PROGNOSTIC ANALYSIS OF SURVIVAL IN SMALL BREAST CANCERS

Background: Routine axillary dissection in patients with invasive smal l breast cancer remains controversial. We previously reported a model for predicting nodal involvement in patients with T1a or T1b breast ca ncer that may guide the practice of selective nodal dissection. The ob jective of this study was to determine whether the prognosticators tha t predict nodal metastases also predict survival. Study Design: This s tudy is a retrospective review of 2,153 women with small invasive brea st cancer (less than or equal to 1 cm) diagnosed between January 1984 and December 1995. Cases were identified from a statewide tumor regist ry, the Hospital Association of Rhode Island, and the tumor registry a t Baystate Medical Center in Massachusetts. The impact on survival of patient age (less than or equal to 40 versus >40 years), nodal status (positive versus negative), tumor size (T1a versus T1b), and tumor gra de (1 versus 2 or 3) were analyzed. Breast cancer-specific survival (B CSS) mas analyzed using the Kaplan-Meier method and the proportional h azards regression method. Results: There were 388 patients with tumors 0.5 cm or less (T1a) and 1,765 with tumors 0.6-1.0 cm (T1b). Nodal st atus was known in 68% of cases (1,461 of 2,153), and tumor grade was r ecorded in 42% of cases (902 of 2,153). In univariate analysis, age, g rade, and nodal status were significant in their association with BCSS . Tumor size did not influence BCSS among patients with small invasive tumors. Women older than 40 years had superior survival compared with younger women (93% versus 78% at 5 years; p = 0.01). Similarly, women with low grade (1) tumors did better than those with higher grade (2 or 3) tumors (98% versus 88% at 5 years; p = 0.03). The 5-year BCSS wa s 96% versus 78% for node-negative versus node-positive disease, and t he 10-year BCSS was 91% versus 62% (p = 0.001). In the multivariate an alysis, age and nodal status remained firmly associated with survival, although grade lost its significance. Conclusions: Small tumor size d oes not affect survival. Although risk profiles for nodal involvement can be constructed to help guide the practice of selective axillary ly mphadenectomy in patients with small invasive breast cancers, these fa ctors cannot serve as a surrogate to nodal status in establishing pati ent prognosis. Nodal status remains the most powerful determinant of s urvival in breast cancer patients, even those with very small tumors. (C) 1998 by the American College of Surgeons.