Donor-site morbidity after osteochondral autologous transplantation for osteochondritis dissecans of the capitellum: a systematic review and meta-analysis

Purpose To determine the rate of donor-site morbidity after osteochondral autologous transplantation (OATS) for capitellar osteochondritis dissecans. Methods A literature search was performed in PubMed/MEDLINE, Embase, and Cochrane Library to identify studies up to November 6, 2016. Criteria for inclusion were OATS for capitellar osteochondritis dissecans, reported outcomes related to donor sites, ≥10 patients, ≥1 year follow-up, and written in English. Donor-site morbidity was defined as persistent symptoms (≥1 year) or cases that required subsequent intervention. Patient and harvest characteristics were described, as well as the rate of donor-site morbidity. A random effects model was used to calculate and compare weighted group proportions. Results Eleven studies including 190 patients were included. In eight studies, grafts were harvested from the femoral condyle, in three studies, from either the 5th or 6th costal-osteochondral junction. The average number of grafts was 2 (1–5); graft diameter ranged from 2.6 to 11 mm. In the knee-to-elbow group, donor-site morbidity was reported in 10 of 128 patients (7.8%), knee pain during activity (7.0%) and locking sensations (0.8%). In the rib-to-elbow group, one of 62 cases (1.6%) was complicated, a pneumothorax. The proportion in the knee-to-elbow group was 0.04 (95% CI 0.0–0.15), and the proportion in the rib-to-elbow group was 0.01 (95% CI 0.00–0.06). There were no significant differences between both harvest techniques (n.s.). Conclusions Donor-site morbidity after OATS for capitellar osteochondritis dissecans was reported in a considerable group of patients. Level of evidence Level IV, systematic review of level IV studies. Electronic supplementary material The online version of this article (doi:10.1007/s00167-017-4516-8) contains supplementary material, which is available to authorized users.


Introduction
Osteochondritis dissecans (OCD) of the capitellum is a disorder of the subchondral bone and articular cartilage [3,23,32]. This condition is primarily seen in teenagers engaged in sporting activities in which the elbow is repetitively exposed to extensive valgus forces, such as baseball and gymnastics [3,23,32]. In early phases, a stable capitellar OCD may cause pain and effusion, while in advanced stages, the OCD may become unstable and cause locking, restricted range of motion, and instability [3,23,32]. Stable OCD may initially be treated nonoperatively with activity modification and physical therapy [17,18], whereas unstable lesions require operative treatment. Several surgical options have been developed over the past two decades including arthroscopic debridement with or without marrow stimulation [13,26,29], fragment fixation [7,10,31], and osteochondral autologous transplantation (OATS) [15,16,24,33].
OATS has become a popular treatment option for large, unstable lesions (diameter > 10 mm) with lateral wall involvement, as well as also for athletes without an acceptable outcome after less invasive techniques [8,15,22,24]. In OATS, a single or multiple osteochondral grafts are harvested from either the less-weight-bearing parts of the femoral condyle [8,11,15,24] or the costal-osteochondral junction [22,30]. The cylindrical donor plug consisting of hyaline cartilage and subchondral bone is then transplanted into the defect area to restore the integrity of the articular surface of the capitellum. A major disadvantage of this procedure is the need to harvest one or multiple grafts from an asymptomatic knee or the rib area in an adolescent athlete and thus the risk for morbidity at the donor site [2,12,27]. Recently, a review conducted by Andrade et al. reported knee donor-site morbidity rates of 17% and 6% for ankle and knee mosaicplasty procedures, respectively [2]. In contrast, two studies involving adolescent athletes who underwent OATS for OCD of the capitellum found no adverse effects related to the donor site [9,35]. Interestingly, Weigelt et al. reported substantial donor-site morbidity in eight of 14 patients treated for capitellar OCD [33]. The vast majority of patients with capitellar OCD are high-demand athletes who are younger than patients with knee and ankle OCD [2,9]. Also, as opposed to knee and ankle OCD, grafts have been harvested from both the femoral condyle and costal-osteochondral junction in the treatment of capitellar OCD [9,30]. Knowing the overall risk for donor-site morbidity following capitellar OATS is relevant in surgical decision making, as well as it is essential to be able to counsel patients about the risk for possible donor-site effects. To our knowledge, the risk of donor-site morbidity after OATS in this particular population is still unknown.
The purpose of this study was to determine the rate of donor-site morbidity following osteochondral autologous transplantation for capitellar osteochondritis dissecans. The hypothesis op the study was that there would be no difference in the proportion of donor-site morbidity between graft harvesting from the femoral condyle or costal-osteochondral junction.

Protocol
The findings of this systematic review were reported according to the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines [14].

Selection criteria
Studies that met the following inclusion criteria were included: (1) osteochondral autologous transplantation for capitellar OCD, (2) reported outcomes related to the donor site, (3) minimum inclusion of 10 patients, (4) minimum follow-up of one year, and (5) written in English. Case reports, reviews, and cadaveric studies were excluded. Donor-site morbidity was defined as the presence of persistent symptoms (≥1 year) after graft harvesting, as well as the need for subsequent intervention to treat complications related to the donor site.

Search strategy
A literature search was performed in the following databases up to November 6, 2016: PubMed/MEDLINE, Embase, and the Cochrane Library. The PubMed/MED-LINE search strategy was adjusted into similar search strategies for other databases (Online Appendix 1). Reference lists of retrieved studies were manually searched for additional studies potentially meeting the inclusion criteria.

Study selection
Study selection was independently performed by two authors. First, title and abstract were screened to identify potentially relevant papers (R.B. and P.O.). These results were verified by two senior authors (L.O. and M.v.d.B.). Subsequently, manuscripts were retrieved when title or abstract revealed insufficient information to determine the appropriateness for inclusion. Disagreement was resolved by discussion or with arbitration when necessary by the senior authors (L.O. and M.v.d.B.) when differences remained.

Data extraction
The main outcome of interest was the presence or absence of donor-site morbidity (persistent symptoms or need for subsequent intervention) after capitellar OATS. The following data were extracted from each study: author names, year of publication, patient demographics, follow-up time, harvest method, graft characteristics, and the use of fillers. The following outcomes related to the donor site were extracted: symptoms (e.g., pain, locking, and instability), physical examination (e.g., effusion and range of motion), complications (e.g., infection, nerve injury, or subsequent treatment), patient reported outcome scores, imaging evaluation [e.g., radiographs, computed tomography, or magnetic resonance imaging (MRI)], and anatomic and histological outcomes. Data extraction was independently performed by two authors (R.B. and P.O.) and verified by two senior authors (L.O. and M.v.d.B.). Articles were not blinded for author, affiliation, and source.

Methodological quality assessment
Two authors (R.B. and P.O.) independently judged the methodological quality of included studies using the checklist for quality appraisal of case series studies that was developed at the Institute of Health Economics (IHE) [5,20]. Each of the 20 criteria of the checklist were answered with either 'yes,' 'no,' or 'partial' or 'unclear.' For estimating the risk of bias for each study, 'partial' responses were considered 'yes,' and 'unclear' responses were considered 'no.' A study with 0-2 'no' responses was considered to have a low risk of bias, 3-5 'no' responses a moderate risk, 6-8 'no' responses a high risk, and 9 or more 'no' responses a very high risk of bias. In case of disagreement, two senior authors (L.O. and M.v.d.B.) were involved to solve the differences.

Statistical analysis
Categorical data were displayed as absolute numbers with frequencies; continuous data were displayed as means with sample range.
The main outcomes of interest were the rate of donorsite morbidity within the knee-to-elbow group and rib-toelbow group. Proportions of donor-site morbidity were calculated for each study using the Freeman-Tukey double arcsine transformation. Subsequently, a random effects model, to account for heterogeneity across studies, was used to calculate weighted group proportions for each harvest technique, and to compare proportions between the two techniques [19,34]. A p value <0.05 was considered significant. Statistical analyses were performed with the use of Stata 13.0 (College Station, TX, USA).

Donor-site morbidity
A detailed summary of donor-site effects per study is given in Table 2. In the knee-to-elbow group, donor-site morbidity after capitellar OATS was reported in 10 of 128 patients, resulting in a donor-site morbidity rate of 7.8%. Knee pain while stair climbing and heavy lifting were reported in nine patients (7.0%) and locking sensations in one patient (0.8%) ( Table 2) [9,16,33]. In the rib-to-elbow group, donor-site morbidity was reported in one of 62 patients, resulting in a donor-site morbidity rate of 1.6%. One case was complicated by a pneumothorax due to damage to the costal pleura, which required insertion of a chest tube [30].
The proportion of donor-site morbidity in the knee-toelbow group was 0.04 (95% CI 0.0-0.15), and the proportion in the rib-to-elbow group was 0.01 (95% CI 0.00-0.06) (Fig. 2). There was no significant difference between the two harvest techniques in terms of proportion of donor-site morbidity (p > 0.05).

Methodological quality
The results of methodological quality assessment of individual studies using the IHE scale are presented in Table 3. According to the criteria of the IHE checklist for critical appraisal of case series studies, the estimated potential  , moderate for nine studies [9,11,15,16,22,24,30,33,35], and high for one study [28]. Solely 'no' responses were awarded to question 3 and 11, which are related to the patient recruitment process and blinding of outcome assessors, respectively. Here, no study recruited patients from multiple centers, nor was outcome assessment blinded. Three more questions, also related to outcome measures (10, measures established a priori; 12, appropriateness of measures; 13, before and after intervention measured), were awarded with 'no' or 'partial' responses in more than half of the included studies. By contrast, most criteria with regard to the study aim (question 1), study population (question 5-7), intervention (question 8), statistical analysis (question 14), results/conclusions (question [15][16][17][18][19], and sources of support (question 20) were awarded a 'yes' response.

Discussion
The most important finding of the present study is that donor-site morbidity after osteochondral autologous transplantation for capitellar osteochondritis dissecans occurred in 7.8% within the knee-to-elbow group and 1.6% within the rib-to-elbow group. In the knee-to-elbow group, knee pain during daily activities (7.0%) and locking sensations (0.8%) were reported; in the rib-to-elbow group, one case was complicated by a pneumothorax. There was no significant difference in proportion of donor-site morbidity between the two harvest techniques. The findings of this systematic review emphasize the importance of associated donor-site morbidity following graft harvesting in the treatment of capitellar OCD. The rate of donor-site morbidity in our study is comparable to the morbidity rate after knee-to-knee OATS (6%), as reported by Andrade et al. [2]. Interestingly, the authors reported a higher rate after knee-to-ankle OATS (17%) [32]. The higher rate compared to our findings may be the result of more grafts that were harvested (three versus two) [2], although the influence of the number of grafts on morbidity is controversial [1,2,25,27]. Patient characteristics may have played a role as well. The vast majority of patients in our study were adolescent (15 years), highdemanding athletes, while patients with talar OCD are typically older (32 years [25]). We hypothesized that patients in our study may have been in a better physical condition before treatment, and therefore faster recovery with less donor-site effects may be expected.
In the present study, the rate of donor-site morbidity ranged from 0 to 57% in studies in whom grafts were harvested from the knee. Weigelt et al. reported morbidity in eight of 14 patients (57%): occasional pain during heavy lifting in seven patients and intermittent locking sensations in one patient [33]. The advanced age after a relative long follow-up (7 years) may be the reason for more morbidity, as well as it may be the result of large grafts (8-11 mm) Fig. 1 Flowchart of search strategy following PRISMA (preferred reporting items for systematic reviews and metaanalyses) guidelines that were harvested [1,2,25,27]. Two included studies focused on donor-site effects in particular [9,21]. Despite encouraging clinical results reported by Iwasaki et al., MRI evaluation revealed alterations in signal intensity in 89% of donor sites, suggesting fibrous filling of donor holes [9]. Long-term follow-up is needed to see if these changes in signal intensity are permanent and clinically meaningful. Nishimura et al. found a delay in recovery of quadriceps muscle strength [21]. This indicates that even if symptoms may resolve quickly after harvesting, the donor knee is at risk for injury due to muscle weakness within the first year. Three studies attempted to prevent postoperative bleeding by filling donor tunnels [9,15,21]. As included studies were limited by small case series, potential beneficial effects remain unclear. Favorable results have been reported using synthetic implants in repair of knee OCD [4,6], although literature lacks large sample comparison studies.
Three studies that were included in this review harvested grafts from either the 5th or 6th costal-osteochondral junction to repair capitellar OCD [22,28,30]. Although hard scar tissue was detected in palpation, radiographs showed new subperiostal bone formation and no long-term symptoms were observed [30]. However, harvesting grafts from the rib area is a technically demanding procedure that has been described by only a few studies. Most surgeons who perform capitellar OATS are more familiar with knee anatomy rather than rib anatomy. Moreover, risk for devastating donor-site morbidity remains as harvesting may be complicated by a pneumothorax due to close proximity to the costal pleura, which may lead to a prolonged hospital stay [30]. If familiar with this technique, this may be an option in the treatment of large capitellar OCDs (>15 mm). As large lesions usually require multiple cylindrical grafts, one may want to avoid the risk for donor-site morbidity of the knee. Using the IHE scale to evaluate the methodological quality of included studies [5,20], the estimated potential risk of bias ranged from low to high, with the majority of studies found to be of moderate risk of bias (9 of 11). Studies scored the lowest on criteria related to outcome measures, as well as on the fact that cases were collected in a single center in each study. Therefore, the findings of this systematic review should be interpreted by taking into account some limitations. First, a major limitation is the incomplete reporting regarding outcomes related to the donor site. As physical examination or imaging was rarely thoroughly reported, studies lack objective assessment of donor-site effects. Additionally, subjective assessment of knee function by means of patient reported outcome measures was only performed in four of eight studies (Lysholm score) [9,16,21,33]. In the remaining studies, no attempt was made to evaluate knee function, nor was physical examination thoroughly described [11,15,24,35], as this was also the case in two rib-to-elbow studies [22,28]. Also, in none of the included studies, either subjective or objective preoperative assessment of the donor site was reported. As data were obtained from studies that evaluated donor-site effects in varying degrees, we hypothesized that donorsite morbidity may be substantially underreported in this population. Also, due to incomplete reporting, we were unable to investigate associations between morbidity and harvest characteristics including donor-site location and the size or number of grafts. Second, included studies are limited by case series with small numbers of patients because capitellar OCD is a rare condition in the general population, and capitellar OATS is a relatively new procedure. Third, after pooling the data of included studies, we found no significant difference between the proportions of donor-site morbidity for the knee-to-elbow group and the rib-to-elbow group (p > 0.05). One may argue whether it is valid to combine study data because of between-study variance; however, we attempted to alleviate statistical heterogeneity with the use of a random effects model. This model has been used previously in systematic reviews who pooled data from case series studies [19,34].
Future studies should comprehensively evaluate effects related to the donor site. Patient's symptoms and physical examination should be reported for each patient. Additionally, the use of a patient reported outcome measure to assess knee function should be a routine part of clinical evaluation, both preoperatively and postoperatively. Radiographs should be performed to evaluate potential progression to osteoarthritis at 1 year and may be performed in cases in which donor fillers were used. MRI evaluation should, because of cost-effectiveness reasons, only be ordered in case of persistent symptoms. Besides the evaluation of donor-site effects, harvest characteristics should be reported in great detail, such as donor location, number of grafts, graft diameter, and depth. Also, alternative harvest methods should be investigated to have no longer the need to violate the integrity of an asymptomatic knee or rib in an adolescent athlete.
The findings of the present investigation demonstrate a considerable risk for donor-site morbidity following capitellar OATS. Although good-to-excellent results related to the elbow have been reported after capitellar OATS, surgeons should be aware of the risk for donor-site morbidity and patients should be counseled about this issue. Knowing the overall risk for donor-site morbidity is also relevant in surgical decision making. Taking this into consideration, surgeons could consider other resurfacing techniques such as allografting or autologous chondrocyte transplantation.

Conclusions
Osteochondral autologous transplantation in the treatment of capitellar osteochondritis dissecans may lead to donor-site morbidity in a considerable group of patients, either after harvesting from the femoral condyle (7.8%) or costal-osteochondral junction (1.6%).