Abstract
Background:
Quadriceps tendon ruptures occur infrequently in the general population. Biomechanical data suggest advantages with the use of suture anchor fixation for major tendon repair. Clinical studies of quadriceps tendon repair have been limited to small case series.
Purpose:
To evaluate clinical failure following primary quadriceps tendon repair with transosseous tunnel or suture anchor repair.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
The Military Health System Data Repository was queried to identify all adult patients who underwent primary quadriceps tendon repair in the Military Health System between 2014 and 2018. Patients were excluded if they had incomplete records, polytrauma, open injury, prior ipsilateral total knee arthroplasty did not undergo quadriceps tendon repair, or underwent revision repair. Follow-up was obtained by manual chart review of both orthopaedic and nonorthopaedic records, recording any evidence of ongoing knee issues. Univariate analysis was performed to determine associations between potential risk factors and repair failure. Multicollinearity was assessed between potential risk factors, and candidate variables were included in multivariate logistic regression models to determine independent risk factors for repair failure.
Results:
Following application of inclusion/exclusion criteria, 245 knees in 234 patients were included. Mean age of the cohort was 52 years. Patients were predominantly male (223/234; 95%) and military retirees (143/234; 61%). Transosseous tunnel repair was the most frequently employed surgical technique (147/245; 60%), followed by suture anchor repair (78/245; 32%). Repair failure requiring revision surgery occurred in 11% of knees (27/245). Surgical-site infection following index surgery was associated with eventual rerupture (P = .02). There was no difference in failure rate between transosseous tunnel repair and suture anchor repair (12.2% vs 9.0%; P = .51). Among knees undergoing suture anchor repair, no difference in failure was found between knot-tying and knotless suture anchor fixation (P = .73).
Conclusion:
We observed no difference in failure of primary quadriceps tendon repair between transosseous tunnel and suture anchor repair types or between the 2 main suture anchor implant subtypes (knotless vs knot-tying anchors). A greater than previously reported rerupture rate was observed, indicating the need for continued investigation into optimal surgical techniques.
Quadriceps tendon ruptures are uncommon injuries, occurring at a rate of 1.37 per 100,000 patients a year. Patients are commonly older men, with mean age reported as in the 60s.8,12,29 The tendon typically fails with a powerful eccentric contraction or low energy falls, often in the presence of predisposing conditions. There is a significant association with chronic tendinopathy and underlying systemic factors such as diabetes mellitus, obesity, chronic steroid use, and renal or thyroid disorders.1,8,12 The most common technique for quadriceps tendon repair is the transosseous tunnel technique, which involves passing suture through bone tunnels in the patella.1,16,29,31 In 2002, Richards and Barber 23 first recommended suture anchor repair as an alternative. They, along with later advocates, suggested that suture anchors may permit a smaller surgical incision, shorter operative times, more limited disruption of the patellar blood supply, and earlier mobilization.1,6,23,26,29
Over the past decade, multiple biomechanical models have evaluated the 2 techniques.10,13,21,27 A greater load to failure for anchor fixation was initially proposed from these studies.13,21,30 Most recently, Onggo et al 19 performed a systematic review of extensor mechanism biomechanical studies, demonstrating decreased tendon-bone gap formation but similar load to failure. Belk et al 3 reported the pooled gapping across the literature at 6.3 mm in transosseous tunnel repairs and 3.5 mm in suture anchor repairs. Failure of suture anchor repair has consistently been at the suture anchor eyelet interface, whereas transosseous repairs fail through tendon failure itself.2,19 Enthusiasm has increased for use of suture anchors since their introduction over a decade ago. A recent survey study of 264 surgeons from 2 national orthopaedic societies reported suture anchors as the primary repair method at 20%, compared with the transosseous bone tunnel gold standard at 70%. 14
Clinical outcome studies of quadriceps tendon repair are sparse. Ciriello et al 7 performed a systematic review through 2012 and found 319 total patients in the literature with a pooled rerupture rate of 2%. They noted consistent quadriceps atrophy, loss of strength, and worse outcomes with delayed (>3 weeks) versus acute repair. Furthermore, direct clinical comparisons between transosseous tunnel and suture anchor repairs have been limited to small case series.5,6,9,17,22,23 In a systematic review of 156 knees comparing transosseous tunnel and suture anchor repairs, Mehta et al 16 found no significant differences in rerupture rate between quadriceps tendon repairs using transosseous tunnels or suture anchors.
The purpose of this study was to investigate the rate of repair failure following primary quadriceps tendon repair with either transosseous tunnel repair or suture anchor repair techniques. In light of biomechanical data, we hypothesized that there would be a greater repair failure rate with use of transosseous tunnel repair. We secondarily sought to define the rate of return to military duty among active-duty personnel.
Methods
The Military Health System (MHS) comprises a network of over 550 health care facilities serving 9.6 million active-duty military personnel, military retirees, and their families. The MHS Data Repository (MDR) contains data on all health care encounters for MHS beneficiaries. Following approval from our local institutional review board, the MDR was queried through the MHS Management Analysis and Reporting Tool (M2) for all patients with an encounter for Current Procedural Terminology code 27385 (which includes quadriceps tendon repair) at an MHS healthcare facility between January 1, 2014, and December 31, 2018.
Patients were eligible for inclusion if they underwent primary repair of a ruptured quadriceps tendon over the specified time period. Patients were excluded if they underwent a procedure other than quadriceps tendon repair (eg, revision repair, patellar tendon repair, quadriceps tendon reconstruction, quadriceps tendon debridement), sustained an open tendon rupture, had polytrauma requiring surgical treatment beyond repair of the quadriceps tendon, had a history of ipsilateral total knee arthroplasty, were <18 years of age at the time of surgery, or did not have an available operative report with procedural details.
The initial query yielded 450 patient encounters potentially eligible for inclusion. Following application of exclusion criteria, a total of 245 knees in 234 patients were included in the cohort. A Strengthening the Reporting of Observational studies in Epidemiology diagram was used for patient selection (Figure 1).
A Strengthening the Reporting of Observational Studies in Epidemiology diagram detailing inclusion/exclusion criteria for the patient cohort. CPT, Current Procedural Terminology; TKA, total knee arthroplasty.
Included patient records were searched for demographic and injury information, past medical and surgical history (to include prior ipsilateral/contralateral tendon rupture and prior quadriceps tendinitis/tendinopathy), intraoperative findings (including repair technique and surgical details), details of surgical construct (suture anchor type, number of anchors/tunnels, and core sutures), postoperative complications, and clinical findings at the time of final examination. Muscle strength, when documented, utilized the Medical Research Council scale for muscle power, which is graded from 0 to 5. 15 The surgeon's operative note was utilized to determine the tear type and location; where this was not available, the preoperative magnetic resonance imaging report was used. Military disposition (return to duty vs separation from service for medical reasons) was also recorded for active-duty military personnel.
Follow-up of included patients was performed opportunistically, and all records (both orthopaedic and nonorthopaedic) were reviewed for all patients following the date of surgery up until the present. Repair failure was defined as the need for revision quadriceps tendon repair or, when reoperation did not occur, documented surgeon assessment that repair had failed. Patients without documentation of failure were presumed to have not failed. As the MHS maintains an integrated electronic health record, follow-up at any MHS facility will be documented, even in cases where patients moved from one locale to another. Clinical follow-up time was documented both with an orthopaedic surgeon and with any MHS health care provider and was reported separately.
Statistical Analysis
Summary data are reported as mean, median, range, interquartile range (IQR), and percentage as appropriate.
Continuous variables were assessed for normality using the Kolmogorov-Smirnov test. Results of this testing showed nonnormal distributions for patient age (P = .004), body mass index (BMI) (P = .009), time to surgery (P < .001), number of suture anchors in patients undergoing suture anchor repair (P < .001), number of tunnels in patients undergoing transosseous tunnel repair (P < .001), and number of core sutures used during repair regardless of repair type (P < .001).
For the primary outcome (repair failure), the cohort was analyzed as a whole and as subgroups based on repair technique. For assessment of return to military duty, only active-duty military personnel were included in the analysis.
Univariate associations with collected variables were assessed using Fisher exact tests, chi-square tests, or Mann-Whitney U tests as appropriate. Significance was set at P < .05. All variables showing an association of P < .20 (selected a priori) were included in a multivariate logistic regression model to determine independent risk factors for rerupture (whole cohort and subgroups) and separation from military service for active-duty personnel (active-duty military patients only). Analysis was performed using IBM SPSS Statistics for Macintosh, Version 27 (IBM Corp).
Results
Demographics
A total of 245 knees in 234 patients (11 patients with bilateral injuries) met inclusion criteria. Median patient age was 52 years (IQR, 16). The majority of patients were male (223/234; 95%), Black race (99/234; 42%), and nonsmokers (193/234; 82%). Median BMI was 30.39 (IQR, 6.55). Median follow-up with the treating orthopaedic surgeon was 182 days (IQR, 244 days), and median follow-up with any medical provider in the MHS was 520 days (IQR, 1448 days). Complete demographic data on the study cohort is given in Table 1.
Demographics and Potential Risk Factors (N = 234 Patients) a
Data are presented as n (%) unless otherwise indicated. BMI, body mass index; IQR, interquartile range.
Approximately one-third of the cohort (76/234; 32%) were active-duty military personnel. Active-duty military personnel were younger at the time of injury (median age: 41 years vs 57 years; P < .001), had a lower BMI (median BMI: 29.77 vs 31.03; P = .045), were more likely to smoke (25% vs 14%; P = .03), and less likely to have a diagnosis of diabetes or prediabetes (1% vs 25%; P < .001).
Reported mechanism of injury varied considerably across the study population. In most cases, patients reported a fall or misstep as causing their injury (131/234; 56%). The remainder of injuries primarily occurred during individual recreational activities such as running or individual physical training (45/234; 19%) or organized sporting activities such as basketball and soccer (31/234; 13%). In 10 patients (4%), the injury was described as occurring spontaneously during normal ambulation. Among active-duty military personnel, 9% of patients were injured while deployed overseas. However, only 1 case occurred during actual military training, with the remainder occurring with recreational sports or falls while on military deploymentwith the remainder occurring with recreational sports or falls in a deployed.
Intraoperative Findings and Repair Techniques
The majority of tears were complete superior pole avulsion injuries (218/245; 89%), with the remainder classified as complete midsubstance ruptures (13/245; 5%) and high-grade partial tears (14/245; 6%).
Median time to surgery was 6 days postinjury (range, 0-299 days; IQR, 14 days). Most knees (147/245; 60%) underwent repair using transosseous tunnels, with the remainder undergoing suture anchor repair (78/245; 32%), direct tendon-to-tendon repair (14/245; 6%), or hybrid fixation using both transosseous tunnels and suture anchors (6/245; 2%). In all cases, a nonabsorbable suture was utilized. We did not identify any repairs in which absorbable suture was employed.
For knees undergoing suture anchor repair, a median of 2 anchors (range, 1-4) and 4 core sutures (range, 2-8) were used. In 2 cases, the number of suture anchors and core sutures was not documented in the operative note and so was excluded from analysis for these variables.
For knees undergoing transosseous tunnel repair, a median of 3 tunnels (range, 1-4) and 4 core sutures (range, 2-8) were used. There was no significant difference in the number of core sutures utilized between suture anchor and bone tunnel repairs. In 16 cases, the number of core sutures and transosseous tunnels was not documented in the operative note and so was excluded from analysis for these variables.
Among knees undergoing suture anchor repair, approximately half utilized knotless suture anchors (40/78; 51%), followed by knot-tying anchors (35/78; 45%). The remaining knee received a combination of knot-tying and knotless anchors.
Concurrent procedures were uncommon, occurring in only 14/245 (6%) of cases. The most common concurrent procedure was debridement of patellar enthesophytes or bipartite patella (5/245; 2%), followed by diagnostic knee arthroscopy (3/245; 1%) and open reduction internal fixation of patellar avulsion fragments (3/245; 1%).
Clinical Outcomes
At the time of final follow-up, thigh atrophy (as documented by the treating surgeon) was present in 94/245 (38%) knees. Median flexion at final follow-up was 125° (IQR, 10°). An extension lag was present in 16% (38/245) of knees, with a median lag of 5° (range, 2°-30°; IQR, 7°). Strength on manual muscle testing was full in 83% (203/245) of knees, 4/5 in 15% (36/245), and ≤3/5 in 2% (6/245).
Complications
The overall complication rate from surgery was 18% (43/245). The most common complication of surgery was quadriceps tendon repair failure requiring revision surgery, which occurred in 11% of knees (27/245). Median failure occurred at 92 days postoperatively (range, 2 days–63 months). Three patients had a postoperative pulmonary embolism, and 2 had a postoperative deep vein thrombosis. There were 10 surgical-site infections, and 1 case of wound dehiscence without infection.
The rate of having any complication following surgery was significantly associated with increased BMI (mean: 31.39 vs 29.61; P < .01) but was not associated with sex, race, smoking status, diabetes/prediabetes, tear type, time to surgery, repair technique, or performance of concurrent procedures at the time of surgical repair.
Primary Outcome—Repair Failure
Repair failure occurred in 27 knees (11% of cohort) and is summarized by repair technique in Table 2.
Knees and Proportion With Repair Failure by Surgical Technique
There was a significant association between surgical-site infection following index surgery and eventual repair failure (P = .02). There was a nonsignificant association meeting threshold for inclusion in the logistic regression model for tear location (P = .14). No other assessed variable was significantly associated with tendon rerupture.
Within the multivariate logistic regression model, both the presence of a surgical-site infection following index surgery (P = .01) and a superior pole avulsion tear pattern (P = .047) were independent predictors of repair failure.
When directly comparing transosseous tunnel repair to suture anchor repair, there was no difference in the likelihood of repair failure (P = .51). Among patients undergoing suture anchor repair, there was no difference in the likelihood of failure when comparing knot-tying anchors, knotless anchors, and combination repairs using both (P = .73).
Return to Military Duty
Of the 76 active-duty military personnel, 71 (93%) were able to return to duty, and 5 (7%) required separation from service due to their knee injury.
On univariate analysis, there was no significant association between any potentially predictive factor and the need for medical separation from military service (Table 3). However, lower BMI (P = .06), longer time from injury to surgery (P = .13), and the presence of any postoperative complication (P = .10) met criterion for inclusion in the binary logistic regression model. Within the regression model, no factor was a significant predictor of the need for medical separation from service.
Univariate Analysis of Factor Association With Need for Medical Separation From Service a
Data are presented as n (%) unless otherwise indicated. BMI, body mass index; IQR, interquartile range; USAF, US Air Force; USMC, US Marine Corps.
Discussion
In a large retrospective series of quadriceps tendon ruptures, we identified an overall repair failure rate of 11%, more than twice as high as most prior reports.7,11,18,24 No difference in the repair failure rate was found between suture anchor repair and transosseous tunnel repair (P = .51). The only significant correlation for rerupture was a surgical-site infection after index surgery (P = .02).
Literature on primary quadriceps tendon repairs focuses on functional outcomes, with minimal analysis on repair failure. A myriad of small case series have reported variable rates of failure ranging from (0% to 8.3%).4-6,17,20,22,25,28 Negrin et al 18 reported a failure rate of 8% in 93 quadriceps tendon repairs at a single institution over 15 years, with 3-year minimum follow-up. Roberts and Ketz 24 identified 113 quadriceps tendon repairs between 2012 and 2016 with a reported failure rate of 5% (6/113), with only inflammatory arthritis identified as an independent risk factor (Relative risk, 17.8). In a systematic review of 11 studies with 298 knees, Ciriello et al 7 reported an overall rerupture rate of 2%. In a subsequent systematic review reporting on 18 studies, Haskel et al 11 identified 379 repairs with 10 repair failures (2.6%), irrespective of fixation techniques.
In our study of 245 quadriceps ruptures, the repair failure rate was 11%, substantially greater than previous reports. Our cohort was somewhat unique, with 32% of patients on active-duty military. These younger and higher-demand patients may have contributed to the greater failure rate.
Biomechanical studies comparing suture anchor with transosseous tunnel repairs suggest decreased gap formation with suture anchors but no difference in load to failure.3,19,21,27 Mehta et al 16 systematically reviewed 8 studies, comparing outcomes of suture anchor and transosseous repair techniques published after 2000. They identified 52 total knees repaired with suture anchors across 4 studies and found no difference in repair failure compared with bone tunnels (3.7% vs 0%; P = .07). Suture anchor fixation was used in 78 of the 245 knees in the present cohort (32%). We also found no difference in repair failure between suture anchor and transosseous tunnel repair (9.0% vs 12.2%; P = .51). This finding should be treated with caution given the comparatively small number of patients with suture anchor repair. Larger prospective cohorts with careful control of implants may be able to detect differences in failure between the techniques.
Kindya et al 13 performed a biomechanical comparison of knotless suture anchor quadriceps tendon repair to knot-tying suture anchors and transosseous tunnel techniques. Their study found improved construct stiffness, lower initial gapping with cyclic loading, and greater load to failure (all P < .05) in favor of knotless suture anchors. In our suture anchor cohort, no repair failure difference (P = .73) was found between knotless (40/76; 53%) and knot-tying anchors (35/76; 46%). Further biomechanical and clinical analysis on implant design is needed to investigate the ideal suture anchor implant.
Limitations
The primary limitation of this study is its retrospective design. Like all retrospective studies, our findings are predicated on the accuracy of data recorded in individual patient records. A total of 64 potentially eligible patients were excluded due to incomplete records, missing operative reports, or no preoperative/postoperative documentation, which could have affected our outcomes.
Another limitation is that we relied on existing documentation to determine which patients had failed primary repair. Our assumption that a repair was intact unless failure was specifically documented could have resulted in underestimation of the overall failure rate. We believe it is unlikely that patients would have no documentation of failure if failure occurred, given that care at MHS facilities is free for beneficiaries. It is unlikely that patients would seek care in non-MHS facilities, where they may incur additional cost, rather than return to their MHS surgeon where care is free. We also believe that given the significant disability that occurs with failure of the knee extensor mechanism, such impairments would be regularly documented even by nonorthopaedic providers. However, due to the retrospective design of this study, we cannot exclude the possibility of nondocumented repair failures.
Additionally, our cohort was primarily military retirees (61%), and so it may not be generalizable to the civilian population. Given the number of patients and facilities involved in the care of this population, we were unable to determine the level of training and experience of operative surgeons. Finally, we were unable to obtain complete data on implant manufacturers, suture material used, or postoperative rehabilitation, which could have influenced outcomes.
Conclusion
We observed no difference in failure of primary quadriceps tendon repair between transosseous tunnel and suture anchor repair types or between the 2 main suture anchor implant subtypes (knotless vs knot-tying anchors). A greater than previously reported rerupture rate was observed, indicating the need for continued investigation into optimal surgical techniques.
Footnotes
Final revision submitted June 18, 2024; accepted June 28, 2024.
One or more of the authors has declared the following potential conflict of interest or source of funding: M.S.K. has received hospitality payments from Medical Device Business Services and education payments from Fortis Surgical. S.H.R. has received education payments from Fortis Surgical. G.C.B. has received education payments from Fortis Surgical and Supreme Orthopedic Systems. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval for this study was obtained from DHA IRB Office at Naval Medical Center Portsmouth (No. 959104).
Data Accessibility Statement
The complete data set may be made available on request to the authors.
