Abstract
Background:
Medial patellofemoral ligament (MPFL) reconstruction is an established treatment for recurrent patellar instability. Techniques requiring patellar drilling (tunnels or sockets for fixation) may increase the risk of iatrogenic patellar fracture.
Purpose/Hypothesis:
The purpose was to compare clinical outcomes and complications after MPFL reconstruction using a superficial quadriceps tendon (QT) graft versus a hamstring tendon (HT) graft) fixed with patellar all-suture anchors. It was hypothesized that the superficial QT technique would yield patient-reported outcomes comparable with the HT technique, with a different profile of patella-related complications.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
A retrospective, 2-center review identified 47 patients (49 knees) who underwent MPFL reconstruction for recurrent patellar instability. A total of 23 knees were reconstructed with a patellar tunnel–free superficial QT graft and 26 with an HT graft fixed to the patella using suture anchors and to the femur at the Schöttle point. Patients with prior patellar surgery, periarticular fractures, or insufficient follow-up were excluded. Preoperative radiographic parameters and patient-reported outcomes (Kujala, Lysholm, Tegner, and visual analog scale [VAS] for pain) were recorded. The primary outcome was the Kujala score at final follow-up; secondary outcomes included other scores and complications (recurrent instability, patellar fracture, positive apprehension, and range-of-motion limitation).
Results:
Groups were comparable in demographics and radiographic measures at baseline. The mean follow-up duration was 59.8 months in the QT group and 50.7 months in the HT group. No significant between-group differences were observed in Kujala, Lysholm, Tegner, or VAS pain scores (all P > .05). Overall complication rates were similar (21.7% vs 15.4%). Complications included redislocation, positive apprehension, flexion limitation, and 1 patellar fracture in the HT group. Patients who developed postoperative complications had lower clinical scores, irrespective of graft type.
Conclusion:
Superficial quadriceps and HT autografts provide similarly favorable clinical outcomes and complication rates in MPFL reconstruction. No patellar fractures occurred in the QT group; however, fracture rates were not statistically different between groups and the study was underpowered for rare events. The QT technique avoids patellar drilling and may be considered in patients perceived to be at higher risk for patellar fracture.
Keywords
Patellar dislocation accounts for approximately 3% of all knee injuries. 2 The incidence of patellar instability is 5.8 per 100,000 in the general population and 29 per 100,000 among children aged 10 to 17 years.2,9 MPFL injury frequently occurs following patellar dislocation. 14 Biomechanical and clinical studies have demonstrated that the MPFL is the primary anatomical structure preventing lateral displacement of the patella.4,9 Therefore, MPFL reconstruction is recommended for the surgical treatment of recurrent patellar dislocations. 10
Hamstring tendons (HTs), superficial QT, and synthetic allografts are among the preferred graft options for MPFL reconstruction.6,13 The most commonly used technique involves fixing an HT graft to the patellar and femoral attachment points of the MPFL using various fixation materials. 15 MPFL reconstruction with HT grafts has generally yielded favorable clinical outcomes22,23; however, placement of implants on the patella remains a major concern because of potential complications such as patellar fracture and implant irritation. 11 Recently, MPFL reconstruction using the superficial QT has gained popularity owing to its biomechanical similarity to the native MPFL and the avoidance of patellar fixation, thereby reducing the risk of iatrogenic patellar fracture.7,17,19
HT and QT autografts are commonly used options for MPFL reconstruction.20,21 While most hamstring-based techniques require patellar drilling for fixation (tunnels or sockets for implants), 13 quadriceps-based pedicled techniques can preserve patellar attachment and avoid patellar drilling.5,6 Comparative evidence between these approaches remains limited and heterogeneous, and the clinical relevance of patella-related complications warrants further clarification.20,21 Furthermore, no consensus has been reached regarding the optimal graft choice.
Therefore, the aim of this study was to compare clinical outcomes and complications after MPFL reconstruction using a patellar drill-free superficial QT graft versus an HT graft fixed with patellar all-suture anchors. We hypothesized that the superficial quadriceps technique would yield patient-reported outcomes comparable with the hamstring technique, with a different profile of patella-related complications.
Methods
Study Design and Ethical Approval
This 2-center retrospective observational clinical study was approved by the institutional review board at the authors’ institution. The study was conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants prior to surgery.
Patient Selection
Patients who underwent MPFL reconstruction between January 2018 and June 2023 at 2 centers were retrospectively reviewed.
The inclusion criteria were as follows:
undergoing surgery using either the HT graft or the superficial QT graft technique,
availability of complete preoperative radiographs (long-leg, tangential, and knee radiographs) and magnetic resonance imaging (MRI) scans,
availability of complete postoperative radiographs (long-leg, tangential, and knee radiographs).
The exclusion criteria were as follows:
a history of previous surgery for patellar dislocation (eg, tibial tubercle transfer),
fractures around the knee joint,
refusal to provide informed consent,
insufficient follow-up data.
The indication for MPFL reconstruction was defined as recurrent, clinically relevant patellar instability (≥2 documented lateral patellar dislocations and/or recurrent symptomatic subluxation episodes despite appropriate nonoperative management) in combination with supportive clinical examination findings (eg, positive apprehension test and/or J-sign).
Based on these criteria, 21 patients (23 knees) in the QT group and 26 patients (26 knees) in the HT group were included in the study (Figure 1).
Flowchart of the study population. In the QT group, 21 patients corresponded to 23 knees because 2 patients had bilateral involvement, whereas in the HT group, 26 patients corresponded to 26 knees. HT, hamstring tendon; QT, quadriceps tendon.
All included cases underwent isolated MPFL reconstruction at the index procedure. The present cohort therefore reflected surgeon-selected isolated MPFL reconstructions, and tibial tubercle–trochlear groove (TT-TG) distance and patellar height were recorded to characterize baseline anatomy rather than to perform powered subgroup analyses.
Graft technique allocation was surgeon/center dependent, not patient dependent. Each surgeon/center exclusively used 1 technique, reflecting the surgeon's established expertise and routine practice.
Demographic (age, sex, side) and radiological (Caton-Deschamps index, TT-TG distance) parameters were compared between the 2 groups to assess baseline homogeneity and to minimize potential selection bias (Table 1).
Baseline Characteristics of the Study Cohorts a
Data are presented as n (%) unless otherwise indicated. BMI, body mass index. Age, sex, and BMI are reported per patient, whereas side, follow-up duration, osteochondral defect, and complications are reported per knee. In the QT group, 23 knee correspond to 21 patients because 2 patients had bilateral involvement.
Mann-Whitney U test.
Pearson chi-square.
Radiological Characteristics
Radiological measurements were performed by an independent orthopaedic surgeon (E.T., Ş.G.) who did not participate in the surgeries. Patellar height was measured on preoperative lateral knee radiographs using the Caton-Deschamps index.
Additionally, the TT-TG distance was calculated from preoperative MRI scans, and the presence of a trauma-related osteochondral defect after the index dislocation was evaluated on preoperative MRI.
Surgical Technique
Superficial QT Group
All surgical procedures were performed by two orthopaedic surgeons (H.A.A. and B.M.A.). All patients were operated on in the supine position under a pneumatic tourniquet. The procedure began with diagnostic knee arthroscopy to evaluate intra-articular pathologies and to assess patellofemoral tracking. An 8-cm longitudinal incision was made over the proximal patella, and the patella and QT were exposed through the subcutaneous tissues. The area to be harvested as the graft was marked, and the most superficial layer of the QT (the superficial QT) was prepared as the graft. A flap measuring approximately 8 to 10 mm in width, 3 to 5 mm in thickness, and 9 to 11 cm in length was elevated proximally, leaving a 3- to 4-mm strip medially intact. The graft was elevated from the upper third of the patella together with 2 to 3 cm of periosteum. Careful dissection was performed to prevent tendon rupture during harvesting. The lateral margin of the graft was elevated 5 mm distal to the medial margin of the patella to prevent folding during medial rotation. Graft dimensions were individualized based on patient size and tendon quality, aiming for a superficial strip typically within the reported range to ensure adequate handling and strength while preserving the remaining QT. Only the most superficial layer was harvested, and patellar attachment was maintained. The prepared graft was passed medially through the second layer between the vastus medialis oblique and the capsule. The femoral attachment site was determined under fluoroscopic guidance according to the Schöttle point.1,3 The knee was flexed to 45°, the graft was inserted into the tunnel, and fixation was performed using a single polyether ether ketone screw. Before fixation, graft tension was evaluated arthroscopically by observing patellar tracking (Figure 2). All layers were closed in an anatomical manner.
Elevation of only (A) the superficial layer of the quadriceps tendon and (B) the superficial quadriceps tendon with its patellar attachment preserved, marked in the blue triangular area (indicated by the green star). (C) Final arthroscopic evaluation of patellofemoral congruence following medial patellofemoral ligament reconstruction. D, distal; P, proximal.
HT Group
All surgical procedures were performed by an orthopaedic surgeon (E.V). All patients were operated on in the supine position under a pneumatic tourniquet. Following diagnostic arthroscopy and management of intra-articular pathologies, a small oblique incision was made over the pes anserinus. The gracilis tendon was identified and released from its distal attachment while preserving the semitendinosus tendon. The gracilis tendon was harvested using a tendon stripper and prepared for implantation. A second longitudinal incision was made over the medial epicondyle, halfway between the patella and the adductor tubercle (Figure 3). The patellar attachment site of the MPFL was exposed, and two 2.9-mm blind sockets were drilled into the proximal two-thirds of the patella. Two all-suture anchors (Anchor Yarn 2.9-mm Needleless; Doratek) were inserted, and the graft was tied over the prepared site without creating transpatellar tunnels. On the femoral side, the Schöttle point was localized under fluoroscopic guidance, and the femoral tunnel was prepared. The graft was passed between the second and third layers and pulled through the femoral tunnel. It was fixed using a bioabsorbable screw at 45° to 60° of knee flexion under moderate tension. Patellar tracking was confirmed arthroscopically and clinically before closure. All layers were closed in an anatomical fashion.
(A) An oblique incision is performed for the harvesting of the gracilis tendon and a longitudinal one for the medial patellofemoral ligament reconstruction. The midportion of the gracilis tendon (B) is positioned over the medial border of the patella and (C) is fixed with 2 all-suture anchors.
Postoperative Rehabilitation
All patients received low–molecular weight heparin prophylaxis for 10 days postoperatively. A hinged knee brace was used for immobilization; full extension was allowed, while flexion was restricted to 90° for the first 2 weeks. Full range of motion was permitted at the end of the sixth week. Weightbearing was restricted for the first 2 weeks, followed by supported ambulation until the sixth week, after which patients began walking without support.
Patients were typically discharged on the first postoperative day.
Outcome Measures
The primary outcome measure was the Kujala patellofemoral score at final follow-up. Secondary outcome measures included the Lysholm knee score, Tegner activity scale, visual analog scale (VAS) for pain, and postoperative complications (recurrent dislocation, patellar fracture, positive apprehension test, and range of motion limitation). Each knee was considered an independent case. Two patients in the QT group had bilateral involvement, while no bilateral cases were present in the HT group. Given the small number of bilateral cases, analyses were performed on a per-knee basis.
Statistical Analysis
Statistical analyses were performed using SPSS Version 24.0 (IBM Corp). Categorical variables were expressed as numbers and percentages, and continuous variables as mean ± SD and median (minimum-maximum) values. Between-group comparisons were made using Pearson chi-square test for categorical variables, the independent t test for normally distributed variables, and the Mann-Whitney U test for nonnormally distributed variables; normality of continuous variables was assessed using the Shapiro-Wilk test.
The relationships between final clinical scores, the presence of trauma-related osteochondral defect, and postoperative complications were evaluated using Spearman correlation analysis. A P value < .05 was considered statistically significant.
Results
The study included 21 patients (23 knees) in the superficial QT group (4 male, 17 female; mean age, 18.3 ± 5.2 years; mean body mass index [BMI], 23.6 ± 2.6 kg/m2) and 26 patients (26 knees) in the HT group (12 male, 14 female; mean age, 19.6 ± 5.7 years; mean BMI, 23.7 ± 2.4 kg/m2). The mean follow-up duration was 59.8 ± 16.8 months in the QT group and 50.7 ± 8.9 months in the HT group (P = .09). Although the proportion of male patients was lower in the QT group, there were no statistically significant differences in demographic characteristics between the 2 groups (all P > .05) (Table 1).
The mean Caton-Deschamps index was 1.2 ± 0.2 in the QT group and 1.1 ± 0.1 in the HT group (P = .07). The mean TT-TG distance was 16.9 ± 3.5 mm in the QT group and 15.5 ± 3.7 mm in the HT group (P = .31) (Table 2).
Radiological Characteristics of the Study Cohort a
TT-TG, tibial tubercle–trochlear groove.
Mann-Whitney U test.
Clinical outcome scores at final follow-up are summarized in Table 3. No statistically significant between-group differences were observed in Kujala, Lysholm, Tegner, or VAS pain scores (all P > .05).
Clinical Scores on Final Follow-up Visit of the Study Cohorts a
VAS, visual analog scale.
Mann-Whitney U test.
Postoperative complications occurred in 5 knees (21.7%) in the QT group and 4 knees (15.4%) in the HT group (P = .76) (Table 1). The distribution included redislocation, positive apprehension, flexion limitation, and 1 patellar fracture in the HT group.
In the QT group, 2 redislocations occurred: one case was associated with increased TT-TG distance and subsequently underwent tibial tubercle osteotomy, and the other was related to femoral fixation issues and underwent revision surgery. In the HT group, 1 redislocation occurred and was managed with revision surgery.
The remaining complications (positive apprehension and flexion limitation) were managed nonoperatively with structured physical therapy, and full range of motion was regained in the flexion limitation cases.
One patient (3.8%) in the HT group sustained a transverse patellar fracture after a minor fall; the fracture line did not traverse the anchor socket locations and the patient underwent surgical fixation.
Correlation analysis revealed that the presence of an osteochondral defect following the initial trauma was not associated with inferior clinical scores (all P > .05). In contrast, postoperative complications were associated with lower Kujala and Lysholm scores and modestly lower Tegner activity scores (Table 4).
Correlation Analysis Between Clinical Scores and the Presence of an Osteochondral Defect and Postoperative Complications
Spearman rho.
Bold P values indicate statistical significance (P < .05).
Discussion
The principal finding of this study is that MPFL reconstruction using either HT or superficial QT grafts achieved comparable outcomes with high patient-reported scores and low complication rates. The consistently high Kujala and Lysholm scores in both groups suggest that graft choice may have a limited impact on clinical outcomes. Nevertheless, the QT technique avoids patellar implantation and patellar tunnel creation, thereby eliminating the risk of iatrogenic patellar fracture and offering a clinically meaningful advantage.
Multiple graft and fixation methods have been described for MPFL reconstruction; however, no consensus exists regarding the optimal graft choice.2,8 Although HT grafts are more commonly preferred in the literature, 12 reconstructions performed with QT grafts have likewise yielded successful results.6,7,17 Sanguanjit et al 21 reported no differences in clinical scores between hamstring and quadriceps grafts, and Runer et al 20 similarly demonstrated comparable functional outcomes between the 2 techniques. The originality of our study lies in the direct comparison of these 2 graft options within a standardized treatment framework across both centers, including uniform preoperative imaging assessment, femoral fixation localization at the Schöttle point, intraoperative confirmation of patellar tracking, and a shared postoperative rehabilitation protocol. Given that patient-reported outcomes were comparable between graft types and that observed failures were associated with factors such as femoral attachment malposition or unaddressed anatomic risk factors (eg, increased TT-TG in a failure case), our findings suggest that surgical execution and patient selection may have greater influence on outcomes than graft selection alone. Although no patellar fractures occurred in the QT cohort, the between-group difference was not statistically significant, and the study was underpowered to draw definitive conclusions regarding rare events such as patellar fracture. Therefore, this finding should be interpreted as an observational signal rather than proof of superiority.
MPFL reconstruction with a superficial QT graft eliminates the need for patellar tunnel creation and thus removes the risk of patellar fracture. Because tunnel entry points—and related stress concentrations—are not created in the patella, patellar biomechanical integrity is preserved. This feature may be particularly advantageous for reducing complications in patients with smaller patellae or in pediatric populations. However, given the relatively thin structure of the QT graft, the risk of graft tearing during preparation should be considered and appropriate surgical care exercised.
The HT graft is a strong biomechanical alternative, with high tensile strength and stiffness, low donor-site morbidity, and a technique that is widely familiar to surgeons. Although our hamstring technique did not involve transpatellar tunnels, it still requires patellar drilling for implant placement (small blind sockets for anchors), which may represent a biomechanical vulnerability in the patella; whether this affects fracture risk requires larger studies. This consideration may be particularly relevant in patients with thinner bone or highly active young individuals. Consequently, each graft offers distinct advantages and risks; selection should be individualized according to patient anatomy, age group, and surgeon experience.
Complications after MPFL reconstruction may include recurrent instability, range-of-motion limitation, a positive apprehension test, and patellar fracture 23 . Shah et al 23 reported an overall complication rate of 26.1%, while Jackson et al 11 noted rates up to 32%. Sappey-Marinier et al 22 reported a redislocation rate of 4.7% after hamstring-based reconstructions, whereas a systematic review of the superficial QT technique reported no redislocations, with a 3% rate of positive apprehension testing. 18 Recurrent instability may occur when isolated reconstruction is insufficient and additional bony procedures are indicated (eg, increased TT-TG distance or high-grade trochlear dysplasia), 16 and postoperative range-of-motion loss may relate to graft tensioning and/or early rehabilitation factors. Techniques involving patellar drilling for fixation (tunnels or sockets for implants) may increase the risk of patellar fracture. In our series, complication rates were consistent with the literature, with no statistically significant differences between groups. No patellar fractures were observed in the QT group; however, the between-group difference was not statistically significant and the study was underpowered to draw definitive conclusions regarding rare events such as patellar fracture. Redislocation occurred in 8.7% (2/23) and 3.8% (1/26) of knees in the QT and HT groups, respectively, and this difference was not statistically significant. In our cohort, redislocation cases were observed in association with factors such as increased TT-TG distance in one case and femoral fixation issues in another; however, given the small sample size, these observations should be interpreted cautiously and do not establish causality. Although TT-TG distance and patellar height were recorded to characterize baseline anatomy, the study was not powered for stratified or multivariable analyses based on these parameters; therefore, no definitive conclusions can be drawn regarding TT-TG distance or patella alta as predictors of redislocation in this cohort.
Trauma-related osteochondral defect was not associated with inferior clinical scores. This finding may be attributable to heterogeneity in lesion size and location and/or concomitant management of intra-articular pathology, including appropriate treatment of osteochondral fragments and early restoration of joint congruity. Although some studies have associated osteochondral defects with lower clinical scores, 5 our results suggest that this relationship may not be uniform across all patients. However, this finding should be interpreted cautiously given the sample size and may not generalize to all patterns or severities of chondral pathology.
In our cohort, patients who developed postoperative complications had significantly lower Kujala and Lysholm scores. This aligns with prior studies reporting that graft failure, patellar fracture, residual anterior knee pain, and recurrent instability adversely affect functional recovery. In particular, technical errors and inaccurately located femoral fixation points have been identified as important contributors to complications. 23 Similarly, a midterm follow-up study showed that complications such as graft failure and patellar fracture were associated with significantly lower clinical scores. 21
Limitations
This study has several limitations. The sample size in each group was relatively small, and follow-up durations varied among patients, limiting statistical power—particularly for rare events such as patellar fracture. In addition, the 2 techniques were performed by different surgeons, which may introduce surgeon-dependent variability. Because graft technique allocation was surgeon/center dependent and nonrandom (each surgeon exclusively used 1 technique), this may introduce confounding related to surgeon experience and center-specific practice patterns. Moreover, the retrospective, nonrandomized design, the absence of objective strength or biomechanical testing, and the lack of standardized grading of trochlear dysplasia severity (eg, Dejour classification) further limit causal inference. However, both centers used standardized core steps including femoral localization at the Schöttle point, graft passage through the appropriate tissue planes, intraoperative tracking assessment with tensioning at 45° to 60° of flexion, and an identical postoperative rehabilitation protocol. In addition, a small number of bilateral knees were analyzed on a per-knee basis, which may introduce partial nonindependence and should be considered when interpreting the statistical comparisons. Despite these limitations, this study provides a direct midterm comparison of 2 commonly used graft options within a standardized treatment framework across both centers.
Our findings indicate that both graft types are safe and effective options for MPFL reconstruction. The superficial QT technique avoids patellar drilling and patellar implants and therefore may be considered in patients perceived to be at higher risk for patellar fracture; however, larger adequately powered studies are needed to confirm any difference in fracture rates. Future large-sample, prospective randomized studies incorporating 3-dimensional motion analysis and comprehensive patient-reported outcomes could further delineate the biomechanical and functional effects of different graft choices. Notwithstanding its limitations, this study provides valuable comparative evidence by directly contrasting the clinical outcomes of MPFL reconstructions performed with hamstring and superficial QT grafts.
Conclusion
Both superficial QT and HT autografts yielded comparable and favorable clinical outcomes in MPFL reconstruction. Although overall complication rates were similar, the absence of patellar tunneling in the QT technique provided an additional safety advantage by reducing the risk of patellar fracture. Redislocation rates were numerically different between groups; however, this difference was not statistically significant. Therefore, no definitive conclusions can be drawn regarding redislocation risk based on graft choice in this cohort. Considering these distinct characteristics, graft selection should be individualized according to the patient's anatomical features and activity level and the surgeon's experience. Large-scale prospective studies with standardized follow-up protocols are warranted to further clarify and establish definitive recommendations for graft selection in MPFL reconstruction.
Footnotes
Final revision submitted January 14, 2026; accepted January 20, 2026.
The authors declared that there are no conflicts of interest in the authorship and publication of this contribution. 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 scientific research and ethics committee of Ankara Etlik City Hospital (approval No: AEŞH-BADEK1-2025-019; May 28, 2025).
