The 2012 Annual Meeting of the American Society of Clinical Oncology (ASCO) is in full swing in Chicago, with a wealth of clinical data being reported across multiple solid and hematological cancers. In this piece, I will focus on two agents that took center stage in the first two days of the meeting the hormonal agent Zytiga (abiraterone) in chemotherapy-naïve metastatic castrate-resistant prostate cancer (mCRPC), and the elegantly designed antibody-drug conjugate T-DM1 in HER2+ breast cancer.
So, let's start with abiraterone. Opening the festivities on Saturday, the interim results of the pivotal Phase III COU-AA-302 trial were presented in a fully packed Arie Crown Theater. In this study, asymptomatic or minimally symptomatic mCRPC patients were randomized to Zytiga plus prednisone or placebo plus prednisone; the co-primary end points were radiographic progression-free survival (rPFS) and overall survival (OS). Zytiga, dosed in combination with prednisone, demonstrated a statistically significantly improved rPFS over placebo plus prednisone (not reached, versus 8.3 months [P<0.0001, HR=0.43]). A trend towards improved OS was reported for patients treated with Zytiga plus prednisone (not reached, versus 27.2 months [P=0.0097, HR=0.75]) but the apparent gain in OS did not reach statistical significance. All secondary end points also favored the Zytiga arm, and no new important safety signals were reported.
The COU-AA-302 study was unblinded earlier in 2012 after a preplanned review by the IDMC concluded that there was compelling evidence of efficacy. Indeed, the difference in rPFS is neat; however, whether this gain reflects a tangible clinical benefit remains unclear, as this was the first Phase III trial that utilized rPFS as a primary end point. The gold-standard primary end point in clinical trials for prostate cancer is OS, and the study was powered to detect a difference in OS between both arms. However, some commentators questioned the decision of unblinding the trial so early, before a statistically significant benefit in OS was even reached. Of course, with longer follow-up, OS might still be attained (it was close to the interim planned boundary, which means that only a few events could have be sufficient to reach statistical significance had the trial not been unblinded). But since the cross-over of patients after unblinding introduces biases and confounds data, the demonstration that Zytiga improves OS of chemotherapy-naïve mCRPC patients may not be unequivocal.
What does this mean for Zytiga, and for other agents in the first-line mCRPC setting. Well, the big question is to assess how Zytiga fares against the personalized immunotherapy Provenge. Based on the IMPACT study, Provenge is associated with an OS of 25.8 months (around 4-month improvement over placebo). By way of comparison, the OS for the placebo arm in the COU-AA-302 was 27.2 months (the OS for Zytiga will be higher considering the positive trend in OS reported above); also, PFS for Zytiga is statistically significant. In the post-docetaxel setting, the improvement in OS over placebo was 4.6 months, and thus it is reasonable to surmise that Zytiga will demonstrate a greater improvement in OS over placebo in the first-line setting. In addition, some interviewed oncologists report using off-label Zytiga in the first-line mCRPC setting. Overall, the indicators point to Zytiga outperforming Provenge in first-line asymptomatic or minimally symptomatic mCRPC making the results from the ongoing Phase III trial of Provenge in the hormone-sensitive metastatic population all the more critical for the commercial future of the vaccine.
And now, let's turn our attention to T-DM1. Amidst much anticipation, the interim results of the pivotal EMILIA trial were unveiled on Sunday at the plenary session. The EMILIA study randomized Herceptin-refractory patients with HER2+ metastatic breast cancer between T-DM1 monotherapy and Xeloda plus Tykerb; PFS and OS were co-primary end points. The efficacy results are exciting T-DM1 showed a statistically and clinically significant improvement in PFS over the combination therapy (9.6 months versus 6.4 months [P<0.0001; HR=0.650). The median OS was 23.3 months in the combination arm and not yet reached in the T-DM1 arm (P=0.0005; HR=0.621); however, the OS rate at 2 years was in favor of T-DM1 (65.4%) compared to Xeloda plus Tykerb (47.5%). Perhaps as importantly, T-DM1 exhibited less toxicity compared to the combination therapy, with a lower incidence of grade 3 or more adverse events (40.8% versus 57%, respectively), a critical consideration in the metastatic setting.
The turbulent Avastin story in breast cancer highlighted the need to demonstrate convincing PFS and OS achievements. In that context, the interim analysis of the EMILIA trial has already shown encouraging results for T-DM1 from a clinical standpoint. More mature data will be eagerly awaited to determine whether the apparent improvement in OS can achieve statistical significance. These results open new avenues for T-DM1: could it be even more effective in combination with another agent. Which other populations of breast cancer could potentially derive benefit from T-DM1. What would be the role of Herceptin in the treatment algorithm if T-DM1 were to gain regulatory approval. The ongoing MARIANNE Phase III study will no doubt shed light on the optimal first-line treatment for HER2+ metastatic breast cancer patients, as this space looks set to be dominated by Roche/Genentech for the foreseeable future.
The success of T-DM1 has broader implications. In 2010, Pfizer's withdrawal of Mylotarg (gemtuzumab ozogamicin) for the treatment of AML from the U.S. market tainted antibody-drug conjugates as a class of agents. The EMILIA study a large, Phase III, randomized trial provides an important proof-of-concept that antibody-drug conjugates are valid agents in oncology and can yield significant clinical improvements. Another recent example is Adcetris (brentuximab vedotin) which was granted FDA approval in August 2011 for the treatment of relapsed/refractory Hodgkin's lymphoma (R/R HL) and R/R systemic anaplastic large-cell lymphoma (ALCL) based on a Phase II study. Other promising antibody-drug conjugates are in late-stage development, such as inotuzumab ozogamicin in non-Hodgkin's lymphoma. The concept of selectively targeting and delivering a cytotoxic agent to tumor cells is starting to become a reality.