I doubt there is a group of lung cancer patients more common but less well studied than the substantial subset of frail and/or very elderly patients with advanced NSCLC. While “elderly” patients, usually defined as age 70, have been evaluated as a subset of the population in larger studies and even been the subject of specific studies just for the elderly, most of this work has shown that fit elderly patients do as well as younger patients getting the same aggressive treatment. What we have relatively little information about is how frailer patients, regardless of age but certainly more likely to be in their mid-70s or older, should be treated.

The available evidence shows that either a carboplatin doublet (I don’t advocate for cisplatin here) or single-agent chemo are feasible and associated with our best results. Among doublets, carbo/taxol (paclitaxel) looks favorable, and I’ve also felt very comfortable with carbo/gemcitabine, a combination that is usually associated with very little in the way of perceived side effects other than fatigue, and most of the adverse side effects are so called “paper toxicity” of low blood counts that the doctor needs to tell a patient about, not the other way around. However, older patients tend to have more of a problem with maintaining adequate blood counts without needing treatment delays, injections of white blood cell stimulators like neulasta (pegfilgrastim) or neupogen (filgrastim), or transfusions of red blood cells). Still, with minimal hair loss and nausea, and no requirement for steroids (which can be a problem in patients with diabetes because it increases blood sugar levels), carbo/gem is an option I’ve commonly turned to. And now with more extensive study, carbo/alimta (pemetrexed) is emerging as a convenient, quite well-tolerated doublet, but one I’d only consider for patients with non-squamous cancers, since converging evidence strongly indicates that alimta is simply not effective against squamous NSCLC (see prior post). I haven’t yet had occasion to use a carbo/alimta doublet in a marginal performance status patient, but if alimta is approved with a platinum as a first line agent by the FDA soon, it will become a leading consideration for my non-squamous patients.

The decision about whether to recommend a single agent or a doublet is really a hard one, for which I use a combination of a read on the general health of a patient and also their concern about side effects. Since either approach is completely reasonable, I lean more toward a single drug if a patient is kind of wobbly, expresses concern about whether treatment for advanced NSCLC is “worth” the side effects, etc. – and despite the findings that fit elderly can do well with treatments commonly used in 60 year-olds, I’m still more likely to recommend a single agent approach for a patient of about 78 or older. But it’s always got to be individualized. I’m actually giving a 76 year-old man cisplatin/navelbine as adjuvant therapy (a setting in which I am particularly wary about causing excess harm) because he appears to be healthier than many 62 year olds, and it’s a curative therapy setting where cisplatin may make a difference. If he were receiving treatment for metastatic disease, I wouldn’t even hesitate to give him a doublet, even projecting his age forward by a few years. But I’ve also recently struggled with the “carbo doublet vs. single agent” question in an aggressively minded, pretty well-appearing 79 year old man, then elected to give single agent therapy and was very glad I did when he had a remarkably hard time, and I turned him from a fit 79 into a frail 79 in a hurry (thankfully, just temporarily).

Single agent approaches can be a great choice for a patient wary about the potential side effects who you think could benefit from chemo if they give it a try. While I don’t want to twist anyone’s arm, I’ve nudged several leery elderly patients to just give single agent therapy a try, most of whom were surprised with how well they felt on treatment, some barely even noticing it. And quite a few have had gratifying responses and gone on to try two or three different single-agent regimens over time.

With regard to which single agent to use, this is practically a no-man’s land. Navelbine is among the best studied for the elderly, and it’s been the most common one I’ve turned to. Both it and gemcitabine are weekly treatments, or usually given as two weeks out of three (sometimes three out of four), which can be an advantage in that it gives an opportunity to get a lot of feedback from the patient and their labs about how they’re tolerating it, so you can make quick adjustments, or they can hit the emergency brake and get off of the program without making a big commitment (I reassure patients that starting chemo entails much less commitment than choosing a cell phone plan, with their extortionate early termination fees). Most of the work in this arena is with a single agent that shows feasibility, activity, and pretty good tolerability with single agents ranging from navelbine to gemcitabine, taxol, or taxotere.

One thing I don’t like as much about navelbine is that it’s a vesicant, which means that it causes a chemical burn in surrounding tissues if it leaks from the vein, so many institutions require patients to have a portacath (“port”) or PICC (peripherally inserted central catheter) line that essentially assures us that the chemo is going directly into a big vein in the chest rather than through a small IV in the arm. Many older patients have rolling, small veins that are especially prone to having exactly the kind of leakage you don’t want.

I may still use navelbine, often with a PICC line or port (our institution doesn’t require it), but I’ve often used gemcitabine. Largely because they require steroid pretreatment and induce hair loss, I haven’t favored the taxanes as much. While a Japanese trial of elderly (not necessarily frail) patients showed better results with taxotere (docetaxel) every three weeks than navelbine (prior post here), I haven’t been wowed enough to adopt that approach. The majority of US oncologists have come to perceive that many patients just have a hard time with taxotere, and there is evidence that the elderly patients did worse with taxotere than with alimta as second line therapy (see prior post). Speaking of which, alimta is also an option I’d definitely consider as a single-agent for a marginal performance status patient with a non-squamous NSCLC, but right now we really don’t have any evidence of it being used in a trial just for poorer risk patients. I’d just surmise that it will look as active and be as well or better tolerated than other options.

Finally, there’s the question of EGFR inhibitors like iressa (gefitinib) and tarceva (erlotinib). Both I (prior post here) and Dr. Laskin (prior post here) have written about a study that showed that poorer performance status patients convincingly do worse with carbo/taxol than with tarceva as first line treatment, so I don’t prefer to “back into” tarceva for patients whose main reason for giving it is that they’re too sick for chemo. I think it’s a fine drug to consider using an EGFR inhibitor first line for patients who have a particular reason to favor it, such as them being a never-smoker or having an EGFR mutation, but to me it appears that the people marginal for chemo aren’t especially well served by starting with tarceva – if they decline, they are out of the running for getting a benefit from chemo, while more patients who fail initial chemo can still try tarceva if they decline. But lest I get too confident about that conclusion, there’s a Japanese trial that directly compared navelbine and iressa in patients over 70 and showed that they performed quite comparably. I suspect, though, that the difference is that this trial wasn’t focusing on frail patients but rather on elderly patients, fit or less fit.

Overall, this brings us back to the early point that elderly isn’t the same as frail or marginal performance status, and that it appears to be consistently more important to tailor the treatment plan based on the performance status of a patient than their chronologic age. But this is still an area where we desperately need more studies: this is a large patient population for which oncologists are largely left to draw their own conclusions because there’s so little real evidence to guide us.