BLOG

What Is a Meaningful Increased Survival?

By Rebecca Bechhold, MD¹ | December 26, 2012

¹Medical Oncologist, Cincinnati, Ohio

Without naming drugs, because there are several that fall into this category, has anyone else noticed that many of the new cancer drugs show an infinitesimal improvement in progression-free survival or overall survival?

Rebecca Bechhold, MD

I mean, we are talking weeks here. And the cost per day of extended life is thousands to tens of thousands of dollars. Someone has to pay for all this. I do not hear my colleagues discussing the hard data with patients, so how are they making the decision to take a personal and financial risk on these agents?

Maybe I am too practical, or am not being honest with myself about what I would choose in a similar situation. I think I know that I do not want to spend a ridiculous amount of my legacy for useless treatment, and I am not about to extend a life that does not allow me to do much more than visit the doctor’s office weekly and sit in a room full of other equally despondent patients waiting for a few minutes of face time with a doctor.

A recent study showed that 77% of patients who had been told their disease was terminal and that they would be on palliative chemotherapy, walked out and told the interviewer they were being treated for cure. So would they even comprehend the data on what is considered “statistically significant,” though perhaps practically irrelevant progression-free survival?

We have started a “treatment planning” visit for palliative chemotherapy patients to sit down with a nurse practitioner and go over the stark details of their illness and expected treatment course, including when hospice care will be appropriate. After 30 years of oncology, I am a firm believer that patients and families deal with concrete information far more rationally than soft, innuendo. If you leave it to their imagination they will always construct a more favorable scenario. Or, if you offer a new drug and no other option and leave out the details of drug efficacy, what choice do they really have?

On the other hand, I have treated patients I thought for sure were going to die before the drugs could work and they had a miraculous—truly!—response that lasted for months to years. But that is a tiny number of patients and I have probably caused far more suffering with desperate measures than healing. But boy, the woman with metastatic breast cancer who stopped taking Herceptin for a year and showed up with liver failure, ascites, and a coagulopathy who I thought would surely die in my office when she refused to go to the hospital was back the next week walking on her own after one dose of Taxol/bevacizumab/trastuzumab. And that was 4 years ago!

So when do we know who should get that Hail Mary drug and who needs to spend more quality time with their family? Performance status? Will to live? True grit? I do not know the real answer to that question, but I think a part of it is the art of medicine. In the meantime, I will try to always be compassionate but completely honest with my patients so they can decide what is meaningful for them.

by Richard Rosenbluth | January 03, 2013 11:39 AM EST

I share the writer's concerns fully. I suspect one has to be in practice for more than 25 years (35 in my case) before one comes to this conclusion. Unfortunately, the younger oncologists are far more treatment-oriented; perhaps their youth prevents their recognition of our shared common, eventual mortality.
I would also go one step beyond the writer - Much of what we do, using conventional chemotherapy (in the treatment of advanced lung and pancreatic cancers, in particular) bears little fruit when considering both overall survival as well as quality of life. I refer interested readers to an excellent op-ed in Wall Street Journal several months ago, titles (I believe) "Why Doctors Die Differently", where the writer of the op-ed exposes the fact that physicians, who are all too aware of what a terminal diagnosis implies, opt for comfort care only, in contrast with the treatments they persist in offering their patients.
In my practice, I always paint a clear and unvarnished picture of the future my patients face. Of course, such information is accompanied by an honest presentation of the benefits (even survival benefits) of palliative care.

by karol sikora | January 03, 2013 12:03 PM EST

Well said Rebecca. We all feel the same on this side of the Atlantic even in this era of personalised medicine. The real challenge is managing patient's expectations. As you say nobody takes chemo for palliation. They all think they are going to be the tail in the survival curve. I suspect you've blown any hospitality at ASCO!

Karol Sikora, Oncologist, London

by Diana Chalmers | January 03, 2013 12:23 PM EST

Thanks for the compassion.
I have personal experience of choice. for breast cancer drugs. Having a history of 'bad' reactions to Statins I have become very wary of medicines. I had breast cancer last year.and now am pretty sure Metformin ( for diabetes at the time) was just suppressing the breast cancer cells, awaiting me giving it up, as I was diagnosed poitve by mammogram within 5 months of giving it up!!! Post op there was a good discussion on what medication there was, the possibility of reactions and effects of giving up if severe problems. I declined the first offered of Metformin as I had 10 years of gut rot from that.. Anastrozole was to be tried , but within 42 days I had every side effect on the packet and more and was unable to walk!!! I now have taken the challenge. If I get another cancer at least I will be young enough to cope with the surgery and radiation.

I too have seen patients walk away when the professionals had thought all was lost. We just had regular morphine and good nursing care. they died in peace or went home, nor stuck in the midway zombie track of prolonged 'life' suffered now. Our aim was QUALITY of life not quantity

by Brian Birch | January 05, 2013 8:43 AM EST

I agree entirely with what Rebecca says.

The problem is that we would like to know the subset of patients who are going to get worthwhile responses to drugs but pharmaceutical companies would prefer us to treat everybody.

What we need is research that enables us to target drug treatments to an appropriate population - perhaps the increasing use of genomics will help in this respect.

Brian (Urologist UK)

by Richard Leff | January 07, 2013 10:01 AM EST

You are right on target describing the dilemma of deciding who and when to treat in the advanced disease setting. I don't completely agree with your interpretation of studies with minimal increase in PFS or OS, however. These studies are designed to prove that the treatment is active and is truly/statistically better than the standard/comparison arm. That is a minimal requirement for all new treatments. Once that is established it is the shape of the KM curve that means more to me. How many more patients are alive and progression-free at an appropriate interval (depending on the disease) is really important even if the median survival isn't moved much. Since the median measures number 50 of 100, if 20 patients benefit dramatically, and others only minimally the median may not be moved much but the KM curve may show significant benefit. (Consider if Herceptin had been tested in all women with breast cancer, not just those who are Her2 positive or if we had given up on erlotinib before we figured out how to identify which patients were moving the curve.) The trick is developing reliable ways of identifying the patients who are the ones who benefit. Biomarkers are certainly beginning to help but in the absence of a reliable marker, sometimes you have to treat the patient to know if he/she will be one who gets the greater benefit. All patients believe they will be one of the chosen who beat the median and, assuming they are appropriate candidates from a medical standpoint, denying them that chance may not be appropriate.