Systematic Review

Bone Grafts and Bone Substitutes for Opening-Wedge Osteotomies of the Knee: A Systematic Review Nicholas J. Lash, F.R.A.C.S., Julian A. Feller, F.R.A.C.S., Lachlan M. Batty, M.B.B.S., Jason Wasiak, M.P.H., and Anneka K. Richmond, B.Sc.

Purpose: To establish the rate of use of various void fillers in the setting of opening-wedge osteotomy around the knee, the types of fixation used, and the rates of delayed union or nonunion related to these variables. In addition, this review addressed short-term to midterm outcomes and complication rates associated with such procedures. Methods: The electronic databases Medline, Embase, and PubMed were searched using the methodology for systematic review as recommended by the Cochrane Collaboration. The search terms used were as follows: knee, osteotomy, knee joint, bone grafting, opening osteotomy, opening wedge, tibial osteotomy, femoral osteotomy, and bone substitute. We screened 1,383 articles and applied exclusion criteria. Fifty-six articles were included. Results: We included 3,033 cases of osteotomy in 2,910 patients. The mean age of patients was 50 years, with a mean follow-up period of 42 months. Male patients comprised 52% of patients. The mean alignment change was 10.8 , shifting the mechanical axis to 5.1 valgus. Delayed union/nonunion rates were 2.6%, 4.6%, and 4.5% for autograft, allograft bone, and synthetic bone substitutes, respectively (P ¼ .03). Delayed union/nonunion rates were significantly lower for autograft compared with allograft (P ¼ .03) and for autograft and allograft compared with synthetic bone substitutes (P < .0001). Non-locking plates (n ¼ 2,148) had a rate of delayed union/nonunion of 3.7% and a mean loss of correction over time of 0.5 . Locking plates (n ¼ 681) had a rate of delayed union/nonunion of 2.6% and a loss of correction of 2.3 . All mean knee outcome scores improved, and an overall complication rate of 14% was found. Conclusions: Opening-wedge osteotomy had good shortterm to midterm outcomes with acceptable complication rates. The lowest rates of delayed union/nonunion were in autograft boneefilled osteotomies. Plate type does not appear to affect osteotomy union or loss of correction. Level of Evidence: Level IV, systematic review of Level I to IV studies.

See commentary on page 731


roximal tibial and distal femoral osteotomies are both widely recognized treatment methods for unicompartmental osteoarthritis of the knee joint. After reports of closing-wedge osteotomy of the proximal tibia by Jackson and Waugh1 in 1961 and Coventry 2 in 1965, closing-wedge techniques were refined and widely

From the OrthoSport Victoria Research Unit, Deakin University, and Epworth Healthcare, Melbourne, Australia. The authors report the following potential conflict of interest or source of funding: J.A.F. receives support from Stryker, Tornier, and Smith & Nephew. Received July 12, 2014; accepted September 3, 2014. Address correspondence to Anneka K. Richmond, B.Sc., OrthoSport Victoria, Level 5, 89 Bridge Road, Richmond, VIC 3121, Australia. E-mail: [email protected] Ó 2015 by the Arthroscopy Association of North America 0749-8063/14606/$36.00


used. In 1987, Hernigou et al.3 reported their experience with an opening-wedge technique for proximal tibial osteotomy. In recent years opening-wedge techniques have, arguably, overtaken closing-wedge techniques in terms of popularity. Using an opening-wedge technique creates a void within the patient’s bone. Modern-day orthopaedic surgeons have a wide range of options available to deal with the gap. Historically, autograft bone, usually from the iliac crest, has been used. Autograft bone is readily available, provides an excellent framework for bony ingrowth, and can provide structural support if cortical bone is used. It also provides osteoprogenitor cells, bone morphogenetic proteins, and growth factors that encourage new bone formation. The potential negative aspects of autograft bone relate to harvest-site morbidity. With the use of bone donation and allograft banks, allograft bone has

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 31, No 4 (April), 2015: pp 720-730


become more readily available. However, processing of the donor bone renders it less structurally sound than autograft bone. In addition, despite rigorous treatment, the potential for disease transmission remains. Bone substitute fillers have been used with increasing frequency. They seek to address the limitations of autograft and allograft bone. Largely calcium and phosphate based, they seek to replicate the porosity of cancellous bone to allow infiltration of capillaries, perivascular tissue, and osteoprogenitor cells to promote new bone formation. In addition to providing a scaffold for cellular migration, depending on their composition, bone substitutes may provide temporary structural support. They have differing compositions of phosphates, carbonates, and sulfates, and this imparts a varied rate of “resorption” as they all ultimately undergo replacement with new host bone as a part of natural bone turnover. Non-calcium and -phosphateebased void fillers are also available. These are predominantly collagen-based matrices, which provide frameworks for host cells to migrate along and into, allowing for gradual bone formation. In addition to collagenous proteins, the matrices also contain non-collagenous proteins such as bone morphogenetic proteins, which encourage osteoinduction through cellular signaling. The science behind bone substitute use is complex and is not within the scope of this review. However, for a comprehensive review of the topic, we refer readers to the article by Moore et al.4 This systematic review set out to establish the rate of use of various void fillers in the setting of openingwedge osteotomy around the knee, the types of fixation used, and the rates of delayed union or nonunion related to these variables. In addition, it addressed short-term to midterm outcomes and complication rates associated with such procedures.

Methods Data Sources and Search Strategy The method for this review was derived from the systematic review methodology adopted by the Cochrane Collaboration.5 A structured literature search was performed in Medline from 1946, Embase from 1980 onward, PubMed, and The Cochrane Library up to issue 1, 2014, using the following keywords: knee, osteotomy, knee joint, bone grafting, opening osteotomy, opening wedge, tibial osteotomy, femoral osteotomy, and bone substitute. The predefined search strategy was designed for maximal retrieval using Medical Subject Headings and free text searching. The thesaurus vocabulary of each database was used to adapt the search terms. The selected time frame was chosen to take into account the development of knee osteotomy techniques and biomaterial/bone substitute availability. In addition to the automated search strategies, reference lists of related journal articles, key journals, and existing reviews were


manually searched for additional studies. No attempt was made to locate unpublished material or to contact authors of unpublished studies. Study Selection Criteria and Procedures All published peer-reviewed studies, including randomized controlled trials, non-randomized comparative studies, cohort studies, and case series, that investigated the use of tibial or femoral osteotomy in the treatment of unicompartmental tibiofemoral osteoarthritis of the knee were considered for inclusion. We excluded in vivo studies, animal studies, noneEnglish-language articles, series with fewer than 5 patients, and case reports. Unpublished manuscripts, narrative or systematic reviews, guidelines, commentaries, and other descriptive articles were also excluded. Records retrieved by the initial search were scanned by a single author (J.W.) to exclude obviously irrelevant studies. Two authors (N.J.L., L.M.B.) subsequently screened titles and abstracts with reference to the exclusion criteria. This generated a list of references for which full-text articles were retrieved. These articles were reviewed independently by 2 authors (N.J.L., L.M.B.), and articles based on closing-wedge osteotomy, dome osteotomy, non-metaphyseal location of osteotomy, or callotasis distraction techniques were also excluded. In all instances, differences of opinion were resolved by discussion between the authors. The references cited by the included articles were then reviewed to capture articles that were undetected by the initial database search. Data Extraction Two authors (N.J.L., L.M.B.) independently applied a standardized data extraction form designed specifically to collect data of interest. Extracted data included details on study designs, patient demographic data, and the following variables: limb alignment measurements, filler type (if used), fixation type (if used), length of time to weight bearing, time to union of osteotomy, incidence of delayed union and nonunion, clinical follow-up data, knee scores (Lysholm, International Knee Documentation Committee, Tegner, and so on), and number and nature of complications. Assessment of Methodologic Quality We assessed the methodologic quality of the articles using the Methodological Index for Non-Randomized Studies (MINORS) including additional criteria for comparative studies, as described by Slim et al.6 This meant that 8 criteria were used to assess noncomparative studies and 12 criteria were used for comparative studies, with each criterion scored using a 3-point scale from 0 to 2, with 0 representing not reported, 1 representing reported but inadequate, and 2 representing reported and adequate. No weighting of criteria was applied. The ideal overall score was 16 for a non-comparative study and 24 for a comparative study.



The assessment was performed independently by individual authors, and differences of opinion were resolved by discussion between the authors.

PubMed. After initial abstract review, 104 articles were identified for potential inclusion. After further review, a total of 32 full-text articles were retrieved for analysis. Reference list review of these 32 articles generated a supplementary list of a further 141 articles that could potentially be included. These additional articles were sourced and reviewed. Fifty-six were eligible for inclusion. A total of 88 articles were therefore submitted for data extraction. However, 32 articles did not specify whether a void filler was used or, if specified, did not give sufficient detail to allow appropriate data extraction. Thus 56 studies met the inclusion criteria with sufficient detail to allow adequate data extraction (Fig 1).7-62 Of the 56 included studies, 30 were prospective and 26 were retrospective. Forty-three were case studies, and 13 were comparative cohort studies. Of the latter, only 1 was a randomized controlled trial. Application of the MINORS score to the included studies rendered a mean score of 10 of 16 for non-comparative studies

Statistical Analysis Summative data are presented for categorical variables. Meta-analysis was not possible because of the heterogeneity of the included studies, low numbers of randomized trials, and inconsistent outcome reporting. Contingency tables were used to compare groups based on fixation device or type of filler for the variables of delayed union or nonunion and complications.

Results Search Results Our database search retrieved 1,383 abstracts: 655 from Medline (1946 to November 2013), 615 from Embase (1980 to week 49 of 2013), and 113 from

Records idenfied through database searching

Addional records idenfied through other sources

(n = 1383)

(n = 0)

Records aer duplicates removed

(n = 432)

Records screened

Records excluded, not applicable to knee osteotomy

(n = 432)

(n = 328) Full-text arcles assessed for eligibility

Full-text arcles excluded, Non English, non opening wedge, skeletal immaturity, biomechanical or animal studies, small case series

(n = 104)

(n = 72) Reference List review of potenally eligible arcles

Studies included in qualitave synthesis

(n = 32)

(n = 141) Full text arcles eligible for inclusion

(n = 24)

Studies included in quantave synthesis

(n = 56)

Fig 1. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) flow diagram illustrating search process.


(median, 10) and 18.5 of 24 for comparative studies (median, 19). Demographic Data A total of 3,248 knees in 3,121 patients were identified. After excluding 211 cases (6.8%) that were lost to follow-up, data for 3,033 knees in 2,910 patients were deemed adequate for inclusion. When stated, 52% of patients (1,251) were male patients and 48% (1,152) were female patients. The mean age of the patients included was 50 years (range, 27 to 69 years; median, 49 years). The mean follow-up period was 42 months (range, 3 to 312 months; median, 24 months). Osteotomy Details Of the 3,033 included knees, 2,926 underwent medial opening-wedge tibial osteotomy and a further 55 underwent lateral opening-wedge tibial osteotomy. Fiftytwo knees underwent lateral opening-wedge distal femoral osteotomy. Thirty-five studies evaluated preoperative and postoperative limb alignment. Thirty-four studies analyzed medial compartment disease and varus deformity correction. The mean preoperative mechanical femorotibial angle was 174.3 (range, 162 to 180 ), or 5.7 varus. The mean postoperative femorotibial angle was 185.1 (range, 180.7 to 190.3 ), or 5.1 valgus. This equates to a mean change in limb alignment of 10.8 . One study investigated lateral knee disease and valgus

Fig 2. Number of cases for each filler type.


deformity and alignment preoperatively and postoperatively. The mean preoperative femorotibial angle was 185.3 , or 5.3 valgus, and after surgery, the mean angle was 178.7 , shifting the alignment to a mean of 1.3 varus. Of the 3,033 osteotomies reported, 2,950 (97%) used some form of fixation. Of the 2,877 medial tibial osteotomies, 2,148 (75%) were stabilized with nonlocking plates or specific high tibial osteotomy plates with non-locking screws and 681 (24%) were stabilized with plates using locking screws. One series of 48 cases used external fixation to maintain an acute intraoperative correction. The 21 cases of lateral tibial opening-wedge osteotomy used non-locking plates for fixation. Lateral femoral osteotomy was performed in 52 knees. In 14 cases fixation was achieved with a locking plate, and in the remaining 38 cases, non-locking plates were used. Union of Osteotomies Although all 56 studies provided details on the filler used (3,033 cases), only 17 studies (1,175 cases) reported time to union of the osteotomy. The types of fillers used in these 17 studies were autograft bone, allograft bone, tricalcium phosphate, calcium phosphate, hydroxyapatite, hydroxyapatiteetricalcium phosphate, coralline wedges, and bioglass. There were a small number of cases in which a combination of filler types was used,



namely autograft combined with a synthetic filler. In addition, there was a group of osteotomies for which no filler was used. The number of cases in which each option was used is shown in Figure 2. The reported mean time to union varied greatly. Autograft bone was associated with the shortest time to union, with a mean of 3.1 months in 155 cases, compared with allograft bone, with a mean time to union of 3.8 months in 468 cases. Calcium phosphate, no filler, and tricalcium phosphate had mean times to union of 25 months, 9 months, and 10.6 months in 219 cases, 209 cases, and 122 cases, respectively. Bioglass had a mean time to union of 4 months, but this was only in 2 cases. Among the 3,033 cases, there were 60 cases of delayed union (2%) and 43 cases of nonunion (1.4%) (Table 1). The 2 most frequently used fillers were autograft and allograft bone. Allograft bone (895 knees) had a delayed union/nonunion rate of 4.6% compared with autograft bone (787 knees) with a rate of 2.6% (P ¼ .03). The next largest group was no filler (526 cases), which had a delayed union/nonunion rate of 1.4%. Although this was significantly lower than that with allograft bone (P < .005), it was not significantly different from autograft. It should be noted that the mean alignment changes for autograft, allograft, and no filler were 9.4 , 12.1 , and 7.6 , respectively. Bone substitute fillers had a varying rate of delayed union/nonunion. Bioglass, coralline wedges, and combined fillers had relatively high rates of delayed union/ nonunion of 21.6%, 14.8%, and 14.3%, respectively. Other bone substitutes had low rates of delayed union/ nonunion: tricalcium phosphate, 3.2%; calcium phosphate, 1.4%; hydroxyapatiteetricalcium phosphate, 1%; and hydroxyapatite, 0%. When all bone substitutes were grouped together (n ¼ 825), there were 37 cases (4.5%) of delayed union/nonunion compared with 59 of 1,682 knees (3.5%) in which either autograft or allograft bone was used (P < .0001).

Union rates for medial opening-wedge tibial osteotomy were also considered regarding both the type of fixation and the type of filler used. Because of the wide variety of specific fixation methods used in the studies, they were grouped into either non-locking plates (Table 2) or locking plates (Table 3). Overall, the rates of delayed union/nonunion were 3.7% for non-locking plates and 2.6% for locking plates. This difference was not significant. External fixation was used in 1 series of 48 knees in association with autograft bone. Delayed union was reported in 3 cases (6.3%). No series reported using autograft bone with locking plates. For those series in which allograft bone was used, the delayed union/nonunion rate was 3.4% when non-locking plates were used and 12.3% when locking plates were used (P < .001). However, there were only 89 knees with the allograftelocking plate combination compared with 698 with the allograftenon-locking plate combination, and all 11 delayed union/nonunion cases came from 1 study of 50 cases. When no void filler was used, there was no difference in the rates of delayed union/nonunion between locking plates (1.4%) and non-locking plates (0%). Lateral distal femoral osteotomies always used fixation with either a locking or non-locking plate. Forty-five knees had an osteotomy secured with nonlocking plates, with either calcium phosphate cement or no filler. There was 1 nonunion (2.2%) in the group that had no filler inserted. In 7 knees with a locking plate and no filler, 2 delayed unions (28.4%) were reported. Loss of Correction Fifteen studies of medial tibial osteotomy recorded limb alignment at subsequent follow-up. In all cases these measurements were made at final follow-up in addition to an early postoperative measurement. This allowed for quantification of changes in alignment over the intervening period. The mean loss of

Table 1. Rates of Delayed Union/Nonunion by Filler Type Type of Filler Allograft bone Autograft bone Bioglass Tricalcium phosphate Calcium phosphate Hydroxyapatiteetricalcium phosphate Hydroxyapatite Coralline wedge Combined fillers No filler Total NOTE. Data presented as number (percent).

Total Cases 787 895 51 384 219 103 13 27 28 526 3,033

Cases of Delayed Union 23 (2.9) 19 (2.1) 0 (0) 7 (1.6) 2 (0.9) 0 (0) 0 (0) 3 (11.1) 2 (7.1) 4 (0.8) 60 (2)

Cases of Nonunion 13 (1.7) 4 (0.5) 11 (21.6) 7 (1.6) 1 (0.5) 1 (1) 0 (0) 1 (3.7) 2 (7.1) 3 (0.6) 43 (1.4)

Total Cases of Delayed Union/Nonunion 36 (4.6) 23 (2.6) 11 (21.6) 14 (3.2) 3 (1.4) 1 (1) 0 (0) 4 (14.8) 4 (14.3) 7 (1.3) 103 (3.4)



Table 2. Rates of Delayed Union/Nonunion for Medial Opening-Wedge Tibial Osteotomy With Non-Locking Plate Fixation Type of Filler Allograft bone Autograft bone Bioglass Tricalcium phosphate Calcium phosphate Coralline wedge Hydroxyapatite Combined fillers No filler Total

Cases of Delayed Union 13 (2.0) 15 (2.0) 0 (0) 7 (2.6) 2 (1) 3 (11) 0 (0) 2 (7.1) 0 (0) 42 (2.0)

Total Cases 698 670 51 274 203 27 113 28 84 2,148

Cases of Nonunion 11 (1.7) 4 (0.5) 11 (21.6) 7 (2.6) 1 (0.5) 1 (4) 0 (0) 2 (7.1) 0 (0) 37 (1.7)

Total Cases of Delayed Union/Nonunion 24 (3.6) 19 (2.6) 11 (21.6) 14 (5.2) 3 (1.5) 4 (14.8) 0 (0) 4 (14.2) 0 (0) 79 (3.7)

NOTE. Data presented as number (percent).

correction was 1.3 in 814 cases. Although the number of knees was small, hydroxyapatiteetricalcium phosphate was associated with a mean 4 loss of correction (Fig 3). When we considered loss of correction in terms of fixation methods, non-locking plates (n ¼ 490) had a mean loss of correction of 0.5 whereas locking plates (n ¼ 200) had a mean loss of correction of 2.3 . It should be noted, however, that the cases in which hydroxyapatiteetricalcium phosphate was used as a filler are included in the locking plate group and may therefore skew the results. External fixation (n ¼ 48) was associated with a mean 0.2 loss of correction, and in 1 series in which no fixation was used (n ¼ 76), there was a mean loss of 3.5 at final follow-up. Outcome Scores Because of the heterogeneous nature of studies and data collection, a broad variety of outcome scores were used. All groups of differing filler type showed improvement in postoperative scores compared with preoperative scores. Table 4 shows the mean preoperative and postoperative outcome scores. Complications Complication rates were reported for 2,821 cases. The overall complication rate was 14%. If subsequent removal of hardware and irritation related to hardware were included, the rate was 23%. Nonunion and delayed union have been separately analyzed in the preceding paragraphs. Fixation failure

occurred in 62 cases. Fifty-nine failures occurred in 2,148 cases using non-locking plates (2.7%) compared with 3 failures in 681 cases using locking plates (0.4%, P ¼ .0001). Autograft harvest morbidity was documented in 15 cases (1.9% of the 782 cases using autograft filler). Superficial wound infections, though infrequent, were more frequent with hydroxyapatiteetricalcium phosphate (6.2%) than with other fillers (0.6%). However, all 7 infections in the hydroxyapatiteetricalcium phosphate group came from 1 study (24 cases), with none being reported in the other 4 studies (89 cases). Three common peroneal nerve injuries were reported in 55 lateral opening-wedge tibial osteotomies (5.5%) compared with 12 injuries in 2,926 medial openingwedge tibial osteotomies (0.4%, P < .005). In all cases of peroneal nerve injury, recovery was reported. Other variables that were reported included osteotomy propagation at the time of surgery (2.5%), fixation causing soft-tissue irritation (4.3%), removal of fixation (5%), deep infection (1%), and conversion to total knee replacement (2.8%).

Discussion This systematic review examined the influence of both the void filler and the type of fixation in the setting of opening-wedge osteotomy procedures around the knee. Specifically, it examined their influence on delayed union, nonunion, loss of correction, clinical outcomes, and complications. Most of the studies

Table 3. Rates of Delayed Union/Nonunion for Medial Opening-Wedge Tibial Osteotomy With Locking Plate Fixation Type of Filler Allograft bone Tricalcium phosphate Hydroxyapatiteetricalcium phosphate No filler Total NOTE. Data presented as number (percent).

Total Cases 89 130 43 419 681

Cases of Delayed Union 9 (10.1) 0 (0) 0 (0) 4 (0.9) 13 (1.9)

Cases of Nonunion 2 (2.2) 0 (0) 1 (2.3) 2 (0.5) 5 (0.7)

Total Cases of Delayed Union/Nonunion 11 (12.3) 0 (0) 1 (2.3) 6 (1.4) 18 (2.6)



Fig 3. Loss of correction over time for varying fillers for all fixation types.

included in this review related to tibial osteotomies, mainly medial. The short-term outcomes of opening-wedge osteotomy were generally reported to be good, whichever knee score was used, with acceptable overall rates of complications. When reported, the postoperative alignment after tibial osteotomy was usually within the 3 to 6 valgus values recommended by Hernigou et al.3 Because the main aim of any bone graft or substitute is to facilitate bone union, the primary outcome measured in this review was the rate of delayed union/nonunion. When we reviewed delayed union and nonunion by filler type, the 2 largest groups were allograft and autograft bone filler, with rates of delayed union/ nonunion of 4.6% and 2.6%, respectively. Bone graft substitutes as a group had an overall rate of delayed union/nonunion of 4.5% compared with 3.5% for allograft and autograft combined. Both of these differences were statistically significant. Overall, autograft bone had superior union results compared with allograft, and likewise, allograft was superior to bone graft substitutes. However, it should be noted that several bone graft substitutes stood out with higher rates of delayed union and nonunion, with bioglass, coralline wedges, and combined fillers having rates of 21.6%, 14.8%, and 14.3%, respectively. By comparison, tricalcium phosphate (3.2%), calcium phosphate (1.4%), hydroxyapatiteetricalcium phosphate (1%), hydroxyapatite (0%), and cases in which no filler was used (1.4%) all had lower rates of delayed union/nonunion. Another factor that can influence union of an osteotomy is the type of fixation used. Although the rate of delayed union/nonunion was 3.7% for cases

with non-locking plates and 2.6% for locking plates, the difference was not statistically significant. Modern designs of locking plates provide increased stability, such that use of a filler may not be required. In cases in which locking plates were used without a filler, the delayed union/nonunion rate was 1.4%. Surprisingly, when allograft bone filler was added, the delayed union/nonunion rate was 12.3%, but this finding relates to 1 study only. Furthermore, few studies documented the defect size in terms of medial opening distance, and it is unknown whether the decision to insert a filler with a locking plate was made because of wider osteotomy gaps. This may be a source of bias. Another outcome that can potentially be affected by fixation and delayed union is loss of correction during the time over which the osteotomy unites. The mean loss of correction was 1.3 in 15 studies. Studies in which non-locking plates were used reported a mean loss of correction of 0.5 , whereas for studies using locking plates, a mean of 2.3 of correction was lost. This result seems counterintuitive given the purported benefit of locking plate technology being to provide enhanced stability. However, when loss of correction was analyzed for different fillers, a high loss of correction was noted (4 ) when hydroxyapatiteetricalcium phosphate was used, This small cohort of 43 patients accounted for 21.5% of the 200 cases in which locking plates were used. If this cohort was excluded, the mean loss of correction was 0.5 , the same as that for nonlocking plate cases. Fixation failure occurred in 59 of 2,148 cases (2.8%) in which non-locking plates were used. By comparison, there were 3 failures among 681 cases (0.4%) in which

FILLERS FOR OPENING-WEDGE OSTEOTOMY OF KNEE Table 4. Functional and Knee Specific Outcome Scores by Filler Type Mean Score Scale Allograft (n ¼ 224) Cincinnati Knee Rating Scale (n ¼ 33) HSS knee score (n ¼ 86) KSS (n ¼ 95) KSFS (n ¼ 95) Lysholm Knee Scoring Scale (n ¼ 105) Autograft (n ¼ 239) HSS knee score (n ¼ 94) JOA score (n ¼ 43) Lysholm Knee Scoring Scale (n ¼ 76) Tegner Activity Level Scale (n ¼ 45) Tricalcium phosphate (n ¼ 247) KSS (n ¼ 66) KSFS (n ¼ 20) HSS knee score (n ¼ 25) KOOS (n ¼ 20) Lysholm Knee Scoring Scale (n ¼ 144) Tegner Activity Level Scale (n ¼ 20) WOMAC (n ¼ 41) No filler (n ¼ 257) HSS knee score (n ¼ 43) KSS (n ¼ 65) IKDC Subjective Knee Form (n ¼ 135) Lysholm Knee Scoring Scale (n ¼ 170) Tegner Activity Level Scale (n ¼ 43) Calcium phosphate (n ¼ 16) KSS Combined filler (n ¼ 20) Lysholm Knee Scoring Scale (n ¼ 20) HSS knee score (n ¼ 20) Hydroxyapatite (n ¼ 24) Lysholm Knee Scoring Scale





59.5 75 95 57

83 91 99 79

65 58 53

85 85 83


35 67 62 193 63 1.9


70 95 96 300 84 3.1



75 49 42

88 74 68













HSS, Hospital for Special Surgery; IKDC, International Knee Documentation Committee; JOA, Japanese Orthopaedic Association; KOOS, Knee Injury and Osteoarthritis Outcome Score; KSFS, Knee Society Function Score; KSS, Knee Society Score; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.

locking plates were used. Fixation failure was 6.75 times more likely in non-locking plate cases. Despite this finding, loss of correction of alignment within the non-locking plate group was no greater than that in locking plate cases, and union rates were similar for cases using either plate type. This may be explained by the osteotomy healing even while fixation may be failing or under duress. One of the main issues in analyzing the rates of delayed union and nonunion was how studies reported


methods of measurement, as well as the definitions of delayed union and nonunion. When autograft bone, allograft bone, or no filler was used, standard descriptions of union were used. The definition of delayed union and nonunion is only separated by a time frame and is subject to interpretation. This varied from 3 to 12 months within the included studies. Another difficulty in defining union of an osteotomy in which a bone substitute has been used is that there are 2 very different adjoining interfaces. Radiologically, this is represented by normal bone adjacent to very radiopaque bone substitute. In an attempt to address this issue, van Hemert et al.59 modified the definition of fracture healing of McKibbin.63 The modified classification analyzes varying stages of incorporation of the bone graft substitute interface through remodeling and, finally, absence of graft and lack of visualization of signs of osteotomy. Because of the varying compositions, different bone substitutes are known to take many months and potentially years to fully remodel and finally disappear. This was manifested in the mean times to union of tricalcium phosphate and calcium phosphate filler cases being 10.5 months and 25 months, respectively. These inherent properties make defining union and bone substitute filler incorporation difficult and are potentially a source of inaccuracy in reporting time to union and rates of delayed union/nonunion. Complication rates were reported for 93% of cases. Overall, the rate of conversion to total knee replacement was 2.8%, with a mean follow-up period of 3.5 years. This appears to be lower than the 14% conversion rate at 2.5 years reported by Naudie et al.64 However, it should be noted that none of the articles included in our review looked specifically at survivorship. Superficial infection rates were highest in cases using hydroxyapatiteetricalcium phosphate (6.2%). A possible explanation for this is that certain fillers using calcium and phosphate apatite can have high osmotic loads, leading to fluid shifts toward an area in which they have been implanted. However, all of the reported superficial infections reported for hydroxyapatiteetricalcium phosphate were in 1 study (24 cases), and there were no superficial infections reported for hydroxyapatiteetricalcium phosphate in the other 4 studies (89 cases). Other fillers had an overall low rate of superficial infection (0.6%), including fillers that are high in calcium phosphates (tricalcium phosphate, calcium phosphate cement). Common peroneal nerve injuries were found in 15 cases. The rates of injury by location of osteotomy were 5.5% (3 of 55) for lateral tibial osteotomy and 0.4% (12 of 2,926) for medial tibial osteotomy. In the 3 cases in the lateral opening-wedge tibial osteotomy group, the injury was either traction neurapraxia or direct trauma to the nerve. Twelve cases of injury to the nerve occurred in association with medial opening-wedge tibial osteotomy.



The mechanism of injury in medially located surgery was not described in the studies in question. The strengths of this review are that more than 3,000 cases were able to be evaluated and that the mean follow-up period was 3.5 years, which is adequate to assess the outcome variables of primary interest (delayed union and nonunion, loss of fixation and correction, function, and complications). Because of inconsistent reporting of conversion to total knee replacement and the time from surgery to conversion, we were unable to calculate or analyze survivorship rates. Limitations A weakness of this review is that it had a high number of case series (43) and a low number of well-designed controlled comparative studies (1). By use of common level-of-evidence measures, most of the studies were Level IV. Using the MINORS methodologic assessment tool, we found that the mean scores for non-comparative and comparative studies were 10 and 19, respectively. This suggests a “fair” level of quality for the included studies. However, the studies varied widely in reporting methods, the level of detail in reporting, the numbers of cases, and the definitions of union of osteotomy. Finally, most of the studies were based on high tibial osteotomy (53), with few studies reporting on distal femoral varus osteotomy (3), which means that the findings of this review apply primarily to tibial osteotomy.

Conclusions Overall, knee score outcomes after opening-wedge osteotomy unicompartmental osteoarthritis of the knee are generally good in the short- to medium-term and complication rates are acceptable. Of currently used fillers for opening-wedge osteotomy gaps, autograft bone had higher union rates than allograft bone, which in turn had higher union rates than synthetic bone graft substitutes. Union rates of osteotomy were similar regardless of whether locking or non-locking plates were used. Loss of correction over time was minimal but, when seen, was higher in locking plate cases. However, the use of hydroxyapatiteetricalcium phosphate substitute with locking plates may make this finding spurious. Fixation failure is more common in cases using nonlocking plates, but union rates and loss of correction did not appear to be affected by this. A definition of union, delayed union, and nonunion for both bone and bone substitutes that is commonly agreed on should be sought to minimize inconsistent reporting and confusion in interpreting published data, and more uniform use of knee outcome scores would be helpful.

References 1. Jackson JP, Waugh W. Tibial osteotomy for osteoarthritis of the knee. J Bone Joint Surg Br 1961;43:746-751.

2. Coventry MB. Osteotomy of the upper portion of the tibia for degenerative arthritis of the knee. A preliminary report. J Bone Joint Surg Am 1965;47:984-990. 3. Hernigou P, Medevielle D, Debeyre J, Goutallier D. Proximal tibial osteotomy for osteoarthritis with varus deformity. A ten to thirteen-year follow-up study. J Bone Joint Surg Am 1987;69:332-354. 4. Moore WR, Graves SE, Bain GI. Synthetic bone graft substitutes. ANZ J Surg 2001;71:354-361. 5. Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins JPT, Green S, eds. Cochrane handbook for systematic reviews of interventions version 5.1.0. Oxford: Cochrane Collaboration, 2011. 6. Slim K, Nini E, Forestier D, Kwiatkowski F, Panis Y, Chipponi J. Methodological Index for Non-Randomized Studies (MINORS): Development and validation of a new instrument. ANZ J Surg 2003;73:712-716. 7. Aoki Y, Yasuda K, Mikami S, Ohmoto H, Majima T, Minami A. Inverted V-shaped high tibial osteotomy compared with closing-wedge high tibial osteotomy for osteoarthritis of the knee. Ten-year follow-up result. J Bone Joint Surg Br 2006;88:1336-1340. 8. Arthur A, LaPrade RF, Agel J. Proximal tibial opening wedge osteotomy as the initial treatment for chronic posterolateral corner deficiency in the varus knee: A prospective clinical study. Am J Sports Med 2007;35: 1844-1850. 9. Asik M, Sen C, Kilic B, Goksan SB, Ciftci F, Taser OF. High tibial osteotomy with Puddu plate for the treatment of varus gonarthrosis. Knee Surg Sports Traumatol Arthrosc 2006;14:948-954. 10. Brinkman J-M, Luites JW, Wymenga AB, van Heerwaarden RJ. Early full weight bearing is safe in openwedge high tibial osteotomy. Acta Orthop 2010;81: 193-198. 11. Brouwer RW, Bierma-Zeinstra SMA, van Koeveringe AJ, Verhaar JAN. Patellar height and the inclination of the tibial plateau after high tibial osteotomy. The open versus the closed-wedge technique. J Bone Joint Surg Br 2005;87: 1227-1232. 12. Brouwer RW, Bierma-Zeinstra SMA, van Raaij TM, Verhaar JAN. Osteotomy for medial compartment arthritis of the knee using a closing wedge or an opening wedge controlled by a Puddu plate. A one year randomised, controlled study. J Bone Joint Surg Br 2006;88: 1454-1459. 13. Dallari D. Enhanced tibial osteotomy healing with use of bone grafts supplemented with platelet gel or platelet gel and bone marrow stromal cells. J Bone Joint Surg Am 2007;89:2413-2420. 14. Das DHPW, Sijbesma T, Hoekstra HJ, van Leeuwen WM. Distal femoral opening-wedge osteotomy for lateral compartment osteoarthritis of the knee. Open Access Surg 2008;1:25-29. 15. DeMeo PJ, Johnson EM, Chiang PP, Flamm AM, Miller MC. Midterm follow-up of opening-wedge high tibial osteotomy. Am J Sports Med 2010;38:2077-2084. 16. Dewilde TR, Dauw J, Vandenneucker H, Bellemans J. Opening wedge distal femoral varus osteotomy using the Puddu plate and calcium phosphate bone cement. Knee Surg Sports Traumatol Arthrosc 2012;21:249-254.

FILLERS FOR OPENING-WEDGE OSTEOTOMY OF KNEE 17. Esenkaya I, Elmali N. Proximal tibia medial open-wedge osteotomy using plates with wedges: Early results in 58 cases. Knee Surg Sports Traumatol Arthrosc 2006;14:955-961. 18. Esenkaya I, Unay K. Proximal medial tibial biplanar retrotubercle open wedge osteotomy in medial knee arthrosis. Knee 2012;19:416-421. 19. Franco V, Cerullo G, Cipolla M, Gianni E, Puddu G. Open wedge high tibial osteotomy. Tech Knee Surg 2002;1: 43-53. 20. Gaasbeek RDA, Nicolaas L, Rijnberg WJ, van Loon CJM, van Kampen A. Correction accuracy and collateral laxity in open versus closed wedge high tibial osteotomy. A oneyear randomised controlled study. Int Orthop 2009;34: 201-207. 21. Gutierres M, Dias AG, Lopes MA, et al. Opening wedge high tibial osteotomy using 3D biomodelling Bonelike macroporous structures: Case report. J Mater Sci Mater Med 2007;18:2377-2382. 22. Hernigou P, Roussignol X, Flouzat-Lachaniette CH, Filippini P, Guissou I, Poignard A. Opening wedge tibial osteotomy for large varus deformity with Ceraver resorbable beta tricalcium phosphate wedges. Int Orthop 2009;34:191-199. 23. Hernigou P, Ma W. Open wedge tibial osteotomy with acrylic bone cement as bone substitute. Knee 2001;8: 103-110. 24. Hoell S, Suttmoeller J, Stoll V, Fuchs S, Gosheger G. The high tibial osteotomy, open versus closed wedge, a comparison of methods in 108 patients. Arch Orthop Trauma Surg 2005;125:638-643. 25. Jacobi M, Wahl P, Bouaicha S, Jakob RP, Gautier E. Distal femoral varus osteotomy: Problems associated with the lateral open-wedge technique. Arch Orthop Trauma Surg 2010;131:725-728. 26. Jung KA, Lee SC, Ahn NK, Hwang SH, Nam CH. Radiographic healing with hemispherical allogeneic femoral head bone grafting for opening-wedge high tibial osteotomy. Arthroscopy 2010;26:1617-1624. 27. Kim S-J, Koh Y-G, Chun Y-M, Kim Y-C, Park Y-S, Sung C-H. Medial opening wedge high-tibial osteotomy using a kinematic navigation system versus a conventional method: A 1-year retrospective, comparative study. Knee Surg Sports Traumatol Arthrosc 2008;17:128-134. 28. Koshino T, Murase T, Takagi T, Saito T. New bone formation around porous hydroxyapatite wedge implanted in opening wedge high tibial osteotomy in patients with osteoarthritis. Biomaterials 2001;22:1579-1582. 29. Koshino T, Murase T, Saito T. Medial opening-wedge high tibial osteotomy with use of porous hydroxyapatite to treat medial compartment osteoarthritis of the knee. J Bone Joint Surg Am 2003;85:78-85. 30. Kraal T, Mullender M, de Bruine JHD, et al. Resorbability of rigid beta-tricalcium phosphate wedges in open-wedge high tibial osteotomy: A retrospective radiological study. Knee 2008;15:201-205. 31. Kuremsky MA, Schaller TM, Hall CC, Roehr BA, Masonis JL. Comparison of autograft vs allograft in opening-wedge high tibial osteotomy. J Arthroplasty 2010;25:951-957. 32. LaPrade RF, Spiridonov SI, Nystrom LM, Jansson KS. Prospective outcomes of young and middle-aged adults


















with medial compartment osteoarthritis treated with a proximal tibial opening wedge osteotomy. Arthroscopy 2012;28:354-364. Lee D-H, Nha K-W, Park S-J, Han S-B. Preoperative and postoperative comparisons of navigation and radiologic limb alignment measurements after high tibial osteotomy. Arthroscopy 2012;28:1842-1850. Lobenhoffer P, De Simoni C, Staubli AE. Open-wedge high-tibial osteotomy with rigid plate fixation. Tech Knee Surg 2002;1:93-105. Marmotti A, Castoldi F, Rossi R, et al. Bone marrowderived cell mobilization by G-CSF to enhance osseointegration of bone substitute in high tibial osteotomy. Knee Surg Sports Traumatol Arthrosc 2012;21: 237-248. Marti CB, Gautier E, Wachtl SW, Jakob RP. Accuracy of frontal and sagittal plane correction in open-wedge high tibial osteotomy. Arthroscopy 2004;20:366-372. Marti RK, Verhagen RAW, Kerkhoffs GMMJ, Moojen TM. Proximal tibial varus osteotomy. J Bone Joint Surg Am 2001;83:164-170. Marti RK, Kerkhoffs GMMJ, Rademakers MV. Correction of lateral tibial plateau depression and valgus malunion of the proximal tibia. Orthop Traumatol 2007;19:101-113. Miller BS, Downie B, McDonough EB, Wojtys EM. Complications after medial opening wedge high tibial osteotomy. Arthroscopy 2009;25:639-646. Miller BS, Joseph TA, Barry EM, Rich VJ, Sterett WI. Patient satisfaction after medial opening high tibial osteotomy and microfracture. J Knee Surg 2007;20:129-133. Naudie DDR. Opening wedge high tibial osteotomy for symptomatic hyperextension-varus thrust. Am J Sports Med 2004;32:60-70. Nelissen EM, Langelaan EJ, Nelissen RGHH. Stability of medial opening wedge high tibial osteotomy: A failure analysis. Int Orthop 2009;34:217-223. Niemeyer P, Koestler W, Kaehny C, et al. Two-year results of open-wedge high tibial osteotomy with fixation by medial plate fixator for medial compartment arthritis with varus malalignment of the knee. Arthroscopy 2008;24: 796-804. Noyes FR. Opening wedge high tibial osteotomy: An operative technique and rehabilitation program to decrease complications and promote early union and function. Am J Sports Med 2006;34:1262-1273. Ribeiro CH, Severino NR, de Paula Leite Cury R, et al. A new fixation material for open-wedge tibial osteotomy for genu varum. Knee 2009;16:366-370. Rouvillain JL, Lavallé F, Pascal-Mousselard H, Catonné Y, Daculsi G. Clinical, radiological and histological evaluation of biphasic calcium phosphate bioceramic wedges filling medial high tibial valgisation osteotomies. Knee 2009;16:392-397. Salzmann GM, Ahrens P, Naal FD, et al. Sporting activity after high tibial osteotomy for the treatment of medial compartment knee osteoarthritis. Am J Sports Med 2009;37:312-318. Megied WSA, Mahran MA, Thakeb MF, Abouelela AAKH, Elbatrawy Y. The new ‘dual osteotomy’: Combined open wedge and tibial tuberosity anteriorisation osteotomies. Int Orthop 2009;34:231-237.



49. Santic V, Tudor A, Sestan B, Legovic D, Sirola L, Rakovac I. Bone allograft provides bone healing in the medial opening high tibial osteotomy. Int Orthop 2009;34: 225-229. 50. Saragaglia D, Blaysat M, Inman D, Mercier N. Outcome of opening wedge high tibial osteotomy augmented with a Biosorb wedge and fixed with a plate and screws in 124 patients with a mean of ten years follow-up. Int Orthop 2010;35:1151-1156. 51. Saragaglia D, Roberts J. Navigated osteotomies around the knee in 170 patients with osteoarthritis secondary to genu varum. Orthopedics 2005;28:s1269-s1274 (suppl). 52. Schröter S, Gonser CE, Konstantinidis L, Helwig P, Albrecht D. High complication rate after biplanar open wedge high tibial osteotomy stabilized with a new spacer plate (Position HTO plate) without bone substitute. Arthroscopy 2011;27:644-652. 53. Schröter S, Mueller J, van Heerwaarden R, Lobenhoffer P, Stöckle U, Albrecht D. Return to work and clinical outcome after open wedge HTO. Knee Surg Sports Traumatol Arthrosc 2012;21:213-219. 54. Spahn G, Kirschbaum S, Kahl E. Factors that influence high tibial osteotomy results in patients with medial gonarthritis: A score to predict the results. Osteoarthritis Cartilage 2006;14:190-195. 55. Staubli AE, Simoni CD, Babst R, Lobenhoffer P. TomoFix: A new LCP-concept for open wedge osteotomy of the medial proximal tibiadEarly results in 92 cases. Injury 2003;34:B55-B62 (suppl 2). 56. Staubli AE, Jacob HAC. Evolution of open-wedge hightibial osteotomy: Experience with a special angular stable device for internal fixation without interposition material. Int Orthop 2009;34:167-172.

57. Takeuchi R, Aratake M, Bito H, et al. Simultaneous bilateral opening-wedge high tibial osteotomy with early full weight-bearing exercise. Knee Surg Sports Traumatol Arthrosc 2008;16:1030-1037. 58. van den Bekerom MPJ, Patt TW, Kleinhout MY, van der Vis HM, Rob Albers GH. Early complications after high tibial osteotomy. A comparison of two techniques. J Knee Surg 2008;21:68-74. 59. van Hemert WLW, Willems K, Anderson PG, van Heerwaarden RJ, Wymenga AB. Tricalcium phosphate granules or rigid wedge preforms in open wedge high tibial osteotomy: A radiological study with a new evaluation system. Knee 2004;11:451-456. 60. Warden SJ, Morris HG, Crossley KM, Brukner PD, Bennell KL. Delayed- and non-union following opening wedge high tibial osteotomy: Surgeons’ results from 182 completed cases. Knee Surg Sports Traumatol Arthrosc 2004;13:34-37. 61. Yacobucci GN, Cocking MR. Union of medial openingwedge high tibial osteotomy using a corticocancellous proximal tibial wedge allograft. Am J Sports Med 2008;36: 713-719. 62. Zarrouk A, Bouzidi R, Karray B, Kammoun S, Mourali S, Kooli M. Distal femoral varus osteotomy outcome: Is associated femoropatellar osteoarthritis consequential? Orthop Traumatol Surg Res 2010;96: 632-636. 63. McKibbin B. The biology of fracture healing in long bones. J Bone Joint Surg Br 1978;60:150-162. 64. Naudie D, Bourne RB, Rorabeck CH, Bourne TJ. The Install Award. Survivorship of the high tibial valgus osteotomy. A 10- to -22-year followup study. Clin Orthop Relat Res 1999;(367):18-27.

Bone grafts and bone substitutes for opening-wedge osteotomies of the knee: a systematic review.

To establish the rate of use of various void fillers in the setting of opening-wedge osteotomy around the knee, the types of fixation used, and the ra...
543KB Sizes 0 Downloads 6 Views