Outcome of conservative and surgical treatment of enchondromas and atypical cartilaginous tumors of the long bones: retrospective analysis of 228 patients

Enchondromas belong to the most frequently found cartilaginous bone lesions [1, 2]. Most enchondromas are asymptomatic but malignant transformation into secondary chondrosarcoma is possible in 1–9% depending on localization, pain, size, and radiologic presentation [1‐5]. Chondrosarcoma grade I is a low-grade malignant tumor characterized by local aggressiveness leading to destruction of cortical bone, increased pain, frequent local recurrences, but usually does not develop metastases; consequently, it was renamed by the WHO to be an ‘atypical cartilaginous tumor’ (ACT) [1, 2, 6]. Several reports discuss metastatic potential with a 2–8% probability depending on localization and recurrence [7, 8]. Histological and radiological differentiation between clinically and radiologically aggressive enchondromas and ACTs is extremely challenging and often not possible, so treatment depends mainly on the clinical behavior without specific differentiation between both entities [1, 4, 9‐13]. Treatments include a wide spectrum from wait and see to various surgical strategies without significant evidence and consensus to support the different strategies [4, 14‐17]. Lesions with high local aggressiveness and pain are often surgically resected, whereas smaller asymptomatic enchondromas are most often treated conservatively [7, 10]. Nevertheless, there is no evidence if aggressive enchondromas and ACTs should be treated surgically or not. Even for ACTs, conservative strategies with radiographic follow-up are proposed to prevent the overtreatment and morbidity associated with surgery despite being a low-grade malignant tumor [4, 9, 18].

Differences of clinical outcome between conservatively and surgically treated patients are unknown and treatment strategies remain highly individual. Therefore, the study objective was to perform the thorough evaluation of radiological and clinical outcomes in our series of 228 cases including recurrences, metastases, and functional and emotional parameters of both groups to distinguish potential differences and to improve decision making for the treatment of these benign to low-grade malignant bone tumors.

All patients diagnosed with either aggressive enchondroma or ACT treated in our university orthopaedic oncology outpatient clinic between 2005 and 2015 were retrieved from a prospectively enrolled database and retrospectively analyzed. Approval was given by the local ethics committee of the University of Heidelberg, Germany. Only aggressive lesions that were suspicious for enchondroma or ACT with the need for continuous follow-up in our oncology department were included. Asymptomatic enchondromas without radiological signs of aggressiveness and without recommended further follow-up through our department were not included. Cases initially suspicious for higher grade II or III chondrosarcomas (extensive cortical bone enlargement, large interruption of the cortex, extensive invasion of the soft tissues) were excluded. Further exclusion criteria were follow-up shorter than 24 months (n = 92), and Ollier’s disease (n = 16). Hence, 228 consecutive patients could be included. Decision of conservative versus surgical treatment was done individually as clinical and radiologic appearance would have justified both strategies. Retrospective analysis of medical histories did not reveal hard criteria for initial decision making as both groups showed specific local pain and radiological signs of aggressiveness (large lesion sizes; endosteal scalloping with partial destruction of the cortical bone; similar calcification patterns). Analysis of both alternatives with conservative follow-up or immediate surgical treatment had been discussed with the patients. After consideration of individual advantages and disadvantages decision was made together with the patient, as there is still no evidence and consensus in the literature, which lesions should be treated surgically or not [4, 9, 14, 15, 18]. Radiological evaluation with x-rays and MRIs was performed initially and at regular intervals between 6 and 24 months until final follow-up in both conservatively and surgically treated cases. Images were reviewed by a musculoskeletal radiologist sub-specialized in orthopaedic oncology. Radiologic reports were well-documented with descriptions of the chondroid matrix, calcification patterns, and destruction of cortical bone (endosteal scalloping). Due to heterogeneous geometrical configurations of the lesions, the size was evaluated by measuring the longest diameter. CT-scans of the lesions were only performed in limited cases. Pulmonary x-rays and in the cases of histologically diagnosed ACTs an additional pulmonary CT scan was obtained to rule out metastases.

Histological evaluation was performed by the pathologists of our university pathology department. Final diagnosis was confirmed in a consensus decision together with the senior pathologist (GM). The main histological criterion to distinguish ACT from enchondroma was permeative infiltrative tumor growth with encapsulation of lamellar bone trabeculae [19]. Further criteria were hypercellularity, cellular atypia, myxoid areas, and cell necrosis. Clinical evaluation with physical examination was performed upon presentation to our musculoskeletal oncology outpatient clinic, which is responsible for all orthopaedic tumor cases at our university. Patient demographics and clinical histories were retrieved including detailed information on surgical treatment, recurrences, and complications. For systematic evaluation of pain, patient satisfaction, and functional limitation at final follow-up, an additional telephone interview was performed with a standardized questionnaire available for n = 59 patients in the surgery group and n = 125 patients in the non-surgery group. Analog scale ratings for pain and patient satisfaction from 0 to 10 were obtained. Clinical function was semi-quantitatively evaluated for the affected body part with the oxford knee score [20], oxford hip score [21], foot and ankle disability score [22], or quick disabilities of arm shoulder and hand score [23]. Score values were recalculated into a 4-point grading with a range from 0 to 3 points (0 = no functional deficit; 3 = high disability with poor function). Additionally, outcome was evaluated by the well-established musculoskeletal tumor society (MSTS) score combining functional parameters with emotional acceptance and pain [24].

The mean follow-up was 82 (median 75) months with a range from 25 to 266 months for the whole series. Mean (median) follow-up of the 153 conservatively treated cases was 88 (79) months with a range from 25 to 266 months versus 70 (57) months with a range from 26 to 217 months for the 75 surgically treated cases. Localizations, demographics, and treatment strategies are presented in Table 1.

We analyzed the outcome after both surgical and conservative treatment of clinically and radiologically aggressive benign to low-grade malignant chondrogenic tumors of the long bones of the upper and lower extremities. The surgically treated cases were evaluated histologically, and diagnosed as benign enchondromas in the majority of the cases (n = 59), while low-grade malignant ACT was diagnosed in only a few cases (n = 9); in the remaining 7 cases, the histology was unable to differentiate between ACT and enchondroma, resulting from the known difficulties in histologically differentiating both entities [1, 9, 10, 13, 19]. Comparison of all conservatively and surgically treated lesions showed excellent clinical results, however, the functional limitations were increased after surgical treatment. Surgically treated patients with histological diagnosis of enchondroma showed worse results for functional limitations, pain, and the MSTS score compared to conservatively treated patients. Sub-analysis of cases with larger tumor size (> 4.4 cm) displayed significantly worse clinical results with regard to pain and the MSTS score in the surgery group compared to the conservative cohort. For surgically treated patients, the complication rate has to be considered before carrying out treatment. Seven of the 75 surgically treated patients had complications requiring revision surgery. The revisions were necessary due to five peri-implant fractures and two intra-articular screws, which suggests lower complication risk in cases receiving the more conservative approach without osteosynthesis but otherwise non-inferior results – a finding, which was statistically significant. This underscores higher complication risks in cases treated with more sophisticated surgical strategies [25‐27].

We only found 3 recurrences in the 75 surgically treated patients. Reasons for this might be the majority of histologically benign lesions in the present study, limited follow-up time, but also surgical strategy. In surgically treated cases, common strategies were performed to reduce the risk for recurrence [28] using rigorous curettage, additional administration of a high-speed burr, the application of H₂O₂ and in most cases filling the former tumor cavity with bone-cement, which has the potential added effect of thermal necrosis of remaining tumor cells due to the exothermic polymerization process. The selected surgical treatment approach, involving the extensive intralesional curettage instead of a wider more aggressive resection is a well-reported method, which results in improved functional outcome and sufficient oncologic safety compared to more aggressive surgical strategies [14, 16, 17, 29], which was also supported by our low recurrence rate. We could not find significant differences between the cases treated surgically using bone-cement and cancellous bone, but this may also be attributed to the lower number of cases with cancellous-bone filling.

Interestingly, all recurrences occurred in cases with initial histological diagnosis of benign enchondroma. Of those, two again were histologically classified as enchondroma upon revision. In both cases, it cannot be confirmed that it was in fact a recurrence and not residual tumor. The third recurrence was histologically classified as ACT upon revision. This verifies the evident risk for malignant transformation of initially benign enchondromas with reported risks from 1 to 9%, [1‐5] with the need for long term follow-up of aggressive enchondromas. Another explanation might be the known difficulties of histological diagnosis.

We did not find transformation into higher grade chondrosarcomas. Limited follow-up time but also selection of lesions located in the long bones of the extremities without axial and pelvis tumors may explain missing further transformations into higher grade chondrosarcomas [1, 7]. Rigorous follow-up of radiologically aggressive lesions, however, seems crucial to identify a potential transformation into ACT or even higher-grade chondrosarcomas. The present study cannot answer the question as to how often radiological follow-up is required. We generally recommend a first follow-up MRI after 3 to 6 months and then every 12 months thereafter. A problem with the conservative strategy might be patient compliance, which is essential for sufficient oncological safety. Ideally, follow-up should be carried out in specialized orthopaedic oncology departments.

We only analyzed lesions of the long bones, since more distally located enchondromas of the small bones of the hands and feet show histological differences and appear to have a lower potential for malignant transformation or metastasizing compared to lesions in the long bones [30, 31].

No metastases were noted in our collective during the follow-up period of 7 years, however, later metastatic occurrence after this time cannot be ruled out. The risk for later metastases should be extremely low, since pulmonary metastases were only described in ACTs of the long bones after local recurrence and transformation into higher grade (grade II) chondrosarcomas [7, 32]. Based on these experiences, we currently do not recommend regular pulmonary imaging for these benign to low-grade lesions unless we have reasonable suspicion for malignant transformation or metastatic disease.

In 20 cases, we performed an initial incision biopsy prior to intralesional curettage. In all cases, our initial histological results were verified by final histology after intralesional curettage, despite the reported risk for sampling error and underestimation of the degree of dedifferentiation [33, 34]. The only case showing a possible misclassification was the one case with an initial histological classification of enchondroma after incisional biopsy and then recurrence 6 years later showing evidence of ACT. The time-intervals, however, support a later malignant transformation from enchondroma into ACT instead of misdiagnosis in the biopsy. Hence, we still recommend incisional biopsy if there is any doubt regarding malignancy. The present study did not include cases with initial suspicion for higher grade chondrosarcomas (extensive cortical bone enlargement, large interruption of the cortex, extensive invasion of the soft tissues). In such cases, different treatment algorithms are required. Here we always recommend initial biopsy and extralesional tumor resection. For cases with radiological diagnosis of aggressive enchondroma or ACT, where a higher-grade malignancy seems radiographically unlikely, the intralesional curettage strategy is favorable. This is confirmed by the fact that we did not find higher grade chondrosarcomas in the immediately resected lesions in the present study.

It remains unclear, what is the most favorable treatment option for clinically and radiologically aggressive enchondromas and ACTs, since scientific evidence for or against surgery is missing for these benign to low-grade malignant tumors. To the best of our knowledge, this is the first study with such a large patient cohort to evaluate this collective and compare surgical and non-surgical outcome. A potential selection bias of this retrospective study, however, has to be discussed as an important limitation. Decision making towards surgery or conservative follow-up could not be controlled in the present study. This is also reflected in the slightly, but statistically larger lesion sizes of the surgically treated cohort. To minimize the selection bias between the surgery and non-surgery groups, we performed further sub-group analyses. We selected sub-groups with tumor sizes smaller and larger than the mean size of the series (4.4 cm). The sub-group with lesions larger than 4.4 cm offered the best comparability between the surgery and the non-surgery groups, as in these cases lesion size was not significantly different between both groups. Histopathologic differences between the surgery and the non-surgery groups cannot be ruled out as most conservatively treated tumors did not receive a biopsy. The conservatively treated lesions did not show considerable increase in tumor size or progression of scalloping until final follow-up. They may be more comparable to the cases that were surgically treated and histopathologically classified as enchondromas.

Which chondroid lesions clearly need surgery and which do not need surgery, cannot be sufficiently answered by the present study, as we only differentiated the lesions in respect to lesion size and histology. Further subgroups depending on initial clinical appearance and radiological signs of aggressiveness could not be obtained as we did not find significant differences in our highly selected study population. Future studies should also compare subgroups of lesions with considerably different aggressiveness, to achieve objective data on which lesions will benefit from surgery.

This is the first study proving benefits in clinical and functional outcome of cases conservatively followed as opposed to those treated surgically. Sampath et al. only recommend surgical therapy for actively growing lesions with more than 6 mm growth within 3 years [15]. Both good clinical outcome and sufficient safety have been described using this watch-and-wait strategy [4, 9, 17, 18]. From a histological standpoint, surgery would not have been needed to achieve sufficient oncological safety in our collective, since the histology revealed that the majority of our cases were benign enchondromas.

Our series of conservatively and surgically treated enchondromas and ACTs displayed excellent clinical outcomes. The conservative approach showed a significantly better functional outcome compared to the surgical treatment, and appears to be a sufficiently safe treatment option if strict clinical and radiographic follow-up is assured. For the cases treated surgically, the feasibility of intralesional curettage was confirmed but overall results did not prove superiority compared with only conservative wait and see strategy.