Background: Tumor size affects the choice of surgical procedure and patient prognosis. It is standardly assessed as the largest unidimensional measurement and, for multifocal disease, as the largest size of the largest focus. We examine some different methods of assessing tumor size: the standard method; the sum of the largest sizes for all foci; surface area; and volume.
Methods: Data for a cohort of 678 primary invasive breast cancer patients accrued from 1971 to 1990 were updated to 1996; there were 571 patients with unifocal disease and 107 patients with multifocal disease. We used step-wise Cox regression to investigate the effects on time to death of the prognostic factors tumor size (estimated in one of the four ways), age, nodal status, ER, PgR, adjuvant radiotherapy, adjuvant hormonal therapy, and adjuvant chemotherapy. We also examined the association between tumor focality and nodal status.
Results: For all patients, tumor size was included in the multivariate model, regardless of estimation method. For patients with multifocal disease, tumor size was included in the final model only when it was estimated as the total surface area (P = .03) or volume (P = .01) of the foci. More multifocal patients were N+ (P = .056).
Conclusions: For patients with multifocal disease, the significance association with mortality for total surface area or volume may imply a biologic relevance or mode of tumor activity for the foci.