Presented in part at the 38th Congress of the International Society of Paediatric Oncology (SIOP), September 17–21, 2006, Geneva, Switzerland, and awarded with the SIOP best poster prize in the discipline “adolescent and young adult oncology.”
We declare that we have no conflict of interest.
Involved reference panel members: Oncology: E. Koscielniak, J. Treuner, and S.S. Bielack, Olgahospital, Klinikum Stuttgart, Pediatrics 5 (Oncology, Hematology, Immunology), Germany. N. Graf, University of Homburg, Department of Pediatric Oncology, Germany. H. Jürgens, University Children's Hospital Muenster, Department of Pediatric Hematology and Oncology, Germany. D. Niethammer, University of Tuebingen, Department of Pediatric Oncology, Germany. K. Winkler, University of Hamburg, Department of Pediatric Oncology, Germany. Pathology:H. Bürger, University of Muenster, Department of Pathology, Germany. G. Delling, University of Hamburg, Institute of Osteopathology, Germany. D. Harms, U. Jänig, and I. Leuschner, University of Kiel, Department of Pediatric Pathology, Germany. G. Jundt, University of Basel, Department of Pathology, Switzerland. P. Meister, University of Munich, Department of Pathology, Germany. A. Schulz, University of Giessen, Department of Pathology, Germany. Radiology: P. Winkler, Olgahospital, Klinikum Stuttgart, Department of Pediatric Radiology, Germany. Radiotherapy: B.F. Schmidt, Katharinenhospital, Klinkum Stuttgart, Department of Radiotherapy, Germany. A. Schuck, University of Muenster, Department of Radiotherapy, Germany. R. Schwarz, University of Hamburg, Department of Radiotherapy. Surgery: G. Gehrke, Henriettenstiftungskrankenhaus, Department of Maxillofacial Surgery, Hannover, Germany. U. Heise, Hamburg, Germany. R. Schmelzle, University of Hamburg, Department of Maxillofacial Surgery, Germany. W. Winkelmann, University of Muenster, Department of Orthopedics, Germany
Centers contributing patients: University Hospital Bonn, Germany. Carl-Thiem-Klinikum Cottbus, Germany. Klinikum Dortmund, Germany. University Hospital, Ulm, Germany. University Hospital Duesseldorf, Germany. University Hospital Greifswald, Germany. University Hospital Hamburg, Germany. University Hospital Heidelberg, Germany. University Hospital Jena, Germany. University Hospital Kiel, Germany. Olgahospital, Klinikum Stuttgart, Germany. University Hospital Tuebingen, Germany
Mesenchymal chondrosarcoma (MCS) is a rare tumor with a strong tendency toward late recurrences leading to reported 10-year survival rates below 50%. The recommended treatment is resection with wide margins; the effectiveness of chemo- and radiotherapy remain poorly defined. As reports about MCS in young patients are scarce, treatment and outcomes of children/adolescents/young adults in the CWS and COSS studies were investigated.
Since 1977, 15 of >7000 CWS and COSS patients ≤25 years had a confirmed diagnosis of MCS.
The median age was 16.6 (range, 1–25) and median follow-up 9.6 years (range, 1–22). Four MCS were osseous and 11 extraosseous. All but 1 patient had nonmetastatic disease. Tumor sites were head/neck (n = 6), paravertebral (n = 3), pelvis (n = 3), limbs (n = 2), and kidney (n = 1). All tumors were resected, but only 8 completely. Thirteen individuals received chemotherapy, 6 were irradiated. Actuarial 10-year event-free and overall survival rates were 53% and 67%, respectively. Four recurrences occurred, all within 4 years from diagnosis (3 local, 1 combined; 2 of these in irradiated patients). One of these patients survived after surgery and radiation for local recurrence. Seven of 8 patients whose tumors were completely resected during primary treatment, but only 4 of 7 patients with incomplete surgery survived disease-free.
Mesenchymal chondrosarcoma (MCS) is a rare malignancy characterized by a biphasic histologic pattern of small undifferentiated round cells intermixed with islands of well-differentiated cartilaginous matrix. Because of its aggressive clinical behavior, MCS is regarded as a high-grade sarcoma both in the grading systems of the French Federation of Cancer Centers Sarcoma Group (FNCLCC) and the National Cancer Institute (NCI), although it would not fulfill the formal histopathologic criteria for this classification.1
MCS differs from classic chondrosarcoma (CCS) as to its 1) rarity (few cases have been published in contrast to CCS, the second most common primary malignancy of bone1, 2); 2) age distribution (MCS typically occurs in young adults, whereas most CCS patients are >50 years old); 3) high malignancy (MCS are characterized by a strong tendency toward late local and metastatic [pulmonary] recurrences leading to 10-year survival rates far below 50%,3–8 whereas most CCS have a more favorable prognosis); and 4) a high proportion of extraskeletal tumors (about 1/3 of MCS occur in soft tissues, whereas extraosseous CCS account for <1% of all CCS1, 9, 10).
Since the first report by Lichtenstein and Bernstein in 1959,11 fewer than 500 cases of MCS have been published, mainly as case reports, within literature reviews, or as retrospective histopathologic or radiologic series without adequate clinical data or information concerning the tumor's clinical behavior, treatment, and outcome.11–22 To the best of our knowledge, there are only 8 studies about MCS including more than 10 patients with sufficient clinical data and follow-up.3–6, 23–26 All 3 series encompassing more than 20 individuals were published more than 20 years ago.3, 6, 23 The paucity of MCS and individualized therapy outside of clinical trials are reasons for the current lack of consensus about the best therapeutic approach. Resection with wide margins is usually recommended, but frequently not feasible (eg, in parameningeal sites). The effectiveness of adjuvant chemo- and/or radiotherapy in addition to surgical resection is not well defined.4, 5, 8, 9, 12, 26
We prospectively collected data on patients with this rare tumor registered with the Cooperative soft tissue sarcoma (CWS) and osteosarcoma (COSS) Study Groups of the German Society of Pediatric Oncology and Hematology (GPOH). Our aim was to describe clinical characteristics, treatment, and outcome of children, adolescents, and young adults with MCS to contribute to a better understanding of its clinical management in the young.
MATERIALS AND METHODS
Fifteen patients with MCS from a study population of more than 7000 patients with a primary diagnosis of soft tissue or bone sarcoma registered in the consecutive CWS and COSS trials conducted since 1980 and 1977, respectively, were analyzed: CWS-8627 (n = 1), CWS-9128 (n = 4), CWS-9629 (n = 5), CWS-2002-P (n = 2), COSS-8230 (n = 1), and COSS-9631 (n = 2). Registered patients were regarded as ‘observational,’ ie, trial treatment was optional. Specific guidance for treatment of MCS was only provided in protocols CWS-91, -96, and -2002-P. Frequently, therapy was administered on an individual basis after consultation with the respective study group centers, because there was—and still is—no consensus about the optimal treatment strategy and the role of systemic therapy. Written informed consent was obtained from patients, their guardians/parents, or both, depending on patient age and according to the respective provisions and the Declaration of Helsinki. All trials were conducted in accord with the regulations of the appropriate ethic's committees and accepted by them. A single patient from our series has been published in a surgical-pathologic case report previously.32 Our report is based on follow-up data as of July 2006.
All diagnoses were confirmed by members of the respective CWS and COSS reference pathology panels (see first-page footnote) according to published criteria. Briefly, these are the presence of a typical biphasic pattern of undifferentiated small round cells blended with islands of hyaline cartilage. The small round cells resemble Ewing tumors and typically show a hemangiopericytomatous vascular pattern.1
Initial Staging and Data Collection
Data collection was performed prospectively as described previously.27, 28, 30 In brief, patient data included age and gender, and tumor-related information contained date of diagnosis, histology, tumor size, anatomical site, and the UICC tumor-node- metastasis (TNM) classification. Tumors larger than 5 cm in greatest dimension were classified as Tb, smaller tumors as Ta. If the tumor originated from bone it was classified as osseous (oMCS), if it originated in soft tissues as extraosseous MCS (eMCS). Because many oMCS had a large soft tissue mantle and many eMCS bony erosion, it was sometimes difficult to distinguish their origin and these MCS were coded with the suffix “mixed” as (oMCS-mixed) and (eMCS-mixed). The initial staging procedures were recommended in the respective CWS and COSS protocols, but compliance with this guidance was not controlled and depended on the decision of the treating centers.
Treatment-Related Data Collection and Follow-up
Treatment-related data included postsurgical group (according to the pTNM-classification: pT1 and pT2: complete resection, pT3: incomplete resection with pT3a: microscopic residual tumor, pT3b: macroscopic residual tumor, and pT3c: adjacent malignant effusion), application of radiotherapy (including mode and dosage), chemotherapy, response, start and end of therapy, and status at the end of therapy. If patients were treated with chemotherapy and/or radiation the combination and dosage depended on the respective trial: prescribed drugs included dactinomycin (AMD), carboplatin (CAR), cisplatin (DDP), cyclophosphamide (CYC), doxorubicin (DOX), etoposide (ETO), ifosfamide (IFO), high-dose methotrexate (MTX), and vincristine (VCR) given in various combinations. Follow-up data collected prospectively included date and anatomical site of recurrence, patient status at last follow-up, and—for deceased patients—cause and date of death. All available information of the CWS and COSS databases and patient files were reviewed by the first author. In case of unclear information, the centers were asked to supply the missing medical, surgical, radiologic, and pathologic reports retrospectively.
Definition of Terms
Response was assessed after induction chemotherapy as the proportion of tumor volume reduction in the CWS27 and as the degree of histologic devitalization according to the Salzer-Kuntschik criteria in the COSS trials.33 Complete remission (CR) was defined as lack of residual tumor upon imaging at the end of primary therapy (based on the evaluation of the treating institution). If there was a residual structure detected by imaging, patients were classified as CR if second-look surgery was negative or if the residual structure remained unchanged during follow-up imaging for at least 6 months after the end of therapy. Time to recurrence was defined as the interval between diagnosis and detection of recurrence. The failure pattern of a first recurrence was classified as either locoregional, metastatic, or combined (ie, concomitant local and metastatic). Locoregional recurrence was defined as recurrence at the primary site or in tissues directly adjacent to the primary anatomical tumor localization including regional lymph nodes. Metastatic recurrence was defined as appearance of tumor at distant sites including distant malignant effusions.
Statistics were calculated using Statistica v. 6 (Statsoft, Tulsa, Okla) and SPSS v. 11.0 (Chicago, Ill). Overall survival (OS) and event-free survival (EFS) were calculated using the Kaplan-Meier estimator.34 For OS the time from diagnosis to death, either from therapy, disease, other, or to last follow-up was calculated. For EFS the time from diagnosis to first recurrence or progression (any evidence of growth of a tumor that was not in clinical CR) or last follow-up was calculated. If there was no event the survival data were censored at last follow-up.
The median age of the 15 patients was 16.6 years (range, 1.4–25.2). Median follow-up was 9.6 years (range, 1.3–21.5); 6 individuals had a follow-up shorter than 5 years (Table 1). Four MCS were osseous and 11 soft tissue MCS. Only 1 patient had (pulmonary) metastatic disease at presentation. All but 1 patient presented with unspecific clinical symptoms like swelling, pain, or diplopia; the exception was an individual who was diagnosed without clinical symptoms during a routine medical checkup (chest X-ray before military service). A positron emission tomography (PET) scan was performed in 1 patient in addition to the regular imaging with x-rays, magnetic resonance imaging (MRI), and computed tomography (CT); the primary as well the metastases were PET-positive.
Table 1. Patient Characteristics, Treatment, and Outcome
Primary tumor sites were head/neck (n = 6), paravertebral (n = 3), pelvis (n = 3), limbs (n = 2), and kidney (n = 1). Thirteen individuals received combination chemotherapy, all with at least 4 different cytotoxic drugs. All tumors were resected, but only 8 completely (ie, without micro- and macroscopic residual disease). Six patients were irradiated with dosages ranging from 44.4 to 54 Gy; 4 of the irradiated individuals had incompletely resected disease. Response to induction chemotherapy could be assessed in 7 patients during primary therapy: chemotherapy caused >50% tumor volume regression or histologic devitalization of >50% in 4 individuals.
Four recurrences occurred among the patients with CR achievement, all within 5 years from diagnosis (3 local, 1 combined; 2 of these in irradiated patients). In 2 patients the recurrences were detected because of clinical symptoms (pain and diplopia) and without clinical symptoms upon imaging during routine follow-up in the remaining 2 individuals. The site of the systemic recurrence was the lungs and the timepoint of recurrence was 1.5 years after primary diagnosis. One patient with recurrent disease survived after macroscopically complete surgery and radiation for local recurrence. One individual developed a secondary malignancy (AML) in first CR, but is currently alive in CR of both malignancies.
Two patients received high-dose chemotherapy followed by autologous stem cell transplantation after conventional chemotherapy during primary treatment based on the decision of the respective treating centers: 1 is currently alive, but follow-up is <2 years in this individual. Response to recurrence treatment could be assessed in a single patient after development of a combined recurrence; with salvage chemotherapy the pulmonary metastases disappeared upon imaging (CT thorax) and the tumor volume of the primary decreased substantially.
Seven of 8 patients whose tumors were completely resected during primary treatment, but only 4 of 7 patients with incomplete surgery, survived disease-free. Three of the 4 incompletely resected but irradiated patients survived, 2 of them event-free. Actuarial EFS and OS at 10 years for all 15 patients were 53% and 67%, respectively (Fig. 1).
MCS is 1 of the rarest sarcoma histiotypes of patients registered in the databases of the cooperative multicenter, multinational trials of the Cooperative soft tissue sarcoma (CWS) and osteosarcoma (COSS) Study Groups of the GPOH. Fewer than 0.2% of more than 7000 sarcoma patients registered had this diagnosis. Although our series encompasses no more than 15 patients, it is among the larger analyses of unselected patients with a detailed description of treatment and follow-up. In particular, it is 1 of the largest reports about MCS in childhood and adolescence, where this sarcoma is even rarer than in adults, although our sample is certainly not representative or population-based even for young MCS patients.
Although there are usually no tumor necrosis and the mitotic count is rather low, MCS is classified as grade 3 in the grading systems of the French Federation of Cancer Centers Sarcoma Group (FNCLCC) and the National Cancer Institute (NCI).35 This classification is based on its highly malignant clinical behavior, with frequent (late) local and metastatic recurrences leading to poor long-term outcomes. Much of the current knowledge about the clinical course of MCS was gained from the largest retrospective series published in 1986, which included information on treatment of 78 patients with adequate follow-up.3 Thirty-seven of these patients had been treated with radiation, chemotherapy, or both after initial surgery. This report, however, had several limitations. Nearly every second patient had been reported previously, most data had been gained from consultation files, and the vast majority of patients were older than 30 years. No information was provided correlating tumor characteristics with the quality of surgery, and the less amenable tumors may thus have been selected for adjuvant therapies. The vast majority of patients in that series developed metastases (mainly pulmonary) after an average follow-up of 4.3 years (maximum interval: 23 years) and 10-year survival was merely 26%. The prognosis was similar for extraskeletal and osseous MCS.3
This poor long-term prognosis was confirmed by other single-center reports, which all had their limitations: either a small number of patients, retrospectiveness, incomplete information about therapy, focus on pathologic-anatomic aspects, or inadequate follow-up.4–6, 8, 23 In addition, all larger reports about MCS were published more than 2 decades ago. More recent publications encompassing at least 10 patients focused on MCS in specific sites, eg, the sinonasal tract,24 jaws,22, 26 mandible,36 orbit,37 or the central nervous system (CNS).21 Prognosis in these locations seemed to be better than the historical results from the larger unselected series (best 10-year survival: 56%26). It is still open to discussion whether this might have been because of advantageous tumor locations similar to the more favorable and unfavorable sites in other soft tissue27 or bone sarcomas,38 whether it could be attributed to more favorable tumor characteristics (eg, smaller tumor sizes), or whether it was the result of the more widespread use of radio- and chemotherapy in more recently treated patients. Again, small patient numbers and heterogeneous cohorts preclude further interpretation.
Despite the rarity of MCS, there is some agreement that resection with wide margins is the ‘gold standard’ of treatment.3, 24–26 There is, however, no consensus as to whether additional radiation and/or systemic treatment are indicated. Proponents of supplemental treatment argue that the propensity of MCS for local and systemic recurrences requires treatment aimed at improving local and systemic control.5, 11 As early as 1981, Harwood et al.5 noted that MCS should be treated with concomitant radiochemotherapy whenever possible, because these had been found to be effective in some tumors. Opponents argue that there is still no convincing evidence concerning the effectiveness of either systemic treatment or radiotherapy.3, 16, 21, 23, 26
Our series of 15 MCS differs from others: the median age was young, most tumors were extraosseous, the patient population was not restricted to selected tumor locations, and most patients received chemotherapy. The outcome of our unselected cohort compares favorably to published results from series encompassing 10 patients or more.3–6, 16, 21, 23, 24, 26, 39 Also, despite long-term follow-up, the characteristic late metastatic recurrences were not observed in our sample. We cannot determine whether the relatively good prognosis was due to favorable patient and/or tumor characteristics or whether it could be attributed to effective therapy. Similar to other sarcomas, in which age is a well-established prognostic factor,40, 41 it has also been suggested as a clinical parameter of prognostic relevance in MCS.5, 7, 42 The young age in our sample might therefore have been an important nontreatment-associated factor related to a more favorable prognosis. As there are so few reports about MCS in children and adolescents,4, 12, 14, 20 it will remain difficult to prove this hypothesis. In the largest published series of 19 young MCS patients aged <21 years from 1983, Dabska and Huvos4 reported a 10-year OS probability of only 20%, but did not supply information regarding treatment. Still, the poor survival in their report makes a more benign biologic behavior of MCS per se in the young unlikely. Concerning treatment, it will be difficult to convincingly prove that the comparatively favorable outcome in our series was due to the widespread use of chemotherapy. Nevertheless, the finding that the majority of assessable tumors in our sample responded to induction chemotherapy argues for the effectiveness of systemic therapy in young MCS-patients.
In conclusion, similar to other pediatric soft tissue sarcomas,43 we suggest that clinical experience gained from adult MCS patients cannot easily be translated into treatment recommendations for pediatric patients. Prognostic factors that are not applicable (like histologic grading for MCS) or unevaluated (eg, histologic grading for pediatric soft tissue sarcoma in general) should not be used for treatment stratification outside prospective clinical trials. From the results of our series we conclude that achievement of local control by surgical resection with wide margins remains of paramount importance in MCS and should be attempted in all patients. Radiotherapy may contribute to local control if oncologically adequate surgical resection is not feasible. Although the use of systemic treatment in MCS remains a matter of debate, it seems to be effective in a considerable proportion of young patients, possibly improving prognosis. We therefore propose to treat MCS according to standard multimodal soft tissue or bone sarcoma regimens and to assess response to induction chemotherapy with the appropriate methods. Patients with MCS should be included in prospective databases so that further information about pathology, clinical behavior, treatment, and outcome of this rare malignancy can be gained.
We thank Erika Hallmen, Iris Veit-Friedrich, and Matthias Kevric for excellent data management, and the contributing hospitals for continuous cooperation with the study centers, and all the patients/parents/guardians for their willingness to participate. This report is dedicated to Joern Treuner and his commitment to the Cooperative Weichteilsarkom Studiengruppe (CWS).