Our tendency in oncology is that once we find a new active drug in cancer, we try to add it to our current standard treatment approach and see if we can do better than what our current standard achieves. More is better. And we knew that the epidermal growth factor receptor inhibitors Iressa and Tarceva could lead to significant shrinkage of some lung cancers. So the lung cancer community was relatively optimistic about the clinical trials that compared chemo alone to the same chemo with Iressa or Tarceva. In fact, each EGFR inhibitor was studied in two different first-line trials, one each with carbo/taxol, for a primarily US-based experience, and also cisplatin/gemcitabine, a common standard for much of the rest of the world. Of these four trials, each with over 1000 patients, the overall results showed no benefit with the addition of either EGFR inhibitor:

(Click to enlarge either)

So how could we add two active anti-cancer approaches and not get any additional benefit? One idea is that perhaps we looked at the wrong group, that we should use targeted therapies in a targeted way. We did see encouraging results in the never-smokers on the TRIBUTE trial, who had a more than doubling of their median survival when tarceva was added to carbo/taxol:

As I described extensively in my prior post about why I don't give chemo and EGFR inhibitors concurrently, I think those results are encouraging, but I'm struck by the fact that the tarceva arm really seems to start doing better several months into the trial, right around the time when they'd be finishing the planned 6 cycles of chemo. So to me it seems that the never-smoker tarceva recipients did well despite receiving chemo and tarceva concurrently, because they really come into their own when the chemo is over (arrows added for the approximate time when 6 cycles of chemo would end):

Now look at how poorly other people do when they receive chemo/tarceva concurrently. Notice how the people with wild-type, non-mutated EGFR on the TRIBUTE trial appear to be on a lower (worse) curve than the people with the normal, non-mutated EGFR who received chemo alone:

Not so hot. And when we look at the patients who didn't develop a rash on the TRIBUTE chemo/tarceva arm, they did worse than the patients who received chemo and a placebo (again, a lower curve):

So many people appear to not only not be doing better, but actually doing worse, with concurrent chemo and EGFR tyrosine kinase inhibitors like Iressa or Tarceva. In fact, it's not unprecedented that a targeted therapy could interfere with chemotherapy. Kathy Albain, who is both a lung cancer and breast cancer expert, led a trial that showed that patients who received chemo and tamoxifen (an estrogen antagonist) did better if they received non-overlapping chemo and then tamoxifen, rather than concurrent chemo and tamoxifen:

In fact, tamoxifen causes tumor cells to stop at a point called G1 in the cell cycle, and this appears to lead these cells to be less sensitive to the damaging effects of chemotherapy. And this same thing appears to happen with EGFR tyrosine kinase inhibitors given concurrently with at least some kinds of chemo. Specifically, some lab work (abstract here) has demonstrated that Iressa causes apoptosis, or programmed cell death, in cell lines with mutant EGFR, but it led to growth arrest in cells with the wild type (non-mutated) EGFR gene. Over the past few years, a few labs have shown that EGFR inhibitors can block the cell killing effects of various kinds of chemotherapy, and the group at Univ. of California, Davis has done work suggesting that the best results are when chemo (taxotere, in one version they've studied) is given before EGFR inhibitors, but that EGFR inhibitors right before taxotere lead to an interference with taxotere-induced cancer cell death:

Other similar work has been done by the group at Albert Einstein School of Medicine, led by Dr. Roman Perez-Soler (abstract here), basically showing that tarceva given just before alimta actually led to antagonism and less effective cancer cell destruction by the chemo.

So how can we move forward with this information? This work suggests that it may be possible to move forward with more intelligent combinations of chemo and EGFR inhibitors, as long as the timing is optimized. This is the concept of "pharmacodynamic separation". Angela Davies led a trial of second-line taxotere combined with tarceva in two ways that were designed to avoid antagonism between the chemo and tarceva: one is chemo every three weeks with weekly high-dose tarceva starting the day after chemo, and the other approach is chemo every three weeks with daily tarceva starting the day after chemo and stopping 5 days before the next chemo is given:

Both of these concepts involve stopping the Tarceva several days before chemo, and some tests of the early patients on these trials showed that blood levels of Tarceva were very low by the time patients received chemo. This approach (called a Ping-Pong trial because patients were alternately assigned to one or other other arm as the trial proceeded) tested the highest doses that were feasible to give in each arm; a total of 42 patients were enrolled with some type of solid tumor, and this included 22 with previously treated NSCLC. For the first arm, taxotere at 70 mg/m2 IV every three weeks was able to be given with Tarceva at 600 mg weekly. However, there were no responses seen among 10 evaluable patients with NSCLC on that arm. For the other arm, it was possible to get doses up to 70 mg/m2 every three weeks for IV taxotere along with tarceva at 200 mg daily from days 2 to 16. On arm B, they had 10 patients evaluable for response, of whom they had 1 with a complete response, 4 with partial responses, and 3 with stable disease (abstract here). These are small numbers, but the results for arm B were encouraging enough to move forward with a larger phase II trial that Dr. Davies led and presented at ASCO 2007 (abstract here). This latter trial enrolled 37 patients with previously treated advanced NSCLC and gave taxotere 70-75 mg/m2 with tarceva at 150-200 mg/day (day 2 to 16 every 21 days). They saw a response rate of 35% (11 of 34 evaluable), which is very encouraging, and the median survival time had not yet been reached. On the other hand, low blood cell counts, infections, and sometimes a combination of fevers with low blood counts were problematic, and because of that, further work with this approach is moving ahead with the addition of growth factor support (neupogen or neulasta).

This work is still early in its development, but it's important to see that labs in NYC and Sacramento are seeing similar results when combining EGFR inhibitors with several types of chemo. Trials combining taxotere or alimta with tarceva are just getting off the ground. As someone who has been pretty definitive about not wanting to combine conventional chemo with EGFR inhibitor therapy, and who still wouldn't be particularly inclined to do it outside of a trial, I think that this approach of pharmacodynamic separation will likely show us the most intelligent way to combine them if there is a smart way to give them in an overlapping manner.