The first interim analysis was conducted at a median follow-up of 23 months after the occurrence of 322 events, including 84 deaths.
At 4 years, disease-free survival was 84% for AC→TH, 80% for TCH, and 73% for AC→T. The analysis showed a relative reduction in the risk of relapse of 51% for AC→TH (P < .0000005) and 39% for TCH (P < .00015) vs the control arm of AC→T. There have been 147 events in the control AC→T arm, 98 in the TCH arm, and 77 in the AC→TH arm. At this time, the 21 events separating the two experimental arms are not statistically significant (P = .16).
Deaths occurred in 36 patients in the control arm, 20 receiving AC→TH, and 28 receiving TCH (only 3 breast cancer deaths separated the two experimental arms). The data are insufficient for evaluating overall survival differences.
Grade 3 and 4 hematologic and nonhematologic toxicities were very similar among the arms. Neurotoxicity and renal toxicities of any grade were not significantly increased by any of the regimens, he said. "Trastuzumab is generally a very well tolerated drug, but its one problem is cardiac toxicity," he said, noting that differences between the experimental arms did emerge with regard to this side effect. "The differences we saw were not due to disproportionate cardiovascular risk factors, which were very well balanced among the arms."
There were no cardiac deaths. Clinically significant cardiac events per the Independent Review Panel occurred at the following rates: 0.95% (10 events) with AC→T, 2.34% (25 events) with AC→TH, and 1.33% (14 events) for TCH, with a significant P value (P = .016) between AC→TH and AC→T. Clinical heart failure occurred in 3 patients, 17 patients, and 4 patients, respectively.
BCIRG 006 also included categories for arrhythmias and ischemia or infarction, cardiotoxicity categories not included in other trials of trastuzumab, Dr. Slamon said. Grade 3-4 arrhythmias were seen in 7 patients on AC→T, 4 patients on AC→TH, and 9 on TCH (5 of these 20 are not yet adjudicated). Grade 3-4 ischemia or infarction occurred in 0, 4, and 1 patient, respectively.
"The interaction between trastuzumab and anthracycline seems to come to the fore here," Dr. Slamon observed. "By our definition of cardiac toxicitygrade 3 and 4 congestive heart failure, arrhythmias and infarctionsthe difference is about twofold for AC→TH."
Regarding left ventricular dysfunctionanother common observation among trastuzumab recipients in generala total of 353 patients experienced a greater than 10% relative asymptomatic decrease in left ventricular ejection fraction (LVEF). This occurred in 9% of patients in the AC→T arm, 17.3% in the AC→TH arm, and 8% in the TCH arm, highly statistically significant for AC→TH vs AC→T (P = .002) or TCH (P < .0001).