In most of the history of lung cancer management, we have been "lumpers" rather than "splitters", tossing together many different kinds of lung cancer together and presuming that they all respond similarly and should be treated similarly. After decades of primarily focusing just on whether a lung cancer was SCLC or NSCLC, with drugs like avastin, it's more important now to recognize squamous cell carcinoma, while there's also been a focus on using EGFR inhibitors for adenocarcinomas (although there is actually good evidence that patients with squamous cancers receive a very similar survival benefit despite a lower response rate). Beyond that, there are also a few "special" subset populations defined by lung cancer histology (microscopic appearance), such as bronchioloalveolar carcinoma, or BAC, which is a very well-differentiated end of the adenocarcinoma spectrum that can have a slower and more favorable prognosis than many other kinds of lung cancer and are often treated a little differently. On the other end of curve is an unusual NSCLC subtype called large cell neuroendocrine carcinoma (LCNEC), which many experts are inclined to treat as a distinct subset because it is generally recognized as having a poorer prognosis than other NSCLC tumors.

Neuroendocrine tumors come from cells that are part of the tissues that develop into neurons and hormone-releasing glands, and these cells are not only in specialized organ tissues like the adrenal glands or the pancreas, but also scattered in other tissues of the body, including the lungs. These cells make proteins that act as hormones, and these types of neuroendocrine tumors can be identified under a microscope by the pathologist performing special stains that demonstrate secretory granules typical of endocrine cells. The specifics aren't important, but rather just to know that neuroendocrine tumors are a distinct class, and in lung tumors there is a complete spectrum of neuroendocrine tumors that range from very favorable to much less favorable. Much of the most pivotal work in defining these types of tumors has been done by expert pathologist Dr. Bill Travis, now at Memorial Sloan Kettering Cancer Center in New York City.

On one end of the spectrum are carcinoids, which have such a favorable prognosis as tumors that they aren't historically considered a carcinoma. The term carcinoid suggests just a similarity to carcinomas. In fact, they are on the most favorable end of the neuroendocrine tumor spectrum. A typical carcinoid has a very slow growth rate, very few dividing cells, little to no chance of spread to lymph nodes, and a very high survival rate. An atypical carcinoid still has more dividing cells, a small but higher risk of lymph node involvement, and a generally quite favorable prognosis but not as excellent as for typical carcinoids. Both typical and atypical carcinoids are seen in a pretty even balance of men and women, with a much lower proportion of smokers than other lung cancers, and are often seen in younger patients, an average age in the 50s in many studies. In contrast, LCNEC and SCLC sit at the high grade end of the spectrum, with many rapidly dividing cells and a much less favorable prognosis. They also have a strong majority of men, up to 80-90% in some series, and are almost always seen in smokers.

The difference at the time of making the diagnosis of LCNEC and SCLC is really the size of the cells, namely that SCLC has, you guessed it, small cells, while LCNEC has large cells (also termed non-small cell in oncology's ungainly lexicon). SCLC has been recognized for its fast tumor growth and strong tendency to spread distantly early in the disease course. In contrast, LCNEC hadn't really been broken out from other NSCLC types until the last few years, but a rather consistent picture has emerged from nearly all of the studies that examine it. They can be challenging to study because LCNEC tumors only account for about 3% of NSCLC, but the studies offer a near consensus view that LCNEC tumors have a prognosis that is clearly worse than other types of NSCLC, and a high risk of recurrence after surgery, even for stage I tumors, where many studies show a long-term survival of only around 30%, vs. an expected 70% range for other NSCLC stage I cancers after resection. Here's the results from one trial (abstract here) for 41 patients with stage I LCNEC compared with the results of 126 other patients with poorly differentiated stage I NSCLC tumors:

(click to enlarge)

In contrast, several trials show the survival of LCNEC tumors to be in the range of SCLC tumors. Here are the survival results from one study (abstract here) that showed a markedly superior survival for typical and atypical carcinoids (typical better than atypical), and LCNEC and SCLC clustered far below, both for all stages (figure on left) and for stage I tumors (on right):

So the question is what can we do about it. If SCLC happens to be detected at an early stage and surgery is performed, whether planned or because SCLC was not surpected, chemo is routinely recommended after surgery because of the high risk for distant spread. Similarly, while adjuvant chemo is often recommended for patients with resected stage II or IIIA NSCLC, it is not as consistently recommended for stage I patients. I would be considerably more likely to recommend adjuvant chemo for a stage I LCNEC than I would for most other NSCLC tumors. I have also recommended cisplatin/etoposide, a regimen routinely used for SCLC, in such cases, since LCNEC appears in many ways to be as similar to SCLC as NSCLC. For non-surgical stage III disease, the approach of cisplatin/etoposide plus radiation is an appropriate choice whether you're using it for NSCLC or SCLC-LD: it's essentially the same approach for both settings. For advanced disease, you could consider using a SCLC regimen, but thus far we've never seen any results to suggest that LCNEC should be treated more like SCLC than other NSCLC subtypes. It's fair to say that we're still learning the best ways to treat LCNEC, but thus far we know that we need to treat it aggressively to combat its own aggressive behavior.