Medial femoral condyle width and osteochondritis dissecans: Cause or eﬀect and the implications for osteochondral allograft transplantation

4 on A student t test statistic was used to compare MFCW and MFCW/TPW between the 4 groups as well as to the age and sex matched controls. males, mean age: 25.7) had a mean MFCW of 29.3 mm. The 115 control patients (74 men, mean age: 25.6) had a mean MFCW of 24.8 mm. Men with OCD had wider MFCs compared to women ( P < .001). For both the OCD and control populations, there was no statistical signiﬁcance between MFCW and age. Overall, when controlling for age and sex, OCD patients had larger MFCWs than controls ( P < .001). Conclusions: Uniform widening of the MFC in OCD patients lends evidence that a wider MFC might be causal in the development of OCD. in ﬂattening


Introduction
Osteochondritis dissecans (OCD) of the knee is a relatively rare cause of knee dysfunction and pain that has been recognized for over 100 years. There are many theories on the pathogenesis of OCD, including biologic and genetic factors, anatomic factors, and repetitive micro-traumatic causes. [1][2][3][4] The most common location for OCD in the knee is the lateral aspect of the medial femoral condyle (MFC). When lesions continue to be symptomatic despite conservative management, surgical options become reasonable, including osteochondral allograft transplantation. 5 While use of non-orthotopic lateral femoral condyle grafts are excellent graft options for MFC lesions, 6 including OCD lesions, 7 most surgeons request size-matched MFC grafts for MFC lesions. The width and size of the MFC has been studied previously and there are gender and race variations. 8 More so, women generally have smaller femoral condyles as compared to men. 9 , 10 Analysis of 5962 donors of distal femoral allografts revealed that the mean medial femoral condyle width (MFCW) in women donors was 23.9 mm, while in men it was 26.1 mm. Overall, the mean MFCWs was 25.6 mm in 5962 donors younger than 35 years of age (Data on file, Joint Restoration Foundation).
The size of the MFC has been linked to medial compartment pathology. [11][12][13][14][15][16][17] Specific MFC morphologic characteristics have not, however, been linked to OCD. While stress changes as a result of an OCD lesion can occur, the question is whether a wider MFC is an anatomic factor that can lead to an OCD lesion.
It has been our experience that the MFC in patients with OCD is larger than normal. In many situations, the width of the patient's condyle exceeds the width of any donor condyle. In those situations, surgeons desiring a sized-match orthotopic condyle may never be offered a suitable graft because of the rarity of large MFC donor grafts.
The purposes of this study were to measure MFCWs and tibial plateaus width (TPW) using magnetic resonance imaging (MRI) in patients with OCD of the MFC and compare MFCWs and MFCW/TPW ratios of patients with OCD to an age and sex-matched, healthy control group.

Materials and methods
We conducted a case control study of patients with OCD of the MFC. From a computer database of patients, we identified 126 patients that were diagnosed with OCD of the MFC from the years 2006 to 2019. Inclusion criteria were patients with isolated OCD of the MFC who had an MRI that was available for review. Exclusion criteria included those who were diagnosed with concomitant pathology in addition to OCD of the MFC, OCD in areas of the knee other than the lateral aspect of the MFC, obvious medial compartment osteoarthritis with marginal osteophytes, and articular cartilage loss on the non-OCD portions of the MFC. Following a review of the MRIs, we excluded 15 patients, giving us 111 OCD patients that were the subject of the study. Aged matched controls that had MRIs of their knees for conditions other than medial compartment pathology were also included in the study. Inclusion criteria for the control group were those with MRIs of their knees for acute isolated ACL tears, patellar instability, and isolated lateral meniscus tears. Any control patients that were noted to have any medial compartment pathology, including medial meniscus tears, chondral lesions in the medial compartment, and medial collateral ligament injuries, were excluded from the control group. 115 patients served as the control group.
We conducted a power analysis to help determine sample size for our study. We used a mean MFC width of 25.6 mm ± 2.50 mm, which was calculated from 5962 donors younger than 35 years of age (Data on file, Joint Restoration Foundation). We defined a clinically significant increase in MFC width to be 4 mm. We chose 4 mm as clinically significant because of donor MFC availability, as donor MFC grafts are rarely available at or greater than 29.0 mm. Based on a 4.0 mm increased MFC width, power of 0.8, and an alpha of 0.05, it was determined that a minimum sample size of 12 would be needed to show significance.
Each OCD and control patient had an MRI performed on a 1.5 Tesla scanner. We analyzed Proton Density Fast Spin Echo and T1 FSE coronal sequences on each patient and control. We selected a coronal MRI image that was in the mid-point of the MFC and which best outlined the medial and lateral bone of the MFC. The image was referenced from anterior to posterior, and the PACS DICOM (Med-Dream Web Dicom Viewer Aycan Medical Systems) sequence correlation tool was used on corresponding MRI sagittal sequences to define the appropriate coronal image for measurement. The MFCW and (TPW were measured using the PACS DICOM software for the study patients and controls ( Image 1 ). Measurements using the PACS system was accurate to 0.01 mm. MFCW/TPW ratio was also calculated.
Patients were grouped by decade with 10 to 20 years in Group 1, 21 to 30 years in Group 2, 31 to 40 years in Group 3, and 41 years and older in Group 4. Each group was further separated by gender ( F = female; M = male).
A student t test statistic was used to compare MFCW and the quotient of MFCW/TPW between the 4 groups as well as to the age and sex matched controls. A post-hoc analysis was performed between each of the OCD and control groups in each of the age-gender groups to further assess the power of statistically significant results.

Results
The 111 patients with osteochondritis of the MFC included 70 males and 41 females with a mean age of 25.7 years. The 115 control patients included 74 males and 41 female patients with a mean age of 25.6 years. In the OCD group, the overall mean MFC width was 29.3 mm. When the OCD patients were grouped by age, Group 1 (ages 10-20 years) had a mean MFC width of 28.9 mm; Group 2 (ages 21-30 years) had a mean width of 29.2 mm; Group 3 (ages 31-40 years) had a mean width of 29.5 mm. There was no statistical difference in MFC width in the OCD patients with respect to age. Men patients with OCD had a mean MFC of 30.4 mm compared to women OCD patients with a mean width of 27.2 mm. Using a student t test statistic, this difference was statistically significant with P < .001 ( Table 1 ).
The control patients had an overall mean MFC width of 24.8 mm, with men having a mean MFC width of 26.5 mm and women having a mean MFC width of 23.6 mm. There was no difference in MFC width in the control group with respect to age, but like the OCD patients, men in the control group had a wider MFC as compared to women ( P < .001). Overall, the MFC width in OCD patients was statistically greater than the controls with a P < .001. This observation was present when controlling for age and sex ( Table 1 , Fig. 2 ).  OCD groups had significantly wider condyles, both absolute and relative to TPW. * denotes statistically significant result using = 0.05.

ARTICLE IN PRESS
We performed a post-hoc power analysis with an alpha of 0.05 for both genders in each of the groups. We calculated a power greater than 0.99 in each except Group 2 Female and Group 4 Male (0.93 and 0.77, respectively). The power analysis of the overall OCD and control groups yielded a power of 100%.

Discussion
The width of the MFC and MFCW/TPW ratio in patients with OCD of the MFC is statistically greater than controls, even when controlling for age and gender. Men were found to have wider MFCW compared to women in both the OCD patients and controls. Widening of the MFC was even seen in Group 1 patients and in patients as young as 10 years of age. It is unknown whether a wider MFC is a risk factor for developing an OCD lesion of the MFC. The widening of the MFC was uniform, as compared to widening that is seen with a marginal osteophyte as a result of meniscus deficiency 18 and articular cartilage pathology. 15 , 16 Our results support the concept that widening of the MFC is a cause of OCD and not the result of OCD. However, further studies are needed to explore this concept.  Our study does have strengths, including the fact that it is the largest study to date in literature of patients with OCD of the MFC. This study also is the first to report the increased size of the MFC in OCD patients compared to control as well as to data available for normal femoral condyle metrics.
The results of this study are clinically relevant to joint restoration surgeons that use osteochondral allograft transplantation in the treatment of OCD of the MFC. Requests for a size-matched orthotopic MFC graft is the accepted standard that most surgeons use. However, finding a size-matched allograft for patients with OCD of the MFC will be challenging, given the sizes of MFC grafts that are typically available for transplant. Analysis of 5962 donors of distal femoral allografts revealed that the mean MFCW in women donors was 23.9 mm, while in men it was 26.1 mm. Overall, the mean MFC width was 25.6 mm ± 2.5 mm (Data on file at Joint Restoration Foundation, Centennial, CO) ( Fig. 1 ). In our analysis, the mean MFC width in patients with OCD was 29.3 mm. Donor MFC allografts are rarely available at or greater than 29.0 mm. Using an undersized MFC graft, particularly if the graft is 4 mm narrower than the patient, can be difficult to achieve an anatomic restoration of the articular surface due to the different surface topography and radius of curvature. If a surgeon desires a size-matched condyle graft, consideration for using a non-orthotopic lateral femoral condyle allograft would lead to more available donors. Approximately 75% of condyles that are wider than 26 mm are lateral femoral condyles (data on file, Joint Restoration Foundation) ( Fig. 1 ). Large lateral femoral condyle are readily available and should be considered as suitable graft options for osteochondral allograft transplant in patients with an OCD of the MFC.

ARTICLE IN PRESS
There are limitations to this study. The study was conducted on MRIs of patients that we identified as having a diagnosis of OCD of the MFC. We acknowledge that there are certainly patients that are diagnosed with an OCD that are much younger than our study patients. Our youngest patient in our OCD group was 10 years of age, yet we did not have many patients that were skeletally immature and truly in the earliest ages of onset of OCD. As such, we are not able to definitively conclude that a larger MFC is a risk factor for development of OCD. However, our patients did not have the typical adaptive bony changes seen in patients with meniscal deficiency and osteoarthritis, including condyle flattening and marginal osteophytes. Additional studies in skeletally immature pediatric patients with OCD of the MFC may help to answer the question as to whether a wider MFC is a cause or effect.
Group 4 Male had very few patients, which was a potential explanation for the post-hoc analysis result in that group. However, we do not feel this small subset changes the overall power and overall conclusions.
We used PACS software to measure the MRI (Med-Dream Web Dicom Viewer, Aycan Medical Systems). The measurement tool for this PACS system is accurate to 0.01 mm. However, while our measuring protocol was not validated, we still feel our ability to accurately measure condyle and plateau widths was accurate. The software had accuracy to 0.01 mm but we chose to round to the nearest tenth of a millimeter.

Conclusion
The MFC is statistically wider in patients with OCD of the MFC compared to normal age and sex matched controls. The uniform widening as opposed to widening from a marginal osteophyte and the presence of widening even in patients as young as 10 years of age supports the concept that the widening of the condyle is a cause more than an effect of OCD. Surgeons performing osteochondral allograft transplantation for symptomatic OCD lesions of the MFC may need to consider a non-orthotopic lateral femoral condyle donor graft if size matching is important, as there is a paucity of large MFC donor grafts that would match the sizes seen in patients with OCD of the MFC.

Informed patient consent
Complete written informed consent was obtained from the patient for the publication of this study and accompanying images.