Graft Choice for Anterior Cruciate Ligament Reconstruction in Women Aged 25 Years and Younger: A Systematic Review

Abstract

Context:

Although anterior cruciate ligament (ACL) tears are relatively common in athletic populations, few studies have systematically reviewed graft choice in young women.

Objective:

To quantitatively and qualitatively examine reported outcomes for graft choice in women aged 25 years and younger undergoing primary ACL reconstruction.

Data Source:

A systematic review was performed using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. An electronic search in the PubMed (includes MEDLINE) and EMBASE databases was completed using a combination of key terms.

Study Selection:

Studies were included if they reported graft choice outcomes in women aged 25 years and younger.

Study Design:

Systematic review.

Level of Evidence:

Level 4.

Data Extraction:

The following information was extracted: title, author, year of publication, number of female patients and age, graft type, follow-up, and patient-reported outcome measures. The following outcome scores were identified as being reported or not reported by each study: graft failure, contralateral ACL (CACL) rupture, IKDC (International Knee Documentation Committee), graft survival (Kaplan-Meier), Lysholm, Tegner, KT-1000, kneeling pain, return to sport, and Lachman.

Results:

Of 1170 identified articles, 16 met inclusion criteria, reporting on 1385 female patients aged 25 years and younger. Comparison of 655 bone–patellar tendon–bone (BPTB) versus 525 hamstring tendon (HT) autografts showed significant differences in mean failure rate between BPTB autografts (6.13% ± 2.58%) and HT autografts (17.35% ± 8.19%), P = 0.001. No statistically significant differences in CACL failure rates were found between BPTB autografts and HT autografts (P = 0.25). Pooled results for IKDC were possible in 3 of the HT autograft studies, showing a mean score of 88.31 (95% CI 83.53-93.08). Pooled Lysholm score results were possible in 2 of the HT autograft studies, showing a mean score of 93.46 (95% CI 91.90-95.01).

Conclusion:

In female patients aged 25 years and younger, BPTB autografts showed significantly less graft failure compared with HT autografts. However, BPTB autografts had comparable patient-reported outcomes compared with HT autografts with the available data. The overall state of evidence for graft choice in female patients aged 25 years and younger is low. Future studies should report statistics by age and sex to allow for further analysis of graft choice for this specific population that is known to be more vulnerable to ACL injury.

The incidence of anterior cruciate ligament (ACL) injuries in young women is increasing, especially in those who play competitive sports, with 1 study reporting a 5.7-fold increase in the rate of ACL reconstruction (ACLR) between 2004 and 2014 in patients 18 years or younger.^{14,25,30,31,40,64,66,71,74} Studies have demonstrated that women possess a greater risk for ACL injury versus men with younger patients being especially at risk for reinjury versus adult populations.^{5,6,18,30,31,66,71} Patient age, graft size, graft type, number of bundles, surgical technique, and more have been ongoing topics of debate for approaches to ACLR in younger populations.^{5,7,19,22,31,34,39,43,49,50,55,58,72,74}

The ACL is cited as the most frequently injured ligaments in the knee that requires surgical reconstruction.^14,25,32 Current literature identifies a clear trend toward ACLR as the preferred treatment option for ACL injuries in young populations, allowing for faster return to athletic activity and decreased risk of further chondral and meniscal injuries.^{14,18,20,22,28,31,36,54,69}

While previous systematic reviews and meta-analyses reporting on graft choice have not shown major differences in terms of graft failure rates, studies have failed to control for both age and sex in patients undergoing ACLR.^{11,23,24,42,65} Studies have shown a higher ACL reinjury rate in younger patients, especially during high-risk sports involving jumping/landing, cutting, and pivoting.^6,62,75 Many studies have attempted to identify the most suitable graft type in different patient populations; however, there is a lack of agreement on the ideal choice for younger populations and particularly in women.^{29,31,44,50,61,71} Hamstring tendon (HT) autografts have commonly been used, but there is an increased research focus on alternative graft choices, including quadriceps tendon (QT) autografts and bone–patellar tendon–bone (BPTB) autografts.^8,31,63,71 More recently, BPTB autografts have shown promising results in younger patients.^6,31,63 Spindler et al ⁶³ reported that in male and female patients aged 14 to 22 years, the incidence of ACL graft revision was 2.1 times higher with HT versus BPTB autograft (95% CI 1.3-3.5; P = 0.004). ⁶³ While there are studies reporting on graft choice in younger patient populations, there is no conclusion on graft choice based on sex in those aged 25 years and younger despite the known differences associated with sex.^31,35,46,59

The aim of this study was to systematically review ACLR outcomes in women aged 25 years and younger after undergoing primary ACLR with different graft types. This article is one of the first attempts to consolidate existing knowledge to provide greater insight on the importance of sex and age in the selection of graft choice for ACLR.

Methods

A systematic search of PubMed (which includes the MEDLINE database) and EMBASE was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. ⁴⁸ Although a protocol was not registered, all stages of the PRISMA search were planned before their execution. The search was independently performed by 2 authors for internal validity on July 16, 2020 and repeated on December 1, 2020. A high precision search was achieved using Medical Subject Heading (MeSH) terms in PubMed (MEDLINE) and associated terms from Emtree in EMBASE to identify women aged 25 years and younger. ²⁷

Studies were identified using search combinations of 1 word from each of the bracketed lists: [“Anterior Cruciate Ligament Injuries”] AND [“Female”] AND [“Anterior Cruciate Ligament Reconstruction” OR “Allograft” OR “Autograft” OR “Bone-Patellar Tendon-Bone Grafting OR “Transplants” OR “Hamstring Tendons” OR “Hamstring Muscles” OR “Patellar Ligament” OR “Quadriceps Muscle”] AND [“Outcome Assessment (Health Care)” OR “Graft Survival” OR “Treatment Outcome” OR “Postoperative Complications” OR “Return to Sport” OR “Recurrence” OR “Sex Factors” OR “Treatment Failure”] AND [“Child” OR “Adolescent” OR “Young Adult” OR “Adult”].

Duplicates were removed in EndNote during the screening stage and remaining articles were screened using title and/or abstract. Studies were excluded in our screening stage if they included female patients >25 years (without age groups stratification); included associated ligament or bone injuries requiring surgery or multiple surgical techniques in addition to a full ACL tear reconstruction; included cadavers or non-human subjects; did not mention graft choice; compared bundle type; were ACL repair studies; case reports; abstracts; revision outcome studies; non-English; meta-analyses; or systematic reviews.

Qualitative Analysis

The following data were collected from each study for descriptive analysis: title, primary author, year of publication, number of female patients and age, graft type, follow-up, and patient-reported outcome measures (PROMs). The following outcomes were identified as being reported or not reported by each study: graft failure, contralateral ACL (CACL) rupture, graft survival (Kaplan-Meier), Lysholm, KT-1000 arthrometer test (KT-1000), Lachman, Tegner activity score, International Knee Documentation Committee (IKDC), kneeling pain, and return to sport.

Quantitative Analysis

The following statistics were calculated or recorded from included studies in our statistical analysis: sample size, graft failure rate, CACL rupture rate, and mean and standard deviation (SD) of PROM scores. Outcome scores were excluded from further quantitative analysis when pooling by graft type if they did not report on graft type, did not report numeric scores, did not report a P value, or did not report variance. Statistical analysis of PROMs was limited to IKDC and Lysholm because these scores were the only standardized scores reported in >2 studies to allow for comparison by graft choice.

Studies with reported SDs related to PROMs were pooled using a random-effects model. This model used an inverse variance approach and I² statistic to assess heterogeneity. Data analyzation was completed using RevMan 5.3 software (The Cochrane Collaboration).

Weighted means and weighted SDs were able to be calculated for rates of graft failure in BPTB autograft, HT autograft, and allograft studies. Weighted means for CACL rupture were also calculated in BPTB autograft and HT autograft studies. Weights were determined by summing the number of knees in each study that reported on a target parameter. A study’s percentage contribution to this sum served as the weight placed on that article’s target parameter.

When comparing means, we first performed a single-factor analysis of variance (ANOVA), and if a statistically significant difference existed, we then conducted paired t tests assuming unequal variance, as the underlying populations of the studies may be different. Tests were conducted using Microsoft Excel. Throughout all aspects of this study, statistical significance was considered P < 0.05.

Results

Systematic Search Results

Of 1170 identified articles, 16 met inclusion criteria reporting on women aged 25 years and younger undergoing an ACLR with graft type identified. Included studies reported on 1385 female patients (BPTB autograft, n = 691; HT autograft, n = 629; QT autograft, n = 4; allograft [any], n = 62). Mean follow-up in the included studies ranged from 8 months ⁶⁸ to 20 years. ⁶⁷ In studies with multiple follow-ups, follow-up duration was specified as the time to final follow-up. Follow-up for 3 studies^16,51,73 reported as a combined male and female data point was not included in statistics.

After screening, a full-text reading of 175 nonexcluded studies was completed to assess eligibility through the PRISMA selection process illustrated in Figure 1. A total of 159 full-text articles were excluded during our eligibility phase because they reported on patients out of the age range (n = 35), did not stratify age groups (n = 95), did not stratify by sex (n = 18), were not in English (n = 8), and were not complete articles (n = 2). One study met all inclusion criteria but was ultimately excluded for use of an uncommon graft technique (n = 1). Living donor HT grafts are not common and not able to be used in multiple countries. All 16 studies that were not excluded during screening or full-text eligibility assessment were included for quantitative or qualitative synthesis.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram showing application of the study criteria, including the number of studies excluded at each stage in relation to the eligibility criteria.

Table 1 reports on the breakdown of graft choice for the 16 included studies. Five studies^{38,47,57,60,68} reported on multiple graft choices. Thirteen^{13,15-17,38,47,51,57,58,60,68,70,73} reported results for HT autograft, 5 studies ^{62,47,60,68,57} on BPTB autograft, and 1 study ³⁷ on QT autograft. Four studies^4,38,47,68 reported on allografts.

Table 1.

Graft choice in the included studies

First Author, Year	No. of Female Patients	Graft Type	Allograft, n	Hamstring Tendon Autograft, n	Bone–Patellar Tendon–Bone Autograft, n	Quadriceps Tendon Autograft, n
Aronowitz, 2000 4	10	Allo	10	—	—	—
Bourke, 2012 13	25	HT	—	25	—	—
Cassard, 2014 15	8	HT	—	8	—	—
Cohen, 2009 16	15	HT	—	15	—	—
Demange, 2014 17	5	HT	—	5	—	—
Kohl, 2014 37	3	QT	—	—	—	3
Larson, 2016 38	17	Allo, HT	4	13	—	—
Nelson, 2016 47	195	BPTB, HT, Allo	42	123	30	—
Perkins, 2019 51	197	HT	—	197	—	—
Salem, 2019 57	256	BPTB, HT	—	81	175	—
Salmon, 2018 58	26	HT	—	26	—	—
Shakked, 2017 60	65	BPTB, HT	—	28	37	—
Shelbourne, 2009 62	413	BPTB	—	—	413	—
Toole, 2017 68	88	BPTB, HT, Allo	6	46	36	—
Wall, 2017 70	4	HT	—	4	—	—
Webster, 2017 73	58	HT	—	58	—	—
Total reported	1385		62 (4.5%)	629 (45.4%)	691 (49.9%)	3 (0.2%)

Allo, allograft; BPTB, bone–patellar tendon–bone autograft; HT, hamstring tendon autograft; QT, quadriceps tendon autograft; —, not recorded in the article.

General study characteristics are summarized in Table 2. Four studies did not report mean age for female patients only, but specified that patients were <20 years old, ⁷³ <18 years old,^13,58 or <19 years old. ⁵¹ Table 3 summarizes included outcome measures in each study.

Table 2.

Patient demographics and follow-up

First Author, Year	No. of Female Patients	Age, a y	Graft Type	Mean Follow-up, b y
Aronowitz, 2000 4	10	13 (11-14)	Allo	2.18 (1-5)
Bourke, 2012 13	25	<18	HT	15 (14.25-16.9)
Cassard, 2014 15	8	12.8 (11.3-15)	HT	2.8 (2-5)
Cohen, 2009 16	15	13.1 (11-14)	HT	M+F: 3.75±1.53
Demange, 2014 17	5	10.7 (8.3-12.2)	HT	18.3 (15-22)
Kohl, 2014 37	3	11 (6.2-14.2)	QT	2.9 (1.6-4.4)
Larson, 2016 38	16	13.4 (12-15)	Allo, HT	4 (2-7)
Nelson, 2016 47	195	14.4	BPTB, HT, Allo	2.9
Perkins, 2019 51	197	<19	HT	M+F: 2.17 (0.5-4.67)
Salem, 2019 57	256	BPTP: 18.5 (15-25) HT: 18.4 (15-25)	BPTB, HT	BPTB: 2.4 (2.1-4.9) HT: 2.5 (2.0-4.2)
Salmon, 2018 58	26	<18	HT	20
Shakked, 2017 60	65	19.2	BPTB, HT	BPTB: 1.75 (1-3.67) HT: 2.68 (1-5.42)
Shelbourne, 2009 62	413	<25	BPTB	1
Toole, 2017 68	88	16.6 ± 2.1	BPTB, HT, Allo	0.68 ± 0.2
Wall, 2017 70	4	8-15	HT	2
Webster, 2017 73	58	<20	HT	M+F: 5.1 (3-7)

Allo, allograft; BPTB, bone–patellar tendon–bone autograft; HT, hamstring tendon autograft; QT, quadriceps tendon autograft; M + F, male and female statistic reported together (when not given as only female statistic).

^aAge is presented as mean (range), mean ± SD, or less than a certain age.

^bFollow-up is presented as mean, mean ± SD, or mean (range).

Table 3.

Outcomes reported by each of the included studies

First Author, Year	Graft Failure	CACL Rupture	IKDC	Lysholm	Graft Survival (Kaplan-Meier)	Tegner Level	KT-1000	Kneeling Pain	Return to Sport	Lachman
Aronowitz, 2000 4				+			+
Bourke, 2012 13	+				+
Cassard, 2014 15	+			+		+
Cohen, 2009 16	+		+	+			+
Demange, 2014 17	+		+
Kohl, 2014 37	+		+	+			+
Larson, 2016 38	+						+
Nelson, 2016 47	+
Perkins, 2019 51	+	+
Salem, 2019 57	+	+	+					+	+	+
Salmon, 2018 58	+				+
Shakked, 2017 60	+	+		+		+	+	+		+
Shelbour,ne 2009 62	+	+
Toole, 2017 68			+
Wall, 2017 70	+		+
Webster, 2017 73									+

CACL, contralateral anterior cruciate ligament; IKDC, International Knee Documentation Committee; +, indicates that an outcome measure was reported.

Graft Failure

Three of the 16 included studies^4,68,73 did not report graft failure in women aged 25 years and younger. Only 1 study ³⁷ reported a failure rate for QT autografts (0%), which included 4 knees (Table 4). Four studies^47,57,60,62 reported failure rates for BPTB autografts for a total of 655 knees with a weighted average of 6.21% ± 1.68%. Eleven studies^{13,15-17,38,47,51,57,58,60,70} reported failure rates for HT autografts for a total of 525 knees with a weighted average of 14.96% ± 1.57%. Two studies^38,47 reported failure rates of allografts for a total of 46 knees with a weighted average failure rate of 18.85% ± 4.10%.

Table 4.

Outcomes of graft choice in the included studies

First Author, Year	Graft Type	Graft Failure	CACL Rupture	Graft Survival (Kaplan-Meier)
Aronowitz, 2000 4	Allo	—	—	—
Bourke, 2012 13	HT	n = 7 (28%)	—	2 y: 92% 5 y: 76% 10 y: 76% 15 y: 72%
Cassard, 2014 15	HT	n = 1 (12.5%)	—	—
Cohen, 2009 16	HT	n = 1 (6.6%)	—	—
Demange, 2014 17	HT	n = 1 (20%)	—	—
Kohl, 2014 37	QT	n = 0 (0%)	—	—
Larson, 2016 38	Allo HT	HT: n = 1 (7.7%) Allo: n = 3 (75%)	—	—
Nelson, 2016 47	BPTB HT Allo	BPTB: 8.9% HT: 11.0% Allo: 13.5%	—	—
Perkins, 2019 51	HT	n = 28 (14.2%)	n = 17 (9%)	—
Salem, 2019 57	BPTB HT	BPTB n = 12 (6.9%) HT: n = 11 (13.6%)	BPTB: n = 13 (7.4%) HT: n = 5 (6.2%)	—
Salmon, 2018 58	HT	n = 8 (31%)	—	2 y: 88% 5 y: 77% 10 y: 77% 15 y: 69% 20 y: 69%
Shakked, 2017 60	BPTB HT	BPTB: n = 1 (2.7%) HT: n = 6 (21.4%)	BPTB: n = 3 (8.1%) HT: n = 1 (3.6%)	—
Shelbourne, 2009 62	BPTB	n = 25 (6.1%)	n = 41 (9.9%)	—
Toole, 2017 68	BPTB HT Allo	—	—	—
Wall, 2017 70	HT	n = 1 (0.25%)	—	—
Webster, 2017 73	HT	—	—	—

Allo, allograft; BPTB, bone–patellar tendon–bone autograft; CACL, contralateral anterior cruciate ligament; HT, hamstring tendon autograft; QT, quadriceps tendon autograft; —, not recorded in the article.

Significant between group effects were found when comparing BPTB autograft, HT autograft, and allograft mean failure percentages in a single-factor ANOVA test (F = 5.28; P = 0.02). Further t test analysis revealed a statistically significant difference in mean failure rate between BPTB autografts (6.13% ± 2.58%) and HT autografts (17.35% ± 8.19%), P = 0.001. No statistically significant differences were found for mean graft failure between BPTB autografts and allografts (P = 0.43) or between HT autografts and allografts (P = 0.54).

CACL Rupture

Four studies reported CACL rates.^47,57,60,62 Three studies^57,60,62 reported CACL ruptures in 625 BPTB autograft patients, with a weighted average failure rate of 9.12% ± 3.15%. Three studies^51,57,60 reported CACL ruptures in 306 HT autografts, with a weighted average failure rate of 7.52% ± 2.72%. No statistically significant differences in CACL failure rates were found between BPTB autografts and HT autografts (P = 0.25).

Kaplan-Meier Graft Survival

Graft survival as reported using the Kaplan-Meier estimate was reported on in 2 studies,^13,58 both reporting on HT autografts in women <18 years old. Mean graft survival at 15 years was 70.5% for n = 51 patients, showing how 29.5% of patients in these 2 studies suffered from graft failure within 15 years of ACLR with an HT autograft.

Patient-Reported Outcome Measures

International Knee Documentation Committee

IKDC scores were reported on in 6 studies^{16,17,37,57,68,70} (Table 5). Only Salem et al ⁵⁷ reported on BPTB in a way that fit inclusion criteria. Because this was the only study that reported numeric IKDC score and a P value, BPTB IKDC scores were not pooled. Three studies^16,57,70 reported numeric IKDC scores with values of SD for HT autografts, showing a mean pooled score of 88.31 (95% CI 83.53-93.08; Figure 2).

Table 5.

Patient-reported outcome measures available for pooling

First Author, Year	IKDC	Lysholm
Aronowitz, 2000 4	—	97.9 ± 1.97
Cassard, 2014 15	—	94.28 ± 4.61
Cohen, 2009 16	91.03 ± 5.08	93.2 ± 3.51
Demange, 2014 17	n = 5 reported A—normal A—normal B—nearly normal A—normal A—normal	—
Kohl, 2014 37	n = 3 reported C—abnormal B—nearly normal B—nearly normal	87 ± 14.42
Salem, 2019 57	BPTB: 86.7 ± 10.57 HT: 86.9 ± 10.06 P = 0.91	—
Shakked, 2017 60	—	BPTB: 86 HT: 85 P = 0.81
Toole, 2017 68	87.7 ± 11.2 P = 0.83	—
Wall, 2017 70	n = 3 reported 87 at 3.7 y 49 at 4.1 y 100 at 4.2 y	—

BPTB, bone–patellar tendon–bone autograft; HT, hamstring tendon autograft; IKDC, International Knee Documentation Committee; —, not reported in the article.

Figure 2.

Random-effects model pooling International Knee Documentation Committee (IKDC) in Wall et al, ⁷⁰ Salem et al, ⁵⁷ and Cohen et al. ¹⁶

Lysholm

Lysholm was reported on in 5 studies^{4,15,16,37,60} (Table 5). Shakked et al ⁶⁰ was excluded because no SD specific to graft type was reported. Kohl et al ³⁷ and Aronowitz et al ⁴ reported on grafts not reported by other studies and were also excluded from pooled statistics. Data from Cohen et al ¹⁶ and Cassard et al ¹⁵ for HT autografts alone were pooled in a random-effects model, showing a mean score of 93.46 (95% CI 91.90-95.01; Figure 3).

Figure 3.

Random-effects model pooling Lysholm in Cohen et al ¹⁶ and Cassard et al. ¹⁵

Tegner Activity Score

Tegner score was reported on in 2 studies.^15,60 Cassard et al. ¹⁵ reported only postoperative Tegner scores and did not report individual SDs for postoperative scores, so this metric could not be pooled. Shakked et al ⁶⁰ reported no significant difference in Tegner scores preinjury (P = 0.56) or at follow-up (P = 0.8) between HT and BPTB autografts.

Lachman, KT-1000, Return to Sport, Kneeling Pain

While statistics were reported and could be pulled from the included studies, variance and P values were not reliably reported, precluding statistical comparison of data. The following outcome measures were reported on for both BPTB and HT autografts: Lachman in 2 studies,^57,60 KT-1000 in 5 studies,^{4,16,37,38,60} return to sport in 2 studies,^57,73 and kneeling pain in 2 studies^57,60 (Table 6).

Table 6.

Patient-reported outcome measures not available for pooling

Author, Year	Tegner Activity Score	KT-1000 a	Kneeling Pain	Return to Sport	Lachman
Aronowitz, 2000 4	—	1.4 ± 0.96 mm	—	—	—
Cassard, 2014 15	8.57 ± 0.53	—	—	—	—
Cohen, 2009 16	—	1.94 ± 0.92 mm	—	—	—
Kohl, 2014 37	—	2.67 ± 1.15 mm	—	—	—
Larson, 2016 38	—	0.4 ± 1.3mm	—	—	—
Salem, 2019 57	—	—	Unable to do: BPTB: 5.6% vs HT: 0% P = 0.04 Extremely difficult: BPTB: 6.4% vs HT: 1% P < 0.05 Moderately difficult: BPTB: 16% vs HT: 10.2% P = 0.33 Minimally difficult: BPTB: 32.8% vs HT: 22.4% P = 0.18 Not difficult at all: BPTB: 39.2% vs HT: 65.3% P = 0.002	BPTB: 67.5% HT: 66.7%	BPTB: n = 71 1A n = 2 (2.8%) HT: n = 30 1A n = 1 (3.3%)
Shakked, 2017 60	Preinjury: BPTB: 7.5 HT: 7.7 P = 0.56 Current: BPTB: 6.2 HT: 6.3 P = 0.81	BPTB: n = 36 15 lb KT-diff: 0.4 20 lb KT diff: 0.5 Max KT diff: 0.5 HT: n=24 15 lb KT-diff: 1.7 20 lb KT-diff: 1.8 Max KT diff: 2.9 P: 15: <0.001 20: 0.002 Max: 0.001	“Usually” or “Always”: BPTB: 39% vs HT: 37.5% P = 0.89 Knee pain that interferes with activities: BPTB: 30.6% vs HT: 41.7% P = 0.38	BPTB: 42.4% HT: 29.2% P = 0.31	BPTB: n = 36 1A: n = 34 (94.4%) 2A: n = 2 (5.6%) 2B: n = 0 HT: n = 24 1A: 15 (63%) 2A: 8 (33%) 2B: 1 (4.2%) P: 1A: 0.001 2A: 0.001 2B: 0.27
Webster, 2017 73	—	—	—	HT: 71%	—

BPTB, bone–patellar tendon–bone autograft; HT, hamstring tendon autograft; —, not recorded in the article.

^aKT-1000, side-to-side difference measurements.

Discussion

Graft failure rates after ACLR have been well established in the literature for both men and women. However, these studies report on a wide age range that generalizes graft failure without stratifying based on a more vulnerable younger female patient population involved in competitive, high pivoting sports.^58,59,75 Younger age and female sex are known risk factors for ACL graft rupture, increasing risk by a magnitude of 2 to 7 times.^{1,3,26,33,41,52} Given the increasing rate of ACL injury in young female athletes and risk for rerupture in this population, there is a need for more high-quality studies to aid in the discussion of graft choice in this specific population. Sixteen studies covering 1385 female patients aged 25 years and younger were included for final quantitative and qualitative analysis. The most important finding of the present study was a significant difference in failure rate between BPTB autografts and HT autografts in women aged 25 years and younger undergoing ACLR (6.13% ± 2.58% vs 17.35% ± 8.19%, P = 0.001). Measures of graft survival and PROMs showed comparable results between BPTB and HT when reported, yet could not consistently be pooled for significant evaluation, leaving a gap in the literature for future studies to fill.

In studies reporting graft failure in adolescents versus adults, Ekeland et al ²¹ reported that the hazard ratio (HR) for graft revision was 5 times higher for individuals aged ≤18 years than for those aged ≥35 years (P < 0.001) and that the HR for graft type was 1.8 times higher for HT than for BPTB grafts (P < 0.001), but 2.8 times higher for individuals aged ≤18 years (P < 0.001). ²¹ Salmon et al ⁵⁸ reported that ACL graft survival with HT autografts at 20 years was 86% for adults and 61% for adolescents (HR = 3.3; P = 0.001), while increased ligament laxity was also found in the adolescent group versus adults. The pooled analysis of reported graft failure in this review on young women corresponds with the literature on BPTB grafts compared with HT autografts in adolescents versus adults.

Graft failure outcomes reported in studies identifying young men and women show optimal results for BPTB autografts compared with HT autografts.^6,31,63 Ho et al ³¹ reported that in 561 cases of boys and girls aged 5 to 19 years, BPTB autografts had the lowest failure rate when compared with soft tissue grafts (6% vs 13%; P < 0.001) with multivariate analysis revealing that graft choice was the primary variable predictive of failure (P < 0.05). Barber-Westin and Noyes ⁶ reported on athletes <20 years old undergoing ACLR with both BPTB and HT autografts and reported that 9% of BPTB autografts and 15% of HT autografts failed (odds ratio [OR] = 0.52; P = 0.002). Our systematic review evaluating young women only maintained the results found in the literature reporting on combined measures in adolescent men and women comparing BPTB and HT autografts.

In this study, BPTB autograft reported a lower weighted graft failure rate of 6.21% over 5 studies, while HT autograft reported a weighted failure rate of 14.96% over 13 studies, with only 2 studies^16,38 reporting an HT failure rate less than 10%. Shakked et al ⁶⁰ assessed young women (15-25 years) and showed that graft failure occurred in 2.7% of patients in the BPTB group versus in 21.4% of patients in the HT group (P < 0.02), with significantly fewer subsequent procedures in the BPTB group. Salem et al ⁵⁷ showed a significant difference in the failure rate in young female athletes aged 15 to 20 years, with reported failure rates for the BPTB group of 6.4% and 17.5% in the HT group (P = 0.02). However, this difference was not observed in women aged 21 to 25 years, ⁵⁷ suggesting that women aged 15 to 20 years undergoing ACLR could be most affected by graft choice.^46,57,75

Compared with pooled gender studies mentioned previously showing a 1.7-fold ⁶ to 2.2-fold ³¹ increased HT failure rate relative to BPTB failure, our systematic review showed a nearly 3-fold relative increased failure rate for HT autografts in young women (6.13% ± 2.58% vs 17.35 % ± 8.19%, P = 0.001). Multiple factors explain this variation in ACL rupture rates between sexes. Generalized joint laxity, hormonal factors, increased posterior tibial slope, smaller ACL cross-sectional area, decreased notch width, and landing tendencies have been suggested to effect rupture rates in women.^2,12,65 Evidence also suggests that musculoskeletal changes in women during maturation correlates with increased risk of ACL tear in girls during a time of rapid growth and hormonal influx.^53,76,77

Comparison of graft choice using PROMs was limited in this study because of insufficient reporting of measures of variance and significance in most of the included studies. Only IKDC and Lysholm scores could statistically be pooled for a limited number of studies. Despite the lower graft failure rates in BPTB autograft patients, HT autografts have been associated with a slightly lower incidence of kneeling pain and harvest site morbidity compared with BPTB autograft regardless of age.^45,57,60 Studies reporting IKDC scores all reported relatively high mean scores with pooled results for HT autografts revealing a mean pooled score of 88.31 (95% CI 83.53-93.08). Only Wall et al ⁷⁰ reported a mean IKDC less than 80 for HTs. Other measures of activity level, including Tegner activity score and return to sport, could not be pooled, yet were reported for HT and BPTB. Of reported measures, BPTB had slightly greater return-to-sport and Tegner scores, yet were not significant.^15,57,60,73 Webster et al ⁷³ reported that 71% of women with HTs returned to sport, yet only 48% returned to level I sports involving jumping and hard pivoting. Studies reporting on Lysholm scores all reported high mean scores and a pooled HT autograft score of 93.46 (95% CI 91.90-95.01).

KT-1000 testing was also reported on for HT and BPTB but could not be pooled. Despite direct graft comparison only being reported in 1 study, side-to-side difference for maximum KT-1000 in BPTB autografts was shown to be significantly lower than HT autografts by Shakked et al ⁶⁰ (P = 0.001). While all results showed relatively normal KT-1000 difference results <3 mm, increased laxity can be noted in non-BPTB grafts. Women have a greater degree of laxity than men after ACLR and that in women, and HT autografts show greater graft laxity than BPTB autografts.^10,12,65,72

Only 1 study ⁶⁸ reported a mean follow-up less than 1 year for PROMs, suggesting that long-term follow-up was possible to evaluate graft choice. The reported PROM results agree with the literature that, excluding results of knee pain and graft failure, PROMs regarding BPTB versus HT autografts appear relatively similar.^9,56,60

There are many studies evaluating ACLR with BPTB versus HT autografts, but further research is necessary to better understand ideal graft choice by sex and age group. Based on the results of this systematic review, BPTB grafts show optimal results in young female patients based on lower graft failure rates alone. However, beyond graft failure rates, there is a paucity of literature, particularly with measures of graft survival and PROMs, regarding graft choice in women aged 25 years and younger. Future studies reporting on ACLR graft outcomes should ensure results delineate findings for women from men and report by age groups instead of a range of patients.

Limitations

While our initial search yielded many potential articles for inclusion, our investigation was ultimately heavily limited by a lack of appropriate age and gender stratification in studies. A meaningful analysis was limited for QTs and allografts because of the low number of reported grafts and outcomes. Additionally, missing information on outcome variance further limited consolidation of data when it was appropriately stratified. There is also a lack of studies with high level of evidence, including a lack of randomized control trials.

Surgical technique was not included in this analysis. Surgical technique influences the success or failure of a particular graft independently of the innate characteristics of the graft itself. This study is limited in its ability to account for surgical technique as studies directly comparing surgical technique that stratify based on graft choice do not report outcome measures needed to be included in this analysis. Therefore, the results of this analysis are unable to be applied to a specific surgical technique and our analysis was completed with the understanding that surgical technique is a major variable that is not controlled for. This limitation further supports the need for more specific study protocols looking at graft choice in a select population while also controlling for and comparing the influence of surgical technique.

Return to sport was not included for as part of the analysis of graft failure. Return-to-sport timing and level of competition are known to influence graft failure. However, the studies that did include return-to-sport parameters (Salem et al, ⁵⁷ Shakked et al, ⁶⁰ and Webster et al ⁷³ ) did not stratify based on sex or age, which limits our ability to control for “return to sport” with regard to its influence on graft failure in our population.

When considering this population (female patients <25 years of age), there exists a large variety of physiology. This study is limited in its ability to account for the potential confounding associated with the variability within the study population.

Conclusion

In female patients aged 25 years and younger, BPTB autografts showed significantly less graft failure compared to HT autografts. However, BPTB autografts had comparable patient-reported outcomes compared to HT autografts with the available data. The overall state of evidence for graft choice in women aged 25 years and younger is low. Future studies should report statistics by age and sex to allow for further analysis of graft choice for this specific population that is known to be more vulnerable to ACL injury.