LOCOREGIONAL RECURRENCE OF BREAST-CANCER FOLLOWING MASTECTOMY - ALWAYS A FATAL EVENT - RESULTS OF UNIVARIATE AND MULTIVARIATE-ANALYSIS

Purpose: The outcome of patients with local-regional breast cancer rec urrence after mastectomy often is described as fatal. However, certain subgroups with favorable prognoses are thought to exist. To determine these favorable subgroups, we analyzed prognostic factors for their i nfluence on postrecurrence survival by univariate and multivariate ana lysis. Methods and Materials: Between 1979 and 1992, 145 patients with their first isolated locoregional recurrence of breast cancer followi ng modified radical mastectomy without evidence of distant metastases were treated at the Department of Radiation Oncology of the University of Wurzburg. Thirty-nine percent of patients (n = 67) had had postmas tectomy radiotherapy, representing 7% of patients who had received rou tine postmastectomy irradiation at our institution. Systemic adjuvant hormonal therapy had been applied in 24% and systemic chemotherapy in 19% of patients. Several combinations were used. Treatment of recurren ces consisted of surgical tumor excision in 74%, megavoltage irradiati on in 83%, additional hormonal therapy in 41%, and chemotherapy in 12% of patients, employing different combinations. Local control in the r ecurrent site was achieved in 86%. Median follow-up for patients alive at the time of analysis was 8.9 years after recurrence. We tested dif ferent prognostic factors, including prior treatment and treatment of recurrence, for their influence on postrecurrence survival, using univ ariate and multivariate analysis. Results: Eighty-two of the 145 patie nts (57%) developed distant metastases within the follow-up period. Me tastases-free rate was 42% at 2 years and 36% at 10 years following re currence. With development of distant metastases, the survival rate de teriorated. Recurrences appeared within the first 2 years from primary surgery in 56% of patients, and in 89% within 5 years. Overall, 2-yea r and 5-year survival rates following local-regional recurrence were 6 7% and 42%, respectively. Univariate analysis revealed statistically s ignificant worsening of survival rates for pT3 + 4 primary tumors, pri mary axillary lymph node involvement, tumor grading 3 + 4, lymphatic v essel invasion, blood vessel invasion, tumor necrosis, negative estrog en (ER) and progesterone (PR) hormonal receptor status, postmastectomy chemotherapy and hormonal therapy, short time to recurrence (<1 year) , combined recurrences and supraclavicular site of recurrence, non-sca r recurrence, size of the largest recurrent nodule >5 cm, multiple rec urrent nodules, no surgical excision of recurrence, small target volum e of irradiation, chemotherapy for recurrence, and no local control wi thin the recurrence site. The 2-year and 5-year survival rates ranged from 68% to 94%, and from 33% to 65%, respectively, in the favorable s ubgroups compared to 2-year and 5-year survival rates ranging from 20% to 59% and 0% to 35%, respectively, in the unfavorable subgroups. Mul tivariate analysis showed that site of recurrence and number of recurr ent nodules have the strongest influence on postrecurrence survival, b ut time to recurrence, age at time of recurrence, local control in rec urrent site as well as primary pT and axillary status, and the presenc e of tumor necrosis in the primary tumor specimen showed additional in dependent influences on survival. Thus, we identified a highly favorab le subgroup of patients with a single chest wall or axillary recurrent nodule (in a patient aged >50 years), a disease-free interval of grea ter than or equal to 1 year, pT1-2NO primary tumor, and,without tumor necrosis, and whose recurrence is locally controlled. This group (12 p atients) had 5- and 10-year survival rates of 100% and 69%, respective ly. Conclusion: We conclude that locoregional recurrence of breast can cer following mastectomy is not always a sign of systemic disease. Our data support previous findings, that subgroups with favorable prognos is exist and they still have a chance for cure, demanding comprehensiv e local treatment. This subgroup can be identified by established prog nostic factors as demonstrated. Subgroups of patients with unfavorable prognostic factors might be candidates for additional systemic therap y immediately following or combined with local therapy because of the forthcoming systemic progression. (C) 1997 Elsevier Science Inc.