One very important trial that has changed how we treat previously treated patients with advanced NSCLC has been a trial coded JMEI by Lilly, which sponsored this phase III randomized trial that directly compared second line treatment with alimta (pemetrexed) against the only FDA-approved second line treatment at that time, taxotere (docetaxel). Essentially identical in activity, this trial led to alimta's approval for lung cancer because it was associated with less frequent and severe drops in blood counts and fewer hospitalzations than with taxotere (abstract here). But from this trial we've also been able to glean additional information on who benefits more or less from second-line treatment.

Dr. Glen Weiss published a follow-up paper (available here) that analyzed several variables in the JMEI trial. Some findings were definitely anticipated: patients with the best performance status of 0 had a far better survival than patients with a modestly or more significantly compromised performance status of 1 or 2 (median survival 12.7 vs. 8.3 vs 2.6 months for 0/1/2, respectively) -- we knew that healthier patients do better and live longer.

(Click on image to enlarge)

We also saw that patients who had previously had an objective response (complete or partial) did especially well, with a median survival of 15.8 months, compared to patients who achieved stable disease (median 10.5 months) or progressive disease (median 4.6 months) on first line chemo. So "responders respond", and the patients who did best early continue to do better, and those with aggressive disease initially tended to continue to do the least well.

And you might expect that the patients who went the longest time between completing first line therapy and starting second line treatment did the best, since they'd have the slowest progressing disease. There was a modest difference there as well, with those who needed to move promptly to second line treatment (< 3 months) having a median survival of 6.9 months vs. a longer survival if you had a longer interval before needing second line treatment (9.2 and 9.3 months for 3-6 month and >6 month interval, respectively).

As has been seen in some other trials, such as the BR.21 trial of tarceva vs. placebo for 2nd or 3rd line therapy (abstract here), patients with adenocarcinomas had the longest median overall survival (9.1 months), followed by "other/mixed" (7.8 months), with squamous a little further behind (6.5 months). And women had a significantly better median survival compared with men, a full two month difference (9.4 vs 7.2 months). Age of younger than 70 vs. 70 or older didn't make a difference in overall survival.

There is significant attention in the paper on a trend toward better survival in the patients who received a combination of a platinum (cisplatin or carboplatin) and gemcitabine compared with a platinum and a taxane (such as taxol) or another platinum-based doublet:

But this difference really didn't emerge as statistically significant.

One other issue came up in the further analysis of the trial, and that was the difference in outcomes with alimta vs. taxotere in patients 70 and over. I'll cover that next.