“Are we theeeeere yet?” Every parent knows that familiar whining from the back seat of the car. Sometimes, I feel as impatient as the kid in the back of the car. Although EGFR mutation helps with predicting tarceva (erlotinib) response, I want markers for cytotoxic (traditional) chemotherapy now. Unlike the parent in the driver’s seat of the car, I’m not entirely sure where we are, or where we’re going. Sometimes I feel as though I’m driving in the dark. Could molecular markers be the headlights?

A new article in the Journal of Clinical Oncology brings us a few more miles down the road. I want to talk about the biocorrelates of the study, but the original point of the study is important enough to pull off our metaphorical road for some brief sightseeing. The investigators randomized patients with a performance status (PS) of 2 to chemotherapy with single-agent gemzar (gemcitabine) or gemcitabine plus carboplatin. Before proceeding, let’s review what PS2 means.

PS Definition
0 Normal activity - Fully active, able to carry out all pre-disease performance without restrictions

1 Symptoms but ambulatory - Restricted in physical strenuous activity, but ambulatory and able to carry out work of a light or sedentary nature (house work, office work).

2 In bed 50% of time - Capable of only limited selfcare, confined to bed or chair for more than 50% of the waking hours.

3 In bed >50% of time - Capable of only limited selfcare, confined to bed or chair for more than 50% of the waking hours.

4 100% bedridden - Completely disabled. Cannot carry out any selfcare. Totally confined to bed or chair.

As you can tell, these definitions are vague and broad. At any rate, they hypothesized that the addition of carboplatin would increase median survival from 2.4 months to 3.8 months. They wanted to power their study to have a 90% chance of finding this difference, if real, and calculated that they needed 220 patients to do it. Unfortunately, they only accrued 170 patients. More fortunately, the patients on both arms lived longer than expected-- 5.1 months in the gemcitabine alone arm (yellow in the graph below) and 6.7 months for combined gemictabine and carboplatin (blue). The p value for this difference was .24, making this a negative study. However, like Popeye without his spinach, the study lacked power. So although the study didn’t pass the magic .05 bar and we can’t conclude that survival is superior with carboplatin, we also can’t conclude that they don’t do better. So, I’ll keep doing what I was doing before this study—personalizing therapy in this group to the individual patient at hand, with a bias towards using carboplatin in patients at the more functional end of PS2.

Okay, we’ve completed the scenic side trip and we’re back to the main story—molecular markers. This study also looked at two proposed chemotherapy sensitivity markers — ERCC1 and RRM1. ERCC1 is involved in repairing DNA, so cancer cells low in ERCC1 may be less able to stand up to the DNA damage caused by platinum compounds. A prior study suggested that ERCC1 may predict for cisplatin efficacy in adjuvant chemotherapy and another study retrospectively suggested utility in advanced disease. RRM1 is the major target of gemcitabine; we have hints that gemcitabine works better in people with low RRM1, but also no proof. The investigators of this study looked at their results by RRM1 and ERCC1 levels and showed correlation between disease response and both RRM1 and ERCC1 as shown in the figure below. The relationships are promising, but the imperfection of the correlation should be noted. Look at a single point on the x axis, for example, just to the right of the “39.1” on the first graph. If you look vertically upward, you will note that there were people with a variety of rates of disease response, disease stability, and disease progression at this level. So the markers can help, but they are far from absolute.

What level of proof do we need to be “there”? In my mind, proof would be a prospective trial showing that my patients will actually live longer and have a higher quality of life with marker-selected chemo over what I’m doing now. The first hints of such an effect came from George Simon’s Phase II MADeIT trial. Dr. Simon conducted MADeIT when he was down in Florida at the Moffitt Cancer Center, but he’s since moved to the Fox Chase Cancer Center in Philadelphia.

In August, Dr. Simon, Dr. West, myself and most oncologists interested in lung cancer were at the World Lung Conference. There, Dr. Simon presented updated survival results from MADeIT and he sat down for an interview with Dr. West. The results from MADeIT were impressive—13.3 month overall survival and 44% response rate without the use of a biologic agent. A phase III trial is ongoing at Fox Chase, Moffitt and many other sites that, if positive, would confirm these findings and get us “there”. Of course, even if use of these biomarkers are confirmed to improve survival, we will still have more work to do—finding better biomarkers, and validating biomarkers for other chemotherapy, such as thymidylate synthetase (TS) for pemetrexed.

So where are we now? My pathology lab, like most, will not run ERCC1 or RRM1, considering it unproven. Most oncologists are not outraged at this, considering the markers promising, but not yet ready for routine clinical application. We’re not there yet, but I have more faith that we’re on the right road.