The Development of an International Registry

Abstract

The International Registry of Lung Metastases was launched in 1990 to gather a large set of data to perform a meta-analysis of the experience accumulated by the major European and American centers of thoracic surgery over a period of 50 years. We sought to evaluate the survival after metastasectomy according to the primary tumor, to define the prognostic factors by multivariate analysis, and to propose a new system of classification based on long-term survival. A comprehensive database was designed to provide a simple and flexible instrument for the registry. This included a single record form for each patient, which is divided into four different sections: patient’s identification, features of the primary neoplasm, description of every metastasectomy performed, and updated follow-up. We made an important distinction between planned sequential surgery and repeat metastasectomy for further recurrence. Patients who underwent planned sequential or staged thoracotomies were considered as one single metastasectomy and not as redo surgery. Between 1991 and 1995, a total of 5290 patients were enrolled, from 18 centers and 9 countries, including the consecutive series of all cases operated by each center within a given period of time. Memorial Sloan Kettering provided the largest series, with 1075 patients treated from 1945 to 1995. Adequate information was available for most of these patients. Only 84 cases (1.6%) were excluded because of an incomplete dataset, thus leaving 5206 patients fully evaluable. The analysis was conducted by an independent agency, the Institute of Drug Development in Brussels. The article published as a result of this work is still widely cited,1 and a number of other analyses have been performed on subsets of patients. The experience is illustrated in the Figure 1 We have applied our classification to the different cancer types. In all cancer types, it brings about a separation of survival trajectories, more widely dispersed in some than in others. It should be noted that the classification incorporates elements identifiable before surgery (interval since the primary resection and the metastasis count) and resectability, which is typically an operative finding. The “disease-free interval” (DFI), defined from the data as the interval between the primary cancer resection and the metastasectomy, is of course an indicator of tumor doubling time while the number of metastases is some measure of aggressiveness of the cancer. It is noteworthy that for most cancer types, those who have unfavorable characteristics have poor survival rates as early as 1 to 2 years. In a recent Memorial Sloan Kettering Cancer Center publication on colorectal metastasectomy, the authors point out that, as a group, patients with multiple metastases and short-time intervals are unlikely to survive for a long period and that we are beginning to define patients who should not be offered metastasectomy.2 Perhaps, we are beginning to be more confident of the limits on which patients should be offered metastasectomy. Conversely, that patients with all three characteristics favorable survive longer, leaves the question of whether it is the selection for surgery, rather than the effect of surgery that is important in survival, a possibility mooted by Aberg.3 It is also notable that the dispersion of survival by our classification is least with breast and melanoma. Breast cancer regarded as being a systemic disease from when first detected. The survival curve may support that the features related to lung metastases are less indicative of survival than in other diseases. Finally, having done the difficult work on data collection for the registry, we would support the plea for better reporting of clinical data.4