At San Antonio this year we presented the results of the DESTINY-Breast02 trial. To just give a little bit of background on that, in 2019 we presented the results of DESTINY-Breast01, this was a phase II single arm study looking at the then new HER2-directed antibody-drug conjugate called trastuzumab deruxtecan, or T-DXd. In that study, the DESTINY-Breast01 trial, we were studying patients who had HER2 positive advanced breast cancer that previously had been treated with T-DM1. This was a very heavily pre-treated population, a median of six prior lines of therapy. In that pre-treated population trastuzumab deruxtecan, or T-DXd, showed fairly robust activity with a response rate of about 60% and progression free survival of around 19 months.
So clearly it was an active drug with durable responses and because of that study the FDA of the United States gave accelerated approval to T-DXd in patients with pre-treated HER2 positive metastatic breast cancer. Because the DESTINY-Breast01 trial was a single-arm phase II study, there was felt to be a need for a confirmatory trial and the DESTINY-Breast02 trial was designed as that confirmatory trial.
What was the study design?
DESTINY-Breast02 enrolled patients with centrally confirmed HER2 positive advanced breast cancer. These were all patients who had had prior therapy with T-DM1. The patients were randomised to trastuzumab deruxtecan as a single agent at standard dose or to treatment of physician’s choice which in this case was either capecitabine with trastuzumab or capecitabine with lapatinib. It was a 2:1 randomisation; about 400 patients were on the T-DXd arm and about 200 on the control arm of physician’s choice. The primary endpoint of the study was progression free survival by independent central review and overall survival was an important secondary endpoint.
What were your findings?
The data we reported at San Antonio, which was the primary analysis of this trial, demonstrated that trastuzumab deruxtecan was superior to treatment of physician’s choice in terms of the primary endpoint of centrally reviewed progression free survival. The hazard ratio was about 0.36 which was highly statistically significant, so a 64% reduction in the risk of progression or death with T-DXd. The median PFS was about 18 months with T-DXd, compared to about 7 months with the treatment of physician’s choice, so a fairly substantial difference. Overall survival, which was analysed formally because the progression free survival endpoint was significant, also demonstrated superiority of T-DXd over the control arm. So here the hazard ratio was 0.66, so a 34% reduction in risk of death with T-DXd.
If you look at the Kaplan-Meier curves in this trial the overall survival curves separate early so there was already a significant difference, or a substantial difference, in overall survival even after 12 months. The median overall survival was about 13 months longer with T-DXd compared to treatment of physician’s choice. So clear improvements in both progression free and overall survival; other secondary efficacy endpoints like response rate, clinical benefit rate and duration of response all favoured T-DXd as well.
In terms of the safety profile of T-DXd in this study, it was almost exactly what we have seen in prior studies. There were no new safety signals; the most common adverse events were nausea with T-DXd and these were generally low-grade events and manageable. Hematologic toxicity was uncommon.
The one potentially serious adverse event that we think about with T-DXd is interstitial lung disease, or ILD, and in this study the overall rate of ILD that was felt to be drug related by an independent review committee was about 10%. So this was similar to what has been seen in other T-DXd studies or even actually a little bit lower than what we’ve seen in other studies. Almost all of these events of ILD were low grade but there were two patients, or 0.5% of the population, unfortunately had grade 5 or fatal ILD that was felt to be drug related. So this is just further evidence that ILD is a known toxicity of trastuzumab deruxtecan and does need to be monitored when patients are on this agent.
What is the clinical impact of these results?
This study clearly confirms that T-DXd is a highly active drug in patients with HER2 positive metastatic breast cancer that had been pre-treated with T-DM1 and other agents. So this was largely a third and fourth line study and clearly T-DXd was very active and quite superior to standard of care type HER2 directed therapies.
But in terms of clinical impact, I don’t think it really has a large impact because we have the data from the DESTINY-Breast03, which was the second line study of T-DXd, which was a randomised trial comparing T-DXd to T-DM1 in a largely second-line population. This study was also updated at San Antonio this year, further confirming the marked superior of T-DXd over T-DM1 in that second line setting. So in the clinic these studies would strongly recommend, and this is consistent with guidelines, that T-DXd be used in the second-line setting for the vast majority of patients with HER2 positive, metastatic breast cancer. Therefore our study in the third and later line setting really doesn’t apply to most patients because they would have already received T-DXd prior to receiving T-DM1. That being said, in those relatively small number of patients who have already had T-DM1 in the second line then clearly T-DXd is a very active drug and should be used in that patient population in the third line. But most patients really should be getting T-DXd in the second line.