Medicine in general has become increasingly evidence-based, which we think is a good thing.  Rather than have doctors use a wide range of approaches to the work-up and treatment of many problems, there are a growing number of guidelines that shape our “standards of care”, the defined best treatment practices.  Oncology more than several other medical fields is very data-driven, looking to the actual evidence from clinical trials to help us define what is the best approach, if there is one.

   Let me say that, in general, I’m a fan of evidence-based medicine and guidelines to define a standard of care.  I consider it a good way to ensure that the vast majority of people receive good treatment rather than a very wide range of options in which many patients would fare far worse.  But there are limitations to this approach.  First, diseases that are too uncommon to be studied well, or less common subgroups of people with common medical problems, can’t be studied well enough to develop good evidence.  Lung cancer is common and therefore well-studied, but tumors of the salivary gland are not.  Pharmaceutical companies aren’t clamoring to develop new agents for very small patient markets (it costs a lot to develop new drugs, and it makes more sense to try to develop them for larger rather than smaller markets); the National Cancer Institute isn’t being pressured by patient advocates with salivary gland tumors to channel funds for clinical trials.  Those trials never happen, and doctors and payers may say, “there’s no good evidence that treating with X helps these people.”  True, but absence of proof isn’t proof of absence.  Or, in the quote from the influential and somewhat sensationalized novel about the back stage of medicine, The House of God, “If you don’t take a temperature, you can’t find a fever”.  Sometimes falling back on the evidence-based medicine can be a cop-out, particularly if you’re depending on good evidence and the question hasn’t been asked properly. 

   Taking from the field of lung cancer, which is reasonably well-studied overall, you haven’t seen much written here about the best way to treat older and sicker patients with locally advanced NSCLC.  Why?  Because there’s been so little research on this, even though a huge proportion of patients in the real world are rather unlike the cherry-picked, healthier patients who have no other cancers or serious medical problems than their lung cancer.  The same could be said for subgroups of patients in every corner of oncology, who fall between the cracks of our neat categories of more straightforward patients. 

   And that leaves us using our judgment very often, which isn’t necessarily a bad thing.  If you don’t have enough evidence to provide a clear answer, use the bits of evidence you do have to develop as compelling a plan as possible, acknowledging the risks and benefits and limitations.  But know that there isn’t any right answer, which is why two or three good smart doctors may develop two or three different proposed treatment plans for a patient even if furnished with the same information.  I have a great disinclination to give chemo and tarceva together, but at Memorial Sloan Kettering several of the doctors routinely do this for never-smokers with NSCLC.  We each point to different pieces of evidence that lead to our recommendations, and neither side is “wrong”.  In the end, you’d like to hope that evidence-based medicine would take “faith” and all of the interpretation out of the process, but it doesn’t. 

   So yesterday I presented a case at a tumor board, actually a case of prostate cancer, which is another of my interests.  The case involves a 50 year-old man who had prostate cancer diagnosed, and after a good work-up was found to be a good candidate for surgery.  Unfortunately, the pathology review after surgery demonstrated some areas of very aggressive-appearing cancer in his prostate gland and one lymph node involved, microscopically.  In the world of prostate cancer, once it gets out to a lymph node, it’s conventionally considered incurable (because cancer cells are essentially always traveling through the bloodstream at that point), although it is often a very chronic disease over many years.  That long course may sound fine to an 80 year-old with several significant medical problems, but “many years” isn’t what most otherwise very healthy 50 year-olds want to hear.  In fact, most patients would start early hormone therapy, and we have no evidence of a survival benefit for chemo and a very unclear role for radiation in that setting to improve outcomes.  We debated all of these points at our tumor board, and I would consider doing everything from a close observation and try to stretch all of our treatments one at a time over many years, to an approach of doing hormones, chemo, and radiation all pretty early in an attempt to kill every last cancer cell.  But other smart, well meaning doctors at this meeting sometimes offered strong opinions about chemo definitely not being standard of care (that’s absolutely right) and a lack of evidence showing that radiation is better if given earlier.  In truth, we just don’t know because most of these questions just haven’t been asked properly.

    We all want to practice sound, data-driven medicine, don’t want to be considered a “cowboy” who is treating wildly beyond the standard of care, partly because it becomes indefensible if something bad happens to a patient of yours when you went out on a limb.  A few patients who receive chemo with avastin will have fatal bleeding, and we know that risk is small but real, about 2-4% for the most selected, avastin-eligible patient population.  If a terrible side effect happens and someone dies of a rare complication, that’s very unfortunate, but we know it can happen.  But if the doctor gave avastin to someone with treated brain metastases and they have a bleed in the brain, that oncologist may find himself in deep doo-doo (technical term referring to fecal material) because avastin’s not approved for these patients.  I’m not saying it isn’t an appropriate consideration, but there are enough people who would say it’s not the standard of care and would be damaging expert witnesses.  Beyond concern about actual legal repercussions, there’s also a general culture in which the most influential thought leaders, the nationally known experts who lecture, are also likely to emphasize the data-based approaches and chide treatment outside of the boundaries of clinical trial evidence.

  But everyone here knows that sometimes the standard of care leads to pretty disappointing results.  In our discussions about DCA, for instance, I certainly appreciate the curse of a lack of evidence and that many people don’t give a flying fig whether there’s incontrovertible proof of benefit.  When the standard of care is no more treatment, or a treatment that most consistently offers a benefit measured in weeks or months, taking a chance on a wild card outside of the standard of care looks like a pretty good bet.  I think that’s part of why naturopathic medicine is so appealing to people in this country, since its practitioners offer much of what has been lost in US-based contemporary allopathic/Western medicine: more time for patients, more personal attention, less constraint by guidelines, and more hope, oftentimes. 

   The situation with DCA, an agent that is going to be involved in clinical trials but can potentially be obtained outside of the trials, is an interesting predicament.  Patients for their own self-interest may prefer to take DCA without the evidence, accepting a risk of harm.  But the evidence to support its use, that could lead to much wider awareness and FDA approval and general use, depends on properly conducted clinical trials.  Such trials remove some of the control from patients, for the good of society in the hope of defining better treatments for the future.  I am a huge fan of clinical trials, but I can certainly understand why people would want to exert more control over what happens to them, if that option is available.  Sadly, one of the reasons that more and more clinical trials are being done in Europe or Asia or Canada rather than the US is because these systems often don’t have certain treatments available except on a trial.  Do you want Avastin?  If you go on a certain trial, you may get it, but otherwise, it’s not likely to be feasible for you.  In the US, if a drug is commercially available, it’s often quite hard to study it in new settings, because people can always lobby to get it whether it’s indicated in that situation or not.

   After our tumor board, in which some of the other doctors were pretty dismissive of more aggressive treatment like chemo for this unfortunate man facing a very high risk prostate cancer, the pathologist (who is the head of his department and very knowledgeable) came up to me and said, “if I had that tumor in my prostate, I’d want you to give me everything you’ve got”.  That’s not standard of care, but one of the big challenges we face is deciding when the evidence is helping us and when we’re being constrained by it.