Hazard rates of recurrence following diagnosis of primary breast cancer

• Published in: Breast cancer research and treatment Vol. 89; no. 2; pp. 173 - 178
• Main Authors: JATOI, Ismail, TSIMELZON, Anna, WEISS, Heidi, CLARK, Gary M, HILSENBECK, Susan G
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer 2005; Springer Nature B.V
• Subjects: Adult; Aged; Biological and medical sciences; Breast cancer; Breast Neoplasms - pathology; Breast Neoplasms - surgery; Cancer research; Cohort Studies; Female; Gynecology. Andrology. Obstetrics; Humans; Mammary gland diseases; Medical sciences; Middle Aged; Neoplasm Recurrence, Local - etiology; Neoplasm Staging; Retrospective Studies; Risk Assessment; Survival Analysis; Time Factors; Tumors; Human; hazard rates; recurrence; Mammary gland diseases; Relapse; Hazard function; hazard functions; Primary; Breast cancer; Diagnosis; Malignant tumor; Survival
• ISSN: 0167-6806; 1573-7217
• DOI: 10.1007/s10549-004-1722-0

We calculated hazard rates for recurrence in patients with primary breast cancer (stage I, II; no adjuvant therapy). Previous publications have indicated a peak in hazard rates for recurrence (or death) at approximately 2-3 years after diagnosis of primary breast cancer. However, there have been conflicting reports concerning the presence of a second peak at 5-7 years after diagnosis. In this study, we estimated hazard functions by the Nelson-Aalen method and fit by cubic-linear and cubic-cubic-linear models to test for the presence of one or two peaks, respectively. We identified two peaks in hazard of recurrence, one at 2 years and another at 5 years. The 5-year peak, though statistically significant, represents very small differences in patient outcome. This additional peak may be an artifact of interval censoring due to a tendency to follow-up patients at specific bench-mark time points.