Soft tissue sarcoma of high grade — a primary systemic disease?

Abstract

Soft tissue sarcoma is a very heterogenous disease. The incidence is rather low (1±2 cases per 100 000 population) and large databases are required to enable homogenous analysis. Major sources for data, although not population-based, are run by the EORTC Soft Tissue and Bone Sarcoma Group, the Memorial Hospital at New York (MSKCC), the French Sarcoma Group and the Scandinavian Sarcoma Group (SSG). The latter has the advantage of having collected data prospectively and derived from a uniform treatment regimen which is applied throughout Scandinavia as well as relying on centralised pathology for more than 15 years [1]. Chemotherapeutic regimens and radiotherapy can easily be standardised, however, it is dicult to compare surgical procedures. How wide is wide when performing a `wide excision' and how `compartmental' is a resection for an extracompartmental lesion? Surgery is always mentioned as the key modality to yield treatment results providing survival free from both local and systemic relapse. In limb sarcoma, two trials of amputation versus limb-sparing treatment did not result in improved survival rates [2,3]. Due to advances in surgical techniques, limb-saving treatment has increasingly become possible [4]. However, the rate of local recurrences is higher after limb-preserving therapy, particularly in high grade lesions, and it is matter of discussion whether this in uences the overall survival of patients. An evaluation of the AJCC (American Joint Committee on Cancer) staging system revealed that 43/100 patients developing locally recurrent sarcoma subsequently died from their disease [5]. Several major reports clearly demonstrated that operative procedures with resections margins that are not tumour-free (R1/2resection) will be followed by local recurrence in the overwhelming majority of patients [6,7]. Does inadequate surgery compromise survival? Early reports have to be interpreted with some caution as they are in uenced by retrospective sampling and inconsistent management of patients (for example, +/y radiotherapy, or chemotherapy as an adjuvant measure). The paper by Trovik and colleagues in this issue (pp. 710±716) evaluated the metastasis-free survival of 559 patients with soft tissue sarcoma of the trunk wall and extremities. All patients had been treated with surgery only and the median follow-up was 7.4 years (range: 0.1±12.5 years). The rate of local recurrences was 18% and using a Cox regression model local recurrence was analysed as a time-dependent variable. As expected, high histopathological grade (relative risk (RR) 3.0; 95% con®dence interval (CI) 1.5±6.3) and an inadequate surgical margin (RR 2.9; 95% CI 1.8± 4.6) were shown to be the main risk factors for local recurrence. High grade was also the main factor associated with metastasis (RR 3.3; 95% CI 1.8±6.3) and tumour size >7 cm was the next major contributing factor (RR 2.3; 95% CI 1.6±3.3). Surgical margin was not a risk factor for metastasis (RR 1.3; 95% CI 0.9±2). If the Cox model was carried out incorporating local recurrence as a time-dependent variable, this was shown to be the highest contributing factor (RR 4.4; 95% CI 2.9±6.8) overriding even the e€ect of histological grade. The authors conclude that the commonly proven association between local recurrence and metastasis must be interpreted as likely to be non-causal. This study is of value as adjuvant radiotherapy after inadequate margins was not used and, therefore, the relationship of local recurrence and subsequent metastasis could be studied. However, there might be some bias as approximately one fourth of the patients of the original data basis were excluded from analysis due to inadequate margins and probably had high-grade lesions. The SSG data are contradictory to an analysis at MSKCC. The study reported by Pisters suggested a contribution of positive surgical margins to tumourspeci®c mortality [6]. However, a subsequent study by Lewis showed that local recurrence was an independent