In their literature survey, Drs. Chao and Goldberg reach the conclusion that surgical metastasectomy is the clear treatment of choice and should be the standard of care for patients with pulmonary recurrences of soft-tissue sarcoma. It is assumed that survival without this operation is negligible, even while there are no survival statistics for sarcoma patients who are eligible for metastasectomy and who choose to forgo this option.
Sarcomas may be quite heterogeneous in their growth rate and biological aggressiveness, and most sarcoma clinicians can recall patients who have lived longer than 5 years with unresected tumor. Whether metastasectomy would have further prolonged their survival, or indeed, whether metastasectomy is “pivotal” to the survival of even some subsets of selected patients remains to be proven. The urge to surgically remove all visible disease and offer the patient a chance for cure is very powerful—perhaps too much so. In the more than 50 years since metastasectomy has been performed, however, no prospective, controlled, randomized trials have been carried out to validate the procedure.
I cannot agree with the authors that, “Although the data are inconclusive and depend on patient selection and the completeness of resection, the oft-quoted overall 5-year survival range of 30% to 40% is real.” This percentage does not refer to any defined subset of sarcoma patients, and it is meaningless unless it does so. It just may be that once soft-tissue sarcoma metastasizes to lungs, less than 1% of unselected patients survive—operated or not.
Patient Selection Variability vs Metastasectomy Outcome
Among the numerous retrospective, uncontrolled, single-institution series used to support the efficacy of metastasectomy, patient selection criteria vary widely. In some studies—but not others—patients with osteosarcoma, patients with low-grade primary tumors, or patients who have had prior chemotherapy or radiotherapy are included. Such patients often have a longer preoperative disease course, and if recurrence is detected earlier in even a few patients, their calculated average survival would be extended—even without surgery. These potential biases can be magnified in studies with few numbers or short follow-up periods.
The possibility is very real that the 30% response rate that Chao and Goldberg quote is artifactual and due to patient selection. Until the procedure is validated, however, there is no scientific rationale for quoting any percentage as a basis for recommending metastasectomy to a patient with newly diagnosed recurrence—even if he or she possesses criteria for a favorable outcome (long disease-free interval, low-grade histology, easily resectable disease).
For example, assume that 40% of patients with resected primary sarcomas eventually have a recurrence. Of that group, the subset of patients with a metastasis-free interval greater than 4 years would make up less than 5% (corresponding to 2% or less of the original unselected group of primary sarcoma patients). Of these patients with favorable tumor biology selected for surgery, the 30% surviving 5 years would comprise only 0.6% of the original unselected primary group: This is not much different than the survival of an unselected, untreated metastatic sarcoma population.
In actuality, the selection criteria for metastasectomy are stringent. In addition to the disease-free interval, histologic, and resectability criteria mentioned above, disease at the primary site should also be controlled, with no extrathoracic disease, and adequate cardiopulmonary function adequate to tolerate one or more thoracotomies. The potential for selection bias is correspondingly great.
Surgery for New, Resectable Lung Metastases
In clinical practice, surgery is almost always immediately considered and recommended to sarcoma patients with newly diagnosed, resectable lung metastases. The detection of lung metastases in patients whose primary tumor had been completely resected usually heralds an abrupt, unfavorable revision in prognosis, which, in turn, soon precipitates a surge of therapeutic activism by a concerned oncologist for a frightened patient. The reasoning appears straightforward: Attempting to treat these patients with palliative chemotherapy or local radiation seems pointless unless symptoms are present, whereas surgery seems to offer the only hope of disease eradication.
However, fully 70% of patients operated on for lung metastases do not benefit from surgery, and it is far from clear whether the remaining 30% would have had an indolent, prolonged course without it.[1,2]
Randomized, controlled trials of metastasectomy are needed. One possibility might be to study survival from initial diagnosis of one group of patients, whose resection is performed immediately upon detection of recurrent disease. This will be compared to a second closely monitored experimental group administered neoadjuvant chemotherapy for 3 months, then resected. Equivalent overall survival in the two groups would suggest that surgical resection may be merely a test for a good prognostic group, rather than a treatment that improves prognosis.
Until such studies are carried out, the best option might be to limit operative treatment of pulmonary metastases to sarcoma patients with agreed-upon favorable criteria, while unambiguously informing patients of the risks and limitations of surgery.
1. Pastorino U, McCormack PM, Ginsberg RJ: A new staging proposal for pulmonary metastases. The results of analysis of 5,206 cases of resected pulmonary metastases. Chest Surg Clin N Am 8(1):197-202, 1998.
2. Verazin GT, Warneke JA, Driscoll DL, et al: Resection of lung metastases from soft-tissue sarcomas. A multivariate analysis. Arch Surg 127(12):1407-1411, 1992.