To begin with, my overall impression is that the preponderance of evidence on adjuvant (post-operative) chemotherapy supports that it can reduce the recurrence risk and improve the survival at five years, which I'd presume to be pretty close to the "cure rate". The benefit isn't uniformly distributed for all patients: higher risk patients, as defined by stage and other additional factors like number of lymph nodes involved and the grade of the cancer, also matter. Our current standards converge on recommendations favoring post-operative chemo for stage II patients and the minority of patients who have "surprise" N2 nodal disease not detected before surgery and are therefore stage IIIA and without evidence of residual disease. These patients derive a clinically significant benefit from chemo that more than offsets the small but real risk of serious side effects.

I consider it important to remember that 0.5 - 1% of patients on most of the clinical trials of adjuvant chemotherapy died from the treatment, and the trial population is non-representative of the broader "real world" population that is sicker and older than the patients who went on these trials, on average. And the most recent updates of one large adjuvant trial (prior post here) raises the specter that the risk from chemo may be not be captured completely in the first five years and that the outcomes may be more disappointing with follow-up beyond five years. Overall, I consider this to be far more of a concern for the patients I would already consider to be marginal for adjuvant chemo -- stage IB, or those who are already pretty frail and may be more harmed than helped by chemo. The emerging story is telling me that it's not just that adjuvant chemo either helps or it doesn't help. While it may help some, there could be detrimental effects, so it's worth being judicious in deciding whether to pursue it: more is not necessarily better.

Overall, I'm highly inclined to recommend it to healthy patients with stage II or IIIA disease, very unlikely to recommend it for patients with stage IA disease, and while I was previously inclined to recommend it for patients with larger stage IB resected cancers (using a 4 cm cutoff, see prior post), work like what I've mentioned above is making me much more inclined to individualize for all stage IB patients rather than reflexively follow a 4 cm cutoff as a threshold for recommending post-operative therapy. I was disinclined to recommend it for smaller tumors already, and I'm even more so now, but I'd want to have a careful discussion and really consider all of the factors for a patient with a 4.5 cm tumor and no nodal involvement.

But having today seen a 4 year survivor after resection of a stage IIB NSCLC who received post-0p chemo, I do definitely believe that chemo can work and decrease the risk for the right patients (this patient may have already been cured, but in my mind, I'm taking credit for it). If I recommend it, my preference is for a cisplatin-based regimen, since the positive trials have all included a cisplatin combination. I think that the carbo-based regimens may be nearly as good, but perhaps not quite, and when you're going for all-out cure, a percent or two matters. So if a patient has the general health, kidney function, etc., I'm going to recommend a cisplatin doublet, reserving carboplatin doublets for patients who have a reason they can't get it, or who start with cisplatin and don't do well with it. I definitely think it's better than no chemo, and it may actually be as good, but I'm not a fan of using carbo/taxol as my standard regimens for patients who are up for anything. Carbo/taxol didn't show a significant survival benefit in the trial that focused on this regimen (admittedly in a lower-risk group for recurrence (stage IB patients), as described in a prior post).

There are three regimens that are included right now in a large adjuvant trial of chemo alone or with avastin, being run across North America: cisplatin/navelbine, cisplatin/gemcitabine, and cisplatin/taxotere. Cisplatin/navelbine is the best studied, the one that has been part of far more of the positive studies than other regimens, but it had largely fallen out of favor in the US for advanced NSCLC and was actually inferior to cisplatin/taxotere in one trial in advanced NSCLC (abstract here). Despite that, I've been impressed that so much of the evidence is with this regimen and have often used cisplatin/navelbine. I've found it to be feasible for many patients, although I've modified the schedule to another commonly used way of giving cisplatin/navelbine as two weekly doses every three weeks, rather than giving weekly treatments with no break (most oncologists I speak with find that schedule to be remarkably difficult or impossible to actually administer as intended). I've sometimes used cisplatin/gemcitabine, a regimen I'm also perfectly happy with. Both cisplatin/navelbine and cisplatin/gemcitabine cause little or no hair loss in most patients and don't require steroid premedication, so that's helpful. I haven't favored cisplatin/taxotere, even though it's certainly a very active combination. I find it hard for many patients to tolerate even if they haven't just undergone major chest surgery, and the folks at Memorial Sloan Kettering actually reported that they found it to be very difficult to administer post-operatively as they had intended (abstract here).

Another issue that is starting to enter into my decision making is tumor histology. Now that there is increasingly compelling evidence that alimta may be particularly active in non-squamous cancers (prior post here), while cisplatin/gemcitabine may be more effective in squamous tumors (prior post here), I am becoming more inclined to tailor my recommendation by histology. There is a plan in place to add cisplatin/alimta to the short list of acceptable regimens in the aforementioned large adjuvant trial, presumably only for patients with non-squamous tumors. Given that this combination is also among the best tolerated, I'd be inclined to strongly consider and likely prefer cisplatin/alimta for non-squamous tumors and cisplatin/gemcitabine for patients with squamous NSCLC.

There's some evidence that molecular markers like ERCC-1 could be helpful in discriminating who should or should not receive adjuvant chemotherapy (prior post here), but I'd really like to see more work done on this before I incorporate it into routine decision-making. Right now, I occasionally obtain it as a tie-breaker in someone for whom I'm really on the fence about the benefit vs. the risk, but I don't think the evidence is strong enough that it should over-ride the standard of care of recommending adjuvant chemo in a healthy patient with stage II or IIIA resected NSCLC. That may well change as we get more information, but right now I'd consider molecular testing to be an up and coming approach that isn't quite ready for routine use in the clinic.