were

observed in either group (authors did not specify what events they expected).

One RCT of 145 females with moderate to severe IUAs compared intrauterine injection of HA gel (MateRegen) with no HA

control.

115

No surgical complications were reported in either group. After surgery, changes in menstrual pattern in the

treatment group (87.7%, 107/122) were similar to the control group (76.4%, 97/123).

One RCT of 274 females undergoing dilation and curettage compared HA gel (MateRegen) with no HA control.

116

The authors

noted that “no serious adverse events were observed during the study period, and there were no prolonged hospitalizations or

reoperations owing to adverse events” in either group.

One RCT of 48 females undergoing curettage compared HA gel (MateRegen) with no HA control.

117

No complications or

adverse events were related to MateRegen. No postoperative infections, prolong hospitalizations, or reoperations occurred in

either group.

Other HA gels:

One RCT of 196 females undergoing laparoscopic surgery compared HA gel (HyaRegen NCH) with saline placebo.

118

adverse events were related to HyaRegen NCH. The authors reported that adverse events were comparable between groups,

mild and spontaneously resolved. No serious adverse events were observed in either group. The authors did not indicate what

adverse events were expected.

One SR examined adverse events and abdominal complaints in RCTs using HA gel (Hyalobarrier Gel) versus control after

endoscopic gynecologic surgery.

119

The SR included 5 RCTs with 335 females, of which 167 were in the Hyalobarrier Gel

group. No adverse events were related to Hyalobarrier Gel. In gel-treated patients, two reported nausea and one reported

vomiting. In control patients, one reported nausea.

Material Performance Study - Hyaluronic Acid

One SR examined pain in clinical studies using HA gel (Hyalobarrier Gel) versus control after laparoscopic gynecologic

surgery.

The SR included 1 RCT with 43 females, of which 25 were in the Hyalobarrier Gel group. The SR reported “Full

conditioning with Hyalobarrier Gel placement resulted in decreased pain scores (p < .001) and faster recovery (p < .0001)

compared with standard laparoscopy.”

One SR examined adverse events and complications in RCTs using HA and HA+CMC barriers versus control after laparoscopic

myomectomy (surgical procedure to remove uterine fibroids).

120

The SR included 8 RCTs with 748 patients, of which 392 were

in the HA group. No serious adverse events or complications occurred in the included studies. The authors did not state what

adverse events or complications they expected.

One SR examined complications in RCTs using HA and HA+CMC barriers versus control after gynecologic surgery.

121

The SR

included 7 RCTs with 748 females. No complications were related to HA barriers.

One SR examined complications in clinical studies using HA and HA+CMC gels versus control after operative hysteroscopy.

122

The SR included 3 clinical studies with 262 females, of which 130 were in the HA groups. No complications were related to HA

gels.

One SR, a Cochrane Review) examined adverse outcomes and pelvic pain in two SRs of RCTs comparing HA+CMC barrier

versus control for adhesion prevention after gynecologic surgery.

123

The SR included 47 RCTs with between 59 and 127

females in HA+CMC and control groups. The available studies reported no adverse outcomes or pelvic pain related to

HA+CMC barrier, but adverse event reporting occurred in only half of the studies.

One SR examined complications in RCTs using HA gel versus control after miscarriage.

124

The SR included 4 RCTs with 625

females, of which 290 were in the HA group. No complications were related to HA gel.

One SR examined complications in RCTs using HA gel versus control after intrauterine operation.

125

The SR included 7 RCTs

with 952 females, of which 455 were in the HA group. No complications were related to HA gel. The authors reported that

none of the included studies reported HA or surgery related complications, including hemorrhage, perforation or cervical

laceration.

One SR examined complications in RCTs using HA gel versus control after hysteroscopic adhesiolysis.

126

The SR included 2

RCTs and 4 clinical studies with 394 females, of which 176 were in the HA group. No complications occurred in the HA gel

group.

One SR examined complications in RCTs using HA gel versus control after gynecologic surgery.

127

The SR included 6 RCTs

with 564 females, of which 286 were in the HA group. Only one RCT reported HA group complications, which were “patient

discomfort and uterine perforation, with no statistically significant differences between the 2 [HA and control] groups.”

One RCT of 65 females undergoing uterine septum resection compared intrauterine injection of HA with no HA control.

128

complications, such as allergic reactions, occurred in the HA group. Menstrual patterns were not affected in either group.

One RCT of 143 females undergoing dilation and curettage compared HA gel with no HA control.

129

No complications occurred

in the HA group. Menstrual patterns and pregnancy rates were not affected in either group.

One RCT of 149 females undergoing dilation and curettage compared HA gel with no HA control.

130

No complications were

specifically related to HA gel. Cervix laceration was experienced by one patient in the HA group (1.3%) and no patients in the

control group (p=1.00). Three patients in the HA group had excessive bleeding (3.9%) compared to one patient in the control

group (1.4%) (p=0.62). Two patients in the HA group had postoperative infections (2.6%) compared to one patient in the

control group (1.4%) (p=1.00). Four patients in the HA group had postoperative pain (5.2%) compared to three patients in

the control group (4.2%) (p=1.00). Uterus perforation was experienced by one patient in the HA group (1.3%) and no

patients in the control group (p=1.00).

One RCT of 124 females undergoing laparoscopic surgery for deep infiltrating endometriosis compared HA gel with sterile

saline placebo.

131

No complications were related to HA gel. Patient pain was significantly reduced in the HA gel group after six

months. HA treatment significantly reduced dysmenorrhea (painful menstruation), dyschezia (difficult or painful defecation),

and dyspareunia (painful intercourse) at 3 and 6 months compared with control patients.

One RCT of 87 patients undergoing colorectal surgery compared HA gel, chitosan, and control groups and reported short-term

(peristomal cutaneous infection, intra-abdominal hemorrhage, and abscess) and long-term (intestinal obstruction and

anastomotic leakage) postoperative complications.

132

No complications were statistically significantly more likely in the HA gel

Material Performance Study - Hyaluronic Acid

group. Short-term (30-day) complications were not seen in any patients. Long-term postoperative complications were similar

among groups (10.5% of HA gel patients and 17.9% of control patients).

One RCT of 89 females undergoing operative hysteroscopy for intrauterine adhesions compared HA gel, intrauterine device

(IUD), and HA gel+IUD groups.

133

No complications occurred in the HA gel group. None of the patients experienced excessive

bleeding, infection, or other complications (not specified by the authors).

Systemic Responses

One SR examined systemic complications reported in 2 RCTs and 1 comparison study using Seprafilm (HA containing

membrane) to prevent intestinal obstruction after gastrointestinal neoplasms surgery.

104

The SR reported aspartate

aminotransferase, alanine aminotransferase (serum indicators of liver function), and blood urea nitrogen (indicator of kidney

function) were not significantly different from controls five- and seven-days post-surgery. However, serum creatinine (another

indicator of kidney function) was significantly higher in Seprafilm patients at 5 days but not at 7 days. The authors noted “that

Seprafilm did not cause acute postoperative inflammation in short term, and had almost no effect on the liver and renal

function of the patients.”

One SR examined systemic complications in 5 RCTs comparing cross-linked HA gel with no treatment in patients undergoing

endoscopic gynecological surgery.

119

Fever was reported in one of 167 HA patients and 2 of 168 control patients.

One RCT of 753 women undergoing caesarean delivery compared Seprafilm with no Seprafilm control.

108

Fever was reported

in similar numbers of patients (4.5% Seprafilm, 3.0% control, risk ratio = 1.5 (95% CI of 0.7 to 3.3)).

One RCT of 196 women undergoing laparoscopic surgery compared a crosslinked hyaluronan gel with saline placebo.

118

The

RCT examined blood chemistry (C-reactive protein, WBC count, hemoglobin, liver enzymes, albumin, blood urea nitrogen,

creatinine, bilirubin, glucose, etc.) at 3 days, 30 days, and 9 weeks and found no clinically significant changes from baseline in

the hyaluronan gel patients. Control patients had clinically significant elevations in white blood cell count and blood glucose

from baseline at 9 weeks.

One RCT of 87 patients undergoing colorectal surgery compared HA gel with chitosan or no treatment.

132

White blood cell

count, liver function, and renal function were comparable across groups at 2 weeks post-surgery.

Overall Quality of Evidence

The quality of evidence is high for local responses since the evidence comes from multiple SRs and RCTs reporting consistent

results that there are few adverse events or complications related to HA-containing products. The quality of evidence is high

for a lack of systemic responses since the evidence is from SRs with multiple RCTs and three individual RCTs.

Anti-adhesion: Nasal packaging

Five human studies (two SRs, three RCTs), no animal studies. For more information, see Table 3 in Appendix D.

Local Host Responses (human studies)

One SR examined complications reported in RCTs using Sepragel or MeroGel, both contain HA, as nasal packing after

endoscopic sinus surgery.

134

The SR included four RCTs with 352 patients. According to the SRs, no adverse events or

complications were reported in two of the included RCTs, and SR authors did not describe the other two RCTs results. No

specifics were provided on what adverse events may have been expected.

One SR examined complications in studies evaluating endoscopic sinus surgery.

135

The SR included 13 studies with 501

patients (5 double-blind RCTs, 5 single-blind RCTs, and 3 prospective comparisons) that used HA as an absorbable packing, a

non-absorbable packing impregnated with HA, or an HA spray. The SR reported “It is clear that hyaluronic acid in all

preparations is safe and well-tolerated by patients, as evidenced by only one adverse event not related directly to hyaluronic

acid, and the overall satisfaction with its use in absorbable nasal packs.”

One RCT of patients (n = 205) undergoing tympanoplasty for adhesive otitis media reported that there were no complications

or adverse events in the group receiving MeroGel (which contains HA).

136

The authors did not specify what complications were

expected after surgery but did report that no patient suffered a sensorineural hearing loss or needed a reoperation.

Material Performance Study - Hyaluronic Acid

One RCT of patients (n = 55) undergoing bilateral endoscopic sinus surgery compared an absorbable crosslinked HA hydrogel

with a gelatin sponge control.

137

The authors reported that no adverse events related to the packing treatment occurred

during the study but did not report what adverse events may have been expected.

One RCT of patients (n = 227) undergoing surgery for unilateral primary chronic dacryocystitis (an infection of the lacrimal

sac, secondary to obstruction of the nasolacrimal duct) compared MeroGel with no MeroGel.

138

The authors did not report any

adverse events related to MeroGel, but did not report what adverse events may have been expected.

Systemic Responses

No systemic responses were reported in the included studies.

Overall Quality of Evidence

The SRs and RCTs were consistent in observing that there are no adverse events or complications related to HA-containing

products. The study authors noted that HA reduces local inflammatory responses and reduces scarring after surgery. The

quality of the evidence is moderate for local responses since the majority of studies are RCTs and is very low for systemic

responses (no studies investigated systemic effects).

Barrier gel for oral lesions

Three human studies (one SR, two RCTs), no animal studies. For more information, see Table 4 in Appendix D.

Local Responses/Device Events (human studies)

One SR examined HA as an adjunct treatment for chronic inflammatory disease.

The SR included 25 controlled studies with

studies examining HA’s effect on gingivitis, chronic periodontitis, dental surgery, and oral ulcers. No adverse events occurred

with HA use. The authors noted that HA “suppresses the immune response preventing excessive exacerbations of

inflammation.” HA’s effect on inflammation may be responsible for faster wound healing and reduction in postoperative

discomfort.

One RCT using a split-mouth design (n = 24) examined HA’s effect on moderate to severe chronic periodontitis.

139

No adverse

events occurred in the HA group. Plaque index, gingival index, papillary bleeding index, and periodontal probing depth

improved in HA-treated and control areas but was significantly better in HA-treated areas.

One RCT using a split-mouth design (n = 28) examined HA’s effect on plaque-induced gingivitis.

140

The authors’ reported “no

adverse effects to the gel were observed on clinical examination and as reported by the patients.” Plaque index, gingival

index, and gingival bleeding were examined after treatment. Inflammation was reduced with all treatments, but the reduction

was greater in HA-treated areas.

Systemic Responses

No systemic responses were reported in the included studies.

Overall Quality of Evidence

The SRs and RCTs were consistent in not observing any adverse events or complications related to HA containing products.

The quality of the evidence is moderate for local response since all studies were RCTs or controlled studies and were

consistent and is very low for systemic responses (no studies investigated).

Bulking agents

Three human studies (three RCTs), no animal studies. For more information, see Table 5 in Appendix D.

Local Host Responses (human studies)

One RCT of pediatric patients (n = 60) with grades III and IV primary vesicoureteral reflux used dextranomer HA to treat the

condition as opposed to surgery.

141

Patients were followed for 1 year. Late local complications included 6 (20%) urinary tract

infections in dextranomer HA patients and 5 (16.7%) in surgery patients.

One RCT of patients with fecal incontinence (n = 206) compared dextranomer HA with a sham injection procedure.

142

Patients

were followed for 1 year. In this study proctalgia (14%, pain due to a spasm of the pelvic floor muscles, the muscles of the

Material Performance Study - Hyaluronic Acid

anal sphincter, or the muscles of the rectum), rectal hemorrhage (7%), and diarrhea (5%) were seen in the treatment group

but were not common in the sham control group (3% proctalgia, 1% rectal hemorrhage, and 4% diarrhea).

One RCT of women (n = 344) with urinary incontinence (reported in a SR) examined mid-urethral dextranomer HA (Zuidex)

injections compared with collagen injections.

143

Patients were followed for 1 year. Dextranomer HA treated patients had higher

rates of injection site complications (16% versus 0%). Injection site sterile abscess (8.4%), injection site mass (4.4%) and

pseudo-cyst formation (2.2%) were only seen in Zuidex-treated women. Zuidex has been withdrawn from the market.

Systemic Responses

One RCT of pediatric patients (n = 60) with grades III and IV primary vesicoureteral reflux used dextranomer HA to treat the

condition as opposed to surgery.

141

Fever was noted in several patients within 2 weeks of the procedure (3 dextranomer HA

patients [10%] and 6 surgery patients [20%]).

One RCT of patients with fecal incontinence (n = 206) compared dextranomer HA with a sham injection procedure.

142

Patients

were followed for 1 year. In this study, fever (8%) was seen in the treatment group but was not seen in the sham control

group.

Overall Quality of Evidence

The quality of the evidence is moderate for local response and for systemic responses since all studies were RCTs and results

were mostly consistent.

Intravesical agents for bladder pain syndrome/interstitial cystitis therapy

Two human studies (1 SR, 1 RCT), no animal studies. For more information, see Table 6 in Appendix D.

Local Host Responses (human studies)

One SR assessed HA and HA plus chondroitin sulphate for reducing the occurrence of recurrent bacterial cystitis.

144

The SR

included 4 comparison studies (2 RCTs) with 143 patients. Bladder instillation of HA was well tolerated and reduced urinary

tract infections.

One RCT of patients with bladder pain syndrome (n = 72) compared intravesical treatment with HA, chondroitin sulfate, and a

combination of both and found there were no serious adverse events or patient withdrawals during the 24-month study in any

patient.

145

The authors reported specifically looking for recurrent febrile urinary tract infections, symptomatic cardiac

arrhythmias, significant nephrotoxicity, and hepatotoxicity that would require discontinuing treatment.

Systemic Responses

No systemic responses were reported in the included studies.

Overall Quality of Evidence

The quality of the evidence is low for local response since only 2 of the 4 studies in the SR were RCTs and the evidence base

is small and is very low for systemic responses (no studies investigated).

Organ spacer

Two human studies (no RCTs, two case series), no animal studies. For more information, see Table 7 in Appendix D.

Local Host Responses (human studies)

One case series of patients (n = 60) undergoing high-dose-rate interstitial brachytherapy for prostate cancer reported that no

adverse effects occurred after an HA injection into the perirectal fat prior to brachytherapy.

146

The authors did not report

which adverse effects they may have expected.

One case series of patients (n = 36) undergoing hypofractionated radiation therapy for prostate cancer reported that

injections of HA were well tolerated with one patient each experiencing persistent pain at the injection site, lower abdomen

pain and hematuria (blood in the urine).

147

Material Performance Study - Hyaluronic Acid

Systemic Responses

One case series of patients (n = 36) undergoing hypofractionated radiation therapy for prostate cancer reported that

injections of HA were well tolerated with one patient experiencing asthenia (abnormal physical weakness or lack of energy).

147

Overall Quality of Evidence

The quality of the evidence is low for both local responses and systemic responses because the evidence base has only two

small studies. Studies of this size are not sufficient to allow a full spectrum of adverse events to be seen.

Protective topical agent for esophageal and gastric lesions in GERD

One human study (one RCT), no animal studies. For more information, see Table 8 in Appendix D.

Local Host Responses (human studies)

One double-blind, placebo-controlled RCT of patients (n = 154) with non-erosive reflux disease treated with proton pump

inhibitors reported no serious adverse events related to using an HA plus chondroitin sulphate formulation to coat the

esophagus after a meal.

148

Mild gastrointestinal and respiratory (cough, rhinitis, throat irritation) adverse events were

somewhat more common in HA-treated patients.

Systemic Responses

The same RCT reported some mild systemic adverse events occurring only in HA-treated patients: 3 nervous system, 1

vertigo, and 1 hypertension.

Overall Quality of Evidence

The quality of the evidence is low for both local responses and systemic responses because the evidence base has only one

study but it was of medium size.

Vocal cord/fold medialization

One human study (1 SR), no animal studies. For more information, see Table 9 in Appendix D.

Local Host Responses

One SR with 14 case series (n = 442) assessed injection laryngoplasty with commercial preparations of HA to treat unilateral

vocal fold paralysis.

149

The authors noted that HA was safe for this procedure and that one patient experienced hematoma

and one patient experienced edema of the aryteno-epiglottic fold and false vocal fold. Some included studies reported local

hypersensitivity and inflammation that the authors believe may have been reactions to proteins produced in the HA

manufacturing process. Outcomes and adverse events will likely vary depending on injection technique and brand of product

used.

Overall Quality of Evidence

The quality of the evidence is low for local responses and very low for systemic responses (no studies investigated). The SR

included only case series and reported on only 442 patients.

ECRI Surveillance Data

There were no relevant reports found in ECRI’s PSO, accident investigation, or PRN databases related to devices composed of

SM-PFAS. Healthcare Technology Alerts search returned 4 manufacturer issued alerts describing problems including

compromised sterility and off label use.

Refer to Appendix F for a list of devices that guided our searches of ECRI Surveillance Data.

Patient Safety Organization

Search Results: ECRI PSO identified 11 reports that involved adhesion barrier composed of HA that occurred between

December 2016 and August 2021; however, none were pertinent to the topic of this report.

Material Performance Study - Hyaluronic Acid

Accident Investigations

Search Results: Zero investigations were recovered from the accident investigations database.

ECRI Problem Reports

Search Results: The search returned zero reports submitted by ECRI members.

Healthcare Technology Alerts

Search Results: The search returned 4 manufacturer issued alerts describing problems with sterility, labeling, and off-label

use, summarized in Table 13.

Table 13: Summary of Regulatory and Manufacturer Alerts - Adhesion Barrier and Bulking Agent Applications

Device Type

# Alerts

Reported Problem

MCN (Barrier,

Absorbable, Adhesion)

2 manufacturer issued

• Defective packaging compromises sterility.

• Recall; off-label use may lead to serious complications.

KHJ (Polymer, ENT

Synthetic-Polyamide

[Mesh or Foil Material])

1 manufacturer issued

• Mislabeled as sterile, when only syringe contents are sterile.

Adhesion Barrier

1 manufacturer issued

• Cannulae may not meet required sterility level.

Potential Gaps

ECRI surveillance searches reflect mostly acute patient incidents that involved medical devices made of Hyaluronic Acid. Areas

of particular concern involve incidents that result in direct tissue exposure to the material if there is moderate to high-quality

evidence of acute or systemic reaction to this exposure, as determined by the systematic review. Topics with very low or low

quality of evidence represent areas of potential gaps in the literature. If the literature revealed areas of new concern (e.g.,

systemic response to long-duration contact) and there is little supporting evidence, these are considered gaps.

For 7 of 9 device categories, no local adverse events were statistically significantly more likely in patients receiving HA devices

compared to patients receiving non-HA devices. This is predominantly because of a lack of studies with strong design that

investigated these events. The exceptions were adhesion barrier and bulking agents.

For 7 of 9 device categories, no systemic adverse events were statistically significantly more likely in patients receiving HA

device compared to patients received non-HA devices. This is predominantly because of a lack of studies with strong design

that investigated these events. The exceptions were adhesion barrier and bulking agents.

Only one study related to HA as a material was identified. However, because treated patients received HA as well as

chondroitin sulfate, one cannot determine whether the reported adverse events were due to HA or chondroitin sulfate or other

factors

For 8 of 9 device categories, only study involving adhesion barriers investigated the cellular or molecular mechanisms for

systemic manifestations.

Material Performance Study - Hyaluronic Acid

Appendix A. Inclusion/Exclusion Criteria and Quality of

Evidence Criteria

Inclusion Criteria

1. English language publication

2. Published between January 2011 and August 2021

3. Human studies (animal studies that provide unique information will also be considered for inclusion)

4. Systematic reviews, randomized controlled trials, cohort studies, case-control studies, cross-sectional studies, case

series

5. Studies that evaluate toxicity/biocompatibility of HA or priority devices that include this material

Exclusion Criteria

1. Foreign language publication

2. Published before January 2011

3. Not a study design of interest (e.g., in vitro lab study, case report, narrative review, letter, editorial)

4. Off-topic study

5. On-topic study that does not address a key question

6. No device or material of interest

7. No relevant outcomes (adverse events or biocompatibility not reported)

8. Study is superseded by more recent or more comprehensive systematic review

Quality of Evidence Criteria

1. Quality of comparison – is there evidence from systematic reviews including randomized and/or matched study

data and/or randomized or matched individual studies?

2. Quantity of data – number of systematic reviews and individual studies providing relevant data, as well as the

proportion of included studies that reported a specific outcome.

3. Consistency of data – are the findings consistent across studies that report relevant data?

4. Magnitude of effect – what is the likelihood of adverse effects compared to controls (with no device, lower dosage,

shorter exposure time), and possibly number of patients likely to have harms.

5. Directness of evidence – do human studies isolate the effect of the device (i.e. can the adverse effects be

attributed to the device)?

6. Is there evidence of a dose response or time response (e.g. adverse effects increase with longer exposure time)?

Material Performance Study - Hyaluronic Acid

Appendix B1. Search Summary – Muscle/Skeletal Applications

Strategies crafted by ECRI’s medical librarians combine controlled vocabulary terms and free-text words in conceptual search

statements that are joined with Boolean logic (AND, OR, NOT).

Most medical bibliographic databases such as Medline and Embase include detailed controlled vocabularies for medical

concepts accessible through an online thesaurus. Controlled vocabularies are a means of categorizing and standardizing

information. Many are rich ontologies and greatly facilitate information transmission and retrieval. Frequently seen examples

of controlled vocabularies include ICD-10, SNOMED-CT, RxNorm, LOINC, and CPT/HCPCS.

Citations in PubMed are indexed with MeSH terms and those in Embase are indexed with terms from EMTREE. These terms

are assigned either by a medical indexer or an automated algorithm. Several terms are selected to represent the major

concept of the article – these are called “major” headings. This “major” concept can be included in search strategies to limit

search retrieval. The syntax in Embase for this is /mj. We have used this convention in our strategies sparingly since indexing

is subjective and we are using a sensitive search approach which errs in the direction of comprehensiveness.

Database providers build functionality into their search engines to maximize the usefulness of indexing. One of the most

frequently used shortcuts is term explosion. “Exploding” in the context of hierarchical controlled vocabularies means typing in

the broadest (root or parent) term and having all the related more specific terms included in the search strategy with a

Boolean OR relationship. We use term explosions whenever feasible for efficiency. Feasibility depends on whether you wish to

include all of the related specific terms in your strategy. For example, in one of our approaches we explode the Emtree

concept mechanics. This explosion automatically added the all the following terms (n=174) and their associated entry terms

(lexical variants and synonyms) to the strategy using an “OR” without the searcher having to type them in. That’s one of the

major advantages to searching using controlled vocabularies. We don’t rely exclusively on controlled vocabulary terms since

there are possible limitations such as inconsistent indexing and the presence of unindexed content. That’s why we also include

free text words in our strategies.

Material: Hyaluronic Acid (HA) - Muscle/Skeletal Applications

Set

Number

Concept Search Statement

Material

Hyaluronic acid (HA) and

derivatives

'hyaluronic acid'/de OR 'hyaluronic acid derivative'/de OR hyaluronan?:ti,ab

OR hyaluronate?:ti,ab OR hyaluronic:ti,ab

HA viscosupplement devices

('adant' OR 'suprahyal' OR 'agili-c' OR 'artflex' OR 'arthrum' OR

'chondrotissue' OR 'cingal' OR 'crespine gel' OR 'curavisc' OR 'durolane' OR

'euflexxa' OR 'nuflexxa' OR 'gel-one' OR 'gel-syn' OR 'gelsyn' OR 'gel-syn-3'

OR 'gelsyn-3' OR 'sinovial' OR 'genvisc 850' OR 'happycross' OR 'hyalgan' OR

'hyalofast' OR 'hyalograft-c' OR 'hyalone' OR 'hyalubrix' OR 'hyalonect' OR

'hyaluron hexal' OR 'hymovis' OR 'hyadd' OR 'monovisc' OR 'neovisc' OR

'orthovisc' OR 'ostenil' OR 'ostenil tendon' OR 'recosyn' OR 'recosyn forte' OR

'recosyn uno' OR 'recosyn md' OR 'sportvis' OR 'rheovital' OR 'rheo vital' OR

'spinevisc' OR 'supartz' OR 'supartz fx' OR 'synocrom' OR 'synocrom mini' OR

'synocrom forte' OR 'synocrom forte one' OR 'structovial' OR 'synojoynt' OR

'hyalrheuma' OR 'synvisc' OR 'synvisc-one' OR 'hylan g-f 20' OR 'triluron' OR

'hyalgen' OR 'trivisc' OR 'visco-3' OR 'viscoseal'):ti,ab,kw,de,tn,dn

HA bone filler devices

('aft dbm' OR 'biosphere flex' OR '45s5 bg' OR '45s5 bioglass' OR 'bioglass

45s5' OR 'confirm dbm' OR 'confirm bioactive' OR (dbx AND bone) OR 'dbx

inject' OR 'dbx strip' OR 'sygnal dbm' OR 'hyaloss' OR 'inqu paste mix' OR

'kinex' OR 'mtf new bone void filler' OR 'scs 17-01' OR

'tactoset'):ti,ab,kw,de,tn,dn

Material Performance Study - Hyaluronic Acid

HA other devices

('agili-c' OR 'biopoly' OR 'chondrotissue' OR 'hyalograft*'):ti,ab,kw,de,tn,dn

HA + general device and

musculoskeletal terms

#1 AND ('viscosupplementation'/exp OR 'intraarticular drug

administration'/exp OR 'musculoskeletal system inflammation'/exp OR

'musculoskeletal injury'/exp/mj OR 'bone matrix'/exp OR 'joint'/exp/mj OR

joint*:ti OR ((intra NEXT/1 articul*):ti) OR intraarticul*:ti OR osteoarthr*:ti

OR 'arthritis':ti)

Combine and Limit by

language and publication

date

(#2 OR #3 OR #4 OR #5) AND [english]/lim AND [2011-2021]/py

Limit by publication type

#6 NOT ('book'/it OR 'chapter'/it OR 'conference abstract'/it OR

'conference paper'/it OR 'conference review'/it OR 'editorial'/it OR

'erratum'/it OR 'letter'/it OR 'note'/it OR 'short survey'/it OR 'tombstone'/it)

Material Response

'biocompatibility'/de OR biocompat* OR tribolog* OR 'bio compat*'

OR 'biological* compat*' OR 'biological* evaluation'

'degradation'/exp OR degrad* OR adsorbable OR split* OR wear OR

deteriorat* OR atroph* OR migrat* OR distend* OR distension OR

'delamination'/exp OR delamina* OR leach* OR filter* OR seep* OR

evaginat* OR subsidence

10.

Leachable* OR extractable*

11.

(swell* OR shrink* OR contract* OR stretch* OR retract* OR extension OR

extend* OR deform* OR creep OR plasticity OR degrad* OR disintegrat* OR

fail* OR fragment* OR debond*) NEAR/3 (hydrogel* OR implant* OR

prosthes* OR prosthetic* OR injectable* OR putty OR putties OR graft OR

device?)

12.

'mechanics'/exp

13.

'device material'/exp/mj

14.

'Biomedical and dental materials'/exp/mj

15.

Combine sets

#8 OR #9 OR #10 #11 OR #12 OR #13 OR #14

16.

HA + Material Response

#7 AND #15

Host Response

17.

Host NEAR/2 (reaction* OR response*)

18.

‘toxicity’/exp OR toxic*:ti OR cytotox* OR teratogenic* OR genotox* OR

‘carcinogenicity’/exp OR carcinogen*:ti

Material Performance Study - Hyaluronic Acid

19.

'immune response'/exp OR 'immunity'/exp/mj OR 'hypersensitivity'/exp OR

'immunopathology'/exp/mj

20.

(immun*:ti OR autoimmun*:ti OR hypersens*:ti) NOT immunofluorescenc*:ti

21.

'inflammation'/exp OR (inflamm* NEAR/3 (tissue OR macrophage* OR

cytokine* OR react* OR respons* OR level* OR sign* OR effect* OR activat*

OR local OR inhibit* OR alleviat* OR reduc* OR decreas* OR induce* OR

synovial))

22.

'foreign body' OR granuloma* OR 'foreign body'/exp OR 'macrophage'/exp

OR 'macrophage*':ti,ab OR fouling OR 'anti-fouling' OR biofilm?

23.

'adhesion'/exp OR 'tissue adhesion'/exp OR 'tissue response' OR 'tissue

reaction' OR 'necrosis':de OR 'necrosis':ti,ab

24.

protrude* OR protrus* OR perforat*

25.

'fibrosis'/exp OR 'seroma'/exp OR 'hematoma'/exp OR 'seroma*' OR

'hematoma*' OR 'thrombosis'/exp OR 'thrombosis'/syn OR 'phlebitis'/exp OR

'phlebitis'/syn OR 'skin irritation'/exp OR 'pruritus'/exp OR 'pruritus' OR

itch*:ti,ab

26.

Combine sets

#17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25

Other combinations

27.

Combine sets

HA + Material

Response+ Host

Response

#16 AND #26

28.

HA general devices +

Host response

#5 AND #7 AND #26

29.

Combine sets

#27 OR #28

30.

HA systematic reviews

#7 AND ('systematic review'/de OR 'meta analysis'/de OR ((meta NEAR/2

analy*):ti) OR 'systematic review':ti)

31.

Final set

#29 OR #30

Material Performance Study - Hyaluronic Acid

Appendix B2. Search Summary – Dermal, Facial and Eye