Glenoid osteochondritis dissecans in 9 baseball players: arthroscopic management and radiographic outcomes Authors: Motoyuki Fujisawa, Masafumi Hara Published: 2026 ABSTRACT AbstractGlenoid osteochondritis dissecans (OCD) is a rare condition seen occasionally in throwing athletes. Early diagnosis using conventional radiography is challenging. The aim of this study was to report on a series of baseball players receiving treatment for glenoid OCD. This retrospective case series included 9 baseball players with glenoid OCD who underwent arthroscopic surgery with a minimum of one-year follow-up. Plain radiographs were routinely obtained in the arm-elevated position (140–150°) to compare glenoid morphology between the throwing and non-throwing shoulders. Computed tomography and magnetic resonance imaging were used for definitive diagnosis and lesion assessment. Surgical procedures were selected based on intraoperative arthroscopic findings, including transarticular drilling for stable lesions and bone marrow stimulation for unstable lesions. The mean age at surgery was 16.6 ± 1.7 years, with a mean follow-up of 30.5 ± 12.7 months. All lesions were located in the posterior glenoid. Radiographic healing was confirmed on post-operative computed tomography at a mean of 12.0 ± 4.2 months, with no progression observed in any case. Eight of 9 patients (89%) successfully returned to play (RTP) at their preinjury competitive level at a mean of 9.2 ± 4.1 months post-operatively. Arm-elevated radiographs comparing both shoulders may serve as a useful adjunctive screening method for detecting glenoid OCD in throwing athletes. Arthroscopic treatment tailored to lesion stability was associated with a favorable RTP rate in this small case series; however, careful consideration of surgical timing is necessary.Keywords: Osteochondritis dissecans, Glenoid, Bone marrow stimulation, Baseball player, Throwing athletes, Arthroscopic treatment Osteochondritis dissecans (OCD) is a condition that most commonly affects the capitellum of the humerus and the knee joint; however, its occurrence in the glenoid of the scapula is extremely rare. Several case reports and small case series have been published, but the largest single-institution report has included 7 cases17; when limited to overhead athletes such as baseball players, the largest series has consisted of 4 cases.21 Glenoid OCD is difficult to diagnose using standard plain radiographs and may therefore be overlooked in young athletes presenting with throwing-related shoulder pain. At our institution, routine evaluation of ball-throwing-related shoulder disorders includes plain radiographic imaging obtained with the shoulder elevated to approximately 140–150°, which has been used primarily to assess subtle posteroinferior contracture of the glenohumeral joint. Using this imaging technique, we have observed cases in which differences in glenoid morphology between the affected and contralateral sides become clearly apparent and are recognized as irregular radiographic findings. Such findings may serve as a useful clue for suspecting glenoid OCD. With regard to treatment, some reports have suggested that conservative management may be effective for stable lesions; however, the number of reported cases remains limited, and no clear consensus has been established regarding optimal treatment strategies. In particular, there are few reports describing the surgical management and post-operative outcomes of glenoid OCD in overhead athletes based on a substantial number of cases. The purpose of this study was to clarify the radiographic findings, treatment strategies, and clinical outcomes in 9 baseball players with glenoid OCD who underwent arthroscopic surgery with a minimum of one-year follow-up. Methods This study was a retrospective case series conducted at a single institution. Between 2011 and 2022, 9 male baseball players and 1 female softball player who presented to our hospital with throwing-related shoulder pain and were diagnosed with glenoid OCD based on imaging studies were identified. Of these, 9 patients who underwent arthroscopic surgery and had a minimum follow-up of one year were included. One additional patient who did not meet the minimum follow-up requirement is presented separately as a supplementary illustrative case to demonstrate the clinical course of conservative management. This study was approved by the Institutional Ethics Committee (approval no. #2018-16-001), and informed consent was obtained from all patients. Radiographic evaluation Plain radiographs were obtained for all patients at the initial visit. In addition to standard anteroposterior (AP) views of the shoulder, radiographs were routinely obtained with the shoulder elevated to approximately 140–150° (arm-elevated position) (Fig. 1). This imaging technique was used to compare glenoid morphology between the throwing and non-throwing shoulders, with particular attention to side-to-side differences and the presence of irregularities. When asymmetry or irregularity of the glenoid contour was identified on arm-elevated radiographs, further evaluation was performed using computed tomography (CT) and/or magnetic resonance imaging (MRI) to confirm the diagnosis of OCD and to assess lesion extent. Lesion size was measured as the maximum diameter on CT images, and pre-operative assessment using MRI was performed according to the International Cartilage Repair Society (ICRS) classification. Radiographs were routinely obtained with the arm elevated at approximately 140–150°. Surgical indications and arthroscopic findings Surgical treatment was indicated for patients who failed to demonstrate symptomatic improvement after a defined period of conservative management, including cessation of throwing and physical therapy, or for those who strongly desired to return to competitive sports. The final treatment strategy was determined intraoperatively, based on arthroscopic findings. During arthroscopy, cartilage stability was assessed by probing. When the cartilage surface was judged to be stable and classified as ICRS grade I or II, transarticular drilling was performed while preserving the cartilage surface. In contrast, when instability was detected by probing and the lesion was classified as ICRS grade III or IV, unstable cartilage was débrided, followed by bone marrow stimulation (BMS) using a microfracture awl. Glenohumeral internal rotation deficit (GIRD) was assessed at initial presentation and defined as a side-to-side deficit of internal rotation greater than 20° compared with the contralateral shoulder. In the 6 most recent cases, posterior capsular release was additionally performed in patients with GIRD to reduce compressive stress on the posterior glenoid. Post-operative rehabilitation and return-to-play criteria Post-operatively, range-of-motion exercises, isometric rotator cuff strengthening, and physical therapy targeting regions outside the affected shoulder were initiated immediately. After improvement in shoulder range of motion and recovery of muscle strength were confirmed on physical examination, a progressive throwing program was initiated at three months or later after surgery. Returned to play (RTP) was defined as return to the same competitive level as before symptom onset and the ability to perform unrestricted, full-effort throwing. RTP was determined based on a comprehensive assessment, including improvement of pain, restoration of shoulder range of motion and strength, and the absence of symptom recurrence during the throwing motion. Results During the study period, 971 patients initially presented with atraumatic throwing-related shoulder pain. Among these patients, 10 ball-throwing athletes were diagnosed with glenoid OCD. Of these, 9 patients who underwent arthroscopic surgery with more than one year of follow-up were included in this analysis. The mean age at surgery was 16.6 ± 1.7 years, and the mean post-operative follow-up period was 30.5 ± 12.7 months. Symptom onset was acute in 4 patients and gradual in 5 patients. The mean duration from symptom onset to surgery was 12.1 ± 11.3 months. The mean maximum lesion diameter on pre-operative CT was 14.0 ± 2.2 mm (Fig. 2, A–F). Open physis was present in 2 patients aged 14 years, whereas physeal closure had occurred in all other patients. Based on pre-operative MRI findings, the ICRS classification was grade I in 2 patients, grade II in 4 patients, grade III in 2 patients, and grade IV in 1 patient. In contrast, concordance between the pre-operative MRI-based ICRS classification and the intraoperative arthroscopic classification was observed in 5 of 9 patients. Three-dimensional CT images in 6 of the 9 surgically treated cases demonstrate subchondral depression of the posterior glenoid. (A–F) Cases 4–9 in Table I. CT, computed tomography. Surgical procedures and radiographic outcomes Based on intraoperative arthroscopic findings, patients in whom the cartilage surface was judged to be stable and classified as ICRS grade I or II were treated with transarticular drilling (n = 5). In contrast, patients in whom instability was identified and lesions were classified as ICRS grade III or IV underwent débridement of the unstable cartilage followed by BMS using an awl (n = 4). Six patients also received posterior capsular release. Post-operative evaluation using CT demonstrated that the mean time to radiographic healing of the OCD lesions was 12.0 ± 4.2 months. No patient demonstrated progressive deterioration of the lesion during the post-operative follow-up period. Returned to play Eight of the 9 patients (89%) achieved RTP. One of these patients achieved RTP with mild residual pain but was able to compete at his preinjury level. One patient was unable to achieve RTP within one year because of persistent pain during throwing but eventually returned to play at 18 months post-operatively. The mean time to RTP was 9.2 ± 4.1 months. The characteristics of the patients are presented in Table I. Table IPatient characteristics and clinical features of glenoid osteochondritis dissecans.CaseAge (yr)PositionLesion size (mm)Symptom duration (months)ICRS grade (MRI)ICRS grade (arthroscopy)Concomitant injurySurgical procedureTime to radiographic healing (months)Time to RTP (months)117Pitcher125IIIII—TAD + PCR1212217Pitcher177IVIVLabrumCD-BMS + PCR1212319Infielder183IIIIIILabrumCD-BMS1218∗417Pitcher1436IIIILabrum, ganglionTAD186516Outfielder145IILabrumTAD68614Pitcher139IIIILabrumCD-BMS + PCR67718Infielder1426IIIILabrumTAD + PCR129817Pitcher1312IIII—CD-BMS + PCR186914Catcher116IIIPTT-RCTTAD + PCR125TAD, transarticular drilling; CD-BMS, chondral débridement and bone marrow stimulation; PCR, posterior capsular release; PTT-RCT, partial-thickness rotator cuff tear; ICRS, International Cartilage Repair Society; MRI, magnetic resonance imaging.∗RTP was achieved at 18 mo post-operatively after persistent pain during the first post-operative year. Patient characteristics and clinical features of glenoid osteochondritis dissecans. TAD, transarticular drilling; CD-BMS, chondral débridement and bone marrow stimulation; PCR, posterior capsular release; PTT-RCT, partial-thickness rotator cuff tear; ICRS, International Cartilage Repair Society; MRI, magnetic resonance imaging. RTP was achieved at 18 mo post-operatively after persistent pain during the first post-operative year. Representative cases Supplementary illustrative case (not included in the study cohort) A 14-year-old male presented with right shoulder pain during throwing. Glenoid OCD was diagnosed based on CT and MRI obtained at the initial visit (Fig. 3, A and B). Because of his young age and the presence of open physes, conservative treatment consisting of cessation of throwing and physical therapy was performed for 6 months. The patient subsequently changed positions from pitcher to fielder and returned to play; however, he became unable to throw again and returned to our clinic 2.5 years later. Follow-up CT at that time demonstrated no improvement of the lesion (Fig. 3C), and surgery was performed at the patient's request. (A and B) CT and MRI of illustrative case at the initial visit. (C) Follow-up CT demonstrates no improvement in the lesion after 2.5 years of conservative management. CT, computed tomography; MRI, magnetic resonance imaging. Case 1 A 17-year-old left-handed pitcher presented with throwing-related shoulder pain. Arm-elevated plain radiographs demonstrated side-to-side differences in glenoid morphology (Fig. 4, A and B), and glenoid OCD was diagnosed on CT (Fig. 4C). Arthroscopic evaluation revealed cartilage softening, and transarticular drilling was performed (Fig. 4D). CT obtained one year post-operatively confirmed lesion healing (Fig. 4E), and the patient achieved full RTP. (A) Radiographs of case 1 in the arm-elevated position. (B) Radiographs demonstrate an irregular glenoid surface compared with the contralateral side. (C) Pre-operative CT images show a subchondral depression. (D) Transarticular drilling was performed in the softening area. (E) Post-operative CT demonstrates that the lesion had healed at 1-year follow-up. CT, computed tomography. Case 2 A 17-year-old left-handed pitcher with experience competing at the national level presented with throwing-related shoulder pain. Arm-elevated radiographs demonstrated side-to-side differences in glenoid morphology (Fig. 5, A and B), and CT revealed an OCD lesion with a maximum diameter of 18 mm (Fig. 5C). Arthroscopic surgery was performed after completion of his final high school tournament, including débridement of the unstable lesion and microfracture (Fig. 5, D and E) combined with posterior capsular release (Fig. 5F). Although complete osseous union was not observed on CT at one year post-operatively (Fig. 5G), the patient demonstrated clinical improvement and achieved RTP at his preinjury level. (A and B) Radiographs of case 2 in the arm-elevated position demonstrate a depressed glenoid surface compared with the contralateral side. (C) Pre-operative CT shows the lesion. (D) Arthroscopic findings revealed unstable cartilage. (E) Microfracture was performed using an awl after débridement of the unstable cartilage, combined with posterior capsular release (F). (G) Post-operative CT demonstrates improvement in the lesion at 1 year follow-up. CT, computed tomography. Case 3 A 19-year-old right-handed third baseman presented with right shoulder pain during throwing. Plain radiographs demonstrated a clear glenoid deformity that was more evident on the arm-elevated view (Fig. 6A) than on the standard AP view (Fig. 6B). The diagnosis of glenoid OCD was subsequently confirmed on CT (Fig. 6C). Arthroscopic findings revealed severe cartilage damage accompanied by a labrum injury (Fig. 6D) and a partial rotator cuff tear (Fig. 6E). CT obtained two years post-operatively demonstrated radiographic improvement of the OCD lesion (Fig. 6F); however, the patient continued to experience pain during throwing and was unable to return to competitive play within one year post-operatively. The patient eventually returned to play at 18 months after surgery. Plain radiographs of case 3 showing a glenoid deformity that appears more pronounced on the arm-elevated view (A) compared with the standard anteroposterior view (B). (C) Pre-operative CT demonstrates an extensive lesion. (D and E) Arthroscopic findings show not only the OCD lesion but also concomitant labrum and rotator cuff lesions. (F) Post-operative CT demonstrates improvement of the OCD lesion at 2-year follow-up. CT, computed tomography; OCD, osteochondritis dissecans. Discussion OCD is a pathological condition affecting the articular cartilage and subchondral bone and is most commonly observed in the elbow and knee; however, OCD of the glenoid is rare.7 OCD involving the glenoid of the shoulder was first reported in 1947.14 Glenoid OCD is considered extremely uncommon.2,9,19 Rossi18 reported that only 13 cases of glenoid OCD had been described in English-language literature; however, additional cases have since been reported.4,5,7,10,12 Furthermore, 7 cases of glenoid OCD in throwing athletes have been reported in Japanese literature.8,21 In the majority of reported cases of glenoid OCD in throwing athletes, the lesions were located in the posterior glenoid. Davis6 and Takatsuji et al21 also reported that glenoid OCD in throwing athletes predominantly occurs in the posterior region, and similarly, all cases in the present series involved posterior glenoid lesions. Shimizu et al20 demonstrated the presence of areas with increased bone density in the posterior glenoid on the throwing side in baseball players, suggesting that stress concentration may occur in this region. In the present study, all 9 throwing athletes with glenoid OCD demonstrated findings of posterior shoulder tightness. This observation suggests that adaptive changes in the posterior shoulder structures associated with repetitive throwing may contribute to the development of this condition. Although the etiology of glenoid OCD related to overuse by repetitive throwing remains unclear, these findings support the hypothesis that shear and compressive forces generated during the throwing motion are amplified by tightness of the posterior shoulder structures and become concentrated on the posterior glenoid. Nevertheless, because this study is a retrospective case series, it does not directly establish a causal relationship. Diagnosis: Early-stage glenoid OCD is difficult to detect on plain radiographs and is therefore likely to be overlooked in young athletes presenting with throwing-related shoulder disorders.4,5,10,25 Previous reports have suggested the usefulness of axillary views or specialized glenoid lateral views for diagnosis.14 However, both our group and other authors have reported that detection of abnormal findings using these imaging techniques remains challenging.5,10 In ball-throwing athletes, we routinely obtain plain radiographs with the shoulder in an elevated position to longitudinally assess subtle glenohumeral contracture and related findings13,16 (Fig. 7; images of the throwing and non-throwing shoulders). Unlike the standard true AP (Grashey) view, the arm-elevated view places the glenoid in an oblique orientation during imaging. In this arm-elevated view, the posterior aspect of the glenoid, where OCD lesions commonly develop, is potentially projected onto the inferolateral portion of the glenoid on plain radiographs. In this series, clear morphological deformities on arm-elevated radiographs, similar to those shown in Figures 5B and 6A, were identified in 6 of the 9 surgically confirmed cases. In the remaining 3 cases, only subtle side-to-side differences, as illustrated in Figure 4B, were observed. These findings suggest that arm-elevated radiographs may provide an initial clue in many cases; however, abnormalities were not uniformly pronounced, and definitive diagnosis required further evaluation with CT and/or MRI. Therefore, this imaging technique should be considered an adjunctive screening method rather than a definitive diagnostic modality. (A and B) Radiographs obtained in the arm-elevated position in patients with throwing disorders often demonstrate subtle limitation of glenohumeral motion due to posteroinferior shoulder tightness compared with the contralateral side. A = throwing side. Within this referral cohort, glenoid OCD was diagnosed in 1.1% of 971 baseball players with throwing-related shoulder pain who underwent arm-elevated radiography. However, this percentage may not reflect the true incidence in the general population because this cohort consisted of athletes referred for persistent throwing-related shoulder pain. Nevertheless, follow-up using arm-elevated radiographs alone is insufficient, and when abnormal findings are suspected, further evaluation with CT and MRI is essential.25 In the present study, lesion size was assessed using CT, and lesion severity was evaluated using the ICRS classification on MRI. Perdiwala et al17 proposed a treatment algorithm based on MRI findings, and treatment indication was determined according to the MRI-based ICRS classification. However, discordance between pre-operative MRI findings and intraoperative arthroscopic findings was observed in approximately half of the cases. This result indicates that imaging findings alone are insufficient for determining the optimal treatment strategy and that intraoperative arthroscopic assessment plays a critical role in final decision-making.17,21 Treatment: The optimal treatment strategy for glenoid OCD has not been established. Grau et al10 reported that OCD in a 16-year-old baseball player improved on MRI following conservative treatment. Takatsuji et al21 suggested that conservative management is effective in younger patients before closure of the secondary ossification center. In addition, Perdiwala et al17 reported that conservative treatment was effective for lesions classified as stable on MRI. However, in our early cases, we encountered patients in whom lesion improvement was difficult to achieve with conservative treatment, including rehabilitation and cessation of throwing. In the supplementary illustrative case (not included in the present series), a 14-year-old patient demonstrated no radiographic improvement on CT despite two years of conservative management. Therefore, in our practice, surgical treatment is considered when no improvement is observed after a defined period of conservative management. The duration of conservative management before surgery varied among patients because sufficient evidence regarding the effectiveness of conservative treatment and stage-based treatment indications for glenoid OCD remains limited. The timing of surgical intervention was influenced not only by persistent symptoms and imaging findings but also by competitive season schedules and the desired timing for return to sport. Rossi18 reported a case of glenoid OCD in a 20-year-old baseball player who underwent loose body removal and drilling, with RTP achieved one year post-operatively. Koike et al12 similarly reported a case involving a 22-year-old professional baseball infielder who underwent loose body removal and drilling of the subchondral bone surface and returned to baseball training four months after surgery. Edmonds7 reported a case of a 15-year-old softball player in whom arthroscopic débridement of fibrous tissue beneath the OCD lesion and labral repair were performed. For ICRS grade III or IV lesions, or for larger lesions, osteochondral autograft transplantation and other surgical options have been proposed.11,24 Wyland23 reported the use of osteochondral autograft transfer, Tingey et al22 described treatment using an autologous bone marrow extracellular matrix graft, and Bonfatti et al3 reported the use of a hyaluronate-based scaffold fixed with fibrin glue; however, clinical outcome data for these techniques in glenoid OCD have not been established. Barnes1 recommended drilling into the subchondral bone to promote fibrocartilage formation. Among the various surgical options available, Perdiwala et al17 reported performing BMS even for ICRS grade III and IV lesions, achieving satisfactory cartilage repair and successful return-to-sport outcomes. They concluded that BMS represents a safe and predictable treatment option. We selected the surgical procedure as transarticular drilling or BMS based primarily on arthroscopic assessment of lesion stability. Using this strategy, 8 of 9 patients (89%) were able to achieve RTP. In OCD of the capitellum, microfracture has been reported to yield poor outcomes for lesions larger than 1 cm, and osteochondral autograft transplantation is often recommended.15 However, because the glenoid is a concave articular surface with a distinct loading environment, favorable clinical outcomes may be achieved with BMS even in relatively large lesions, as observed in the present study. In more recent cases, posterior capsular release was performed in conjunction with arthroscopic treatment. Because all patients demonstrated GIRD, improvement of posterior capsular tightness may reduce compressive stress on the posterior glenoid.6,13 Nevertheless, the effectiveness of posterior capsular release could not be independently evaluated because it was introduced in later cases and the small sample size precluded subgroup analysis. Therefore, the favorable outcomes observed in this study may reflect not only treatment of the OCD lesion itself but also improvement of posterior capsular tightness and restoration of shoulder motion. This study has several limitations. First, it is a retrospective case series conducted at a single institution with a limited sample size. Second, because no control group was included, direct comparisons with conservative management or alternative surgical techniques could not be performed. Third, radiographic assessment and determination of RTP may have involved a degree of subjectivity. Fourth, patient-reported outcome measures were not collected because of the retrospective nature of the study, limiting detailed assessment of subjective shoulder function and patient satisfaction after treatment. Conclusions In the present study, glenoid OCD associated with throwing motion was identified in throwing athletes using plain radiographs obtained with the shoulder in an elevated position. This imaging technique may serve as an adjunctive screening method in patients suspected of having glenoid OCD. Although the optimal treatment strategy for glenoid OCD remains unclear, in this case series, 8 of 9 patients (89%) who underwent arthroscopic treatment including BMS achieved RTP at their preinjury competitive level within one year. Arthroscopic treatment including BMS was associated with a favorable RTP rate in this small case series; however, careful consideration of surgical timing is required, as some patients required a prolonged rehabilitation period. Declaration of generative AI and AI-assisted technologies in the writing process AI language assistance (ChatGPT, OpenAI GPT-5.2) was used for English language editing and manuscript refinement. All clinical content, data interpretation, and final approval were performed by the authors. Acknowledgments The authors are grateful to their physical therapists and radiologic technologists of Hisatsune Hospital. Disclaimers: Funding: No funding was disclosed by the authors. Conflicts of interest: The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article. Patient consent: Obtained. Footnotes This study was approved by the Institutional Ethics Committee (approval no. #2018-16-001). References