2015 ISAKOS Biennial Congress ePoster #1507

A Novel Healing Classification for Osteochondritis Dissecans of the Knee Provides Substantial Inter-Observer Reliability

David Ramski, MD, Philadelphia, PA UNITED STATES
Theodore Ganley, MD, Philadelphia, PA UNITED STATES
James L. Carey, MD, MPH, Philadelphia, PA UNITED STATES

The Children's Hospital of Philadelphia and Hospital of the University of Pennsylvania , Philadelphia, PA, USA

FDA Status Not Applicable

Summary: Evaluation of radiographic healing patterns of patients with osteochondritis dissecans of the knee revealed a novel system for healing classification.




Osteochondritis dissecans (OCD) is a localized process in which a focus of subchondral bone and adjacent articular cartilage (progeny fragment) separates from the surrounding bone (parent). To our knowledge, patterns of healing have not been formally delineated on the standard AP, lateral, and notch view knee radiographs. The goal of this study was to examine temporally associated patterns of healing to first evaluate practicality of the classification system and second to examine associations between healing pattern and age, gender, lesion location, treatment type, and physeal patency.


489 patients were retrospectively screened from 2006-2010 for a total of 41 consecutive evaluable knee OCD lesions (average age 12.8 years, range 7.8-17.1, average follow-up 75.1 weeks), representing all cases for a single surgeon from 2006-2010 that met inclusion criteria. Patients were included if they were under 18 years of age, had at least three consecutive images (including either lateral or notch view radiographs), and had an OCD lesion of the knee. Images were arranged in sequential order for rating with only time from initial presentation provided to the readers. Two fellowship trained orthopaedic sports medicine specialists classified the images according to lesion location, healing type, and physeal patency. Healing patterns were rated as resolution of the boundary between progeny fragment and parent bone (that is, from distinct boundary to indistinct boundary), increasing radiodensity of the progeny fragment (that is, from radiolucent to the same radiodensity as the parent bone), combined (features of both boundary resolution and increasing radiodensity patterns), or not applicable (see Figure 1). Repeat readings were conducted three weeks later.


The average patient age was 12.8 years (SD 2.1, range 7.8-17.1) and average follow-up time was 75.1 weeks (range 14-276); the average time between radiograph studies was 22 weeks. Gender distribution was 35 males and 6 females. The most common patient type was male, between the ages of 13-17, with open physes, receiving operative treatment, with a medial femoral condyle lesion. Both the inter-observer and the intra-observer reliability of the proposed healing classification were substantial with intra-class correlation coefficients of 0.67 (95% confidence interval 0.55-0.79). These are considered “good” by the Altman standard for reliability coefficients. The per cent agreement across all healing ratings combined for both attending physician raters was 0.78. Boundary healing and radiodensity healing was observed in all ages, genders, lesion locations, treatment types, and physeal patency states. The rating of “not applicable” was not used.


This study presents a valuable three-category paradigm – boundary resolution, increasing radiodensity of progeny fragment, and combined – for the assessment of OCD lesion healing. The proposed novel system of healing classification on radiographs demonstrated good inter- and intra-observer reliability. Healing patterns were not significantly associated with any particular age, gender, lesion location, treatment type, or physeal patency status.