Last Month in Oncology with Dr. Bishal Gyawali


Me-too deja vu

I read the report of a phase 3 RCT of a “new” breast cancer drug but I had the feeling that I had already read this before. Later I realized that this was indeed a new trial of a new drug, but that I had read a very similar report of a very similar drug with very similar results and conclusions. This new drug is a PARP inhibitor called talazoparib and the deja vu was related to another PARP inhibitor drug called olaparib tested in the same patient population of advanced breast cancer patients with a BRCA mutation. The control arms were the same: physician choice of drug, except that physicians couldn’t choose the one drug that is probably most effective in this patient population (carboplatin). The results were nearly the same: these drugs improved progression-free survival, but didn’t improve overall survival. In another commentary, I had raised some questions on the choice of control arm, endpoint and quality of data about the olaparib trial when it was published last year. This current talazoparib trial is so similar to the olaparib trial that you can literally replace the word “olaparib” with “talazoparib” in that commentary and all statements will stay valid.

The oncology version of half-full, half-empty glass

The PARP inhibitors olaparib and niraparib are also approved in ovarian cancer based on improvement in progression-free survival (PFS), without improving overall survival (OS). If a drug doesn’t improve OS but improves only PFS, it should also improve quality of life to justify its use. According to two new reports, these drugs do not appear to improve quality of life. The niraparibtrial reported that the patients were able to “maintain” their quality of life during treatment while the olaparib trial reported that olaparib did not have a “significant detrimental effect” on quality of life. I find it remarkable that a drug that isn’t proven to improve survival is lauded for not significantly worsening quality of life … at $10,000 a month!

It is also important to recognize that these drugs were tested as maintenance therapy against placebos. For “maintenance therapies,” as explained in this paper, improving PFS alone is not an important endpoint. That’s why I am also not excited about this new trial of sorafenib maintenance in ovarian cancer. A drug has to be very ineffective to fail to improve even PFS as a maintenance therapy against placebo.


What if I told you that there is a drug that doesn’t improve overall survival or disease-free survival, and also worsens quality of life, but is FDA approved? This is the story of adjuvant sunitinib in renal-cell cancer (which I have written about multiple times, for example here and here). In the latest report from the S-TRAC trial, adjuvant sunitinib significantly worsened quality of life versus placebo. However, the conclusion reads: “Patients on sunitinib did report increased symptoms and reduced HRQoL, but these changes were generally not clinically meaningful, apart from appetite loss and diarrhoea, and were expected in the context of known sunitinib effects.” If reporting lack of improvement in quality of life as “no detrimental effect” is the “glass is half-full” view then this assertion that the reduced quality of life doesn’t matter because it was expected has to be “an empty glass is not empty; it’s actually full of air” view towards cancer drugs .

Brain mets: checkmate?

Checkmate 204 was a phase 2 study that tested nivolumab plus ipilimumab for melanoma patients with brain metastases and found an intracranial clinical benefit rate of 57%. Although whole brain radiation therapy or stereotactic radiosurgery would achieve higher response rates, a justification for using immunotherapy instead could be that, unlike radiotherapy, these drugs will also treat extracranial lesions and have shown improved survival in melanoma patients in general. They would also avoid the side effects of radiotherapy to the brain. However, there are important caveats to consider: first, the fact that only 57% patients had intracranial benefit means that the other patients still needed radiotherapy; second, the patients included in this trial were asymptomatic; third, this drug combination has its own risks including death; and fourth, we don’t know that combining nivolumab plus ipilimumab is better than nivolumab alone.

Making Precision Oncology Precise Again

Much of what is being sold as precision oncology is actually imprecise in that drug and target combinations are being prescribed based on preclinical data, bioplausibility, retrospective data, and single arm studies showing a few responses…in short, everything but survival outcomes in RCTs. There are two main harms with this approach: the over-expectation but under-deliverance with precision approach and the pessimism surrounding the whole precision approach despite some remarkable success like trastuzumab and in HER2 positive breast cancer patients. The ESMO Scale for Clinical Actionability of molecular Targets (ESCAT) is a welcome step in the right direction because it will at least help to bring uniformity in reporting these targets. Kudos to ESMO for trying to systematize the targets in oncology.

Let me take a selfie

recent RCT that showed that H. pylori eradication therapy lowered the rates of metachronous gastric cancer also showed a non‐significant increased risk of all‐cause mortality among patients randomized to eradication therapy (Hazard Ratio 1.95; 95% CI, 0.72 to 5.27; P = 0.19). The possibility that there could also be an association with an increased risk of all‐cause mortality is an important public health concern. To address this possibility, we performed a meta‐analysis of RCTs evaluating H. pylori eradication therapy focusing on the risk of all‐cause mortality. Reassuringly, we did not find a statistically significant increase in the risk of all‐cause mortality associated with H. pylori eradication therapy. However, continued real‐world surveillance of patients treated with eradication therapy as well as longer‐term follow up of these RCTs remains necessary.

Dr. Gyawali is a research fellow at the Program On Regulation, Therapeutics And Law (PORTAL) at Brigham and Women’s Hospital/Harvard Medical School. The opinions expressed herein are his own. This piece was originally published in ecancer here.

8 replies »

  1. Well-written. A good read. True that if reporting lack of improvement in quality of life as “no detrimental effect” is the “glass is half-full” view then this assertion that the reduced quality of life doesn’t matter because it was expected has to be “an empty glass is not empty; it’s actually full of air” view towards cancer drugs .

  2. Or more generally, does everyone use approximately the same societal resources throughout life and death?

  3. I wonder if one could look at all deaths and all their detailed associated costs, direct and indirect and opportunity costs, that one might find that everyone costs about the same amount to die?

    E.g. a young man killed in a skiing accident. This seems low cost, but he has lost years of productive employment, perhaps productive children, perhaps lost consumer demand over 50 or more additional years and hundreds of other lost opportunity costs….like his missing artistic contributions to his family.

    This total of his costs might roughly equal the total costs of an 80 year old who is treated for prolonged prostate ca disease using precision approaches with DNA studies, etc.

    The point of this thought exercise is that, if one could show this, then the society could issue everyone, fairly and equitably, at the time of inception, vouchers to cover catastrophic medical insurance policies for life….and the detailed underwriting would not be necessary because everyone would roughly incur the same costs throughout life.

    But, alas, it probably is not true.

  4. In any given year, the sickest 5% of patients account for 50% of medical claims but they are not the same people from one year to the next. Lots of people have an expensive event like a heart procedure and then recover. Some die after expensive care and their healthcare costs obviously cease upon death. A relatively tiny percentage need expensive medical care every year. Long term custodial care is a different issue.

    As for new high cost cancer drugs that don’t seem to be any better than less expensive drugs already on the market, why do payers agree to pay for them? Just because a drug wins FDA approval for outperforming a placebo doesn’t mean we should pay for it, especially if it’s significantly more expensive than alternative therapies already available. Drug companies can’t profit from a new drug if doctors don’t prescribe it. Can they?

  5. We continue to witness the unfettered expression, aka institutional codependency, of the Big Pharma business model that represents an abdication of their institutional social responsibility other than profit margin. Meanwhile, 50% of the population is offered 5% of our nation’s health spending for their HEALTH. Shame on us, all of us!

    Hats off to Doctor Gyawali and THCB’s Zoya Khan