A few months ago I wrote about the preliminary reported results from the AVAPERL trial, which started patients with previously untreated  advanced nonsquamous NSCLC up to four cycles of cisplatin/Alimta (pemetrexed)/Avastin (bevacizumab), then randomized patients who hadn't progressed after four cycles to either maintenance Alimta/Avastin or Avastin alone.  At the Eurpean Society for Medical Oncology (ESMO) 2011 meeting, the investigators (Barlesi and colleagues) presented early results that showed a very significant improvement in progression-free survival (PFS) from the beginning of all treatment, at 10.2 vs. 6.6 months (HR 0.50, p < 0.001).  For those of you who understand things better visually, here's the curve plotting the PFS outcome for the two different groups, with a very impressive difference:  

  (click on image to enlarge)

   The same results, plotted from the time of randomization to combination vs. single agent Avastin as maintenance therapy, are even more striking:

   I didn't have any information about overall survival (OS) back in September, but here's the preliminary OS results, with numbers from the time of starting all treatment:

So while these are only preliminary results, the difference of a 25% better OS with continuation of Alimta is impressive to me, especially considering that the arm with the worse outcome, receiving maintenance Avastin alone, has a median survival of nearly 16 months, which we would consider to be quite excellent compared the outcome of other advanced NSCLC trials (12 months for cisplatin/Alimta on one large phase III randomized trial, 12 months with carboplatin/Taxol (paclitaxel) with Avastin in another).  So the inferior arm on AVAPERL has still done very well: the superior arm is just doing meaningfully better.

   My take home messages from this work, though it's from just one trial without a huge number of patients, is that:

1) Maintenance Avastin on its own may or may not add anything, but adding another agent with good activity against NSCLC (at least Alimta) is likely better; it's unclear whether the Avastin is adding anything to the combination.  That'll be tested directly in the ECOG 5508 trial that is ongoing now.

2) Clearly the value of the Alimta hasn't been exhausted by continuing it after four cycles of initial therapy in the first line setting.  In patients who aren't progressing and are tolerating it well, more is better.

   I'll discuss some other evidence along these lines in my next post.