Post-operative, also known as adjuvant chemotherapy, is the established method for delivering systemic therapy to improve long-term outcomes beyond what surgery alone can deliver. An alternative approach, though, is to give treatment prior to surgery. This gives the potential advantage of treating potential micrometastatic disease at the earliest opportunity, identifying the response to treatment given rather than treating "blind" with no disease to follow, and potentially delivering treatment more reliably because more patients will receive treatment as planned than in the post-operative setting, where many patients may drop out of consideration for chemo as they contend with recovering from surgery and potentially having complications.

Still, the higher profile pre-operative chemotherapy studies haven't been positive, or at least not positive enough. But it's worth reviewing what's being considered a negative study.

First, there was the SWOG 9900 trial, designed with a plan to enroll 600 patients to detect the differences they hoped to, but closing after only 354 were enrolled, in the face of a series of positive post-operative trials making it very difficult to continue to enroll patients on a trial of pre-op chemo vs. up front surgery, with no plan for adjuvant chemotherapy. This trial, for patients with stage IB - IIIA NSCLC (without N2 nodal disease, a higher risk group), gave 3 cycles of pre-operative carboplatin/Taxol (paclitaxel) and was reported as showing a non-significant trend toward more favorable survival in the chemotherapy recipients. Results are illustrated on the curves for progression-free and overall survival shown below:

(click on images to enlarge)

Though these differences were not quite statistically significant, they both reflected an approximately 20% improvement with chemotherapy compared with surgery alone. Though some have speculated that these results may have been more favorable if a cisplatin-based regimen were used, the magnitude of relative improvement here with carbo/taxol seems very comparable to the results we've seen in positive adjuvant trials with cisplatin-based regimens, but the trial was definitely underpowered, closing with just under 60% of their intended patient enrollment completed.

This year at ASCO, we saw the long-term results of a French trial that used an older and arguably obsolete chemo regimen of mitomycin C, ifosfamide, and cisplatin (termed MIC or MIP, for cisplatin, or platinum, respectively) before surgery, compared with surgery alone. It actually had a fairly complex design with a few variants, shown below:

This study had actually been published in preliminary form all the way back in 2002, but the updated results with median follow-up beyond 13 years was presented by Dr. Virginie Westeel. This study did enroll higher risk patients with stage IIIA N2 NSCLC, who in North America would be excluded from trials that included surgery alone because of the unacceptably high risk of recurrence without chemotherapy before or after surgery. Importantly, not only were these patients enrolled on this study, but they were disproportionately assigned to the chemotherapy arm (42% vs. 28%, a nearly significant difference, at p = 0.065). This difference would be presumed to disfavor patients assigned to chemotherapy compared to the group assigned to surgery, with an arguably better prognosis.

What the results demonstrated was a non-significant trend toward more favorable survival in the recipients of pre-operative chemotherapy, with an 8-9% absolute difference that is maintained beyond 5 years and even out to 10 years:

The differences in disease-free survival are of a similar magnitude, just slightly greater, but enough to actually now be statistically significant:

So why would there be a difference between disease-free survival and overall survival so far out? The two reasons I can think of for that small difference would be death from competing medical problems over a long period of time, or death potentially as a complication or long-term consequence of chemotherapy.

Dr. Westeel and her colleagues also did a multivariate analysis in which the role of chemotherapy was separated from the differences in T stage, N stage, age, and other factors (remember that the chemotherapy arm had been assigned a higher risk population). Isolating that variable of chemotherapy revealed a statistically significant difference of an approximately 40% improvement.

In the commentary at ASCO, the new data from the French study were shown along with the SWOG 9900 trial, highlighting that the improvement in survival, in absolute terms, is very comparable to the benefits of chemotherapy in the post-operative setting. I believe this is quite true, but that both of these trials were underpowered to show a statistically significant benefit compared to the larger trials of adjuvant therapy.

But the best test of one strategy against another would be a trial in which patients in the same trial are randomized to either pre-operative or post-operative chemotherapy. That's what we'll turn to next.