After a challenging year, Exelixis (EXEL) is finishing 2010 on a positive note, with the help of promising data for its lead agent, XL184. Last week at the EORTC conference, the company published data from a large phase II study in multiple cancer types. The results came at a crucial time for Exelixis, as many were questioning the value of XL184 following BMS’ (BMY) decision to opt-out of its development. As discussed in a previous post, when a partner like BMS dumps a late stage clinical asset after a licensing payment of over $150M, the alarm bells start ringing.
XL184 inhibits several targets, primarily VEGFR2 and Met. As a multi-targeted kinase inhibitor, the drug has some overlap with other drugs, primarily VEGFR inhibitors. There are two drugs that inhibit VEGFR (in addition to other targets) currently in the market as well as a long list of VEGFR inhibitors in late stage clinical testing from GSK (GSK), AstraZeneca (AZN), BMS and Aveo (AVEO). The main question regarding XL184 is whether the drug has a differentiated clinical profile in comparison to other VEGFR inhibitors. Based on recent data presented at EORTC the answer is a resounding “YES”.
Results for XL184 were from a large phase II trial in 9 tumor types, which comprised of a single arm portion followed by a randomized placebo-controlled stage (randomized discontinuation trial). This design is considered suitable for drugs that do not lead to meaningful tumor shrinkage but could still stabilize the disease for prolonged periods of time. In the first stage all the patients receive the drug, but only those who achieve stable disease (limited decrease or increase in tumor burden) move on to the randomized stage, where the drug’s ability to keep patients without progression is assessed.
The results presented were from the single agent portion, which is considered less reliable. Nevertheless, in some indications even the single agent portion generated impressive results. XL184 had various levels of anti-tumor activity in all the indications including lung, ovarian, liver and skin cancers, however, these findings were eclipsed by groundbreaking results in prostate cancer.
Based on standard criteria, XL184 led to multiple cases of tumor shrinkage in prostate cancer patients, some of which were long lasting. The objective response rate was modest (15%, confirmed and unconfirmed response rate) while the disease control rate was encouraging (71% of evaluable patients). It is important to note that the response rate in randomized discontinuation trials is typically lower than standard phase II studies since some patients with initial tumor shrinkage who do not qualify as responders switch to placebo at the time of randomization.
A new type of anti-cancer effect
Given the fact that these results are from the single arm portion it is hard to reach definitive conclusions with respect to XL184’s value compared to other drugs. But when investigators started looking at bone scans, they saw a different picture. The most important finding was the complete or partial resolution of bone lesions in 19 out of 20 evaluable patients with bone mets. This finding by itself, which is preliminary and should be corroborated in larger studies, could represent a breakthrough in the treatment of prostate cancer.
Resolution of bone involvement is not represented in standard endpoints such as objective responses or progression free survival but could have an enormous clinical significance. The vast majority of patients ultimately develop bone mets, which lead to severe morbidities and death.
There are currently no standard metrics for evaluating bone metastases simply because no agent to date had a meaningful effect on existing bone mets. This is probably why investigators chose to put bone scans for each and every evaluable patient in the poster. The images provide only qualitative evidence, but a quick look is enough to understand XL184 is doing something remarkable in these late-stage refractory patients. The only quantitative data was from biomarkers related to bone metabolism, which also showed a dramatic effect.
There are sporadic documented cases of bone mets resolution with single agent Sutent or Nexavar (both inhibit VEGFR) , but none of the reports come close to XL184’s impact. The difference can be attributed to XL184’s ability to inhibit both VEGFR and Met, and implicate MET as an important target in prostate cancer, especially in combination with VEGFR inhibition. Altogether, XL184 data imply that simultaneous inhibition of both Met and VEGFR is a valid approach in certain indications but not in every clinical setting.
More validation for Met
XL184 is the first Met inhibitor to be evaluated in prostate cancer. Exelixis’ results will probably motivate other companies with Met inhibitors to test their drugs for prostate cancer as single agents or in combination with VEGF inhibitors. As the only Met inhibitor in development in prostate cancer, Exelixis enjoys a first mover advantage in this indication. Interestingly, Arqule’s (ARQL) Met inhibitor, ARQ 197, led to a partial response in a prostate cancer patient in phase I (the trial included only one prostate cancer patient). In 2007 the company had originally planned to conduct a phase II trial similar to that of XL184, but eventually decided to focus on lung and liver cancers, which was not such a bad decision after all.
As I discussed in a previous post, multi-targeted kinases came under some pressure lately, but it appears that in some settings they represent a promising approach. ARQ 197 does not inhibit VEGFR, so its effect in prostate cancer, especially on bone metastases is still an open question.
The XL184 data is another validation of Met as an important target in cancer. This comes after Arqule’s and Genentech’s efficacy signals in lung cancer. A registration phase III trial for ARQ197 in lung cancer started to recruit patient earlier this month. On top of the promising results in prostate cancer, XL184 generated an intriguing efficacy signal in ovarian cancer in the form of a 32% response rate. The effect was observed also in platinum resistant patients and compares favorably to other agents in this indication. Results from the randomized portion could mark ovarian cancer as another indication for which Met is a relevant target.
The BMS riddle
Given the quality and breadth of XL184’s data set at EORTC, BMS’ decision to abandon the drug becomes even more puzzling. BMS probably had access to some of the data when it made the decision, which is hard to reconcile with XL184’s performance. Another intriguing step was a partnership deal with China based Simcere for the development of an internal Met/VEGFR inhibitor announced last month. This agent will be the second internal Met inhibitor BMS promoted to phase I (their first Met program, BMS-777607 completed phase I and is currently not in active development). In any case, BMS’ decision will probably remain an overhang until Exelixis finds a new partner for XL184. Based on recent data, it should be able to strike a lucrative deal already next year.