We define whether someone has cancer, and what type of cancer it is, by a piece of tumor tissue. This evidence of cancer under a microscope is considered critical, so much so that there is a general oncology dictum of “no meat, no treat”, requiring a tissue diagnosis before starting a treatment with potentially significant side effects.

Up until very recently, the only important decision-making point once you established that someone has lung cancer was whether it is small cell lung cancer or non-small cell lung cancer (NSCLC). While there are several subtypes of NSCLC, there was not much reason to clarify whether the type of non-small cell was adenocarcinoma (the majority in the US), squamous cell (the majority in Europe, only about 30% in US), large cell carcinoma (about 5-10% in most series), a rare type like large cell neuroendocrine tumors (in the range of 5%, generally felt to have a worse prognosis than other types), or NSCLC not otherwise specified (NOS). Cancers can also have a mix of different components, such as adenosquamous, just in case it seemed too easy to understand otherwise.

At the same time, cancers can be designated as well differentiated, which look pretty close to normal non-cancer cells, or moderately differentiated, or poorly differentiated, which look very mutated and unlike the non-cancer cells they originated from. In truth, however, with a moderately to poorly differentiated cancer, even expert pathologists (the medicine specialists who evaluate tissue under the microscope and provide the cancer diagnosis) can differ not only with each other but even with some of their own prior readings (perhaps relating to whether they drank regular or decaf coffee that day) about whether a cancer is one subtype or another.

Why does this matter now? For years we’ve often diagnosed cancers with a very tiny amount of tissue, like a CT-guided fine needle aspiration, or FNA, of the lung cancer. That often provides just a few cancer cells, enough to generally determine whether a cancer is small cell or non-small cell, but often not enough tissue to say whether a cancer is one subtype or another. Why might this matter? For several reasons. First, the newly FDA-approved drug Avastin for lung cancer is actually only appropriate for patients who do not have predominantly squamous cancers (patients with a mix of squamous and other cancers were permitted to receive Avastin on the trials if the squamous component was not more than 50% of the cancer material). However, patients with squamous cancers appear to have a MUCH higher risk of bleeding complications, and approximately 20% of the patients on the key Avastin trial had NSCLC NOS, meaning that there was either not enough tissue or the cancer was too poorly differentiated to say whether the cancer was squamous cell carcinoma. Although I believe that Avastin is an important addition to our treatment arsenal for lung cancer, I would be very concerned about possibly giving it to someone with a high risk for bleeding complications because we got too small of a biopsy to determine the level of risk.

Secondly, although right now there is no clear role for molecular tests such as EGFR mutation testing and other tests of certain proteins on tumors that might predict the chemo or targeted agents a person might respond well or poorly to, this seems to be the direction in which things are heading. In the next few years, EGFR mutations and other tests like ras (rhymes with mass) mutations (another important protein in the cascade of signals that control signals for growth and division of cancer cells) may be important in helping predict which patients will be very well served by certain therapies and which ones will receive only the side effects with little or no chance of tumor shrinkage and survival benefit. We may also be able to clarify from tumor tissue whether patients are more likely to respond to chemo drug A than chemo drug B. More and more clinical trials with interesting and promising drugs are requesting or requiring tumor tissue to enroll on the trial, and correlating outcomes with certain new drugs with characteristics of the tumor will likely increase our understanding and optimize treatment plans for the future.

Because of these issues, I would consider it preferable for patients to have a core biopsy of a tumor or metastasis, or to have a lymph node or other small bit of tumor removed from a minor surgery to make a diagnosis. This also may be a reason that patients are asked to undergo another biopsy procedure if they wish to enroll on some clinical trials. Although we are always reluctant to have patients undergo any extra invasive procedures, having enough tissue to really study the tumor is likely to be very valuable, for both the patient receiving treatment and the future of the lung cancer field.