The Summary
A patch may make the cuff-repaired shoulder tendon look better on MRI. However, sound evidence is lacking that patient-reported outcomes are clinically significantly better for patients having cuff repairs with a patch than those with unpatched repairs.
Rotator cuff repair is one of the most common operations in orthopaedics. However, the repaired tendon does not always stay attached to the bone: on MRI, roughly a quarter of repairs show a retear within a year or two. This is especially the case for the most common type: the degenerative tear, the repair of which is at risk for failure because of the limited quantity and quality of the muscle, tendon, bone unit required for cuff function.
Biologic patches are commonly used with the goal of improving the outcomes of cuff repair. For example, the bovine bioinductive patch (Regeneten) is a thin sheet of purified collagen from cow tendon. The surgeon fastens it over the repaired tendon, and the body is meant to grow new tendon-like tissue into it. In August 2025 the American Academy of Orthopaedic Surgeons gave the patch its highest grade of endorsement, a “strong” recommendation, saying it can lead to lower retear rates and better patient-reported outcomes.
The first half of that statement is supported. The second half appears to be unsupported by evidence.
Two trials, two different questions
The AAOS recommendation rests on two randomized trials of the same implant. The larger one, from Spain and supported in part by the manufacturer, randomized 124 patients with fully repairable tears 1 to 4 cm across (on average about 2 cm) to a standard repair with or without the patch. At two years, 12% of patched repairs had retorn on MRI, compared with 35% of unpatched ones. But the patients’ own ratings of pain and function were the same in both groups at every visit. The authors titled their two-year report “a lower retear rate but similar outcomes.”
The smaller trial, of 60 patients at one center, asked a different question: could the patch replace a repair altogether in small tears (2.5 cm or less, with the rotator cable intact)? Those patients did score better: 8 points higher on the ASES shoulder score, but this difference is not clinically significant (less than the minimal clinically important difference).
In other words, neither trial showed an important difference in comfort and function.
Why a healed tendon does not necessarily lead to a better shoulder
A meta-analysis of 20 randomized trials of biologic augmentation explains the paradox. Shoulders whose repairs stayed intact did score better than those that retore. But about three of four repairs heal anyway without augmentation. Augmentation prevented a retear in only about 12 of every 100 shoulders. Thus if augmentation were applied to all cuff repairs, the average gain would be 1 to 2 points per patient on the 100-point ASES and Constant shoulder scores, far too small a benefit for the patient to notice.
Other studies suggest that a retear itself has less impact on patient comfort and function than the appearance on MRI might imply. Pooled data from more than 3,000 repairs found that patients with retorn tendons scored only 6 to 7 points lower than those whose tendons healed, and most were satisfied.
The cost
The patch adds time and expense to surgery: estimates range from $5,000 to $10,000 per case. Non-conflicted studies have also reported a downside that the AAOS guideline does not mention: stiffness. In one matched comparison, 9% of patched shoulders needed a second operation for inflammation and stiffness, compared with none of the unpatched ones. In another, six of 32 patched shoulders and none of 32 controls were reoperated for stiffness.
“What’s New in Sports Medicine” and “What’s New in Shoulder and Elbow Surgery”
The 2026 “What’s New” reviews in the Journal of Bone & Joint Surgery report the same pattern for other add-ons to cuff repair. Injected skin cells, drilling the bone and an osteoporosis drug each improved the tendon’s appearance on MRI without a clinically meaningful change in patient reported outcomes. Neither review mentions patches.
Readers of patch reports should be wary of summaries. A 2026 study found that 13 of 16 published reviews of patch augmentation contained “spin” in their abstracts, presenting results more favorably than the data supported.
What patients and surgeons can ask
None of this means the patch never helps. For massive tears or poor-quality tendon, where retears are most likely, it may yet prove worthwhile over the long haul, but most trials have followed patients for only one to two years. Until a trial shows that patients have better long-term comfort and function, the burden of proof falls on the proponents of the device. We can reasonably ask: Will it make the shoulder feel and work better? What will it add to the cost? What is the risk that it will make the shoulder stiff?
The details
Many shoulder surgeons are using bioinductive patches in rotator cuff repair. The evidence offered for patches is based mostly on MRI images, histology and retear rates. Evidence of a clinically significant improvement in patient comfort and function with their use is lacking. Recent JBJS “What’s New” articles review a series of randomized trials in which adjuncts to cuff repair improved healing on MRI without a clinically significant improvement in patient-reported outcomes [1,2]. One states that optimizing the biological environment after cuff repair “remains a challenge” [2]. Neither article mentions patch augmentation.
What the 2026 “What’s New” articles report
Both updates review a year of randomized trials in rotator cuff surgery [1,2]. The cuff trials they chose share a pattern: the adjunct improves the tendon on MRI, and the patient outcomes are not different.
• Autologous dermal fibroblast injection. Retear at 6 months fell from 24.3% to 5.6%, with no difference in ASES, Constant or SST scores [3].
• Bone marrow stimulation. Incomplete healing at 2 years fell from 54.7% to 33.9%, but ASES scores were 89.2 and 87.8 (p = 0.498) [4].
• Zoledronic acid. Retear at 6 months fell from 35% to 15%, with what the authors of the update called “modest” clinical improvement [1,5].
If a lower retear rate reliably meant a better shoulder, at least one of these trials should have shown a difference in patient-reported outcome at the minimal clinically important difference (MCID). None did.
What the pooled randomized trials show
Our recent meta-analysis of 20 Level I randomized trials (1,327 shoulders) of biologic augmentation confirms this pattern [6]. Augmentation lowered the retear rate from 26.7% to 14.8% (risk ratio 0.62). For patch augmentation alone (153 shoulders), the retear rate was 17.6%, a risk ratio of 0.55 [6]. Yet no patient-reported outcome improved by the MCID. The pooled differences were 0.9 points for the ASES and 2.2 points for the Constant score. In the patch subgroup the Constant score was only 3.9 points higher, and the ASES was not different [6].
Why does a healed tendon not translate into a better result? Intact repairs did score 14.3 ASES points and 10.5 Constant points higher than retorn repairs [6]. But because about 73% of unaugmented repairs heal, augmentation prevents a retear in only about 12 of every 100 shoulders. Spread across all the patients who receive it, the expected gain per patient is 1.3 to 1.7 points [6].
Other meta-analyses suggest that even this gap may overstate what a retear costs the patient. Across 43 study arms (3,350 repairs, 25% retorn at a median of 18 months), healed repairs scored only 6.1 Constant points and 7.1 ASES points higher than retorn repairs, both values are below the commonly suggested MCIDs. The authors concluded that most patients may expect satisfactory outcomes despite a retear [7]. An earlier meta-analysis of Level I and II studies found no difference in ASES between intact and retorn repairs; strength was the clearest difference [8].
“Healed” on MRI may also not mean healed where the tendon was fixed. In a study using tantalum markers in the repaired tendon, all 13 repairs retracted from their point of fixation during the first year, by 16 mm on average, yet only 4 showed a defect on MRI [9]. The MRI retear rate is thus an imperfect measure of the tendon’s condition, and a weak predictor of how the patient will do.
A second 2026 meta-analysis, of 9 randomized trials (556 patients) of biologic grafts, found a similar structural effect from augmentation: a retear risk ratio of 0.31 and tendons 1.2 mm thicker [10]. Its abstract reports the patient-reported outcomes only as improvement from before to after surgery, which occurs with or without a graft, and concludes that clinical improvement was “less consistent” [10].
The AAOS strong recommendation for patches
The 2025 AAOS guideline states that “the use of bioinductive tendon implants to augment rotator cuff repair or as an alternative to non-augmented repair can lead to lower retear rates and better patient-reported outcomes” [11]. It grades the evidence High and the recommendation Strong. By the guideline’s own definition, Strong requires two or more high-quality studies “with consistent findings” [11]. The two studies are randomized trials of the same bovine collagen implant (Regeneten).
Trial | Patients and comparison | Imaging | Patient-reported outcomes |
Ruiz Ibán et al. [12,13] | 124 posterosuperior tears 1–4 cm across (mean about 2 cm), retraction ≤3 cm, fully repairable; transosseous-equivalent repair with vs. without patch | Retear at 2 years 12.3% vs. 35.1% | No difference in ASES, Constant, pain or EQ-5D; ASES MCID reached by 91% vs. 88% |
Camacho Chacón et al. [14] | 60 supraspinatus tears ≤2.5 cm with intact rotator cable; patch instead of repair vs. sutured repair | Better tendon histology at 6 months | ASES 88 vs. 80 at 2 years, a difference below the MCID |
The guideline’s rationale acknowledges that in the larger trial “functional outcomes did not demonstrate any differences” [11]. That trial’s 2-year report, which the guideline does not cite, is titled “a lower retear rate but similar outcomes” [13]. The smaller trial asked a different question: can a patch replace a repair in a small tear with an intact rotator cable? Its 8-point ASES difference is below the MCID [14]. The “better patient-reported outcomes” in the recommendation thus rest on a single 60-patient, single-center trial of a different operation. The two trials do not have consistent findings on the outcome that matters to patients.
Neither trial studied the tears for which augmentation is most often proposed. The larger trial enrolled only fully repairable tears 1 to 4 cm across, with tears averaging about 2 cm; it excluded tendons retracted more than 3 cm and muscles with Goutallier grade 3 or 4 fatty infiltration [12,13]. The smaller trial enrolled only tears of 2.5 cm or less with an intact rotator cable [14]. Massive tears and tears with poor-quality tendon, where retears are most common, were not tested.
Figure 1. ASES and Constant scores over time, with and without the patch, in the two trials behind the AAOS strong recommendation. When the patch was added to a repair, the two groups tracked each other at every time point [12,13]. When the patch was used instead of a repair, the groups separated [14]. Ruiz Ibán values were read from the published figure; Camacho Chacón values are medians from the published tables.
The guideline lists the risks of these implants as longer operating time, possible infection and added cost. It does not mention stiffness or reoperation. It closes by calling for high-quality studies “to definitively establish the benefits of these devices” [11].
What else we know about patches
• Partial-thickness tears. In a manufacturer-funded randomized trial of 41 patients, a patch without repair produced better WORC scores at 6 weeks and 3 months than takedown and repair. There was no difference in any patient-reported outcome at 6 or 12 months. The patch group was told to wear the sling for 1 week, the repair group for 4 to 6 weeks [15].
• Meta-analysis. Improvement and the proportion of patients reaching the MCID after patch-augmented repair were similar to historical standard repair. The overall complication rate was 15.5% after full-thickness and 16.2% after partial-thickness repairs. There were no concurrent controls [16].
• Matched cohort. At a minimum of 2 years, 47 patched repairs and 94 matched unpatched repairs had similar ASES (81 vs. 85), SSV (79 vs. 85) and pain scores. Reoperation for inflammation and stiffness was 9% (4/47) with the patch and 0% (0/94) without it. The authors stopped using the implant [17]. In the accompanying commentary, McAllister Nolan noted that these grafts are “heavily promoted to surgeons by industry” and recommended avoiding them for this indication [18].
• Partial-thickness repairs. In a propensity-matched comparison, stiffness in the first 12 weeks occurred in 8 of 32 patched shoulders and 1 of 32 controls. Six patched shoulders, and no controls, required reoperation for stiffness [19]. This happened even though the controls more often had a tear takedown and repair (18 vs. 5 of 32), an operation expected to cause more stiffness, not less [20].
• Human dermal allograft in large tears. One randomized trial reports a longer follow-up. Sixty patients with 3 to 5 cm tears had repair with or without a human dermal allograft; 43 were seen at a mean of 5.7 years [21]. Retear was 9.1% with the allograft and 38.1% without it. The trial’s primary outcome, the Constant score, did not improve more with the allograft (p = 0.053). The ASES improved 13 points more with the allograft (p = 0.047), but the allograft group started 5 points lower [21]. This is a different product from the bovine implant in the AAOS strong recommendation, and the guideline rates the dermal allograft evidence as moderate [11].
A patch placed on the bursal surface of the cuff adds tissue to the humeroscapular motion interface, the space that must glide for the shoulder to move. The stiffness signal from independent centers is consistent with that.
Why the word “strong” matters
A strong recommendation from our Academy carries weight beyond the operating room [6]. A patient with a retear after an unpatched repair could point to it. A surgeon who left the patch out because the patient-reported data did not support it would have to defend that choice. Meanwhile the patient who receives a patch, becomes stiff and needs a second operation may ask why the implant was used.
The cost is real. An economic analysis co-authored by an employee of the manufacturer assumed that the implant adds $3,641 to each repair; our meta-analysis cites $5,000 to $10,000 per case [6,22]. That analysis estimated an added $232,468 per 100 patients to obtain 18 more MRI-healed tendons [22]. That is a cost per healed tendon, not per patient with greater improvement.
A 2026 analysis, labeled Level I evidence, concluded that the patch is “the dominant, cost-effective treatment strategy” over 10 years, saving money and adding 1.3 quality-adjusted life years per patient [23]. This is a computer model built on assumptions, not an outcome study: no patient was followed for 10 years. Its result follows from three assumptions. First, its yearly cycles appear to apply the one-year retear rates from the Ruiz Ibán trial (26% without the patch, 8% with it) as the chance of a new retear in every year; in the trial itself, only 9% more control patients retore between the first and second years [12,13]. Second, it assigns quality of life by tendon status, 0.7 if intact and 0.6 if torn, taken from an earlier publication, instead of using the quality-of-life scores measured in the trial, which did not differ between groups [13]. Even if every unpatched tendon were torn and every patched tendon intact for all 10 years, those values could yield a gain of less than 0.9 quality-adjusted life years. Third, it assumes that 16% of patients with a retorn tendon progress each year to reverse arthroplasty. This study received grant funding from the manufacturer [23].
Who pays also matters. In one ambulatory surgery center, use of a biologic (a dermal allograft or bioinductive implant) was the largest single driver of charges for cuff repair, adding $17,791 to a risk-adjusted $19,728 [24]. Ambulatory cuff repair is increasingly paid under fixed-rate, episode-based bundles, in which the surgeon and facility bear the financial risk of added costs [25]. When the payment for the repair is fixed, the patch comes out of the margin, and a surgeon who owns a share of the center faces a three-way pull. On one side are a strong guideline recommendation and its medicolegal weight; on another, the surgeon’s own finances; and in between, trials showing no difference the patient can feel. Contracts that pay for implants separately, or consulting relationships with manufacturers, can tilt the balance the other way. None of these is a clinical reason to use a patch, or to withhold one.
Questions to ask of any patch study
Abstracts deserve particular care. Among 16 systematic reviews comparing patch-augmented with unaugmented repair, 13 (81.3%) had spin in the abstract [26]. Spin was not associated with funding (P = .36), but each methodological flaw on the AMSTAR-2 checklist was associated with 61% more types of spin [26]. One example: a 2024 meta-analysis of 5 randomized trials of dermal patches concluded in its abstract that patches “improve outcomes.” Its discussion acknowledges that the differences, 2.8 Constant points and 5.6 ASES points, did not reach the MCIDs of 4.6 and 11.1 [27].
When the results of a well-controlled study on the effectiveness of patches in cuff repair are presented, we should ask:
1. What was the difference in patient-reported outcome between groups, and did it reach the MCID?
2. What proportion of patients in each group reached the MCID or the patient acceptable symptom state?
3. How many had stiffness, capsulitis or reoperation, compared with controls?
4. Was the control a repair alone, something less (debridement), or something more (takedown and repair)?
5. Were the rehabilitation protocols the same in both groups?
6. Who paid for the study?
Bottom line
What is demonstrated: (1) a bovine patch placed over a sound repair lowers the MRI retear rate; (2) a patch used instead of a repair combined with faster rehabilitation may return some patients to daily activities sooner; and (3) patches are associated with more stiffness and reoperation in unconflicted series.
What is not demonstrated: that adding a patch improves comfort or function by an amount that is clinically significant.
Most trials follow patients for only one to two years; the mean follow-up in our meta-analysis was 18 months [6]. Whether fewer retears translate into better function later, or into less cuff tear arthropathy, is not known.
The author of this post has no financial relationship with any orthopaedic company.
Follow on twitter: https://twitter.com/shoulderarth
Follow on facebook: https://www.facebook.com/frederick.matsen
Follow on LinkedIn: https://www.linkedin.com/in/rick-matsen-88b1a8133/
References
[1] Rao AJ, Saltzman BM, Harrison AK. What’s new in shoulder and elbow surgery. J Bone Joint Surg Am. 2026;108(20):1574-7. doi:10.2106/JBJS.26.00744
[2] Al-Hourani K, Khan S, Boufadel P, et al. What’s new in sports medicine. J Bone Joint Surg Am. 2026;108(11):763-70. doi:10.2106/JBJS.25.01633
[3] Kim YK, Kim YT, Won YS, et al. Efficacy of an autologous dermal fibroblast injection in reducing the retear rate after arthroscopic rotator cuff repair: a prospective randomized controlled trial. Am J Sports Med. 2025;53(3):592-9.
[4] Ye Z, Chen C, Chen H, et al. Bone marrow stimulation yielded similar clinical outcomes with improved tendon healing and abduction strength compared with decortication for arthroscopic rotator cuff repair in shoulder active patients. Arthroscopy. 2025;41(8):2740-9.e3.
[5] Schanda JE, Ullrich R, Boesmueller S, et al. Intraoperative zoledronic acid for arthroscopic rotator cuff repair: short-term results from a prospective, randomized, placebo-controlled phase II trial. Am J Sports Med. 2026;54(2):342-52.
[6] Checketts JX, Daniels SD, Schiffman CJ, Matsen FA III, Hsu JE. Biologic augmentation improves rotator cuff integrity but not clinically significant patient outcomes: a meta-analysis of randomized controlled trials. JBJS Open Access. 2026;11:e26.00248. doi:10.2106/JBJS.OA.26.00248
[7] Holtedahl R, Bøe B, Brox JI. The clinical impact of retears after repair of posterosuperior rotator cuff tears: a systematic review and meta-analysis. J Shoulder Elbow Surg. 2023;32:1333-46. doi:10.1016/j.jse.2023.01.014
[8] Russell RD, Knight JR, Mulligan E, Khazzam MS. Structural integrity after rotator cuff repair does not correlate with patient function and pain: a meta-analysis. J Bone Joint Surg Am. 2014;96:265-71. doi:10.2106/JBJS.M.00265
[9] McCarron JA, Derwin KA, Bey MJ, et al. Failure with continuity in rotator cuff repair “healing.” Am J Sports Med. 2013;41(1):134-41. doi:10.1177/0363546512459477
[10] Amin V, Ghayyad K, Oshikoya O, Goltz D, Huffman GR. Biologic graft augmentation is associated with lower retear rates and greater tendon thickness compared with arthroscopic rotator cuff repair without augmentation: a systematic review and meta-analysis of randomized controlled trials. Arthroscopy. 2026. doi:10.1002/arj.70619
[11] American Academy of Orthopaedic Surgeons. Management of Rotator Cuff Injuries: Evidence-Based Clinical Practice Guideline. Published August 18, 2025. https://www.aaos.org/rccpg2025
[12] Ruiz Ibán MA, Garcia Navlet M, Moros Marco S, et al. Augmentation of a transosseous-equivalent repair in posterosuperior nonacute rotator cuff tears with a bioinductive collagen implant decreases the retear rate at 1 year: a randomized controlled trial. Arthroscopy. 2024;40(6):1760-73.
[13] Ruiz Ibán MA, García Navlet M, Moros Marco S, et al. Augmentation of a posterosuperior cuff repair with a bovine bioinductive collagen implant shows a lower retear rate but similar outcomes compared with no augmentation: 2-year results of a randomized controlled trial. Arthroscopy. 2025;41(10):3869-79. doi:10.1016/j.arthro.2025.03.057
[14] Camacho Chacón JA, Roda Rojo V, Martin Martinez A, et al. An isolated bioinductive repair vs sutured repair for full-thickness rotator cuff tears: 2-year results of a double blinded, randomized controlled trial. J Shoulder Elbow Surg. 2024;33(9):1894-904. doi:10.1016/j.jse.2024.03.043
[15] Wang A, Breidahl W, Ek ET, et al. Early functional recovery is improved in patients treated with bioinductive collagen implant augmentation compared with standard arthroscopic repair of high-grade partial-thickness rotator cuff tears: a prospective randomized trial. Orthop J Sports Med. 2026;14(3):23259671261418675. doi:10.1177/23259671261418675
[16] Warren JR, Domingo-Johnson ER, Sorensen AA, et al. Bioinductive patch as an augmentation for rotator cuff repair, a systematic review and meta-analysis. J Shoulder Elbow Surg. 2024;33(11):2515-29. doi:10.1016/j.jse.2024.05.002
[17] Haft M, Li SS, Pearson ZC, et al. No short-term clinical benefit to bovine collagen implant augmentation in primary rotator cuff repair: a matched retrospective study. Clin Orthop Relat Res. 2025;483:442-52. doi:10.1097/CORR.0000000000003247
[18] McAllister Nolan B. CORR Insights®: No short-term clinical benefit to bovine collagen implant augmentation in primary rotator cuff repair: a matched retrospective study. Clin Orthop Relat Res. 2025;483:453-4. doi:10.1097/CORR.0000000000003281
[19] Yeazell S, Lutz A, Bohon H, et al. Increased stiffness and reoperation rate in partial rotator cuff repairs treated with a bovine patch: a propensity-matched trial. J Shoulder Elbow Surg. 2022;31(6 Suppl):S131-5. doi:10.1016/j.jse.2022.02.003
[20] Yeazell S, Lutz A, Bohon H, et al. Response to Bushnell et al regarding: "Increased stiffness and reoperation rate in partial rotator cuff repairs treated with a bovine patch: a propensity-matched trial." J Shoulder Elbow Surg. 2022;31:e572-3. doi:10.1016/j.jse.2022.06.028
[21] Lee GW, Kim JY, Lee HW, Yoon JH, Noh KC. Clinical and anatomical outcomes of arthroscopic repair of large rotator cuff tears with allograft patch augmentation: a prospective, single-blinded, randomized controlled trial with a long-term follow-up. Clin Orthop Surg. 2022;14(2):263-71. doi:10.4055/cios21135
[22] McIntyre LF, Nherera LM, Schlegel TF. Resorbable bioinductive collagen implant is cost effective in the treatment of rotator cuff tears. Arthrosc Sports Med Rehabil. 2023;5(2):e367-74. doi:10.1016/j.asmr.2023.01.002
[23] Hurley ET, Ruiz Ibán MÁ, Oeding JF, et al. Bio-inductive collagen implant augmentation for arthroscopic rotator cuff repair is cost-effective in medium to large tears for reducing retears: a secondary analysis of a randomized controlled trial. Arthroscopy. 2026;42:73-82. doi:10.1002/arj.70000
[24] Morris JH, Malik AT, Hatef S, Neviaser AS, Bishop JY, Cvetanovich GL. Cost of arthroscopic rotator cuff repairs is primarily driven by procedure-level factors: a single-institution analysis of an ambulatory surgery center. Arthroscopy. 2021;37(4):1075-83. doi:10.1016/j.arthro.2020.11.033
[25] Lu Y, Labott JR, Salmons HI IV, et al. Identifying modifiable and nonmodifiable cost drivers of ambulatory rotator cuff repair: a machine learning analysis. J Shoulder Elbow Surg. 2022;31:2262-73. doi:10.1016/j.jse.2022.04.008
[26] Al-Badaineh M, Josan H, Imran B, et al. Spin in systematic reviews on patch-augmented rotator cuff repair: prevalence, predictors, and methodological implications. J Shoulder Elbow Surg. 2026 [Epub ahead of print]. doi:10.1016/j.jse.2026.07.030
[27] Hurley ET, Crook BS, Danilkowicz RM, et al. Acellular collagen matrix patch augmentation of arthroscopic rotator cuff repair reduces re-tear rates: a meta-analysis of randomized control trials. Arthroscopy. 2024;40(3):941-6. doi:10.1016/j.arthro.2023.09.025
