Right now we use the same conventional staging system for BAC as with other lung cancers. I don't have a great alternative just yet. I can tell you that as the lead investigator on several BAC trials, there are huge differences in the natural history of their cancer, regardless of what our treatment does. For instance, the trials for BAC are generally for metastatic/recurrent disease, which can mean anything from a single 1 cm spot coming up 4 years after a lobectomy for lung cancer (expect years and years of doing well, whether on treatment or not), to both lungs filled with BAC lesions (still can be variable, but often a lot less favorable). That's all still in the same patient group, making it pretty hard to figure out what the treatment is really doing. A trial will look great even if you give sugar cubes if it's filled with people who have a small amount of
slow-moving BAC.

Using "response rates", or the frequency of the measured tumors shrinking by 50% or more is a challenge because BAC can appear more like wispy clouds that don't have clear edges than discrete masses you can draw circles around. So the wisps or cloudy infiltrates in the lungs can become less dense from a useful treatment, and you might breathe easier, but it may not be a response because the edges don't change.

Here’s a typical BAC appearance on CT. Can you draw a circle around it to measure how it changes?

(click to enlarge)

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.