KneeDaily

9.2.1

Graft Selection for Anterior Cruciate Ligament Reconstruction: The Peroneus Longus Tendon

Yongtao Maov1.0Published 38 references
Evidence through 2026.10.04Chinese original
Contents

The peroneus longus tendon (PLT) is harvested proximal to the lateral malleolus and does not involve the knee's extensor apparatus or the knee flexor muscles. In trials comparing it with hamstring tendon, the mean diameter of a woven full-thickness PLT was 8.6–9.1 mm; a four-strand construct of the anterior half measured about 8 mm. Harvest time was shorter than for hamstring tendon (Saeed et al. reported 7.46 versus 10.28 minutes). Published clinical studies come mainly from Asia: of the 20 studies listed in Table 1 of Soleymanha et al. (2026), 16 were from Asian countries, including 11 from India.

Taken together, the current evidence indicates that the peroneus longus tendon should not be the routine first-choice graft for primary isolated ACL reconstruction; for patients preparing to return to jumping, cutting, or contact sports, the peroneus longus tendon should not be chosen whenever periarticular knee autograft is available.

Note: "Should not be first choice" reflects the current lack of evidence; it does not mean the graft cannot be used. As valid evidence accumulates, it may become possible for it to serve as a first choice.

Compared with hamstring tendon (HT), subjective knee function and clinical laxity after PLT reconstruction are similar from 18 months to 5 years, but graft failure—the most important outcome—has accumulated only a small number of events. The pooled odds ratio is 0.96, with a 95% confidence interval of 0.43–2.15; the ends of that interval correspond respectively to substantially fewer and substantially more failures with PLT. Direct comparison with bone–patellar tendon–bone (BPTB) consists of only one randomized trial with 50 patients per group; comparison with quadriceps tendon (QT) consists of only one 36-patient quasi-experimental study. Neither is adequate to compare failure rates.

The cost of harvest is shifted from the knee to the ankle and foot. Foot and ankle scores are generally near the ceiling, yet instrumented dynamometry has repeatedly measured reduced eversion strength; at one year postoperatively, compared with a semitendinosus group, side-to-side differences in ankle plantarflexion strength, eversion strength, and horizontal single-leg hop were clearly larger.

The appropriate use of the peroneus longus tendon is as an alternative when periarticular knee autograft is unavailable, insufficient, or when multiple grafts are needed; adult patients with low activity demands who wish to preserve periarticular knee tendons may also choose it after understanding the ankle cost.

1. Knee Outcomes: Scores Similar to Hamstring Tendon; Failure Risk Not Yet Reliably Compared

Randomized Trials Comparing with Hamstring Tendon

Almost all available comparative data use HT (four-strand semitendinosus–gracilis or semitendinosus alone) as the control; the principal trials are shown in the table below.

Study Design and population Follow-up Principal knee outcomes (PLT/HT) Rerupture or author-reported failure (PLT/HT)
Bi et al., 2018 Anterior-half PLT versus semitendinosus, all-inside reconstruction; randomized 124 patients (62 each), all completed follow-up Mean 30 months IKDC 89.3/90.4; KT-1000 side-to-side difference 1.85/1.71 mm 0/62 and 1/62
Saeed et al., 2023 Doubled full-thickness PLT versus four-strand HT; randomized 232 (PLT 121, HT 111), 158 analyzable (PLT 85, HT 73) 24 months IKDC 94.66/94.54 (score analysis PLT 83, HT 69) Rerupture 2/85 and 4/73, P=0.305
Agarwal et al., 2023 Original text calls it a prospective cohort with patients randomly assigned; PLT 98, HT 96 1 year IKDC 94.13/95.12 Conclusions state no rerupture; 1 PLT patient with clinical instability after reinjury (anterior drawer grade III, pivot shift positive), not confirmed by imaging or arthroscopy
Asif et al., 2024 Called randomized; 60 per group Results section 12 months; Abstract and Discussion state 24 months IKDC 94.34/90.9, original reports a difference (Results) No counts found in the original
Butt et al., 2025 Randomized 64 (PLT 31, HT 33), 60 completed 5 years; 59 men, low activity demand 5 years IKDC 89.9/89.7 None in either group
Khalil et al., 2025 Randomized 76 (38 each); 5 reruptures excluded, 71 analyzed 18 months IKDC 92.78/92.31 2/38 and 3/38
Malik et al., 2026 Randomized 50 (25 each) 6 months IKDC 82.6/78.9, P=0.04 0/25 and 1/25 (author-reported graft failure, undefined)

Trials with follow-up of at least 18 months found no between-group differences at final follow-up in IKDC, Lysholm score, or clinical laxity testing. Early results differed: Malik et al. used 6 months as final follow-up and reported higher IKDC and Lysholm in the PLT group (P=0.04 and 0.03); Saeed et al. also reported higher scores in the PLT group at 6 months, with the difference gone by 24 months; Asif et al. reported higher PLT scores at 12 months in the Results, but that paper's follow-up duration is internally inconsistent (see below). These trials were powered on subjective scores, failure events were few, and how several trials handled failure cases directly affects their conclusions.

Saeed et al. (2023) is the largest. It was a single-center, single-surgeon Pakistani trial that assigned grafts by computer-generated random numbers; patients knew their allocation. Of 232 patients, 158 were finally analyzable; 74 (31.9%) did not enter analysis. The Discussion states that 69 of these were contacted only by telephone, reported no symptoms, and had no in-person scoring or examination. The trial was not prospectively registered; the authors also state that they did not measure KT-1000, knee flexion–extension strength, or ankle inversion–eversion strength. There was no external funding; all items on the ICMJE conflict-of-interest forms of the six authors were marked "None."

Twenty-four-month IKDC and Lysholm were analyzed for only 83 PLT and 69 HT patients—2 and 4 fewer, respectively, than the analyzable numbers in each group, exactly the number of reruptures in each. The original does not state how the score analysis handled failure cases. Among 135 athletes, time to return to sport was 201.32±17.71 days in the PLT group and 235.19±23.82 days in the HT group, a difference of about 34 days. By the authors' definition (harvest-site pain, weakness, or incision dysesthesia persisting to 24 months), donor-site complications were 27/73 in the HT group and 9/85 in the PLT group; AOFAS ankle–hindfoot score in the PLT group fell from 94.11 preoperatively to 93.61 (P=0.004).

The failure-rate data from this trial cannot support equivalence of risk between the two grafts. Two versus four events cannot distinguish difference from equivalence; nearly one third of patients did not enter analysis, and whether any of them failed cannot be answered from the original; the high 24-month scores also exclude the rerupture patients.

The 34-day earlier return to sport must likewise be understood under the study protocol: it is the time athletes actually returned to sport; the rehabilitation protocol separately specified assessment with a single-leg hop test at 6 months for return to noncontact sport, with contact sport deferred to 8 months; both surgeon and patient knew the allocation; graft rupture was reported only as a total count, without analysis by return time or by athletic exposure after return. Earlier return to sport does not equal safer return. What can be used is that, among patients who entered the score analysis, 24-month knee scores in the PLT group were similar to those in the HT group; periarticular knee donor-site symptoms were fewer in the PLT group, and harvest was faster.

Khalil et al. (2025) randomly allocated 76 patients; the 5 graft reruptures that occurred within 18 months (PLT 2, HT 3, all during contact sport 9–14 months postoperatively) were excluded from the study results, and function was finally reported for 71 patients, on which basis the authors concluded that PLT is a "safe and effective" graft. Graft rerupture is a treatment outcome of reconstruction; IKDC after deleting these patients describes only those who did not fail and cannot represent the full randomized cohort. The 24-month scores of Saeed et al. have the same problem, except that that paper retained the follow-up counts of the rerupture patients. The reliable data left by this trial are the rerupture counts calculated on the randomized denominators—PLT 2/38, HT 3/38—and the 18-month function of those who completed follow-up.

Butt et al. (2025) is the randomized trial with the longest follow-up. Sixty-four patients were randomized and 60 completed 5-year follow-up; all but one were male. Sample size was calculated from an IKDC minimal clinically important difference of 11.5 points, requiring 15 per group. Preinjury sport participation was 8/30 in the HT group and 7/30 in the PLT group; apart from one PLT patient who was a national-level athlete, all were recreational athletes. Five-year IKDC scores were similar, and neither group had a rerupture. For the zero reruptures, the authors speculated (the original uses "assumed") that low activity demand and few sport participants were responsible; the study did not record athletic exposure, this explanation is unverified, and zero events cannot show that failure risk is the same for the two grafts. The authors also state explicitly in the Discussion that the study cannot evaluate graft failure and rerupture risk, that the conclusions cannot be applied to more active populations, and that no ankle strength or functional testing was performed. What this trial supports is that in adult men with low activity demand, 5-year knee function after PLT is similar to that after HT.

Bi et al. (2018) compared all-inside reconstruction with anterior-half PLT versus semitendinosus; all 124 patients completed at least 24 months of follow-up. Diameter after four-strand weaving was 8.0±0.7 mm for anterior-half PLT and 8.1±0.9 mm for semitendinosus. The authors' hypothesis was equivalence, yet the statistics used only ordinary between-group tests without a prespecified equivalence margin; P>0.05 cannot prove equivalence. The only rerupture occurred in the semitendinosus group after high-energy trauma at 27 months postoperatively; the body of the text calls the patient male and a table footnote calls the patient female. Results for the anterior half cannot be applied directly to full-thickness tendon; see Section 3.

Among the 194 patients of Agarwal et al. (2023), one in the PLT group was reinjured in a traffic accident more than 6 months postoperatively and at 1 year had an anterior drawer test of grade III and a positive pivot shift, with ongoing instability in daily life. The Discussion attributes this patient's positive pivot shift to the reinjury and also writes that there was "no rerupture during follow-up"; the full text does not report it as a failure. The original has no MRI or arthroscopic data, so whether the graft ruptured cannot be confirmed, but this is a clear clinical failure. Analyses that count only rerupture will miss such patients; whether it should enter a given meta-analysis depends on that analysis's prespecified failure definition.

The follow-up duration in Asif et al. (2024) is internally inconsistent. The Abstract states that 120 patients were followed for 24 months and that IKDC and Lysholm differed significantly between groups; the Results section writes final follow-up as 12 months and reports IKDC 94.34±1.75 in the PLT group and 90.9±7.56 in the HT group; the Discussion then writes that the groups were similar at 24 months (P>0.05). The Conclusions recommend PLT as "superior" to HT. Passages in the Discussion are nearly verbatim with Saeed et al. (2023), including the same harvest time of 7.46 minutes, 2.8 minutes shorter than HT; it cannot be determined whether these figures were measured in this trial. Neither the claimed 24-month superiority nor the return-to-sport advantage from this paper can be used.

The title of Rhatomy et al. (2019, Knee Surg Sports Traumatol Arthrosc) is that the peroneus longus tendon "can be recommended as a superior graft to hamstring tendon." Allocation was by surgical period (HT from October 2015 to June 2016, PLT from July 2016 to February 2017), without randomization: HT 28, PLT 24, follow-up 1 year. Score improvement did not differ significantly between groups; PLT diameter was 8.8 mm and HT 8.2 mm (P=0.012). The authors state that knee stability was not assessed, and the full text has no failure counts.

The donor-site data in that paper contain internal errors: the body states that thigh circumference decreased by 20 mm in 1 PLT patient and by 10 mm in 4, yet Table 7 gives a group mean for all 24 of 2.5±0.5 mm; if the body figures are correct, the SD would be at least about 5.3 mm. In Table 6, the 19 and 5 patients with reductions of <10 mm and 10–20 mm are labeled 68% and 32%; for 24 patients these should be 79.2% and 20.8%. The "superior" in the title is not supported by this study's results, and the specific figures for milder thigh atrophy cannot be used.

What this set of trials supports is that in predominantly male adults with low-to-moderate activity demand, subjective knee function and clinical laxity after PLT reconstruction from 18 months to 5 years are similar to those after HT; the score advantage at 6 months was not maintained at longer follow-up. Neither "PLT failure rate is the same as HT" nor "PLT is superior to HT" is supported by these trials.

Meta-analyses of Failure Rates

Kumar et al. (2025) pooled 10 comparative studies, 949 patients; only 4 of them, with 534 patients (PLT 261, HT 273), reported failure. The authors' pooled graft failure rates were originally reported as 4.2% and 4.3%, OR 0.96 (95% CI 0.43–2.15), mean follow-up 19.5 months; Figure 6 includes the clinically unstable case from Agarwal et al. above that was not imaging-confirmed. At event rates around 4%, the odds ratio is close to the risk ratio; this interval roughly means that PLT failure risk could range from less than half that of HT to more than twice as high. Similar percentages and a nonsignificant P value show only that a difference was not detected. In the Discussion the authors interpret the absence of difference in stability testing as functional noninferiority of PLT, and further state that although PLT has not been compared directly with BPTB, its failure profile is "closer to BPTB." The paper set no noninferiority margin and has no BPTB data; both inferences lack a basis. Its pooled estimate and interval can still be cited.

Park et al. (2024) included only randomized trials (6 trials, 683 patients). The study-characteristics table entered the Saeed trial by the randomized numbers 232 (121/111) and did not present the fact that 31.9% did not enter analysis; missing-outcome bias was still rated low risk.

Desouza et al. (2026) searched through December 2025 and included 4 randomized trials, 407 patients. The Results state that graft diameter was larger in the PLT group (8.4 mm) than in the HT group (8.13 mm) and that the "difference was statistically significant," yet the same sentence gives MD 0.28 mm, 95% CI −0.44 to 1.01, P=0.45 (p. 7, Fig. 8); the Discussion again calls it "significantly thicker." This conclusion contradicts the paper's own statistics and cannot serve as evidence of a diameter advantage.

In the AOFAS pool, the Shair and Vijay studies are marked as not estimable with weight zero; the only studies actually contributing weight are the Bi and Saeed trials, 147 patients. The Abstract and Conclusions state that failure rates were similar between groups, but the Results and the 10 figures contain no corresponding event counts, denominators, or effect sizes that a reader can verify. The donor-site complication OR of 0.12 depends on the Vijay trial's HT count of 12/22; the Results of the publisher's full text of the Vijay trial record 3 infections and 6 kneeling pain cases in the HT group, with no count corresponding to 12 and no statement whether the two symptoms occurred in the same patients; this OR cannot for now be treated as a verified effect size. Of 143 reports requiring full-text retrieval at the screening stage, 92 could not be obtained (pp. 3–4).

Comparisons with Bone–Patellar Tendon–Bone and Quadriceps Tendon

The American Academy of Orthopaedic Surgeons (AAOS) 2022 ACL guideline recommendation on autograft source is that when reconstructing with autograft in skeletally mature patients, the surgeon may prefer BPTB to reduce the risk of graft failure or infection, or prefer HT to reduce anterior knee pain and kneeling pain (high-quality evidence; moderate-strength recommendation). The guideline regards graft rerupture as potentially the most important outcome, especially in young patients returning to high-level sport; QT is listed as a newer option needing further study; the full text does not address PLT. Absence of a PLT recommendation in the guideline cannot be read as opposition to PLT.

Direct comparison of PLT with BPTB comes from the Russian single-center trial of Goncharov et al. (2023). One hundred eight patients were randomized with sealed envelopes; the Methods state 53 in the PLT group and 55 in the BPTB group; enrollees exercised at least 3 times per week. Three and five patients, respectively, were lost to follow-up, leaving 50 in each group who completed 2-year follow-up. Within 2 years, graft rupture was 4/50 (8%) in the PLT group and 3/50 (6%) in the BPTB group, P=0.69; rupture was diagnosed from history, examination, KT-1000 anterior tibial translation >5 mm, and MRI. Lysholm, IKDC, and KT-1000 were similar between groups. Pain or discomfort in the knee during sport was 6/50 in the PLT group and 14/50 in the BPTB group (original P=0.4). Two patients in the PLT group sustained intraoperative sural nerve injury and still had sensory loss at 2 years.

The Abstract of that paper swaps the group names, writing 53 in the BPTB group and 55 in the PLT group, so that the 4 ruptures and the AOFAS results of the PLT group are listed under the BPTB heading; reading only the Abstract misattributes the failure counts. In the Limitations the authors acknowledge that age, sex, and fixation differed between groups (mean age 33.4 years in the PLT group, 27.9 years in the BPTB group). With 50 per group and 7 failures in total, the study is insufficient for a reliable comparison of failure rates; the direction of the numbers is unfavorable to PLT, but PLT cannot on that account be judged to fail more often.

Between PLT and QT there is only the 36-patient study of Islam et al. (2025): nonrandomized, allocated by odd/even admission number, 18 per group, follow-up 1 year, knee scores similar between groups; likewise unable to answer the failure-rate question.

For young patients preparing to return to high-intensity sport, what needs comparison with PLT is BPTB and QT. On high-quality evidence the AAOS guideline holds that BPTB has fewer graft failures than HT; PLT versus HT has not shown equal failure rates, and direct comparison with BPTB is only the 100 patients above. In the author's view, given this evidence, PLT should not replace BPTB, which has failure-risk evidence behind it. This is a clinical judgment from indirect material; it does not mean that PLT failure rate has been shown to exceed that of BPTB.

Graft Diameter

In most trials PLT was thicker than four-strand HT: Saeed et al. 8.81 versus 8.17 mm, Khalil et al. 8.64 versus 7.63 mm, Asif et al. 9.13 versus 8.26 mm, Rhatomy et al. 8.8 versus 8.2 mm; the mean difference pooled by Kumar et al. was 0.46 mm (P=0.03). Four-strand anterior-half PLT was comparable to semitendinosus (Bi et al. 8.0 versus 8.1 mm). Several PLT papers cite the association between diameter and revision risk in prior ACL reconstruction cohorts to support PLT (for example, Khalil et al. citing the Swedish National Knee Ligament Register), yet Khalil et al.'s own trial found no reinjury difference despite larger diameter. There is no direct evidence on whether a PLT graft taken from another tendon follows the same diameter–failure relationship. Larger diameter can avoid using an undersized HT graft; that is a technical advantage of PLT and cannot serve as substitute evidence of a lower failure rate.

Single-Arm Cohort Combining Lateral Extra-Articular Tenodesis

Arora et al. (2026) reported 500 consecutive patients, all with complete ACL injury and pivot shift grade 2 or 3, who underwent three-strand PLT reconstruction combined with modified Lemaire lateral extra-articular tenodesis (LET). Four hundred eighty-two (96.4%) completed 2-year follow-up; 354 (70%) participated in competitive or semicompetitive sport, of whom 68% played contact sport.

Within 2 years, 18 patients developed subjective or examination laxity, of whom 6 had MRI-confirmed complete graft rupture (6/500, 1.2%); after 2 years there was no active follow-up, and among 22 who presented for new injury a further 3 ipsilateral reruptures were found (1 at 2.5 years, 2 at 3 years). The "rerupture rate of 1.8%" in the Abstract and Conclusions is these 9 cases divided by 500, combining systematic 2-year follow-up with passive presentation after 2 years; it cannot be written as a 2-year failure rate, nor treated as a 3-year failure rate for the whole cohort.

Other reporting in that paper is also problematic. The proportions returning to sport (92%) and returning to the preinjury level (81%) give no numerators or denominators. In Table 2 the mean differences for Lysholm score fall outside their own 95% CIs (for example, 1-year mean difference −12.456, 95% CI −17.680 to −15.223); these intervals and P values cannot be used. One place in the original calls them "500 consecutive kabaddi patients," another lists kabaddi, football, wrestling, and noncontact sports; the population description is inconsistent. Harvest-incision infections numbered 4 superficial and 2 deep.

These data provide one favorable observation: in a population with marked pivot shift and predominantly contact athletes, complete graft rupture within 2 years after PLT combined with LET was 6/500. The study had no control group; the contribution of LET cannot be separated from that of PLT, so this result cannot be attributed to PLT itself and cannot be compared with other grafts.

2. Donor-Site Ankle and Foot Function: Scores Near Ceiling; Objective Tests Show Reduced Eversion and Push-Off

The peroneus longus is one of the principal evertors of the foot; its tendon inserts on the plantar surface at the base of the first metatarsal and the medial cuneiform and contributes to plantarflexion of the first ray. What harvest may affect is eversion strength, first-ray plantarflexion, push-off, and dynamic lateral stability; these functions require instrumented dynamometry, jumping, and balance testing to evaluate.

Limitations of AOFAS and FADI Scores

The AOFAS ankle–hindfoot score is a clinician-based score published in 1994 that mixes pain, function, and alignment. The American Orthopaedic Foot & Ankle Society's 2022 position statement states explicitly that, because of inadequate reliability and validity, it does not endorse this scoring system and does not encourage its use as the sole evaluation tool. AOFAS also responds to knee status: in Agarwal et al. (2023) the HT group had no ankle surgery, yet AOFAS rose from 96.0 preoperatively to 99.80 postoperatively, changing in parallel with the PLT group (96.2 to 99.05), which suggests that the portion of the preoperative score affected by the knee injury recovers as knee function recovers; change in AOFAS is difficult to attribute to the ankle alone.

The FADI main scale has 26 items and a maximum of 104 points; FADI Sport has 8 items and a maximum of 32; each is converted to a percentage, with 100% indicating no disability. The main scale mainly asks about daily activities; the Sport subscale targets more demanding athletic tasks and is more sensitive in high-function populations. Hale and Hertel (2005) validated it in 50 recreational athletes, 31 of whom had chronic ankle instability. Most PLT studies report only AOFAS or the FADI main scale.

Ceiling effects of the scores are clearest in Masoud et al. (2026): in 42 patients the AOFAS median was 100 (interquartile range 100–100) preoperatively and at 3 and 6 months postoperatively, while over the same period the median number of single-leg heel rises fell from 9.5 to 6.5 per minute. A high AOFAS or FADI mean can show that those assessed have good daily foot and ankle function; it cannot exclude declines in strength and athletic performance.

Data Errors in the Soleymanha et al. (2026) Meta-analysis

The Soleymanha et al. meta-analysis is titled to the effect that foot and ankle function is good after harvest. It searched through 5 March 2025, included 20 studies and 1024 patients, and reported pooled AOFAS of 96.0, 96.1, and 97.2 at 6, 12, and 24 months and 24-month FADI of 99.4. The Conclusions hold that donor-site morbidity after PLT harvest is mild, that foot and ankle scores are consistently high across age groups, and that this supports the safety of the PLT graft; the Discussion further states that this result may allay concerns about donor-site morbidity (p. 11). Two verified errors affect the pooled data in this paper.

Mismatch of de Oliveira numbers and subgroups. de Oliveira et al. (2024) initially enrolled 147 patients, of whom 75 had full-thickness and 72 anterior-half harvest; 147 is the full set assessed at 6 months postoperatively, with only 102 and 63 assessed at 12 and 24 months. Scores were reported by harvest method; AOFAS of 98.72, 99.85, and 99.92 are the full-thickness group's means at the three time points (original p. 2 Methods and Fig. 1; p. 7 Table 2).

Figure 6 of Soleymanha et al. uses these three means at all three time points of 6, 12, and 24 months, yet pairs each with a denominator of 147; Figure 7 likewise enters FADI with 147. The full-thickness subgroup means are thus paired with the whole-cohort denominator that includes the anterior-half group. The 147 at 12 and 24 months also exceeds the 102 and 63 actually assessed at those times. This error enters the three AOFAS pools of Figure 6 and the two FADI pools of Figure 7; in the 24-month FADI pool (Fig. 7b), de Oliveira carries 99.3% of the weight under a fixed-effect model and 34.7% under a random-effects model.

The original study is also not a pure ACL-reconstruction cohort and does not match this meta-analysis's inclusion criterion of "ACL reconstruction with PLT as the sole autograft." The original table lists 126 isolated ACL reconstructions, 6 ACL with medial collateral ligament reconstruction, 3 with posterolateral corner reconstruction, and 3 each of isolated posterolateral corner reconstruction, medial patellofemoral ligament reconstruction, and quadriceps repair; these procedures sum to 144, which does not match 147.

Single-arm studies counted as randomized trials. The Abstract and Results state that 9 randomized clinical trials were included and evaluated with RoB 2. Joshi et al. (2021) is explicitly a prospective interventional study in its Methods: all 48 nonathletes received PLT reconstruction, with no control group and no random allocation. That meta-analysis writes the bias concern for the Joshi study as "inadequate reporting of the randomization process" (p. 6)—a misclassification of study design.

In addition, the time-point subgroup analysis states that it "involved 1235 patients" (p. 8), more than the 1024 included overall. The 1235 is the cumulative count of assessments across time points; the same cohorts appear repeatedly (de Oliveira's 147 appears three times) and are not independent patients. From this the original reports an overall mean of 96.4 across time points and tests of subgroup differences, without stating how correlation from repeated measures on the same patients was handled; this overall result cannot be verified. Separate pooling by time point is itself unproblematic. The "across age groups" in the Conclusions comes from meta-regression on study mean age; mean ages of the included studies ranged from 23.4 to 33.5 years. In the Limitations the authors acknowledge that AOFAS has a ceiling effect that makes subtle functional deficits hard to detect (p. 12).

Until the numbers and study classifications are corrected, the pooled estimates of this meta-analysis should not be used to prove that PLT harvest is safe; even after correction, AOFAS and FADI means cannot answer questions about strength and athletic performance. Scores from the primary studies can still describe the daily foot and ankle function of those assessed. The authors declare no conflicts of interest.

Ankle Eversion Strength

Of the 9 instrumented strength studies cited in this chapter, 5 measured lower eversion strength on the harvest side than on the contralateral side, preoperative baseline, or a semitendinosus control (Kotian, Shao, Mirza, Angthong, and Kehribar et al.), with magnitudes from a few percentage points to more than ten; 4 did not detect a significant decline (Rhatomy, Ertilav, Karapinar, and Masoud et al.).

Kotian et al. (2026) has the most complete design: 31 patients underwent isokinetic testing with a Humac NORM 1 day preoperatively and at 6 months and 1 year postoperatively. Peak eversion torque at 60°/s on the donor side fell from 16.81±4.75 Nm to 13.00±2.79 Nm and was 14.35±2.64 Nm at 1 year (overall P<0.001); on the healthy side the sequence was 16.71, 15.87, and 15.35 Nm (overall P=0.386). The Abstract uses P=0.386 (eversion) and P=0.685 (inversion) to state that there was no difference at 1 year; these two P values are in fact the overall comparisons across the three time points on the healthy side and cannot be used to show that the donor side had recovered (p. 4, Tables II and III). The Discussion states that the post hoc comparison of 1 year with preoperative was nonsignificant, but gives neither the test method nor the P value.

Calculated from the group means in the original tables, the donor side was 22.7% lower than preoperative at 6 months and still 14.6% lower at 1 year; the healthy side was also about 8% lower at 1 year than preoperative, and at 1 year the donor side was about 6.5% lower than the healthy side. What can be confirmed is that eversion strength declined clearly 6 months after harvest and partially recovered by 1 year; "complete recovery at 1 year" has no evidence. In the Conclusions the authors hold that stronger evidence is needed before PLT can be a first-choice graft, and that it can be used in selected cases such as multiligament injury or revision.

Among the 21 full-thickness harvests of Shao et al. (2020), 7 were ACL reconstructions, 13 medial patellofemoral ligament reconstructions, and 1 pectoralis major tendon transfer augmentation; mean follow-up was 31.8 months. Peak eversion force was 121.5±34.0 N on the donor side and 133.0±35.3 N on the contralateral side (P<0.001); the individual donor/contralateral ratio averaged 91.5%±7.6%. Whole-foot plantarflexion force and square hopping (completed by 18) showed no side-to-side difference; AOFAS was 98.7 preoperatively and 98.5 at final follow-up. On MRI every patient had tissue resembling tendon, but the reported 82.9% is the bilateral ratio of the combined cross-sectional area of the long and short peroneal tendons; the two tendons could not be distinguished at most levels and cannot be written as "PLT regeneration 82.9%"; this ratio also did not correlate with the eversion strength ratio. With scores unchanged and MRI showing tendon-like tissue, eversion strength can still be about 8% lower than the contralateral side two and a half years postoperatively.

Mirza et al. (2024) measured 20 patients with a handheld dynamometer; eversion force was lower than the healthy side at 1, 3, and 6 months postoperatively, and at 6 months was 7.665 versus 7.920 lbf (P=0.023), a difference of about 3%. Of 7 volunteers who underwent plantar pressure testing, 5 showed lateral shift of the point of maximum pressure at 3 or 6 months. The authors explain the pressure shift by first-ray plantarflexion weakness, but did not measure first-ray strength directly and did not measure ankle instability.

Of the 24 patients of Angthong et al. (2015), 10 volunteers underwent Biodex isokinetic testing at a mean of 7 months postoperatively (2.0–10.5 months); peak eversion and inversion torque at 60°/s and 120°/s were all lower than the healthy side (P=0.008, 0.001, 0.013, and <0.001). The original reports the differences as "% of N-m" without bilateral raw values or a calculation formula, so the specific magnitude cannot be determined. The study excluded patients with high ankle demands such as football and dance; the Abstract also writes "anterior drawer test normal in 83.3%" as an ankle result, whereas in the body this figure belongs to the knee; ankle anterior drawer testing was normal in all. On the basis of their donor-site results the authors do not recommend PLT as a first choice and consider it only when multiple grafts are needed for multiligament reconstruction.

Among studies that did not detect a significant decline, Rhatomy et al. (2019, Orthop J Sports Med) performed isometric testing at 6 months in 31 patients and found no side-to-side difference in eversion (65.87 versus 66.96 N, P=0.546) or first-ray plantarflexion (150.64 versus 152.12 N, P=0.680); a Methods subheading calls it isokinetic testing, but the measurement was isometric. Ertilav et al. (2024) followed 50 nonathletes for 4 years; isometric eversion force changed from 69.06 N preoperatively to 67.94 N, and whole-foot plantarflexion from 154.40 N to 153.26 N, with no significant difference versus preoperative or contralateral; ankle dorsiflexion, plantarflexion, eversion, and inversion range of motion all decreased slightly (for example, eversion from 18.52° to 16.84°).

In the 20 patients of Karapinar et al. (2025) at about 1 year, eversion force on the harvest versus contralateral side was 85.10 versus 94.40 N (P=0.096), and Flamingo single-leg balance time was 67.5 versus 86 seconds (P=0.049). That paper applied independent-samples tests to bilateral data from the same patients without treating them as paired, and the plantarflexion values in the body do not match Table 2; the raw observation that balance time was shorter on the donor side can be retained, but P=0.049 needs verification on paired data before it can be treated as a significant difference.

Rhatomy, Ertilav, and Karapinar et al. each had 20–50 patients, with differing populations, methods, and observation times; failure to detect a significant difference cannot prove that the harvest side has no strength loss. The eversion results of Masoud et al. are given below.

After full-thickness PLT harvest, multiple studies have measured reduced eversion strength. In longitudinal data, Mirza et al. from 1 to 6 months postoperatively and Kotian et al. from 6 months to 1 year observed shrinking of the deficit; Shao et al. still measured about an 8% deficit at two and a half years. Kehribar et al. measured both ankle strength and jump performance in the same patients (see next subsection), but the independent association between the eversion deficit and athletic performance, and the relationship of the deficit to ankle sprain or ACL reinjury risk, have not been analyzed in any study.

Ankle Strength and Single-Leg Hop Testing Compared with a Semitendinosus Group

Kehribar et al. (2026) measured both ankle strength and jump performance in a PLT group and a semitendinosus group. The data come from a retrospective series of a single senior Turkish surgeon from July 2023 to June 2025; age 18–35 years, isolated ACL reconstruction. Of 98 eligible patients (semitendinosus 55, PLT 43), 1:1 propensity-score matching on age, BMI, preoperative Tegner, IKDC, Lysholm, and AOFAS yielded 33 pairs; 3 in each group did not attend testing, leaving 30 per group tested at about 13 months postoperatively. Both groups used the same four-strand graft construct, suspensory fixation, and rehabilitation protocol.

IKDC, Lysholm, Tegner, and AOFAS showed no between-group differences. The group × side interactions for isometric ankle plantarflexion and eversion strength both had P<0.001; as shown in the figures, side-to-side differences were 14.91% and 12.97% in the PLT group and 1.91% and 2.00% in the semitendinosus group. Side-to-side differences on five horizontal single-leg hop tests were 14.06%–21.34% in the PLT group and 3.95%–7.85% in the semitendinosus group (interaction P=0.002 or <0.001). How the percentages in the figures were calculated is not stated in the original.

The semitendinosus group likewise underwent ACL reconstruction with the same rehabilitation protocol, yet ankle strength side-to-side differences were only about 2%. The recovery process common to both groups after reconstruction cannot readily explain the additional ankle asymmetry in the PLT group; this contrast supports attributing the extra loss to harvest. The study had no preoperative objective testing and did not record athletic exposure or the amount of rehabilitation actually completed, so the contribution of harvest cannot be fully separated from other between-group differences such as disuse; for distinguishing the effect of harvest, it remains the most direct comparison in the material of this chapter.

Statistical presentation of the remaining results in that paper needs correction.

The percentage differences in Table 2 use the operated side as the denominator. Taking single-leg drop-jump height as an example, the PLT operated side was 5.83 cm and the healthy side 10.68 cm; the 83.19% in the table is (10.68−5.83)÷5.83. The same difference with the healthy side as denominator is (10.68−5.83)÷10.68＝45.4%. The arithmetic of 83.19% is not wrong, but it cannot be written as an "83% decline in vertical jump." The percentage differences in the table also take absolute values: in the semitendinosus group RSImod on the operated side was slightly higher than on the healthy side, yet the table still lists a positive difference.

The Methods state that exploratory outcomes (the 16 jump metrics in Table 2 and a further 5 balance metrics) were Bonferroni-corrected, yet the Results declare significance at uncorrected P<0.05. Correcting the 16 jump metrics alone as one family changes the single-leg drop-jump interaction P from 0.009 to 0.144; correcting the 5 balance metrics as one family changes the anteroposterior sway interaction P from 0.013 to 0.065. The Discussion also calls RSImod significantly reduced, whereas its interaction P in Table 2 is 0.115. The title's claim of "impaired balance and jump performance" is only partly supported.

The study had no preoperative functional testing, did not record sport type, competitive level, or athletic exposure after return, and did not report reinjury, so it cannot be used to judge whether these asymmetries increase the risk of ankle sprain or ACL rerupture.

The between-group differences in ankle plantarflexion strength, eversion strength, and horizontal single-leg hop do not depend on the marginal P values above and can be used: in 18- to 35-year-olds undergoing isolated ACL reconstruction, at one year postoperatively asymmetry in ankle strength and horizontal single-leg hop was larger in the PLT group than in the semitendinosus group, whereas knee scores and AOFAS did not differ.

Heel-Rise Endurance, Gait, and Long-Term Foot Structure

Masoud et al. (2026) prospectively observed 51 eligible full-thickness harvest patients in Egypt (distal stump side-to-side sutured to the peroneus brevis tendon), aged 18–50 years, no professional athletes; 42 completed (9/51, 17.6%, did not enter analysis), all receiving a uniform ankle-specific rehabilitation program. Median single-leg heel rises fell from 9.5 per minute preoperatively to 7.0 at 3 months and 6.5 at 6 months (P<0.001); median first-ray plantarflexion force rose from 55.5 N to 73.6 N.

The title states that eversion strength increased, but the overall test for eversion force had P=0.125, and the paired comparison of preoperative with 6 months had P=0.021, above the authors' own Bonferroni threshold of 0.0167; an increase in eversion force has no statistical support. The increase in first-ray plantarflexion force lacks a control and occurred concurrently with the specific training, so how much is recovery from post-injury disuse, how much is training effect, and how much is compensation after harvest cannot be determined. In the Discussion the authors state that the single-leg heel-rise test mainly reflects calf endurance and neuromuscular control and is also affected by fatigue and postoperative disuse, so its decline cannot be equated with a diagnosis of ankle instability; calcaneal pitch decreased by a mean of about 0.7°, and the other three arch angles showed no change; the authors also state that this should not be taken as evidence of arch collapse. What this study can show is that in ordinary adult patients who completed specific rehabilitation, single-leg heel-rise performance at 6 months was below preoperative, while the AOFAS median remained at the ceiling.

The gait study of Kotian et al. (2026) matches their isokinetic study's 31 patients exactly in number, sex composition, age, follow-up time, and mean height and weight; the authors and institution are also the same, so the two reports should be treated as one cohort rather than a total of 62 patients. Single-leg stance time lengthened from 0.40 s preoperatively to 0.45 s at 6 months and 0.44 s at 1 year, but the original table labels stride length in cm while giving a value of 1.53, and cadence and step time both increase together; parameters within the table contradict one another, so precise gait effects cannot be used. That cohort added peroneal muscle training only after finding eversion weakness at 6 months.

In the 65-patient gait study of Zhang et al. (2024), support-phase inversion–eversion angles of 12.88° and 9.06° are incompatible with t=43.56 and P=0.017 in the same row of Table 7; multiple rows have this problem, and its inferential statistics cannot be used.

Long-term data come from Velioglu et al. (2025): of 64 full-thickness PLT reconstructions by the same surgeon from June 2000 to February 2007 (distal stump not sutured to the peroneus brevis), 26 returned for review and 38 (59.4%) were lost; of those reviewed, 25 were men and 1 a woman, follow-up 19.5 years (12–23 years). AOFAS was 98.8, FAAM activities of daily living subscale 99.6, and sports subscale 98.1; bilateral weight-bearing radiographs showed no differences in Meary angle, calcaneal pitch, first–second intermetatarsal angle, hallux valgus angle, or arch height index; 2 had sensory loss in the sural nerve distribution.

The study had no strength, jump, or gait testing; the status of the 59.4% lost to follow-up is unknown; comparison of the two feet of the same patient used independent-samples tests; failure to detect a difference also cannot prove structural equivalence. Among those who returned for review, static foot structure and self-reported function were good more than a decade after full-thickness PLT harvest; this observation can refute the claim that "full-thickness harvest necessarily leads to late arch collapse." A further retrospective study of 327 patients with mean follow-up of 42 months (Patel et al., 2026) reported no decline in AOFAS, FADI, or pain scores; patients were predominantly of low-to-moderate activity; the above data are from that paper's Abstract.

Nerve Injury

The harvest incision lies along the posterior border of the fibula proximal to the lateral malleolus, adjacent to the sural nerve. Sensory disturbance in the sural distribution is fairly common and mostly transient: Dogruoz et al. (2026) recorded 6/30 and 5/30 during follow-up (distal fixation versus no-fixation groups), Kotian et al. 2/31 (resolved within 3 months), Angthong et al. 2/24, and Velioglu et al. 2/26 at long-term follow-up.

Common peroneal nerve injury is uncommon but has been recorded. As the harvester is advanced proximally it approaches the common peroneal nerve near the fibular head; Khalil et al. stopped the harvester 4–5 cm distal to the fibular head to reduce this risk. Arora et al. (2024) prospectively observed 600 patients; Table 2 shows subjective nerve symptoms in 24 at 3 weeks and 4 at 6 months postoperatively (sural 3, common peroneal 1), with objective abnormalities falling to 0 by 6 months. The original writes 4/600 as 0.01% and "one in a thousand," and 24/600 at 3 weeks as 0.04%; the actual figures are 0.67% and 4%. The Discussion states that persistent symptoms occurred in only 1 case, contradicting the 4 in the table; the authors speculate that residual subjective symptoms were malingering, without a basis. The observation that objective nerve abnormalities resolved within 6 months in this study can be cited; the incidence figures in the original cannot.

Saeed et al. had 1 common peroneal nerve injury in the PLT group that recovered by 6 months. Yadav et al. (2022) reported a 25-year-old wrestler who on waking from anesthesia after PLT reconstruction had foot drop and sensory loss on the dorsum of the foot; exploration found intraneural hematoma and muscle contusion of the common peroneal nerve, with recovery 3 months after release. A case report proves that such injury can occur and cannot be used to estimate incidence; the statement in the Soleymanha et al. Abstract that "no serious nerve injury was seen" applies only to the studies they included.

3. Full-Thickness Harvest, Anterior-Half Harvest, and Distal Tenodesis

Almost all of the donor-site material in the preceding section comes from full-thickness harvest. Full-thickness tendon after weaving can reach 8.6–9.3 mm in diameter (9.32 mm in the Kotian et al. cohort); the anterior half needs four-strand weaving to reach about 8 mm (Bi et al.), or can serve as a supplement to another graft.

Material on anterior-half harvest comprises three studies. In the randomized trial of Bi et al. (2018), knee outcomes after all-inside reconstruction with anterior-half PLT were similar to those with semitendinosus; AOFAS changed from 99.4 preoperatively to 99.1 postoperatively; the authors reported no serious ankle instability, pain, or strength loss, but performed no instrumented dynamometry. That trial supports the feasibility of anterior-half reconstruction with good knee outcomes; it cannot serve as evidence ranking donor-site safety of anterior-half versus full-thickness harvest.

The 20 patients of Zhao et al. (2024) received supplementary anterior-half PLT because the woven diameter of the primary graft was ≤7 mm; at 6 months postoperatively, foot length, arch index, cadence, and step length showed no side-to-side difference, whereas the proportion of push-off phase on the harvest side (17.10% versus 18.34%, P<0.001), landing angle (11.06° versus 13.02°), and push-off angle (44.60° versus 49.80°) differed. The study excluded those with poor functional recovery at 6 months; the Abstract states that stance-phase differences were significant, whereas Table 4 has P=0.083; the Discussion writes a 4.2° difference in push-off angle, whereas the table shows 5.20°. The original calls the primary graft the "popliteal tendon," an irregular name from which its exact identity cannot be determined. Even with only the anterior half harvested, side-to-side differences in push-off-phase gait were still visible at 6 months.

de Oliveira et al. (2024) included both full-thickness (75) and anterior-half (72) harvest; the two methods were chosen by the surgeon as needed, without randomization. The Abstract states that score differences between groups at every time point were statistically significant (P=0.001); that P value in fact comes from the Friedman test of the three time points within each group in Table 2; between-group comparisons of Lysholm, IKDC, and AOFAS at each time point in Table 3 have P values of 0.070–0.739.

Dispersion values in the two tables also contradict each other. For the same anterior-half group's 6-month AOFAS, the SD is 1.87 in Table 2 and 2.87 in Table 3; for 12-month FADI, 0.01 in Table 2 and 0.31 in Table 3. At least one of each pair is wrong, so the reported dispersion is unreliable, and any result that uses these SDs for precise estimation, inverse-variance weighting, or verification cannot be used. Between-group comparisons in Table 3 used the Mann–Whitney test and do not depend directly on the tabulated SDs (p. 8, table note); the Lysholm, IKDC, and AOFAS comparisons at each time point did not reach significance, which shows only that a difference was not detected; FADI did differ between groups at 6 months (P=0.015).

Both groups' scores were near the ceiling; this qualitative observation can be retained. The study cannot prove equivalence of the two harvest methods and should not be treated as a precise safety estimate. The study population was predominantly football players (76.87%); at 12 months, 4/102 (3.92%) of ACL reconstruction patients had stability failure; the original does not state their graft type.

There is no direct objective comparison of strength or athletic performance between full-thickness and anterior-half tendon. Anterior-half harvest retains the posterior half and its continuous insertion toward the first metatarsal; by anatomical inference, loss of eversion and first-ray plantarflexion should be less than after full-thickness harvest; this inference has not yet been confirmed by measurement. The author's opinion is that when PLT is truly needed, if the required graft volume allows (supplementing a thin HT graft, or four-strand all-inside reconstruction), prefer the anterior half; take full thickness when larger volume is needed.

The role of suturing the distal stump to the peroneus brevis (distal tenodesis) is likewise unclear. In the 200 patients of Arora et al. (2025), patients chose whether to have tenodesis (tenodesis 128, no tenodesis 72; the tenodesis group was younger); at 1 year AOFAS and FADI showed no between-group differences apart from a trivial difference at 6 weeks. In the 60-patient retrospective comparison of Dogruoz et al. (2026), scores and weight-bearing radiographic arch angles showed no differences; strength was assessed only by manual testing, and the original states explicitly that there was no instrumented measurement. Neither nonrandomized comparison detected a score benefit from distal tenodesis, and neither performed an equivalence test; the effect of tenodesis on eversion strength, first-ray function, and athletic performance was not measured, so absence of functional benefit cannot be established. The long-term cases of Velioglu et al. had no distal tenodesis, and foot structure remained good.

4. Clinical Recommendations

Primary isolated ACL reconstruction: there is as yet no basis for recommending the peroneus longus tendon as the routine first choice. First choice should remain among BPTB, HT, and QT, selected by age, sport, kneeling demands, and prior injury. PLT harvest is fast, woven diameter is relatively large, and periarticular knee tendons are preserved—these advantages are real—but they are not enough to offset the inadequate failure-rate comparisons and the ankle cost.

High-intensity athletes: whenever periarticular knee autograft is available, do not choose the peroneus longus tendon. This includes patients preparing to return to jumping, cutting, sprinting, and contact sport, and those in sports that depend on ankle eversion and foot push-off (such as football, basketball, volleyball, handball, and dance). These patients most need a low failure rate, yet PLT versus BPTB has only one comparison with 50 per group (2-year rupture 4/50 versus 3/50), there is no failure-rate comparison with QT, and comparison with HT has accumulated only a small number of events. At one year the asymmetry in ankle strength and single-leg hop remained clear against a semitendinosus control. The Angthong et al. sample excluded such patients; in the 5-year trial of Butt et al., sport participants numbered fewer than 10 per group. Soleymanha et al., who argue PLT safety from foot and ankle scores, and Kehribar et al. both advise caution in high-level athletes in their Discussions. The 34-day earlier return to sport reported by Saeed et al. cannot serve as a reason for these patients to choose PLT.

Adult patients with low activity demand: PLT may serve as an alternative, with preoperative counseling on the ankle cost. The randomized trial of Butt et al. (almost entirely men, low activity demand) showed 5-year knee function similar to HT; Ertilav et al. detected no significant eversion-strength decline at 4 years in nonathletes; long-term review subjects of Velioglu et al. had good foot structure. For such patients who wish to avoid harvesting hamstring or patellar tendon, PLT is an acceptable choice. Counseling should cover eversion-strength decline from six months to one year postoperatively, sensory disturbance in the sural distribution, and uncommon common peroneal nerve injury. Subjects in existing studies are predominantly male; data on female patients are scarce.

When periarticular knee autograft is unavailable, insufficient, or multiple grafts are needed: the peroneus longus tendon is a reasonable choice. Examples include revision when ipsilateral HT or BPTB has already been harvested in prior surgery and QT has also been used or is unsuitable; multiligament reconstruction requiring multiple grafts; and intraoperative HT diameter insufficient, needing a supplementary graft. The premise of this recommendation is that other suitable common autografts are truly limited; the alternatives to PLT are then often contralateral or further periarticular knee harvest, an undersized graft, or allograft, and the ankle cost can be accepted. When HT diameter is insufficient, anterior-half PLT can be used as a supplement (the cases of Zhao et al. are of this type). The authors of Kotian et al. and Angthong et al. also place the appropriate use of PLT in such cases. The comparative material of this chapter comes from primary isolated ACL reconstruction; this recommendation rests on the trade-off above. In revision and multiligament populations, comparative material of PLT versus other grafts is scarce and insufficient by itself to support this recommendation.

Preexisting lateral ankle instability, peroneal tendinopathy, or other foot and ankle functional impairment: do not harvest the peroneus longus tendon. The peroneus longus contributes to foot eversion, first-ray plantarflexion, and dynamic lateral ankle stability; after full-thickness harvest multiple studies have measured reduced eversion strength. Adding this loss to an ankle that already has insufficient function is, in the author's view, inadvisable, and the author suspects that harvest may worsen preexisting lateral instability. Butt, de Oliveira, Angthong, and Masoud et al. all listed prior ankle injury, instability, or surgery among exclusion criteria; there are no data on outcomes after harvest in such patients, and the exclusion criteria themselves cannot be used to estimate their risk. For those with prior ankle surgery, judge by current lateral ankle stability and by peroneal tendon and foot–ankle function; do not exclude solely on the basis of surgical history. Soleymanha et al. also advise caution in those with preexisting lateral ankle instability in their Discussion.

When using the peroneus longus tendon: Prefer the anterior half when volume allows (reasons in Section 3; smaller donor-site loss has no direct measurement yet); protect the sural nerve when exposing the incision, and do not advance the harvester too far proximally. The author recommends incorporating ankle strength and balance training early postoperatively (the Kotian et al. cohort added peroneal training only after finding eversion weakness at 6 months); no comparative study has assessed whether early training reduces the deficit. Return-to-sport assessment should include, beyond knee testing, donor-side ankle eversion and plantarflexion strength and single-leg heel-rise testing; a ceiling AOFAS or FADI score cannot substitute. Asymmetry on single-leg hop may partly arise from the ankle and should be interpreted together with ankle strength results.

Evidence that could change this recommendation: A randomized trial with graft failure (rupture or revision) as the primary outcome, adequate sample size, follow-up of at least 2 years with complete inclusion of failure cases, comparing PLT with BPTB or HT in a young athletic population, while also measuring donor-side ankle strength and jump performance. If such a trial showed that PLT failure rate is no higher than the control, that the ankle deficit is small, and that sport is unaffected, PLT could be elevated to an equal choice for ordinary athletic populations; if the failure rate is higher or the ankle deficit persists, the recommendation should be tightened further.

Literature search through 4 October 2026.

References

  1. [1]

    Agarwal A, Singh S, Singh A, Tewari P. Comparison of Functional Outcomes of an Anterior Cruciate Ligament (ACL) Reconstruction Using a Peroneus Longus Graft as an Alternative to the Hamstring Tendon Graft. Cureus. 2023;15(4):e37273. PMID 37168157; DOI 10.7759/cureus.37273.

  2. [2]

    American Academy of Orthopaedic Surgeons. Management of Anterior Cruciate Ligament Injuries: Evidence-Based Clinical Practice Guideline. 2022. Full guideline text. Full text.

  3. [3]

    American Orthopaedic Foot & Ankle Society. Position Statement: Patient-Reported Outcomes Measures. Approved by the AOFAS Board of Directors, July 29, 2022. Full text.

  4. [4]

    Angthong C, Chernchujit B, Apivatgaroon A, Chaijenkit K, Nualon P, Suchao-in K. The Anterior Cruciate Ligament Reconstruction with the Peroneus Longus Tendon: A Biomechanical and Clinical Evaluation of the Donor Ankle Morbidity. J Med Assoc Thai. 2015;98(6):555-60. PMID 26219159.

  5. [5]

    Arora M, Shukla T, Singla M. Risk of Nerve Injury Following Peroneus Longus Harvest is Very Low: A Prospective Cohort of 600 Patients. Indian J Orthop. 2024;58(9):1232-1238. PMID 39170659; DOI 10.1007/s43465-024-01180-w.

  6. [6]

    Arora M, Tapasvi S, Shah J. Anterior cruciate ligament reconstruction using peroneus longus with lateral extra-articular tenodesis has excellent functional outcomes with a high return to sport rate: A prospective cohort study of 482 patients over 2 years. Knee. 2026;60:104404. PMID 41722485; DOI 10.1016/j.knee.2026.104404.

  7. [7]

    Arora M, Tapasvi S, Shukla T, Garg S, Shah J. Peroneus Longus to Peroneus Brevis Tenodesis Does Not Improve Ankle Functional Outcomes: A Prospective Cohort Study. Indian J Orthop. 2025;59(4):558-565. PMID 40276794; DOI 10.1007/s43465-025-01361-1.

  8. [8]

    Asif N, Khan MJ, Singhal U, Anwer A, Firoz MA, Rashid M. Randomized Controlled Study Comparing Hamstring Graft and Peroneus Longus Tendon Graft in Arthroscopic ACL Reconstruction. J Orthop Case Rep. 2024;14(11):240-245. PMID 39524299; DOI 10.13107/jocr.2024.v14.i11.4980.

  9. [9]

    Bi M, Zhao C, Zhang S, Yao B, Hong Z, Bi Q. All-Inside Single-Bundle Reconstruction of the Anterior Cruciate Ligament with the Anterior Half of the Peroneus Longus Tendon Compared to the Semitendinosus Tendon: A Two-Year Follow-Up Study. J Knee Surg. 2018;31(10):1022-1030. PMID 29421837; DOI 10.1055/s-0038-1627466.

  10. [10]

    Butt U, Vuletic F, Shaikh MAA, Amanullah, Rehman GU, Shah IA, Stålman A, Khan ZA. 5-years outcomes following arthroscopic anterior cruciate ligament reconstruction comparing quadruple hamstring and peroneus longus tendon autografts: a randomized control trial. Arch Orthop Trauma Surg. 2025;145:85 (published online 2024-12-23). PMID 39714511; DOI 10.1007/s00402-024-05639-1.

  11. [11]

    de Oliveira DE, Horita MM, Malpaga JMD, de Padua VB, de Gusmão Canuto SM, Jorge PB. Evaluation of the peroneus longus tendon as an autologous graft in knee surgery. Sci Rep. 2024;14(1):26341. PMID 39487269; DOI 10.1038/s41598-024-76779-y.

  12. [12]

    Desouza C, Nighot A, Sahu A, Shetty V. Peroneus longus autograft versus hamstring tendon autograft in primary anterior cruciate ligament reconstruction: a systematic review and meta-analysis of randomised controlled trials. Eur J Orthop Surg Traumatol. 2026;36(1):311. PMID 42542493; DOI 10.1007/s00590-026-04907-7.

  13. [13]

    Dogruoz F, Sari MK, Ertan MB, Ergun A, Gurcan S, Kose O. Ankle Function and Donor-Site Morbidity Following Peroneus Longus Graft Harvesting with or Without Tenodesis to Peroneus Brevis in Anterior Cruciate Ligament Reconstruction. J Clin Med. 2026;15(7):2577. PMID 41976877; DOI 10.3390/jcm15072577.

  14. [14]

    Ertilav D, Ertilav E, Dirlik GN, Barut K. Donor Site Morbidity after Removal of Full-Thickness Peroneus Longus Tendon Graft for Anterior Cruciate Ligament (ACL) Reconstruction: 4-Year Follow-up. Acta Chir Orthop Traumatol Cech. 2024;91(3):170-174. PMID 38963896; DOI 10.55095/achot2024/023.

  15. [15]

    Goncharov EN, Koval OA, Bezuglov EN, Vetoshkin AA, Goncharov NG, Encarnación Ramirez M, Nurmukhametov R, Montemurro N. Outcome of Primary Anterior Cruciate Ligament Reconstruction with Peroneus Longus and Bone–Patellar Tendon–Bone Autografts: A Clinical Comparative Study. Surgeries. 2023;4(3):434-445. DOI 10.3390/surgeries4030043.

  16. [16]

    Hale SA, Hertel J. Reliability and Sensitivity of the Foot and Ankle Disability Index in Subjects With Chronic Ankle Instability. J Athl Train. 2005;40(1):35-40. PMID 15902322.

  17. [17]

    Islam MN, Chowdhury AZ, Mahmud CI, Dey D, Ferdous MGS, Diba ANR, Rasul A. Comparative Study between the Functional Outcome of Arthroscopic Reconstruction of Anterior Cruciate Ligament Tear by Quadriceps and Peroneus Longus Tendon Autograft. Journal of Orthopedics and Sports Medicine. 2025;7(3):428-434. DOI 10.26502/josm.511500224.

  18. [18]

    Joshi S, Shetty UC, Salim MD, Meena N, Kumar RS, Rao VKV. Peroneus Longus Tendon Autograft for Anterior Cruciate Ligament Reconstruction: A Safe and Effective Alternative in Nonathletic Patients. Niger J Surg. 2021;27(1):42-47. PMID 34012241; DOI 10.4103/njs.NJS_22_20.

  19. [19]

    Karapinar SE, Tas A, Dincer R, Atay T, Baydar ML, Kumbul H. An Evaluation of the Effect on Ankle Stability of the Peroneus Longus Tendon Graft Used in Anterior Cruciate Ligament Reconstruction. Orthop J Sports Med. 2025;13(9):23259671251371229. PMID 40980563; DOI 10.1177/23259671251371229.

  20. [20]

    Kehribar L, Okutan AE, Akdemir E, Karabağ ÜA, Akın E, Anıl B, Salkılıç EK, Yılmaz AK. Peroneus longus autograft in anterior cruciate ligament reconstruction impairs ankle strength, balance and jump performance despite preserved knee function: a one year comparative study. Int Orthop. 2026. PMID 42624924; DOI 10.1007/s00264-026-06998-7.

  21. [21]

    Khalil MH, Zawam SH. Comparative study of Peroneus longus tendon autograft versus Hamstring tendon autograft in arthroscopic anterior cruciate ligament reconstruction. Int Orthop. 2025;49(6):1365-1372. PMID 40053066; DOI 10.1007/s00264-025-06468-6.

  22. [22]

    Kotian RN, Ajoy SM, Panduranga R, Visweswara RD, Galagali DA, Singh I, Patil S, Jagadish S. Altered gait biomechanics after peroneus longus tendon harvest for anterior cruciate ligament reconstruction: truth or myth? Bone Joint J. 2026;108-B(1):54-61. PMID 41475367; DOI 10.1302/0301-620X.108B1.BJJ-2025-0487.R1.

  23. [23]

    Kotian RN, Ajoy SM, Panduranga R, Visweswara RD, Kumar P, Galagali DA, Patil SC, Rafi Vaderi BM. Decreased evertor strength following peroneus longus tendon harvest for anterior cruciate ligament reconstruction: an isokinetic muscle strength testing study. Bone Joint J. 2026;108-B(6):766-772. PMID 42219170; DOI 10.1302/0301-620X.108B6.BJJ-2025-0274.R1.

  24. [24]

    Kumar K, Rao V, Panda AK, Joshi D. Peroneus Longus Tendon Versus Hamstring Tendon Autograft for Primary Anterior Cruciate Ligament Reconstruction: A Systematic Review and Meta-analysis of Comparative Studies. Orthop J Sports Med. 2025;13(9):23259671251374313. PMID 40980562; DOI 10.1177/23259671251374313.

  25. [25]

    Malik S, Biswas G, Singh SK, Sundaresan N, Bera AP, Kyada M. Comparative Evaluation of Peroneus Longus Tendon and Hamstring Tendon Autografts in Primary Anterior Cruciate Ligament Reconstruction: A Randomized Controlled Trial. J Orthop Case Rep. 2026;16(5):320-328. PMID 42131028; DOI 10.13107/jocr.2026.v16.i05.7294.

  26. [26]

    Masoud M, El-Assal MA, Ibrahim MA, Abdelzaher H, Khalifa AA, Sabra M, Fadle AA. Potentially impaired dynamic ankle control despite increased eversion and first-ray plantarflexion strength after peroneus longus tendon harvest for anterior cruciate ligament reconstruction: an observational study. Int Orthop. 2026;50(6):1341-1353. PMID 42177357; DOI 10.1007/s00264-026-06850-y.

  27. [27]

    Mirza K, Menezes RJ, Acharya PU, Austine J, d'Almeida VR, Kamath A. Donor-site morbidity following arthroscopic anterior cruciate ligament reconstruction using peroneus longus tendon autograft. Eur J Orthop Surg Traumatol. 2024;34(6):3171-3180. PMID 39039170; DOI 10.1007/s00590-024-04046-x.

  28. [28]

    Park JY, Fernandes A, Park SY, Lim H, Sayudo IF, Leibovitch L, Machinski E, Ha J. Comparative effectiveness of peroneus longus tendon (PLT) autografts versus hamstring tendon (HT) autografts in anterior cruciate ligament reconstruction: a comprehensive systematic review and meta analysis. Eur J Orthop Surg Traumatol. 2024;34(5):2691-2699. PMID 38755499; DOI 10.1007/s00590-024-03984-w.

  29. [29]

    Patel K, Mehta N, Taraviya A, Dhillon MS. ACL Reconstruction with Peroneus Longus Tendon Autograft Does Not Compromise Ankle Function: A Retrospective Study of 327 Patients With ≥ 2-Year Follow-Up. Indian J Orthop. 2026. DOI 10.1007/s43465-026-01884-1 (Abstract only).

  30. [30]

    Rhatomy S, Asikin AIZ, Wardani AE, Rukmoyo T, Lumban-Gaol I, Budhiparama NC. Peroneus longus autograft can be recommended as a superior graft to hamstring tendon in single-bundle ACL reconstruction. Knee Surg Sports Traumatol Arthrosc. 2019;27(11):3552-3559. PMID 30877316; DOI 10.1007/s00167-019-05455-w.

  31. [31]

    Rhatomy S, Wicaksono FH, Soekarno NR, Setyawan R, Primasara S, Budhiparama NC. Eversion and First Ray Plantarflexion Muscle Strength in Anterior Cruciate Ligament Reconstruction Using a Peroneus Longus Tendon Graft. Orthop J Sports Med. 2019;7(9):2325967119872462. PMID 31632995; DOI 10.1177/2325967119872462.

  32. [32]

    Saeed UB, Ramzan A, Anwar M, Tariq H, Tariq H, Yasin A, Mehmood T. Earlier Return to Sports, Reduced Donor-Site Morbidity with Doubled Peroneus Longus Versus Quadrupled Hamstring Tendon Autograft in ACL Reconstruction. JB JS Open Access. 2023;8(4):e23.00051. PMID 38058511; DOI 10.2106/JBJS.OA.23.00051.

  33. [33]

    Shao X, Shi LL, Bluman EM, Wang S, Xu X, Chen X, Wang J. Satisfactory functional and MRI outcomes at the foot and ankle following harvesting of full thickness peroneus longus tendon graft. Bone Joint J. 2020;102-B(2):205-211. PMID 32009424; DOI 10.1302/0301-620X.102B2.BJJ-2019-0949.R1.

  34. [34]

    Soleymanha M, Soleymani Nejad A, Keyhani S, Vosoughi F, LaPrade RF, Tollefson LV. Peroneus longus tendon harvest for ACL reconstruction yields good functional outcome of the ankle: A systematic review and meta-analysis. Knee Surg Sports Traumatol Arthrosc. 2026;34(9):3233-3246. PMID 41025611; DOI 10.1002/ksa.70079.

  35. [35]

    Velioğlu K, Öner K, Aslan FG, Okutan AE, Kerimoğlu S, Turhan AU. Harvesting the Full Thickness Peroneus Longus Tendon Is Not Associated With Structural Foot Impairments: A 12- to 23-Year Follow-up Study. Orthop J Sports Med. 2025;13(4):23259671251320659. PMID 40297044; DOI 10.1177/23259671251320659.

  36. [36]

    Yadav U, Nemani M, Devgun A, Malik M, Agrawal GK. Iatrogenic Foot Drop After Anterior Cruciate Ligament Reconstruction With Peroneus Longus Tendon Autograft: Report of a Rare Case. Cureus. 2022;14(6):e26476. PMID 35919368; DOI 10.7759/cureus.26476.

  37. [37]

    Zhang S, Cai G, Ge Z. The Efficacy of Anterior Cruciate Ligament Reconstruction with Peroneus Longus Tendon and its Impact on Ankle Joint Function. Orthop Surg. 2024;16(6):1317-1326. PMID 38650179; DOI 10.1111/os.14060.

  38. [38]

    Zhao Z, Tang L, Chen J, Bai X, Chen Y, Ng L, Zhou Y, Deng Y. The effect of harvesting the anterior half of the peroneus longus tendon on foot morphology and gait. J Orthop Surg Res. 2024;19(1):69. PMID 38225652; DOI 10.1186/s13018-023-04429-6.