Factors affecting patellofemoral contact pressure during medial patellofemoral ligament reconstruction. A systematic review of cadaveric biomechanical testing Authors: Junxing Shao, Yu Gao, Zhengwen Si, Modi Yang Published: 2025 ABSTRACT BackgroundMedial patellofemoral ligament (MPFL) reconstruction has become very popular for treating patellar instability. However, it is unclear how patellofemoral (PF) contact pressure changes after MPFL reconstruction and how to restore the native PF contact pressure.PurposeTo identify the effect of intra-operative factors (graft types, tunnel locations, pre-tension levels, and graft fixation angles) of MPFL reconstruction on PF contact pressure in a review of human cadaveric studies.MethodsStudies in PubMed and Web of Science databases were searched from database inception to June 2023. The inclusion criterion was cadaveric biomechanical studies of MPFL reconstruction that provided data on the PF contact pressure. The Quality Appraisal for Cadaveric Studies (QUACS) scale was used to perform comprehensive assessment of the methodological quality of the included studies.ResultsA total of 14 biomechanical studies comprising 135 knees were identified and reported for the overall or medial and lateral PF contact pressures after MPFL reconstruction. Five studies selected at least two different grafts for reconstruction and reported that all graft types could restore the native PF contact pressure. In terms of femoral tunnel location: Schöttle point (7 studies), adductor tubercle (1 study), proximal or distal to the anatomic origin (2 studies), the femoral tunnel location should be closer to the native insertion; otherwise, the postoperative PF contact pressure could increase significantly. The graft pre-tension was most commonly 2 N or 10 N, while fixation of the graft was commonly at 30° or 60° of knee flexion. Overtensioning of the graft increased the medial PF contact pressure and patellar tilt; a fixation angle deviating significantly from the commonly used 30° or 60° (e.g.0° or 90°) may lead to an increase in PF contact pressure, particularly on the medial facet, however, the effect is not uniform across all studies and may be influenced by graft type and tension.ConclusionIn cadaveric biomechanical testing, MPFL reconstruction can effectively restore the native PF contact pressure irrespective of the graft type, but changes in the femoral tunnel location can severely affect the postoperative PF contact pressure. Fixation of the graft at 30° or 60° of knee flexion with 2 N or 10 N of pre-tension would be more conducive to restoration of the original PF pressure. Introduction Considering the high incidence and recurrence of patellar dislocation, medial patellofemoral ligament (MPFL) reconstruction is a cost-effective option for restoring the stability of the patellofemoral (PF) joint with minimal trauma to the patient [1, 2], and is recommended by expert guidelines in cases of recurrent instability [3].The current MPFL reconstruction techniques are becoming the preferred option for treating patellar dislocation, and they have the advantage of a low post-operative recurrence rate and allow for early return to sports [4]. However, a high proportion of patients experience PF pain after MPFL reconstruction, which is associated with changes in PF contact pressure [5–7]. At a minimum 10-year follow-up, Elvire Servien et al. [8] found that one-third of patients have patellofemoral arthritis after isolated MPFL reconstruction using the gracilis tendon autograft. Similarly, Marcheggiani et al. reported a 35% rate of patellofemoral arthritis at the 5-year follow-up [9]. They suggested that the technical aspects of surgery may have affected the PF joint pressure, which in turn led to patellofemoral osteoarthritis. If the reconstructed PF contact pressure is very high, patients will experience knee discomfort and impaired knee mobility in the early stages and osteoarthritis in the later stages [10–12].This is of particular concern as acute patellar dislocation events are frequently associated with substantial damage to the MPFL itself, as well as osteochondral injuries to the lateral trochlea and medial patella [13]. If the reconstructed PF contact pressure is very low, then the reconstructed MPFL may not serve as a good checkrein and patients will have a persistently positive patellar apprehension test [14–16]. In patients with unclosed epiphysis, very low PF contact pressure indicates that the femoral trochlea has inadequate mechanical stimulation, which can affect the development of the femoral trochlea in the distant future [17–20]. Therefore, although the MPFL reconstruction techniques are simple and easy to perform, restoration of the native PF pressure is extremely important. As it is not possible to measure PF joint pressures in the living body directly, the results of cadaveric studies and biomechanical tests form an important basis. Identification of the native PF contact pressure and assessment of the effect of intra-operative factors (graft types, tunnel locations, pre-tension levels, and graft fixation angles) on the PF contact pressure provide us with a reference for performing a “perfect” MPFL reconstruction surgery [21–24]. Currently, there are no clinical studies on postoperative PF pressure in patients after MPFL reconstruction, and the biomechanical findings vary and are inconsistent. The primary aim of this study was to identify the effect of intra-operative factors (graft types, tunnel locations, pre-tension levels, and graft fixation angles) of MPFL reconstruction on PF contact pressure in a review of human cadaveric studies. Our initial hypothesis was that MPFL reconstruction could restore the native PF contact pressure irrespective of the graft type, but non-anatomical femoral and patellar tunnel positions, unsuitable graft fixation angles, and inappropriate graft pretensions could significantly affect the postoperative PF contact pressure. Methods Article screening Literature search for the relevant content was performed in the following two digital databases: PubMed and Web of Science on June 1, 2023. The study was registered in the PROSPERO international prospective register of systematic review (CRD42022342387). The purpose of literature search was to identify the reports providing information on the PF contact pressure of intact knees and the change after MPFL reconstruction by using different grafts, different tunnel locations, different knee flexion angles at the time of graft fixation, and different graft pretensions. The search string included the following terms: “patellar dislocation” OR “patellar instability” AND “MPFL reconstruction” OR “medial patellofemoral ligament reconstruction” AND “patellofemoral contact pressures” OR “contact pressures”, resulting in a total of 1235 articles. All hits were exported into EndNoteX9 and then checked for duplicates. Two independent reviewers independently screened the title and abstract of each search-identified article and then reviewed the full text of each article that was selected on the basis of the inclusion and exclusion criteria. Inclusion criteria were as follows: (1) peer-reviewed biomechanical studies using human cadaveric knee specimens; (2) analysis includes graft type, tunnel location, pre-tension level, and graft fixation angle (any one or more); (3) Recording of PF Pressure and flexion angle; (4) reporting of means, sample sizes, and SDs. The exclusion criterion was animal experiments or finite element studies. Studies on MPFL repair or studies that measured only patellar tracking/MPFL length change/patellar tilt were excluded. Then the two reviewers jointly made the final decision regarding whether to include or exclude the article. Quality assessment The articles were evaluated for quality using the Quality Appraisal for Cadaveric Studies (QUACS) scale [25], which is a validated tool for assessing the quality of cadaveric studies. The scale consists of a checklist encompassing 13 items, with higher scores correlating with higher quality of the study. The included studies were scored by two independent raters. A discussion on the differences in scores was performed between reviewers for resolution, and the total score and mean were calculated. Data abstraction Two reviewers independently abstracted the relevant data from included articles and recorded these data on a predetermined Microsoft Excel spreadsheet. PF pressure values, graft types, fixation angles, graft pretensions, femoral/patellar tunnel positions, and outcomes were extracted and cross-checked for accuracy; for cases in which the outcome measures were not reported in the text, numerical data were extracted by digitizing the graphs using WebPlotDigitizer4.5. Data analysis A kappa (k) value was calculated for each stage of article screening to evaluate the inter-reviewer agreement; and k >0.60 indicated substantial agreement, 0.21 ≤ k ≤ 0.60 indicated moderate agreement, and k < 0.21 indicated slight agreement [26]. However, the included studies showed significant heterogeneity, due to numerous surgical techniques, different graft types, different fixation angles, different graft pretensions, different femoral/patellar tunnel positions. Some studies measured the overall PF pressure, and other studies divided the PF joint into medial and lateral facets and tested the mean pressure on the medial and lateral PF facets. Thus, descriptive statistics was calculated using Microsoft Excel. Results Article screening A total of 1235 articles were initially identified from the PubMed and Web of Science databases, and after screening for the inclusion and exclusion criteria, 14 cadaveric studies were included in the analysis. The study selection process is shown in Fig. 1. There was substantial agreement between reviewers on the title and abstract (k = 0.74), and full-text screening (k = 0.75). Fig. 1Flowchart displaying the studies selection process Flowchart displaying the studies selection process Quality assessment The mean QUACS score of the included studies was 85.2 ± 6.2 (range, 77 to 92). The mean QUACS score for each study is shown in Table 1. A study was considered to be of high quality if 10 or more than 10 out of the 13 items (10/13, 77 points) on the scale were fulfilled. Data abstraction After applying the inclusion and exclusion criteria, 14 cadaveric studies comprising 135 knees were ultimately analyzed in this study (Table 1). Almost all studies reported the information about the graft types, tunnel locations, pre-tension levels, graft fixation angles, and the native PF contact pressure. Only one study did not report the graft fixation angles, and one study did not provide information on pre-tension. Table 1Characteristics of included studiesStudyYearCountryKneesAgePressure sensorsGraft LypeFixation anglesGraft tensionQUACSMehl et al.[27]2020U.S.A.1054.4 ± 8.9K-Scan 4000 sensor,TekscanGTFT30°2 N 5 N 10 N92Sakamoto et al.[33]2020Japan971.2 ± 14.2K-SCAN 2, 4010;Tekscan,PST0°, 30°, 60°, 90°2 N92Lorbach et al.[28]2018Germany12 – I-Scan; TekscanGT15°, 30°, 45°, 60°, 75°, 90°2 N77Stephen et al.[34]2016U.K.959, 23–77I-Scan 5051; TekscanGTQTFLT60°2 N 10 N85Stephen et al.[35]2014U.K.874, 46–88I-Scan, 5051;TekscanGT0°, 30°, 60°2 N 10 N 30 N77Beck et al.[36]2007U.S.A.841–67I-Scan, 5051;TekscanST30°2 N 10 N 40 N77Melegari et al.[37]2008U.S.A.1181, 60–99I-Scan; TekscanQT – 10 N77Zumbansen et al.[29]2020Germany12 – I-Scan; TekscanGT15°, 30°, 45°, 60°, 75°, 90°2 N85Stephen et al.[38]2015U.K.861 ± 10I-Scan 5051; TekscanGT30°2 N85Rood et al.[10]2015Netherlands779, 67–89I-Scan, 4205;TekscanGT30°Slight tension85Black et al.[39]2020U.S.A.852, 34–69Sensor, 5040; TekscanSTAT30°2 N92Lorbach et al.[30]2017Germany16 – I-Scan; TekscanGTFLT30°2 N85Ambra et al.[31]2021Brazil846I-Scan 5051; TekscanGTQT30°2 N92Spang et al.[32]2020USA.971, 56–78Sensor, 5040; TekscanGT30°2 N92Data are presented as mean ± standard deviation or range. Fixation angles are listed in degrees of knee flexion and separated by commas for clarity;-, not reported; N, Newton; QUACS, Quality Appraisal for Cadaveric Studies scale; GT, gracilis tendon; ST, semitendinosus tendon; AT, adductor tendon; QT, quadriceps tendon; FLT, fascia lata tendon; FT, fiber tape; PST, polyester suture tape Table 1Characteristics of included studiesStudyYearCountryKneesAgePressure sensorsGraft LypeFixation anglesGraft tensionQUACSMehl et al.[27]2020U.S.A.1054.4 ± 8.9K-Scan 4000 sensor,TekscanGTFT30°2 N 5 N 10 N92Sakamoto et al.[33]2020Japan971.2 ± 14.2K-SCAN 2, 4010;Tekscan,PST0°, 30°, 60°, 90°2 N92Lorbach et al.[28]2018Germany12 – I-Scan; TekscanGT15°, 30°, 45°, 60°, 75°, 90°2 N77Stephen et al.[34]2016U.K.959, 23–77I-Scan 5051; TekscanGTQTFLT60°2 N 10 N85Stephen et al.[35]2014U.K.874, 46–88I-Scan, 5051;TekscanGT0°, 30°, 60°2 N 10 N 30 N77Beck et al.[36]2007U.S.A.841–67I-Scan, 5051;TekscanST30°2 N 10 N 40 N77Melegari et al.[37]2008U.S.A.1181, 60–99I-Scan; TekscanQT – 10 N77Zumbansen et al.[29]2020Germany12 – I-Scan; TekscanGT15°, 30°, 45°, 60°, 75°, 90°2 N85Stephen et al.[38]2015U.K.861 ± 10I-Scan 5051; TekscanGT30°2 N85Rood et al.[10]2015Netherlands779, 67–89I-Scan, 4205;TekscanGT30°Slight tension85Black et al.[39]2020U.S.A.852, 34–69Sensor, 5040; TekscanSTAT30°2 N92Lorbach et al.[30]2017Germany16 – I-Scan; TekscanGTFLT30°2 N85Ambra et al.[31]2021Brazil846I-Scan 5051; TekscanGTQT30°2 N92Spang et al.[32]2020USA.971, 56–78Sensor, 5040; TekscanGT30°2 N92Data are presented as mean ± standard deviation or range. Fixation angles are listed in degrees of knee flexion and separated by commas for clarity;-, not reported; N, Newton; QUACS, Quality Appraisal for Cadaveric Studies scale; GT, gracilis tendon; ST, semitendinosus tendon; AT, adductor tendon; QT, quadriceps tendon; FLT, fascia lata tendon; FT, fiber tape; PST, polyester suture tape Characteristics of included studies K-Scan 4000 sensor, Tekscan GT FT K-SCAN 2, 4010; Tekscan, GT QT FLT ST AT GT FLT GT QT Data are presented as mean ± standard deviation or range. Fixation angles are listed in degrees of knee flexion and separated by commas for clarity;-, not reported; N, Newton; QUACS, Quality Appraisal for Cadaveric Studies scale; GT, gracilis tendon; ST, semitendinosus tendon; AT, adductor tendon; QT, quadriceps tendon; FLT, fascia lata tendon; FT, fiber tape; PST, polyester suture tape PF contact pressure in intact knees The original PF contact pressures in these 14 cadaveric studies were extracted, and they are summarized in Table 2. All included studies tested the PF contact pressure of intact knees by means of a Tekscan pressure-sensitive film. Seven included studies [10, 27–32] tested the mean overall PF contact pressure, and the remaining studies [31, 33–39] divided the PF joint into medial and lateral facets and tested the mean pressure on the medial and lateral PF facets. Because of the different cadaveric conditions, dissection methods, tension forces on the quadriceps, and pressure measurement software used in each study, specific data on the native knee PF contact pressure varied from study to study, but the variation tendency remained the same. The PF contact pressure was close to zero in intact knees at 0° of knee extension, and the overall pressure on the lateral PF joint facet was greater than that on the medial PF joint facet. The PF contact pressure gradually increased during 0°–30° of flexion, reached the maximum at 30°–60° of knee flexion, and then generally decreased with the angle of knee flexion. Once the MPFL was torn or incised, the PF contact pressure was reduced. Table 2Patellofemoral contact pressure (mean value, Mpa) of the intact kneesPF pressureMedial PF pressureLateral PF pressureStudyKneesageFlexion at 0°Flexion at 30°Flexion at 60°Flexion at 90°Flexion at 0°Flexion at 30°Flexion at 60°Flexion at 90°Flexion at 0°Flexion at 30°Flexion at 60°Flexion at 90°Mehl et al.[27]1054.4 ± 8.90.0010.1060.1000.063––––––––Sakamoto et al.[33]971.2 ± 14.2––––0.0930.0960.0940.080.1010.1000.0970.084Lorbach et al.[28]12–0.160.460.710.73––––––––Stephen et al.[34]959, 23–77––––0.370.600.540.560.410.870.690.57Stephen et al.[35]874, 46–88––––0.370.680.460.570.420.790.620.57Beck et al.[36]841–67–––––0.530.470.65–0.720.700.66Melegari et al.[37]1181, 60–99–––––0.581.061.10–0.541.211.32Zumbansen et al.[29]12–0.160.470.720.73––––––––Stephen et al.[38]861 ± 10––––0.380.410.520.560.530.410.320.33Rood et al.[10]779, 67–890.0040.020.020.02––––––––Black et al.[39]852, 34–69–––––0.120.130.11–0.160.210.19Lorbach et al.[30]16––0.300.440.43–0.180.230.22–0.150.210.22Ambra et al.[31]846–0.530.720.49–0.370.720.40–0.660.750.57Spang et al.[32]971,56–78–0.130.190.15––––––––-, not reported; PF pressure, patellofemoral contact pressure; Medial PF pressure, medial patellofemoral contact pressure; Lateral PF pressure, lateral patellofemoral contact pressure. Table 2Patellofemoral contact pressure (mean value, Mpa) of the intact kneesPF pressureMedial PF pressureLateral PF pressureStudyKneesageFlexion at 0°Flexion at 30°Flexion at 60°Flexion at 90°Flexion at 0°Flexion at 30°Flexion at 60°Flexion at 90°Flexion at 0°Flexion at 30°Flexion at 60°Flexion at 90°Mehl et al.[27]1054.4 ± 8.90.0010.1060.1000.063––––––––Sakamoto et al.[33]971.2 ± 14.2––––0.0930.0960.0940.080.1010.1000.0970.084Lorbach et al.[28]12–0.160.460.710.73––––––––Stephen et al.[34]959, 23–77––––0.370.600.540.560.410.870.690.57Stephen et al.[35]874, 46–88––––0.370.680.460.570.420.790.620.57Beck et al.[36]841–67–––––0.530.470.65–0.720.700.66Melegari et al.[37]1181, 60–99–––––0.581.061.10–0.541.211.32Zumbansen et al.[29]12–0.160.470.720.73––––––––Stephen et al.[38]861 ± 10––––0.380.410.520.560.530.410.320.33Rood et al.[10]779, 67–890.0040.020.020.02––––––––Black et al.[39]852, 34–69–––––0.120.130.11–0.160.210.19Lorbach et al.[30]16––0.300.440.43–0.180.230.22–0.150.210.22Ambra et al.[31]846–0.530.720.49–0.370.720.40–0.660.750.57Spang et al.[32]971,56–78–0.130.190.15––––––––-, not reported; PF pressure, patellofemoral contact pressure; Medial PF pressure, medial patellofemoral contact pressure; Lateral PF pressure, lateral patellofemoral contact pressure. Patellofemoral contact pressure (mean value, Mpa) of the intact knees -, not reported; PF pressure, patellofemoral contact pressure; Medial PF pressure, medial patellofemoral contact pressure; Lateral PF pressure, lateral patellofemoral contact pressure. Effect of different graft types on PF contact pressure Eight studies selected the same autograft (gracilis tendon) for MPFL reconstruction, and only one study [33] chose artificial material (polyester suture tape) as a graft. Five included studies [27, 30, 31, 34, 39] selected at least two different graft types for comparison, the study [34] selected three different types of grafts: double-stranded gracilis tendon, quadriceps tendon, and tensor fascia lata allograft to replace the intact MPFL. The final conclusion showed that when the femoral tunnel was positioned at the anatomical site, the PF contact pressures were restored to their original state, regardless of the use of gracilis tendon, quadriceps tendon, and tensor fascia lata allograft reconstruction (P>.050). Another study [31] used two different types of grafts: gracilis tendon and quadriceps tendon, and found that the two surgical procedures for reconstructing the MPFL produced PF contact pressures that were identical to those seen in the native knee (P>.050). In study [27], on comparing fresh frozen gracilis tendon allografts for MPFL reconstruction and FiberTape for MPFL suture tape augmentation, the authors found no significant differences in the mean medial PF contact pressure, peak medial PF contact pressure, medial contact area, or mean lateral PF contact pressure at any knee flexion angle. In the adductor tendon group of the other study [39], the distal adductor tendon was left attached to the adductor tubercle and the free proximal end was pulled across the patella and secured with a screw at force of 2 N. Eventually, the authors discovered that the adductor transfer became tighter in flexion, resulting in potential graft stretching and failure. In study [30], the author found that the PF pressure distributions were closely restored when the MPFL was anatomically reconstructed using either a gracilis graft or fascia lata graft. Effect of different patellar /femoral tunnel positions on PF contact pressure With regards to patellar tunnel positions, almost all studies selected the proximal two–thirds of the medial patellar margin as the tunnel location, except three included studies [31, 34, 37], which selected the medial part of the quadriceps tendon as a graft, and it retained the native patellar insertion of the quadriceps tendon. But the femoral tunnels were quite variable; seven included studies [27–30, 32, 38, 39] selected the Schöttle point [40]as the femoral tunnel location and three included studies [34, 35, 39] selected different femoral tunnel positions for comparison. In the study [34], femoral tunnels positioned 10 mm proximal or distal to the anatomic origin (previously marked anatomic center of the MPFL attachment) resulted in significant increases in the mean medial PF contact pressure, medial patellar tilt, and medial patellar translation during knee flexion or extension (P< .050). Similarly in study [35], the femoral tunnels at 5 mm proximal or distal to the intact MPFL insertion increased the peak and mean medial PF contact pressures and medial patellar tilt (P< .050). In another included study [39], when the femoral insertion of the MPFL was located on the adductor tubercle, higher medial PF pressures (P = .001) at 100° of flexion were observed. Effect of different pre-tension levels applied to the femoral side of the MPFL on PF contact pressure Thirteen included studies [27–39] mentioned the level of MPFL graft tightening, and the majority of them used pre-tension of 2 N on the femoral side. Four included studies [27, 34–36] tested the PF pressures at different levels of tightening. In study [27], the author tested the PF pressure at 2 N, 5 N, and 10 N pre-tension levels. When the pre-tension was 2 N, the PF pressure and the degree of patellar tilt were closest to the knee in the intact MPFL state; and when the level of pre-tension was increased, the medial PF pressure could gradually increase and the patella could gradually tilt to the medial side. Another comparative study [34] tested the PF pressure at 2 N and 10 N pre-tension with three different graft types. There was a trend that with the use of all graft types, the mean medial PF pressure (P< .001) and peak medial PF pressure (P = .002) could rise, the lateral PF pressure (P = .001) could reduce, and patellar tilt (P = .009) could occur when 10 N of graft tension was applied. In the other included study [35], the PF pressure at 2 N, 10 N, and 30 N pre-tension levels was the test. A graft tensioned to 2 N, anatomically positioned MPFL reconstruction restored the intact medial and lateral PF pressures and patellar tracking (P>.050); and grafts tensioned to 10 N or 30 N caused significant increases in the medial PF pressure (P< .050) and patellar tilt (P< .050). The study [36] measured PF contact pressures at 3 levels of tension applied to the graft (2 N, 10 N, and 40 N). The final outcome revealed that the application of 2 N and 10 N to the MPFL did not significantly affect the mean medial PF pressures in comparison with the intact state (P< .05); however, at 40 N of tension, the mean medial PF pressure was increased significantly and lateral patellar translation was decreased (P< .05). Effect of different knee flexion angles at the time of graft fixation on the PF contact pressure. The optimal knee flexion angle for graft fixation remains a topic of discussion, as findings across studies are not entirely consistent. While the majority of studies included in this review suggest that fixation at 30° or 60° of flexion effectively restores native PF contact pressures [27, 34, 36, 30–32, 38, 39], some studies reported that a wider range of angles (e.g., 60° and 90° [33] or a specific double-bundle combination [29]) could also achieve satisfactory results. This variability is likely influenced by the choice of graft material.For instance, synthetic grafts like polyester suture tape are less elastic and may tolerate higher flexion angles for fixation without over-constraining the joint [33], whereas autografts might perform better within a narrower, more anatomical range [28, 35]. Therefore, it is challenging to prescribe a single optimal angle for all scenarios.Based on the aggregated biomechanical evidence, fixation between 30° and 60° of knee flexion appears to be a safe and effective range that minimizes the risk of elevating medial PF pressure for most common graft types. Surgeons should consider the specific properties of the graft being used when deciding on the fixation angle. Discussion There is almost no way to measure the PF pressure inside the living body. Summarising and reviewing relevant cadaveric biomechanical studies could provide ideas on how to restore the original PF pressure during MPFL reconstruction by controlling intra-operative factors.The imperative to restore native biomechanics is underscored by the high prevalence of chondral injuries found in patients with recurrent patellar dislocation, highlighting the long-term consequences of abnormal patellofemoral joint mechanics [41].Two important findings have emerged from this systematic review. First, changes in PF contact pressure during flexion and extension of the intact knee were clarified, which provided the basis for subsequent studies. Secondly, under cadaveric conditions, we investigated the effect of intra-operative factors (graft types, tunnel locations, pre-tension levels, and graft fixation angles) of MPFL reconstruction on PF contact pressure and would like to provide some suggestions on how to restore the native PF contact pressure. Effect of different graft types on PF contact pressure Currently, the choice of the graft for the treatment of patellar dislocation is debatable, and a variety of graft types have been used [42–44]. Most of the included studies reported that the PF contact pressures were restored to their original state, regardless of the use of different graft types. Only one study [39] derived a different conclusion: the adductor tendon group became tighter in flexion, perhaps leading to loss of motion, discomfort, graft stretching, and graft failure. The author provides the following explanation for this occurrence: the adductor tubercle is not the anatomical insertion of the MPFL on the femur. This implies that the graft type is not the key in MPFL reconstruction surgery and has little effect on the postoperative PF contact pressure and motion, but other factors, such as the position of the femoral tunnel and the angle of fixation, can have a large impact. A variety of graft types have been used for MPFL reconstruction, with semitendinosus, gracilis, adductor, quadriceps tendon, and patellar tendon being the most prevalent. The ultimate load of the MPFL was 178 N in a study by Matthew D. LaPrade [45]; thus, the surgeon should use a graft with similar properties while reconstructing these ligaments. Noyes et al. [46] reported an ultimate load of 1216 N for the single loop semitendinosus tendon and a load of 838 N for the gracilis tendon, and they stated that the patellar and quadriceps tendons are stronger than the hamstring tendons [47]; hence, these tendons can be used as a replacement for the MPFL due to their strength. Many approaches have been explored, such as using polyester suture tape and FiberTape, which eliminate the need for tendon harvesting and reduce additional damage to the body, however there may be biocompatibility issues with this method. In certain research, the quadriceps tendon has garnered preference due to the simplicity of its surgical procedure and its ability to maintain the natural attachment to the patella, effectively emulating the native MPFL. Nonetheless, it is important to note that the quadriceps muscle tends to be less pliable. Consequently, individuals with shorter quadriceps muscles or those exhibiting increased Q angles should exercise caution and carefully weigh this option before proceeding. Additionally, fascia lata grafts have been used for their sheet-like structure, which closely imitates the MPFL’s anatomy and biomechanics. This variety of graft options allows for tailored reconstruction to meet the specific needs of the patient and the desired biomechanical outcomes. Therefore, other factors, such as the operator’s experience habits, the potential of the graft to reconstruct the MPFL, the patient’s own graft condition, the patient’s activity level, and his or her own needs and preferences, may be more important in determining which graft is most suitable for use as a material for reconstructing the MPFL. Effect of different patellar /femoral tunnel positions on PF contact pressure On comparing the studies included in this paper, it was found that the femoral tunnel location had a greater effect on PF pressure after MPFL reconstruction compared to the patellofemoral tunnel location, and the postoperative PF pressure was hardly changed if the femoral tunnel location was the Schöttle point. The native MPFL has previously been identified as being “close to isometric” [48, 49], and it makes a “circular motion” to maintain the normal patellar track and PF contact pressure during flexion and extension of the knee joint. The position of the patellar tunnel and femoral tunnel will directly determine the length of the MPFL, and thus, it will indirectly affect the PF contact pressure. MPFL length alterations varied by up to 2.5 mm on moving from the proximal to distal patellar attachments. Moving the MPFL’s femoral attachment 5 mm proximally or distally affects the MPFL length measurements, typically by 12 mm [50], indicating that the femoral tunnel position has a greater impact on the MPFL length than the patellar tunnel. Non-anatomical reconstruction leads to an important change in the length of the MPFL, and therefore, it alters patellofemoral kinematics and increases the PF contact pressure [51–53]. This explains the results obtained from the 14 studies included in our review: the femoral tunnel position, as compared to the position of the patellar tunnel, played a key role in MPFL post-reconstruction PF pressures and stability, and the femoral tunnel position should be closer to the intact insertion site; otherwise, the postoperative PF contact pressure would change significantly. Other published studies have also confirmed that reproduction of the anatomic femoral attachment has been shown to be an ideal position for better clinical outcomes [35, 54, 55]. A systematic review including 10 studies with 1287 knees found that femoral tunnel malposition (>10 mm from the Schöttle’s point) and more severe trochlear dysplasia (types C and D) appeared to have the maximum effect on inducing higher rates of recurrent dislocation [56]. Another study found that the most common indication for MPFL-R revision surgery was malposition of the femoral tunnel (38.1%) [57]. Therefore, compared to the patellar tunnel position, the femoral tunnel should be placed as close to the anatomic position as possible; otherwise, it will affect the postoperative PF contact pressure and have long-term consequences for joint health [51].This critical importance of anatomical femoral tunnel placement is strongly supported by clinical evidence, with systematic reviews identifying femoral tunnel malposition as a predominant technical risk factor for MPFL reconstruction failure and recurrent instability [58]. Effect of different pre-tension levels applied to the femoral side of the MPFL on PF contact pressure Compared to the graft types and patellar/femoral tunnels, pre-tension of the graft appears to be the most direct factor associated with the operative PF contact pressure. Four comparative studies included in this paper showed that when a pre-tension of 2 N or 10 N was applied to the graft, the postoperative PF contact pressure and patellar tilt were closest to those in the native knees; when the graft tension was further increased, the patella would inevitably " tilt " towards the medial facet, then the medial PF contact pressure would increase. Currently, the MPFL has been described as a “checkrein” by some researchers [59] because it slackens in early knee flexion and is almost isometric throughout the knee range of motion [53], implying that it should be gently tightened while it is being reconstructed. Therefore, a pre-tension of 2 N or 10 N to stabilize the patella in the trochlear groove is sufficient and overcorrection should be avoided. Sandro F et al. [60] revealed that a higher postoperative pain score was associated with less lateral patellar mobility in comparison to the unoperated knee. Therefore, sufficient lateral patellofemoral joint play should be retained following MPFL reconstruction. Sufficient lateralization might also allow for compensation for a malpositioned femoral tunnel, and therefore be more forgiving. In cadaveric biomechanical experiments, precise control of pretension can be achieved with pulley sets. However, usually, there is no good way to quantify pretension in actual surgery, and the presence or absence of lateral release can also have an effect on pretension. Therefore, it is no longer advisable to perform undertensioning of MPFL grafts while fixing them onto the femur. Effect of different knee flexion angles at the time of graft fixation on PF contact pressure Currently, knee flexion angles at the time of MPFL fixation are controversial, and most scholars fixed the graft in 0°–90° of knee flexion [61, 62]. Thus, an objective recommendation of the optimal knee flexion angle at final graft fixation on the femoral side may be important to restore the intact PF pressure and knee kinematics. First of all, for different MPFL graft types and graft properties, the fixation angles will be different. Similar to the results obtained in this paper, several previous studies have also reported the optimal fixation angle during MPFL reconstruction using different grafts. Deie et al. [63] reported that the optimal fixation angle was 30° of knee flexion when using a semitendinosus tendon graft; Kita et al. [64] reported that the optimal fixation angle was 45° when using a double-looped semitendinosus tendon; Nomura et al. [65] recommended fixation at 60° when using a polyester tape. The properties of the graft determine the fixation angles; for example, artificial materials are more rigid than autografts [66, 67]. When an artificial material, such as polyester tape, is fixed at a low knee flexion angle due to its high strength and poor flexibility, the ligaments cannot lengthen when the knee is moved to a high flexion angle, resulting in an increase in PF contact pressure. On the contrary, autograft tendons fixed at a low flexion angle of 30° or 60° are good enough for restoring the MPFL function: they show isometric behavior over the first 60° of flexion, followed by shortening of the ligament at high flexion angles [66, 68, 69]; thus restoring the native PF contact pressure. Secondly, the anatomy and biomechanics of the MPFL also affect the fixation angles. The MPFL provides a restraint rate of approximately 23% to 80% against patellar transfer [70, 71], and it plays a major role during knee flexion from 0° to 30° [72, 73]. Once the patella slips into the trochlear groove, the medial and lateral trochlear facets play their limiting role [74, 75]. That is why most surgeons fix the MPFL at 30° of flexion of the knee. Another explanation is that because the patella is not centered in the trochlear groove at full extension and it is difficult to estimate the length of the MPFL, fixation of the graft may be better performed during knee flexion between 30° to 60° when the patella is stabilized in the trochlea. Several other studies [62, 76, 77] have recommended tensioning the graft at 60° of flexion because fixation at lower flexion angles carries a higher risk of over-tensioning the graft as the patella is much more mobile. In general, fixation of grafts at 30° or 60° provides good restoration of the original PF pressure; moreover, the fixation angle should be adjusted depending on the type of graft. Limitations Several limitations of this systematic review are worth noting. There exist inherent differences between cadaveric models and living patients. Cadaveric models are unable to fully replicate the physiological and biomechanical environment of a living patient, which includes aspects like muscle contraction and relaxation, soft tissue elasticity, and the lubrication provided by joint fluids. All of these factors may affect the contact pressures within the PF joint. Cadaveric models may not accurately predict pain perception, the recovery of joint function, and the long-term stability in patients post-surgical repair.Furthermore, individual differences among living patients, including anatomical variations such as patellar shape [78], muscle strength, activity levels, and body weight, may significantly influence surgical outcomes. Therefore, the effects of surgical procedures on the knee joint can be more fully and realistically simulated in the future by improving knee modeling approach. By employing parametric finite element analysis and conducting a global sensitivity assessment, we can more precisely simulate how surgical factors influence ligamentous integrity [79]. Additionally, incorporating the initial strain of the ligament into our models is crucial for a more accurate prediction of the ligament’s mechanical response under both physiological and pathological conditions [80]. Moreover, the impact of pressure on cartilage degradation can be elucidated through the development and application of a multiscale constitutive model specifically tailored for cartilage tissue [81]. These advancements will collectively contribute to improve the knee modeling approach. Additionally, employing advanced imaging techniques and conducting dynamic biomechanical analyses should be considered to enhance our understanding of how surgery affects the function of the patellofemoral joint.Besides, the reviewed cadaveric studies used healthy knees and the specimens were very old. Patellar dislocations, are known to occur most commonly during dynamic activities and in the 10- to 17-year age group [82]. Further, patella alta, femoral trochlear dysplasia, genu valgum, external deviation of the tibial tubercle, and femoral deformities can lead to patellar dislocation [83–85]. In particular, trochlear dysplasia, which can directly affect the patellar track, can lead to changes in the PF contact pressure [86–88]. Therefore, the reported data may contain age/development related bias. Finally, considering the heterogeneity of the studies included in this review with variations in surgeons, cadaveric conditions, surgical technique, graft types, graft pre-tensions, and fixation angles, the outcomes could not be compared using statistical methods; instead, we selected four major aspects to compare these studies. Conclusion In cadaveric biomechanical testing, MPFL reconstruction can effectively restore the native PF contact pressure irrespective of the graft type, but changes in the femoral tunnel location can severely affect the postoperative PF contact pressure. Fixation of the graft at 30° or 60° of knee flexion with 2 N or 10 N of pre-tension would be more conducive to restoration of the original PF pressure.