Association of Cartilage Cap Thickness with Tumor Grade and Outcomes in Secondary Chondrosarcoma

Abstract

Introduction:

Although increased cartilage cap thickness (CCT) is associated with secondary chondrosarcoma in patients with osteochondroma, no studies have assessed its relationship with histologic grade or outcomes after primary resection. Thus, the objectives of this study were to determine if CCT is associated with 1) histologic grade in secondary chondrosarcoma and 2) oncologic outcomes after primary resection of secondary chondrosarcoma.

Methods:

Patients who had a history of osteochondroma and were diagnosed with a biopsy-proven secondary chondrosarcoma at a single institution between January 2000 and December 2023 were included. CCT was measured on T2-weighted imaging or via direct measurement of the resected gross specimen. The primary outcome measure was histologic grade. Secondary outcome measures were overall survival, local recurrence, distant recurrence, and margin positivity. Grade and margin positivity were evaluated using logistic regression models; time-to-event outcomes were assessed using Kaplan-Meier and cumulative incidence methods and Cox models.

Results:

Forty-five patients (30 male; median age, 39 years) were included. Thirty (67%) had grade 1 secondary chondrosarcoma, 12 (27%) had grade 2, and 3 (6.7%) had grade 3. Mean CCT was smaller in grade 1 chondrosarcoma (35 mm; standard deviation [SD], 18 mm) than in grades 2 or 3 disease (76 mm; SD, 30 mm) (P < 0.001). Thicker cartilage cap was associated with increased odds of high (grade 2 or 3) versus low histologic grade (odds ratio, 1.11; 95% confidence interval, 1.05 to 1.21; P < 0.001). There were no statistically significant associations between CCT and overall survival, local recurrence, distant recurrence, or margin positivity.

Conclusions:

Although CCT may be a useful diagnostic marker for secondary chondrosarcoma, thickness itself is not a reliable prognostic marker for predicting long-term patient outcomes. However, its positive association with tumor grade may have implications for surgical planning.

Type of Study:

Diagnostic Study

Level of Evidence:

III

INTRODUCTION

Osteochondroma is the most prevalent benign bone tumor and is characterized by a sessile or pedunculated osseous lesion along the surfaces of bones covered by a cartilage cap.¹ These lesions are most often seen at the sites of tendon insertion in skeletally immature individuals and may be associated with inflammatory, mechanical, or neurovascular compression-related symptoms.² Although the diagnosis is benign, these lesions have the potential for malignant transformation to secondary chondrosarcoma; approximate transformation rates of up to 10% among patients with multiple hereditary exostoses (MHE) and 1% among those with solitary lesions have been reported.^1,3

Secondary chondrosarcoma originating from solitary or multiple osteochondromas is treated with surgical resection with wide margins for optimal local control and survival.⁴ Timely diagnosis of malignant transformation is often difficult due to the clinical and radiologic challenge of distinguishing benign osteochondromas from secondary chondrosarcomas. Historically, cartilage cap thickness (CCT) has been used as a marker for suspected malignant transformation, and thicknesses ranging from 1.5 to 3 cm have been recommended as thresholds for further work-up.^5,6 However, there remains no definitive CCT cut-off value that is pathognomonic for malignancy, and there has been significant debate regarding its use as a suitable marker for assessing malignant potential.^5,7 Previous studies have demonstrated that CCT can be measured with good interobserver reliability using computed tomography (CT) imaging, magnetic resonance imaging (MRI), ultrasound, or direct gross measurement of the histopathologic specimen.^6–9

Although the value of CCT in diagnosing secondary chondrosarcoma in patients with osteochondromas is well-established, no studies have investigated the clinical utility of measuring CCT for patients with secondary chondrosarcoma. In a radiologic case series of 34 patients with secondary chondrosarcoma, Bernard et al. reported a potential relationship between greater CCT and histologic grade, though their findings were not statistically significant.⁶ Furthermore, no studies have examined the relationship between CCT and oncologic outcomes after primary resection of secondary chondrosarcoma. Thus, the objectives of this study were to determine if CCT is associated with 1) histologic grade in secondary chondrosarcoma and 2) oncologic outcomes after primary resection of secondary chondrosarcoma.

METHODS

Study Population

We conducted a retrospective analysis of patients who had a history of osteochondroma and were diagnosed with a biopsy-proven secondary chondrosarcoma at a single tertiary cancer center between January 2000 and December 2023. Biopsies were performed using a core needle, incisional, or excisional biopsy technique. Patients were excluded if there was no MRI available for direct measurement of the CCT, no MRI radiology read of the CCT, and no documented histopathologic measurement of the CCT, or if the biopsy was not obtained within six months of the MRI or histopathologic measurement.

Data Collection

Demographic data were retrospectively retrieved from electronic medical records. Extracted variables included age at biopsy, gender, and anatomic location of tumor. Pathologic and oncologic outcome data were also obtained, including histologic grade, surgical margins, date of local recurrence, and distant recurrence. CCT was measured on T2-weighted imaging using the Visage Imaging software (Visage Imaging Inc, San Diego, CA, USA) as described by Bernard et al⁶ (Figure 1). For a subset of the samples (n=20), CCT was measured by three orthopaedic oncology fellows to evaluate interobserver variability. When an MRI file was not available for CCT measurement, a documented radiologist’s read of the CCT was included. A direct histopathologic measurement of the resected gross specimen was included if neither an MRI file nor a radiology read was available.

Figure 1.

Axial T2-weighted MRI illustrating the measurement technique for the cartilage cap.⁶ The cartilage crevasses are excluded by connecting adjacent peaks in the tidemark (dotted red lines), and the measurement is performed at the thickest portion of the cartilage cap perpendicular to the tidemark.

Outcome Measures

The primary outcome measure was histologic grade, which was determined by a musculoskeletal pathologist using a combination of cellularity, nuclear size, atypical mitosis, and the presence of an abundant hyaline cartilage matrix or mucomyxoid matrix as previously described.^11,12 The secondary outcome measures were overall survival, local recurrence, distant recurrence, and margin positivity.

Statistical Analysis

Descriptive statistics were used to summarize the characteristics of the study cohort. The interobserver reliability of CCT measurements was assessed using intraclass correlation coefficient (ICC); an ICC below 0.5 implies poor agreement, between 0.5 and 0.75 implies moderate agreement, between 0.75 and 0.90 implies good agreement, and above 0.90 implies excellent agreement.¹⁰ We evaluated the relationship of CCT with histologic grade (2/3 vs. 1) and margin status (positive vs. negative) using logistic regression models that adjusted for tumor location and age at surgery; results are reported as odds ratios (ORs) with 95% confidence intervals (CIs). Kaplan-Meier methods were used to estimate overall survival from biopsy. Patients alive at the end of the study period were censored at their last visit date. We estimated the cumulative incidence of local recurrence treating death as a competing event. Cox proportional hazards models were used to model overall survival, local recurrence, and distant recurrence as a function of CCT; the models of overall survival and local recurrence adjusted for tumor grade and age at surgery, but the small number of distant recurrence events precluded multivariable adjustment. For the Cox models of local recurrence and distant recurrence, we used a cause-specific Cox model and censored individuals without recurrence events. Results are reported as hazard ratios (HRs) for overall survival and cause-specific hazard ratios (for local recurrence and distant recurrence, with 95% CIs. Statistical significance was defined as P < 0.05. Data analysis was performed using R statistical software (Version 4.3.1, Indianapolis, IN, USA).

Ethical Approval

The study was conducted in accordance with the Declaration of Helsinki and received approval from our institutional review board (#16-1123). Informed consent requirement was waived due to the retrospective and observational nature of the study.

RESULTS

Study Sample

A total of 45 patients (30 male, 15 female) with a median age at biopsy of 39 years (interquartile range [IQR], 30 to 45) were included in the study (Table 1). Thirty-one patients (69%) had MRIs available for direct CCT measurement. Radiology reports with CCT measurements were utilized for nine patients (20%), and histopathologic measurement of the gross specimen was utilized for five patients (11%). The median follow-up from biopsy was 5.9 years (IQR, 3.3 to 8.3) among individuals alive at the end of the study period. Eight patients (18%) had tumors in the chest wall (18%), five (11%) in the lower extremity, 18 (40%) in the pelvis, four (9%) in the spine, and five (11%) in the upper extremity. The median CCT was highest for tumors in the chest wall (68; IQR, 27 to 83) and pelvis (54; IQR, 30 to 66), followed by those in the spine (38; IQR, 18 to 51), lower extremity (34; IQR, 29 to 50), and upper extremity (26; IQR, 22 to 26). Thirty patients (67%) had grade 1 secondary chondrosarcoma, 12 patients (27%) had grade 2 disease, and three (6.7%) had grade 3. Two patients (4.4%) had dedifferentiated histology. Forty-four patients (98%) underwent primary wide resection at the current institution and were included in the outcomes analyses; negative margins were achieved in 40 of these patients (91%). The remaining patient (2%) was lost to follow-up after the biopsy and was not included in the outcomes analysis.

Table 1.

Characteristics of the study sample

Variable	Total (n=45)	Grade 1 (n=30)	Grade 2 or 3 (n=15)	P a
Age (years)	40.5 ± 13.5	36.9 ± 11.9	47.7 ± 14.1	0.009
Male	30 (66.7)	20 (66.7)	10 (66.7)	>0.9
Dedifferentiated histology	3 (6.7)	1 (3.3)	2 (13.3)	0.3
Tumor location				0.6
Chest wall	8 (17.8)	4 (13.3)	4 (26.7)
Lower extremity	10 (22.2)	8 (26.7)	2 (13.3)
Pelvis	18 (40.0)	12 (40.0)	6 (40.0)
Spine	4 (8.9)	2 (6.7)	2 (13.3)
Upper extremity	5 (11.1)	4 (13.3)	1 (6.7)
Location				0.5
Appendicular	18 (40)	13 (43)	5 (33)
Axial	27 (60)	17 (57)	10 (67)
Cartilage cap thickness (mm)	48.4 ± 29.9	34.5 ± 17.9	76.3 ± 30.0	<0.001
Positive margins	4 (9.1)	2 (6.9)	2 (13.3)	0.6

Values are given as mean ± SD or as n (%)

^aComparisons between grade 1 and grades 2/3 are via Wilcoxon rank sum exact test, Pearson’s chi-squared test, Fisher’s exact test, and Wilcoxon rank sum test. Statistically significant values are expressed in bold.

CCT and Histologic Grade

The interobserver reliability of CCT was excellent, with an ICC of 0.99 (95% CI, 0.98 to 1.00). Mean CCT was smaller in grade 1 chondrosarcoma (35 mm; standard deviation [SD], 18 mm) than in grades 2 or 3 disease (76 mm; SD, 30 mm; P < 0.001) (Figure 2). Grade 2 and 3 tumors were combined into a single group due to the small number of grade 3 tumors. Each increase of 1 mm in CCT was associated with a 11% increase in the odds of higher (grade 2 or 3) histologic grade (OR, 1.11; 95% CI, 1.05 to 1.21; P < 0.001) (Table 2).

Figure 2.

Boxplot of the distribution of CCT by histologic grade, depicting larger CCT among high-grade tumors than low-grade tumors (P < 0.001, Kruskal-Wallis rank sum).

Table 2.

Association between CCT and histologic grade from multivariable logistic regression model

Characteristic	Odds ratio for high (2/3) versus low (1) grade	95% CI	P
1-unit increase in CCT	1.11	1.05 to 1.21	<0.001
1-year increase in age	1.08	1.00 to 1.19	0.04
Tumor location			0.5
Lower extremity	reference	reference	–
Pelvis	0.47	0.02 to 13.1	–
Other (chest wall, spine, upper extremity)	1.89	0.08 to 67	–

CCT = cartilage cap thickness, CI = confidence interval.

Statistically significant values are expressed in bold.

CCT and Oncologic Outcomes

Among the 44 patients who underwent resection at our institution, there were 10 local recurrence events, 3 distant recurrence events, and 7 deaths. Overall survival was 93% (95% CI, 86% to 100%) at 1 year, 91% (82% to 100%) at 2 years, and 81% (70% to 95%) at 5 years (Figure 3). The appendix presents Kaplan-Meier curves for overall survival by tumor location (Appendix, Figure 1), tumor grade (Appendix, Figure 2), and both location and grade (Appendix, Figure 3).

Figure 3.

Kaplan-Meier estimates of overall survival.

The cumulative incidence of local recurrence was 9.1% (2.9% to 20%) at 1 year, 14% (5.6% to 26%) at 2 years, and 26% (13% to 40%) at 5 years (Figure 4). By 3 years, the cumulative incidence of distant recurrence was 7.2% (1.8% to 18%). Only four patients had positive margins; their median CCT was 41 (IQR, 27 to 51), compared with 45 (IQR, 26 to 66) among patients with negative margins. In Cox proportional hazards and logistic regression models, CCT was not associated with overall survival (HR, 1.01; 95% CI, 0.98 to 1.04; P = 0.7), local recurrence (HR, 0.98; 95% CI, 0.95 to 1.02; P = 0.3), distant recurrence (HR, 1.02; 95% CI, 0.99 to 1.05; P = 0.14), or margin positivity (OR, 0.98; 95% CI, 0.93 to 1.02; P = 0.5). However, we observed statistically significant associations between higher tumor grade and overall survival (HR, 4.41; 95% CI, 1.33 to 14.7; P = 0.02) and local recurrence (HR, 5.71; 95% CI, 1.81 to 18.1; P = 0.003) (Table 3).

Figure 4.

Cumulative incidence of local recurrence.

Table 3.

Results of Cox proportional hazards models examining associations of CCT with overall survival and local recurrence.

Characteristic	Univariable			Multivariable
	HR	95% CI	P	HR	95% CI	P
Overall survival
CCT	1.02	1.00 to 1.04	0.069	1.01	0.98 to 1.04	0.7
Age at biopsy	NA	NA	NA	1.03	0.97 to 1.09	0.4
Tumor grade	NA	NA	NA	4.41	1.33 to 14.7	0.016
Local recurrence
CCT	1.01	0.99 to 1.04	0.4	0.98	0.95 to 1.02	0.3
Age at biopsy	NA	NA	NA	0.98	0.92 to 1.04	0.4
Tumor grade	NA	NA	NA	5.71	1.81 to 18.1	0.003

CCT = cartilage cap thickness, CI = confidence interval, HR = hazard ratio, NA = not applicable.

Statistically significant values are expressed in bold.

DISCUSSION

Secondary chondrosarcoma arising from primary osteochondromas may be challenging to distinguish from benign osteochondromas.^5,7 A thicker cartilage cap is associated with the diagnosis of secondary chondrosarcoma; however, its reliability as a marker for a more aggressive clinical and histologic pathology has not been demonstrated. In this study, we explored the association between CCT and histologic grade in secondary chondrosarcoma and between CCT and oncologic outcomes after primary resection of secondary chondrosarcoma. Our analysis found that greater CCT is associated with a higher tumor grade in secondary chondrosarcoma. However, no statistically significant associations were evident between CCT and overall survival, local recurrence, or distant recurrence when we adjusted for tumor grade and age (overall survival, local recurrence).

A major finding in this study was that CCT was strongly associated with histologic grade in secondary chondrosarcoma. On average, CCT was more than 40 mm thicker in grades 2 and 3 tumors than in grade 1 tumors. Although a radiologic series by Bernard et al. demonstrated a pattern consistent with this result, their study of 34 patients with secondary chondrosarcoma may have been underpowered and did not find a statistically significant assocation.⁶ Thus, the findings from this larger study contribute evidence in support of the long-held clinical suspicions that an unusually thick cap is not merely a marker of malignant transformation but also a surrogate for biologic aggressiveness.

In our study, despite being associated with higher histologic grade, CCT was not significantly associated with overall survival, local recurrence, or distant recurrence, though our analysis is limited by the small number of events. In the models that adjusted for grade, we observed a statistically significant relationship between grade and overall survival and local recurrence, consistent with Laitinen et al.’s finding that grade 2 or higher pathology in secondary pelvic chondrosarcoma is strongly associated with increased local recurrence and decreased overall survival.⁵ This is also consistent with reports that patients with low-grade chondrosarcomas in general have a significantly lower rates of local recurrence and distant recurrence and longer overall survival than their counterparts with intermediate- and high-grade chondrosarcomas.^13–15 This uncoupling suggests that although higher CCT may be a useful diagnostic marker for higher grade tumors in secondary chondrosarcoma, the thickness itself is not a reliable prognostic marker for predicting long-term patient outcomes.

Nevertheless, this study adds an additional datapoint to guide the counseling of patients on the management of secondary chondrosarcoma. Previous research has demonstrated that delayed diagnosis and definitive treatment of higher-grade tumors is associated with increased rates of distant metastasis and poorer overall survival.^15,16 Thus, patients with particularly thick cartilage caps may benefit from prompt discussions regarding expedited surgical planning, as these cases are more likely to represent higher grade tumors.

This study has several limitations. First, the retrospective single-center design and the referral patterns inherent to a tertiary sarcoma center may have introduced selection bias and over-represented more complex cases. Second, the sample size and number of events were modest, particularly for grade 3 tumors, which limited our statistical power. Furthermore, although the interobserver reliability for MRI measurements of CCT was excellent, 14 patients required extracting CCT from radiology reports or gross pathology, which introduced heterogeneity in measurement technique. However, previous data have demonstrated that both MRI and direct measurements are reliable for measuring CCT.^6–8 In addition, this study lacked data on adjuvant systemic therapy or radiotherapy, which may confound outcomes analyses. Finally, although we had a median follow-up of 5.9 years, this still may have been insufficient to capture late recurrences typical of chondrosarcomas.

Despite these limitations, to our knowledge this is the largest study to date that directly evaluated CCT as both a diagnostic and prognostic factor in secondary chondrosarcoma. Although no specific CCT threshold exists that is pathognomonic for malignancy, we found that a larger CCT is associated with a higher histologic grade in secondary chondrosarcoma. Future studies through multicenter collaborations are needed to confirm these findings. In addition, studies that incorporate advanced radiomics and molecular profiling may help refine treatment algorithms and personalize surveillance schedules for patients with secondary chondrosarcoma.

CONCLUSION

In secondary chondrosarcoma, increased CCT is associated with a higher tumor grade, and a higher tumor grade is associated with decreased overall survival and increased local recurrence. However, there were no statistically significant associations observed between CCT with overall survival, local recurrence, distant recurrence, or margin positivity. While thickness itself is not a reliable prognostic marker for predicting long-term patient outcome, its positive association with tumor grade may have implications for patient counseling and surgical planning.