Search

Background: Trastuzumab (T) forms the backbone of the current first-line treatment of advanced HER2+ breast cancer. In the present study we sought to evaluate the impact of the time interval between the end of adjuvant trastuzumab and distant recurrence (TDRI), upon a) overall survival (OS), b) the duration of trastuzumab-based regimens (T) in the advanced setting. Patients & Methods: a) The first hypothesis involved 187 pts on HERA treated with adjuvant T and diagnosed with distant recurrence after the scheduled treatment period of 1 year at 4 years median follow-up. The impact of TDRI on OS from distant recurrence was evaluated in a Cox model adjusted for tumor size, nodal and hormone receptor status at time of enrollment into the HERA trial. b) The second hypothesis involved 90 patients retreated with T in the metastatic setting. A questionnaire was sent to investigators retreating pts with T upon distant recurrence; 144 of 156 questionnaires were returned (93%), and 90 pts were selected based on available clinical information and consent for subsequent studies (14 pts were randomized to observation and received adjuvant trastuzumab after crossover). Results: a) There was no statistically significant relationship between TDRI following 1yT and OS from distant recurrence: HR 0.99, p = 0.46; HR 0.99, p = 0.27 from the stratified and unadjusted Cox proportional hazard models, respectively. The median OS from distant recurrence was numerically longer among pts with a TDRI of ≥ 12 months (n = 103) than < 12 months (n = 84) but not statistically significant (23.7 vs. 17.8 months, p = 0.47). b) The median duration of T-based regimens for metastatic disease was 8.8 months (n = 88), 9.4 months in the ER+ cohort (n = 36) and 8.8 months in the ER- cohort (n = 52); (n = 2, missing). Sites of first distant recurrence and subsequent disease progression for the 90 pts were hierarchically distributed as below. First distant recurrence n (%)Subsequent progression n (%)CNS6 (6.7%)22 (24.4%)Visceral45 (50.0%)39 (43.3%)Skeletal25 (27.8%)9 (10.0%)Soft Tissue14 (15.6%)9 (10.0%)No event-11 (12.2%) Conclusions: In the HERA trial, TDRI following 1yT did not impact on OS from distant recurrence in pts experiencing distant disease recurrence. This may be due to the benefit of anti-HER2 regimens in subsequent lines of therapy. Although our sample size is limited, the present analysis suggests that pts experiencing a shorter interval (≤ 12 months) from the end of adjuvant trastuzumab to distant recurrence derive similar benefit from trastuzumab-containing regimens in the metastatic setting when compared to the cohort with an interval > 12 months. CNS relapse in HER2-postive breast cancer remains a clinical problem. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P2-15-04.

Background: Retrospective studies of HER2−positive metastatic breast cancer (BC) showed an incidence of CNS metastases of 21% to 34%. We investigated the incidence and clinical aspects of CNS relapse (CNS-R) in patients (pts) enrolled in the HERA trial, a prospectively randomized adjuvant trial in node + or high-risk node - HER2−positive early BC pts. Methods: 3401 pts were randomized into the 1-year trastuzumab (1yT) or the observation (obs) arms of HERA (Piccart-Gebhart et al, 2005, Gianni et al, 2011). The cumulative incidences of first disease-free survival (DFS) events in the CNS vs other sites were estimated using competing risk analysis. The database of the main study had a clinical cut-off date of 9th June 2008. To obtain additional information regarding CNS-R (including occurrence of CNS-R after first DFS event), a specific CNS-directed questionnaire was sent to investigators of pts who were deceased as of July 2009. Information collected included the date of CNS-R, whether it was symptomatic, the type of CNS-R (brain metastases (BM) or meningeal carcinomatosis (MC)), methods of diagnosis, and treatments at the time of CNS-R. Results: 1yT significantly reduced the risk of other DFS events (p=0.000017, Gray's test), but not of CNS-R (p=0.55) as first event (see table). During the first year of follow up, CNS-R accounted for 15 (14.9%) of the 101 first DFS events in the 1yT arm and 15 (7.7%) of the 194 first DFS events in the obs arm. The analysis of baseline patient and tumor characteristics associated with CNS-R as first event confirmed known risk factors such as young age ( 413 of the 481 questionnaires (85.9%) were returned. 217 of the 413 deceased pts had a CNS-R diagnosed prior to death (52.5%), with more events occurring in the 1yT arm (see table). By contrast, the incidence of CNS-R as first DFS event was balanced across the arms. Based on the survey data, CNS-R was symptomatic in 189 pts (87.1%) with no differences between arms. BM were present in 211 pts (97.2%), absent in 5 (2.3%), and missing information in 1 (0.5%). MC was diagnosed in 25 pts (11.5%), absent in 187 (86.2%), missing information in 5 (2.3%). Frequencies for BM and MC were very similar in both arms. Conclusion: This retrospective analysis of a prospective large study shows more than 50% incidence of clinically diagnosed CNS-R in HER2−positive BC pts who have died. CNS-R was symptomatic in most pts. CNS-R at any time was less frequent in the 1yT arm (88 vs 129). There is no evidence that adjuvant trastuzumab increases the incidence of CNS-R. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-17-01.