Allograft Augmentation of HamstringAutograft for Younger Patients UndergoingAnterior Cruciate Ligament Reconstruction

Clinical and Cost-Effectiveness Analyses

Cale A. Jacobs,*y PhD, Jeremy M. Burnham,yz MD, Eric Makhni,§ MD, MBA,Chaitu S. Malempati,y MD, Eric Swart,k MD, and Darren L. Johnson,y MDInvestigation performed at University of Kentucky, Lexington, Kentucky, USA

Background: Younger patients and those with smaller hamstring autograft diameters have been shown to be at significantlygreater risk of graft failure after anterior cruciate ligament (ACL) reconstruction. To date, there is no information in the literatureabout the clinical success and/or cost-effectiveness of increasing graft diameter by augmenting with semitendinosus allografttissue for younger patients.

Hypothesis: Hybrid hamstring grafts are a cost-effective treatment option because of a reduced rate of graft failure.

Study Design: Cohort study (economic and decision analysis); Level of evidence, 3.

Methods: We retrospectively identified patients younger than 18 years who had undergone ACL reconstruction by a single sur-geon between 2010 and 2015. During this period, the operating surgeon’s graft selection algorithm included the use of bone–patellar tendon–bone (BTB) autografts for the majority of patients younger than 18 years. However, hamstring autografts (ham-string) or hybrid hamstring autografts with allograft augment (hybrid) were used in skeletally immature patients and in those whomthe surgeon felt might have greater difficulty with postoperative rehabilitation after BTB graft harvest. Patient demographics, grafttype, graft diameter, the time the patient was cleared to return to activity, and the need for secondary surgical procedures werecompared between the hamstring and hybrid groups. The clinical results were then used to assess the potential cost-effective-ness of hybrid grafts in this select group of young patients with an ACL injury or reconstruction.

Results: This study comprised 88 patients (hamstring group, n = 46; hybrid group, n = 42). The 2 groups did not differ in terms ofage, sex, timing of return to activity, or prevalence of skeletally immature patients. Graft diameters were significantly smaller in thehamstring group (7.8 vs 9.9 mm; P\ .001), which corresponded with a significantly greater rate of graft failure (13 of 46 [28.3%] vs5 of 42 [11.9%]; P = .049). As a result of the reduced revision rate, the hybrid graft demonstrated incremental cost savings ofUS$2765 compared with the hamstring graft, and the hybrid graft was the preferred strategy in 89% of cases.

Conclusion: Driven by increased graft diameters and the reduced risk of revision, hybrid grafts appear to be a more cost-effec-tive treatment option in a subset of younger patients with an ACL injury.

Keywords: ACL; graft failure; complication; adolescent

The published rates of hamstring autograft failure inyounger patients within the first 5 years after anterior cru-ciate ligament (ACL) reconstruction are alarmingly high,ranging from 5% to 29%.3,10,11,15,18,22,25,33 Roughly 90% ofyounger, active patients attempt to return to high-levelactivities,28,33 thus potentially placing greater strain onthe graft and increasing the likelihood of reinjury. Notonly is there a potentially greater demand placed on thegraft, but both clinical and animal studies have demon-strated that hamstring and other synovialized autografts

take longer to mature than nonsynovialized patellar ten-don or quadriceps tendon autografts.17,21 This early returnto high-demand activities during a period of time when thegraft has not yet matured may play a role in the increasedgraft failure rate in this age group. Furthermore, signifi-cantly greater anterior laxity has been reported in teenagepatients at a mean of 26 months after ACL reconstructioncompared with older patient groups, implying that ham-string autograft elongation during the first 2 years afterACL reconstruction may be greater in younger patients.19

In addition to the increased risk of graft failure for moreactive teenage patients, those with graft diameters less than8.5 mm have also been shown to be at significantly greaterrisk of graft failure and inferior patient-reported out-comes.18,20 Biomechanical studies have demonstrated that

The American Journal of Sports Medicine, Vol. XX, No. XDOI: 10.1177/0363546516676079� 2016 The Author(s)

1

increasing graft diameter results in increased tensilestrength of the graft and reduced anterior laxity.2,34 Clini-cally, this has translated to a significantly lower likelihoodof revision ACL reconstruction as graft diameter is increasedfrom 7 to 9 mm.30

As such, increasing graft diameter may offer cliniciansa potential method to mitigate the increased risk of graftfailure for younger patients. In response to the inferiorresults with small diameter hamstring autografts pub-lished in 2012, we began augmenting our hamstring auto-grafts with semitendinosus allografts in select patientpopulations with the goal of increasing graft diameter.Although these so-called hybrid grafts created using differ-ent allograft tissue and graft preparation techniques havedemonstrated poor clinical results,6 our anecdotal clinicalexperience has been more favorable. Therefore, this studyaimed to (1) evaluate the clinical success of increasing graftdiameter by augmenting hamstring autografts with semi-tendinosus allograft tissue and (2) determine whetherhybrid hamstring grafts provide a cost-effective alternativeto hamstring autografts in this subset of patients. Wehypothesized that the initial costs of the allograft tissuewith the hybrid graft would be offset by a significant reduc-tion in the need for revision ACL reconstruction, thus pro-viding a cost-effective alternative to hamstring autograftsin this subset of younger patients with an ACL injury.

METHODS

Patients

The University of Kentucky institutional review boardapproved this study (protocol no. 16-0077-P1H). This studyincluded all patients younger than 18 years who hadundergone primary ACL reconstruction between 2010and 2015 with hamstring autograft, either with or withoutallograft semitendinosus augmentation. Throughout thestudy period, the operating surgeon’s (D.L.J.) graft selec-tion algorithm involved the use of bone–patellar tendon–bone (BTB) autografts in the majority of patients youngerthan 18 years. Hamstring autografts were selectively uti-lized in skeletally immature patients and in those whomthe senior author determined to be at high risk of not tol-erating the postoperative rehabilitation associated witha BTB autograft. Preoperative factors associated with thepatient not being able to tolerate BTB postoperative reha-bilitation included failure to achieve adequate range ofmotion during preoperative physical therapy, demon-strated poor pain tolerance, or poor engagement in preop-erative physical therapy. After the inferior results of

smaller diameter hamstring autografts published in2012,15 hamstring autografts continued to be used in thesame patient population. However, beginning in August2012, these grafts were consistently augmented with semi-tendinosus allograft tissue to increase the overall graftdiameter.

To compare clinical results of the 2 grafts, we identifiedpatients younger than 18 years who had undergone ACLreconstruction by a single surgeon between 2010 and2015 with either a hamstring autograft (hamstring) orhybrid autograft with allograft augment (hybrid). Similarsurgical technique, pre- and postoperative rehabilitation,and objective return-to-play criteria were used for bothgroups as directed by the single senior surgeon. Patientsundergoing revision procedures, with multiligament inju-ries, or with follow-up less than 18 months were excluded.

Surgical Technique

Gracilis and semitendinosus autograft harvest was identi-cal between those receiving either hamstring or hybridgrafts. The gracilis and semitendinosus tendons were firstharvested via longitudinal incision approximately 4 cm inlength, starting 2 cm distal and 2 cm medial to the tibialtubercle. Dissection was continued to the sartorial fasciaand the gracilis and semitendinosus tendons were pal-pated. The sartorial fascia was then incised horizontallybetween the 2 tendons. The gracilis tendon insertion sitewas visualized deep to the sartorial fascia and subse-quently whipstitched with a nonabsorbable No. 2 TiCronsuture (Medline Industries). The insertion was thensharply incised and adhesions around the gracilis tendonwere released. A closed tendon stripper was used to per-form the harvest. After the gracilis was harvested, thesemitendinosus tendon was harvested in a similar manner.

After harvest of the gracilis and semitendinosus ten-dons was completed, both tendons were brought to theback table. Muscle tissue was removed and a No. 2 TiCronsuture was used to whipstitch the proximal ends of bothtendons. For those receiving a hybrid graft, the semitendi-nosus allograft was similarly whipstitched on both ends.The allograft was passed first, then the autograft aroundthe allograft, with the end goal of having as little of theallograft tissue exposed to the knee joint milieu in vivo.Allograft tissue was provided by LifeNet Health Inc, andall grafts were sterilized using the Allowash XG process(LifeNet Health Inc) including 1.2 to 1.9 Mrad irradiation.6

The allograft strands were then placed in the middle of theautograft strands, and the entire graft was doubled over,forming 6 strands with the semitendinosus allograft in

*Address correspondence to Cale A. Jacobs, PhD, Department of Orthopedics and Sports Medicine, University of Kentucky, 740 S Limestone, RoomK426, Lexington, KY 40536-0284 (email: cale.jacobs@ uky.edu).

yDepartment of Orthopedics and Sports Medicine, University of Kentucky, Lexington, Kentucky, USA.zDepartment of Orthopaedic Surgery, University of Pittsburgh, Pittsburgh, Pennsylvania, USA.§Department of Orthopedic Surgery, Rush University, Chicago, Illinois, USA.kColumbia University Medical Center, New York, New York, USA.E.M. is now affiliated with Division of Sports Medicine, Department of Orthopaedic Surgery, Henry Ford Health System, Detroit, Michigan, USA.

One or more of the authors has declared the following potential conflict of interest or source of funding: D.L.J. has intellectual property with, is a paidconsultant of, and has received royalties and research support from Smith & Nephew.

2 Jacobs et al The American Journal of Sports Medicine

the middle (Figure 1). A cortical suspension fixation device(Endobutton; Smith & Nephew) was added to the loopedend while the free ends were pulled to tension using a grafttensioning device (Graftmaster; Smith & Nephew). Begin-ning at the looped end of the graft, a Cobraid (Smith &Nephew) suture was used to bundle the tendons together.The suture was run down toward the nonlooped end of thegraft, stopping several centimeters from the loose ends,allowing the tendon ends to be separated later for backuptibial fixation. The suture was then run back down the oppo-site side of the graft. The graft was soaked in mineral oilunder tension until ready for insertion into the knee, withthe mineral oil acting as a lubricant to help pass the graft.31

For both hamstring and hybrid grafts, the femoral tun-nel was drilled using an accessory medial portal, and bothtibial and femoral tunnels were reamed to the nearest0.5 mm equal to or greater than the measured diameterof the graft. Femoral fixation was achieved using an Endo-button as the suspensory fixation, and tibial fixation wasachieved using a BioRCI (Smith & Nephew) bioabsorbableinterference screw in the tibial tunnel. For backup tibialfixation, the free tendon ends were wrapped arounda bicortical 4.5-mm large fragment screw, which wasthen tightened down with a 20-mm Geofit washer (DePuyMitek).

Clinical Data Collection and Statistical Analyses

For the current analysis, ‘‘graft failure’’ was operationallydefined as patient-reported instability that affected thepatient’s ability to perform activities of daily living or sportingactivities, pathological laxity during the physical examination(positive Lachman test, marked anterior laxity, or a positivepivot-shift test compared with the healthy knee), or evidenceof a failed graft on magnetic resonance imaging or duringarthroscopy. Any patient that met these criteria was

considered to have a failed graft regardless of whether thepatient chose to undergo revision ACL reconstruction.23,32

Patient demographics, graft type, graft diameter, the timethe patient was cleared to return to activity, as well as thepresence and timing of graft failure were recorded. This infor-mation was located by reviewing all orthopaedic clinical notesand operative notes from each patient’s electronic healthrecord. In an attempt to minimize the number of patientslost to follow-up, we also reviewed the Kentucky electronichealth information exchange database for any knee-relatedoffice visits or surgeries for which the patient’s knee was theprimary complaint. The state-based exchange does not containinformation on the state’s entire population, but reporting ismandatory for all patients with Medicaid and is optional forall others. The exchange has more than 1300 medical facilitiesactively submitting and exchanging live data.

Continuous variables were compared with independentt tests and categorical variables were compared betweenhamstring and hybrid groups using 1-tailed chi-square or

Figure 1. Augmentation of the hamstring autograft witha semitendinosus allograft. (A) Gracilis and semitendinosusautografts are combined with a semitendinosus allograft(asterisk). The allograft strand is positioned on the interiorof the graft. (B) When the graft is doubled, a 6-strand graftis created, with 2 allograft strands on the inside surroundedby the 4 hamstring autografts on the outside, leaving the allo-graft tissue less directly exposed to the knee joint milieu invivo. Free tendon ends are visible on the nonlooped side ofthe graft. These tendon ends are later wrapped arounda bicortical 4.5-mm screw and secured with a spiked washerfor backup tibial fixation.

Figure 2. The simplified decision tree used to compare thebase case cost-effectiveness of hamstring autografts withor without being augmented with semitendinosus allografttissue. ACL, anterior cruciate ligament.

AJSM Vol. XX, No. X, XXXX Hybrid Hamstring Grafts in Younger Patients 3

Fisher exact tests as appropriate. A binary regression (for-ward entry, Wald method) was used to determine whethera model containing a combination of patient sex, age, bodymass index (BMI), duration of follow-up, graft type, orgraft diameter could accurately predict graft failure. Forall analyses, an alpha level of P � .05 was considered sta-tistically significant.

Cost-Effectiveness Analysis

The clinical results were then used to compare the potentialcost-effectiveness and value provided by hybrid grafts. Thisstudy made the following assumptions: (1) the subset ofpatients was similar to the patients included in the clinical

portion of this study (ie, patients younger than 18 years whowere suspected of potentially not tolerating BTB postopera-tive rehabilitation), (2) patients underwent equivalent pre-operative treatment with equivalent costs, (3) patients didnot suffer a contralateral ACL injury during the studyperiod, and (4) those with complications or graft failuresoccurring after a tertiary surgery remained at the currenthealth state (primary surgery, index ACL reconstruction;secondary surgery, revision ACL reconstruction or reopera-tion because of infection or arthrofibrosis; and tertiary sur-gery, second revision ACL reconstruction or secondreoperation for infection or arthrofibrosis) (Figure 2).

The base case cost-effectiveness was assessed usingMonte Carlo microsimulation of a simplified decision-tree

TABLE 1Baseline Variables, Sensitivity Analysis Ranges and Distributions, and Sourcesa

Variable Baseline Value Probabilistic Sensitivity Analysis Distribution Source

Probabilities of Complications After Primary ACL Reconstructionb

Revision for graft failureHamstring 0.283 Beta Current studyHybrid 0.119 Beta Current study

ArthrofibrosisHamstring 0.043 Beta Current studyHybrid 0.048 Beta Current study

InstabilityHamstring 0.240 Beta Genuario et al12

Hybrid 0.240 Beta Genuario et al12

InfectionHamstring 0.013 Beta Brophy et al4

Hybrid 0.013 Beta Brophy et al4

Probabilities of Secondary Complications(Assumed to Be Similar for Both Hamstring and Hybrid)b

After revision ACL reconstructionRerevision for graft failure 0.095 Beta Reinhardt et al26

Arthrofibrosis 0.068 Beta Nwachuwku et al24

Instability 0.095 Beta Reinhardt et al26

Infection 0.022 Beta Leroux et al16

Second reoperation for arthrofibrosis 0.151 Beta Nwachuwku et al24

Second infection 0.286 Beta Calvo et al7

Utility (QALY)c

Well state 1 Normal Genuario et al12

Arthrofibrosis 0.770 Normal Genuario et al12

Instability 0.790 Normal Genuario et al12

Infection 0.965 Normal Genuario et al12

Well after revision 0.98 Normal Genuario et al12

Secondary complication 0.700 Normal Genuario et al12

Cost, US$d

Hamstring 4072 Gamma Genuario et al12

Hybrid 5273.56 Gamma Current studyInfection 9805 Gamma Genuario et al12

Arthrofibrosis 3000 Gamma Genuario et al12

Revision 20,000 Gamma Genuario et al12

aHamstring indicates hamstring autografts, whereas hybrid indicates hybrid hamstring autografts with allograft augment. ACL, anteriorcruciate ligament; QALY, quality-adjusted life-year.

bProbability parameters were assigned beta distributions about the baseline value, with the standard deviation assumed to be 20% of themean.9

cUtility parameters were assigned normal distributions about the baseline value.9dCost parameters were assigned gamma distributions about the baseline value.9

4 Jacobs et al The American Journal of Sports Medicine

model created with 1000 theoretical patients assignedequally to hamstring or hybrid cohorts (Figure 2). Thesetreatment arms were divided into outcome arms based onprobabilities, utilities, and costs derived from both the cur-rent clinical results and from the literature (Table 1). Spe-cifically, we included the probabilities of a successfuloutcome as well as revision owing to graft failure, persis-tent instability, arthrofibrosis requiring reoperation, andinfection. These values were taken directly from our clini-cal results or literature published on patients youngerthan 18 years when possible, but we used adult valuesif adolescent-specific references could not be identified(Table 1). The incremental cost of the hybrid graft wasUS$1201.56 greater than the hamstring graft based onpricing at our facility. All other costs were consistentwith the previous study by Genuario et al.12 Terminal out-comes were assigned a health state/utility score and a soci-etal cost, and the threshold for cost-effectiveness was set asan incremental cost-effectiveness ratio of $50,000/quality-adjusted life-year (QALY).12

To better estimate uncertainty within each model, prob-abilistic sensitivity analysis of 1000 first-order cases wasused to vary probabilities, utilities, and costs within the 2models. Similar to the methods of Crall et al,9 probabilityparameters were assigned beta distributions, utility param-eters were assigned normal distributions, and cost parame-ters were assigned gamma distributions. All parameterdistributions were centered about the baseline value andstandard deviations were assumed to be 20% of the mean.We also determined the percentage of cases in which eachstrategy (hamstring or hybrid graft) would be the preferredstrategy based on the willingness to pay of $50,000.

RESULTS

Clinical Results

This study included a total of 88 patients: 46 in the ham-string group and 42 in the hybrid group (Table 2). The 2groups did not differ in terms of age, sex, or the prevalenceof skeletally immature patients. However, BMI was signif-icantly greater in the hybrid group and the duration of

follow-up was significantly greater in the hamstring group.Graft diameters were significantly smaller in the ham-string group (median 7.8 vs 9.9 mm; P \ .001), which cor-responded with a significantly greater rate of graft failure(13 of 46 [28.3%] vs 5 of 42 [11.9%]; P = .049). Neither thetiming of when the patient was cleared to return to activitynor the time to graft failure differed between groups (Table2). The logistic regression results indicated that a modelcontaining graft type, BMI, and follow-up durationexplained 58% of the variability in the graft failure rates(Nagelkerke R2 = 0.58, P \ .001). Both BMI and follow-up duration were inversely related to graft failure,

TABLE 2Comparison of Patient Demographics and Complications Between the Hamstring and Hybrid Groupsa

Variable Hamstring Group (n = 46) Hybrid Group (n = 42) P Value

Sex (female/male), n 26/20 18/24 .10Skeletally immature, n 31 24 .16Age, y 14.6 6 1.7 15.0 6 1.8 .32Body mass index, kg/m2 23.0 6 4.8 26.6 6 7.2 .01Graft diameter, mm 7.8 6 1.2 9.9 6 0.8 \.001Cleared to return to activity, mo 7.9 6 1.7 8.3 6 1.5 .31Follow-up, mo 46.7 6 21.9 31.2 6 13.8 \.001Reoperations for arthrofibrosis 2 (4.3) 2 (4.7) .66Graft failure 13 (28.3) 5 (11.9) .049Time between surgery and retear, mo 21.4 6 16.0 18.1 6 11.6 .22

aData are reported as means 6 SD or n (%) unless otherwise indicated. Bolded P values indicate a statistically significant differencebetween groups (P � .05).

Figure 3. Scatterplot of the 1000 cases modeled during theprobabilistic sensitivity analysis demonstrating consistentlyreduced costs associated with the hybrid graft. QALY,quality-adjusted life-year.

AJSM Vol. XX, No. X, XXXX Hybrid Hamstring Grafts in Younger Patients 5

suggesting that patients with a lower BMI in the ham-string group were more likely to fail, and that these fail-ures tended to occur sooner after surgery and not later.

Cost-Effectiveness Results

The hybrid graft was the dominant strategy in the base casebecause it provided an incremental cost savings of $2765.49and was slightly more effective (0.94 QALY vs 0.92 QALY).Probabilistic sensitivity analysis results were similar, withthe hybrid graft demonstrating mean costs of $8259.51 6

$1289.55 and effectiveness of 0.94 6 0.13 compared with$10,736.31 6 $1975.85 and effectiveness of 0.92 6 0.11 ofthe hamstring graft (Figure 3). The hybrid graft was thepreferred treatment strategy in 86% of the simulated cases.The cost advantage of the hybrid graft was driven by thereduced revision rate compared with the hamstring groupin the current study; however, if the revision rate for thehamstring graft could be reduced from 28% to 12% in thissubset of younger patients, the hamstring graft wouldbecome the dominant strategy because of the added costassociated with the allograft tissue.

DISCUSSION

The rate of reinjury after ACL reconstruction is alarminglyhigh in the young athlete who returns to level 1 sports,suggesting that the standard of care may not be sufficientin this high-risk patient population. Although primaryACL reconstruction has been demonstrated to be cost-effective,13,14,27 revision ACL reconstruction leads to worseoutcomes29,35 and higher costs.5 As such, cost-effectivealternatives during primary ACL reconstruction must beexplored to better protect long-term knee health for thissubset of high-risk patients. Although there are manymodifiable factors that can potentially improve graft

survivorship for younger patients (eg, improved injury pre-vention and screening, biologic interventions, developmentof evidence-based return to play criteria), optimizing thebiomechanical characteristics of the graft is one area thatwarrants continued evaluation. As such, this study aimedto (1) evaluate the clinical success of increasing the graftdiameter by augmenting hamstring autografts with sem-itendinosus allograft tissue and (2) determine whetherhybrid hamstring grafts provide a cost-effective alternativeto hamstring autografts in this subset of patients. Theresults supported our hypothesis that the initial costs ofthe allograft tissue with the hybrid graft would be offsetby the value provided by the significant reduction in theneed for revision ACL reconstruction.

As previously discussed, there are biomechanical advan-tages to increased graft diameter, including increased ten-sile strength of the graft,2 with the potential for reducedmeniscal stress, decreased joint laxity, and less articularcartilage contact stress.34 Clinically, larger hamstring graftdiameters have been associated with reduced risk of graftfailure and the need for revision ACL reconstruction.8,30

In the current study, the use of the hybrid graft resultedin consistently greater graft diameters compared with treat-ment with hamstring autograft tissue alone, and the hybridgroup also demonstrated a reduced need for revision ACLreconstruction.

Although incorporating allograft tissue within a hamstringautograft can provide surgeons with the ability to consis-tently increase graft diameter,1 results of hybrid graftshave been mixed.6 In contrast with our findings, previousauthors have reported increased failure rates when augment-ing hamstring autografts with allograft tissue. Interestingly,the number of graft failures did not dramatically differbetween studies, with 5 of 42 (11.9%) requiring revision forgraft failure in the current study compared with 4 of 29(13.8%) in the study by Burrus et al.6 However, Burruset al reported that an additional 8 of 29 hybrid grafts (28%)became structurally compromised or incompetent. Therewere several methodological differences between the 2 stud-ies that may explain, at least in part, the discordance in clin-ical results (Table 3). First, the current study used a morehomogeneous patient population in terms of allograft tissueused, femoral tunnel drilling technique, and femoral and tib-ial fixation utilized. Furthermore, the final location of theallograft tissue in vivo differed between studies. In the cur-rent study, allograft tissue was folded within the graft and,although not standardized, Burrus et al reported that thesurgeons attempted to rotate the graft so that the allografttissue was anterior to the autograft tissue. We cannot specu-late as to whether there was an individual effect or combinedeffects of the different patient populations, fixation methods,tunnel locations, or anterior location of the allograft tissue,which may be related to the increased rate of clinical failuresreported by Burrus et al. Future studies in this area are cer-tainly warranted. Furthermore, we restricted the use of thehamstring or hybrid graft to a selected subset of our ACLreconstructions, as discussed below.

This study was not without limitations. First, the sam-ple size of this retrospective study was fairly small and didnot include patient-reported outcomes. Larger prospective

TABLE 3Methodological Differences Between the Current Study

and Previous Reports of Hybrid Hamstring Grafts

Variable Current Study Burrus et al6

Hamstring graft/hybrid graft, n 46/42 29/29Age of hybrid group, y 15.0 6 1.8 26.6 6 10.3Skeletally immature, n 24 0Allograft tissue utilized

Semitendinosus 42 (100) 20 (69)Tibialis anterior 0 (0) 9 (31)

Femoral tunnel drilling techniqueAnatomic 88 (100) 30 (53)Transtibial 0 (0) 28 (47)

Femoral fixationSuspensory 88 (100) 30 (52)Aperture 0 (0) 18 (31)Combined 0 (0) 10 (17)

aData are reported as means 6 SD or n (%) unless otherwiseindicated.

6 Jacobs et al The American Journal of Sports Medicine

studies are necessary to validate the current findings andexamine potential effects on patient-reported outcomes.Second, we studied a very select subset of patients youngerthan 18 years. Future studies are needed to both confirm theclinical success of this procedure, as well as to evaluatewhether the value provided by hybrid grafts is demonstratedin other patient populations. Third, although the 2 groupswere homogeneous in terms of sex, age, surgical technique,femoral and tibial fixation, and postoperative rehabilitation,the length of follow-up was significantly longer and BMI wassignificantly lower in the hamstring group. The regressionresults suggested that shorter follow-up and lower BMIwere included in the model predicting graft failure. Assuch, group differences in these factors may have influencedthe reduced graft failure rates demonstrated by the hybridgroup. Finally, this analysis was based on the US healthcare system, and the cost-effectiveness results may differ inother health care environments. Furthermore, allograft aug-mentation may not be an available option in other healthcare systems or geographic locations.

CONCLUSION

Augmenting hamstring autografts with semitendinosusallograft tissue resulted in a significantly reduced rate ofgraft failure in a subset of patients with ACL injury whowere younger than age 18 years. The value associatedwith the decreased revision rate translated into an overallcost savings of $2765 per patient, suggesting that hybridgrafts offer a cost-effective alternative to hamstring auto-grafts in this subset of high-risk patients. However, if thefailure rate of hamstring autografts can be reduced to12% in this subset of patients, hamstring autografts wouldprovide the more cost-effective treatment option. Futurestudies are therefore needed to determine whether othersurgical techniques or perioperative interventions can pro-vide such a reduction.

REFERENCES

1. Alvarez-Pinzon AM, Barksdale L, Krill MK, Leo BM. Hybrid graft ante-

rior cruciate ligament reconstruction: a predictable graft for knee sta-

bilization. Orthopedics. 2015;38:e473-e476.

2. Boniello MP, Schwingler PM, Bonner JM, Robinson SP, Cotter A,

Bonner KF. Impact of hamstring graft diameter on tendon strength:

a biomechanical study. Arthroscopy. 2015;31:1084-1090.

3. Bourke HE, Gordon DJ, Salmon LJ, Waller A, Linklater J, Pinczewski

LA. The outcome at 15 years of endoscopic anterior cruciate ligament

reconstruction using hamstring tendon autograft for ‘isolated’ anterior

cruciate ligament rupture. J Bone Joint Surg Br. 2012;94:630-637.

4. Brophy RH, Wright RW, Huston LJ, Nwosu SK, Spindler KP; MOON

Knee Group. Factors associated with infection following anterior cruci-

ate ligament reconstruction. J Bone Joint Surg Am. 2015;97:450-454.

5. Brophy RH, Wright RW, Matava MJ. Cost analysis of converting from

single-bundle to double-bundle anterior cruciate ligament recon-

struction. Am J Sports Med. 2009;37:683-687.

6. Burrus MT, Werner BC, Crow AJ, et al. Increased failure rates after

anterior cruciate ligament reconstruction with soft-tissue autograft-

allograft hybrid grafts. Arthroscopy. 2015;31:2342-2351.

7. Calvo R, Figueroa D, Anastasiadis Z, et al. Septic arthritis in ACL

reconstruction surgery with hamstrings autografts. Eleven years of

experience. Knee. 2014;21:717-720.

8. Conte EJ, Hyatt AE, Gatt CJ, Dhawan A. Hamstring autograft size can

be predicted and is a potential risk factor for anterior cruciate liga-

ment reconstruction failure. Arthroscopy. 2014;30:882-890.

9. Crall TS, Bishop JA, Guttman D, Kocher M, Bozic K, Lubowitz JH.

Cost-effectiveness analysis of primary arthroscopic stabilization ver-

sus nonoperative treatment for first-time anterior glenohumeral dislo-

cations. Arthroscopy. 2012;28:1755-1765.

10. Faltstrom A, Hagglund M, Magnusson H, Forssblad M, Kvist J. Pre-

dictors for additional anterior cruciate ligament reconstruction: data

from the Swedish national ACL register. Knee Surg Sports Traumatol

Arthrosc. 2016;24:885-894.

11. Faunø P, Rahr-Wagner L, Lind M. Risk of revision after anterior cru-

ciate ligament reconstruction is higher among adolescents. Orthop J

Sports Med. 2014;2:2325967114552405.

12. Genuario JW, Faucett SC, Boublik M, Schlegel TF. A cost-effective-

ness analysis comparing 3 anterior cruciate ligament graft types:

bone-patellar tendon-bone autograft, hamstring autograft, and allo-

graft. Am J Sports Med. 2012;40:307-314.

13. Gottlob CA, Baker CL Jr. Anterior cruciate ligament reconstruction:

socioeconomic issues and cost effectiveness. Am J Orthop (Belle

Mead NJ). 2000;29:472-476.

14. Gottlob CA, Baker CL Jr, Pellissier JM, Colvin L. Cost effectiveness

of anterior cruciate ligament reconstruction in young adults. Clin

Orthop Relat Res. 1999;367:272-282.

15. Kamien PM, Hydrick JM, Replogle WH, Go LT, Barrett GR. Age, graft

size, and Tegner activity level as predictors of failure in anterior cru-

ciate ligament reconstruction with hamstring autograft. Am J Sports

Med. 2013;41:1808-1812.

16. Leroux T, Wasserstein D, Dwyer T, et al. The epidemiology of revision

anterior cruciate ligament reconstruction in Ontario, Canada. Am J

Sports Med. 2014;42:2666-2672.

17. Ma Y, Murawski CD, Rahnemai-Azar AA, Maldjian C, Lynch AD, Fu

FH. Graft maturity of the reconstructed anterior cruciate ligament 6

months postoperatively: a magnetic resonance imaging evaluation

of quadriceps tendon with bone block and hamstring tendon auto-

grafts. Knee Surg Sports Traumatol Arthrosc. 2015;23:661-668.

18. Magnussen RA, Lawrence TR, West RL, Toth AP, Taylor DC, Garrett

WE. Graft size and patient age are predictors of early revisions after

anterior cruciate ligament reconstruction with hamstring autograft.

Arthroscopy. 2012;28:526-531.

19. Marchand JB, Ruiz N, Coupry A, Bowen M, Robert H. Do graft diam-

eter or patient age influence the results of ACL reconstruction? Knee

Surg Sports Traumatol Arthrosc. 2016;24:2998-3004.

20. Mariscalco MW, Flanigan DC, Mitchell J, et al. The influence of ham-

string autograft size on patient-reported outcomes and risk of revi-

sion after anterior cruciate ligament reconstruction: a Multicenter

Orthopaedic Outcomes Network (MOON) cohort study. Arthroscopy.

2013;29:1948-1953.

21. Mayr HO, Stoehr A, Dietrich M, et al. Graft-dependent differences in

the ligamentization process of anterior cruciate ligament grafts in

a sheep trial. Knee Surg Sports Traumatol Arthrosc. 2012;20:

947-956.

22. Morgan MD, Salmon LJ, Waller A, Roe JP, Pinczewski LA. Fifteen-

year survival of endoscopic anterior cruciate ligament reconstruction

in patients aged 18 years and younger. Am J Sports Med. 2016;

44:384-392.

23. Noyes FR, Barber-Westin SD. Revision anterior cruciate surgery with

use of bone-patellar tendon-bone autogenous grafts. J Bone Joint

Surg Am. 2001;83:1131-1143.

24. Nwachukwu BU, McFeely ED, Nasreddine A, et al. Arthrofibrosis

after anterior cruciate ligament reconstruction in children and adoles-

cents. J Pediatr Orthop. 2011;31:811-817.

25. Reid D, Leigh W, Wilkins S, Willis R, Twaddle B, Walsh S. A 10-year

retrospective review of functional outcomes of adolescent anterior

cruciate ligament reconstruction [published online ahead of print July

9, 2015]. J Pediatr Orthop. doi:10.1097/BPO.000000000000 0594.

26. Reinhardt KR, Hammoud S, Bowers AL, Umunna BP, Cordasco FA.

Revision ACL reconstruction in skeletally mature athletes younger

than 18 years. Clin Orthop Relat Res. 2012;470:835-842.

AJSM Vol. XX, No. X, XXXX Hybrid Hamstring Grafts in Younger Patients 7

27. Saltzman BM, Cvetanovich GL, Nwachukwu BU, Mall NA, Bush-

Joseph CA, Bach BR Jr. Economic analyses in anterior cruciate lig-

ament reconstruction: A qualitative and systematic review. Am J

Sports Med. 2016;44:1329-1335.

28. Shelbourne KD, Gray T, Haro M. Incidence of subsequent injury to

either knee within 5 years after anterior cruciate ligament reconstruc-

tion with patellar tendon autograft. Am J Sports Med. 2009;37:

246-251.

29. Spindler KP, Huston LJ, Wright RW, et al. The prognosis and predic-

tors of sports function and activity at minimum 6 years after anterior

cruciate ligament reconstruction: a population cohort study. Am J

Sports Med. 2011;39:348-359.

30. Spragg L, Chen J, Mirzayan R, Love R, Maletis G. The effect of autol-

ogous hamstring graft diameter on the likelihood for revision of ante-

rior cruciate ligament reconstruction. Am J Sports Med. 2016;44:

1475-1481.

31. Tomlinson DP, Altchek DW, Davila J, Cordasco FA. A modified tech-

nique of arthroscopically assisted AC joint reconstruction and prelim-

inary results. Clin Orthop Relat Res. 2008;466:639-645.

32. van Eck CF, Schkrohowsky JG, Working ZM, Irrgang JJ, Fu FH. Pro-

spective analysis of failure rate and predictors of failure after ana-

tomic anterior cruciate ligament reconstruction with allograft. Am J

Sports Med. 2012;40:800-807.

33. Webster KE, Feller JA, Leigh WB, Richmond AK. Younger patients

are at increased risk for graft rupture and contralateral injury after

anterior cruciate ligament reconstruction. Am J Sports Med.

2014;42:641-647.

34. Westermann RW, Wolf BR, Elkins JM. Effect of ACL reconstruction

graft size on simulated Lachman testing: a finite element analysis.

Iowa Orthop J. 2013;33:70-77.

35. Wright R, Spindler K, Huston L, et al. Revision ACL reconstruction

outcomes: MOON cohort. J Knee Surg. 2011;24:289-294.

For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.

8 Jacobs et al The American Journal of Sports Medicine