A Tiny Cancer Trial Generates Huge Headlines
Is the media overstating the importance of the recent rectal cancer trial with a 100% remission rate?
Wet blanket. Debbie Downer. Buzzkill (that one hurts). I am often called these things, because I tend to throw shade at sunshine headlines proclaiming medical miracles. So, when the New York Times tripped off a cascade of upbeat stories this week with its reporting on a 100% cure rate in a small trial of patients with rectal cancer, my inner cynic frowned.
Despite the reality check subtitle, the article’s mood was one of unfettered optimism. The results were “astonishing.” The cancer had “vanished.” This was a trial without precedent in the world of oncology. In the words of one of the study’s lead investigators, Dr Luis A. Diaz Jr.:
Pretty exciting stuff! To be fair, it really is… for people with a relatively rare subtype of a not-very-common cancer, caught early in their disease process. I am concerned, though, that people reading this headline might feel like we have reached a sudden revolution in overall cancer care. We have not.
There is real cause for optimism, however, and not just for the few thousand people who might be diagnosed with an early stage rectal cancer this year with the right genetic mutation. It helps to understand how this one medication — dostarlimab — fits within the broader context of cancer immunotherapy, to better see how its applications might be limited, despite an increasingly promising future for cancer care in general.
One thing to be wary of when it comes to medical skeptics (and cynics): we tend to be smug, because skeptics tend to have a fantastic track record over time in the medical industry. I struggle to name more than a couple genuine “game changing” pharmaceutical advances per decade in the four decades I have been watching medicine with a backstage pass. In 1998, the year I started medical school, the same New York Times ran an article quoting Dr James Watson (yes, the “Watson & Crick” Watson) that Dr Judah Folkman’s approach to blocking the growth of blood vessels in tumors would “cure cancer in two years.”
I remember being very excited at the time. It’s been twenty-four years now, though, and while the medicines developed by Dr Folkman helped improve survival for several classes of cancer, Dr Watson’s enthusiasm proved a bit over-ebullient.
That said, I would argue that the greatest single medical innovation of the 2010s was the arrival of immune checkpoint inhibitors, the class of drugs to which the newer dostarlimab belongs. They radically improved the prognosis for people with advanced malignant melanoma and non-small cell lung cancer in a way we rarely see in oncology (or medicine as a whole).
To understand the remarkable way in which they work helps explain why they will not cure your aunt’s pancreatic cancer, unfortunately. If we look at cancer from the simplified viewpoint of unchecked cellular proliferation flying under the radar of our immune response to “not-self” antigens, we might assume that tumor cells that mutate a lot would be easier targets for immune T cells to identify and destroy. However, tumors have their own evolutionary adaptations, and that includes the ability of the antigens on the surface of tumor cells to bind certain “checkpoints” on our T-cells, checkpoints which can trigger T cells to die via apoptosis rather than attack the tumor cell (as well as promote the types of T cells which suppress immune responses).
It’s a lot more complicated than that, but you might get the picture: certain tumor types have a growth advantage (high mutation rates) that is also an immune disadvantage (easy recognition by T cells) which they can neutralize by suppressing T cells (by binding checkpoints). In the example of rectal cancer, some tumors feature cells short on proof-reading functionality; these “mismatch repair-deficient,” or MMRd, cell lines tend to acquire a truckload of mutations and are the ideal target for a checkpoint inhibitor medication.
Immune checkpoint inhibitors essentially flip the T-cell self-destruction switch back to “off.” We have these checkpoints — and there are many more than the PD-1 (“programmed cell death type 1”) receptors targeted by dostarlimab — for good reason, to prevent the sort of attacks on our own cells that lead to auto-immune disease. Some checkpoint inhibitors are famous for triggering auto-immune problems, like the severe colitis common with the CTLA-4 inhibitor, ipilimumab. In general, though, the PD-1 inhibitors are fairly well tolerated; certainly a world better than the ubiquitous and often severe side effects associated with traditional radiation and chemotherapy.
I realize this all sounds pretty great, especially for shareholders of GlaxoSmithKline, the company making dostarlimab. Precision medicine. Well tolerated. Unattractive alternatives. It also sounds a little too good to be true. Everyone fully cured? No serious side effects? Really?
The study itself was just published in the New England Journal of Medicine. Twelve patients with stage 2 or 3 rectal cancer (no metastases) had completed 6 months of treatment with dostarlimab and had 6 months of follow-up. They are still enrolling more patients, with a goal of 30. All had complete remission of their cancer, with no tumor visible on MRI, PET scan, or endoscopy. Three quarters of patients had mild-to-moderate side effects like rash, itching, nausea, or fatigue, with no placebo group for comparison. None had severe adverse reactions, below the 10-20% rate most expected.
I don’t see huge holes to poke in this study. I mean, it’s not a randomized controlled trial, obviously, but we don’t exactly expect a high rate of spontaneous remission of rectal cancers. I do wonder at the enrollment process for the study, which is not described in detail. While I don’t expect deliberate foul play, I have to ask: was every single available MMRd patient with stage 2 or 3 rectal cancer included in the study? I mean, everyone likes a study to succeed, especially when it’s funded by the maker of the drug, and when the two lead investigators have either invented multiple genetic tests for tumors, like Dr Diaz, or receive funding from and sit on the advisory board of GSK, like Dr Andrea Cercek.
It’s possible that the study participants were a particularly robust bunch. My extensive research on the matter reveals that the four patients pictured in Sloan Kettering’s promotional article would be right at home in a Gap photo shoot. They look radiant! I’d buy those jeans in a heartbeat.
Maybe out in the real world, where rectal cancer patients might be old, obese, diabetic, or have advanced heart disease, dostarlimab might not fare quite so well.
It’s also not clear whether that 100% remission figure is all that “astonishing.” An accompanying NEJM editorial from Dr Hanna Sanoff raised the point that traditional care in this cohort has reached 77% 3 year survival, not that far off the 100% mark described over 1-2 years in the new study. Dr Sanoff also describes trial evidence that nearly 60% of patients offered chemoradiation therapy followed by chemotherapy were able to avoid surgery at 3 years. She notes that in a group of colorectal cancer patients with this MMRd mutation but metastatic disease, 1 year survival was a less-sunny 55%, and 30% of those with an initial improvement lost signs of response by the 3 year mark.
I take these concerns as appropriate caution in the face of an exciting small study. However, I think a patient newly diagnosed with stage 2 or 3 rectal cancer that tests positive for the MMRd mutation would be mad not to pursue a treatment approach that has the potential to avoid rectal surgery and achieve a high rate of remission with far less grueling side effects. It’s a clear win for medical science!
Is it really headline news, though? To see the big picture a little better, I called up my cousin, Steve Grossman, who is the research oncologist in the family. In fact, he’s the “Cancer Physician in Chief” at USC Norris, with a research specialty of gastrointestinal cancers, so he knows a bit more than I do about this subject. When I started in on the hype around this study, he sighed, and said that while we may have cured a few thousand people with far fewer long term complications than the current standard of care… no, we have not “cured cancer.”
Put bluntly, there might be a lot of people with colorectal cancer; but a smallish number of those have rectal cancer; and only 5-15% of people with rectal cancer have the “right” mutation to respond to checkpoint inhibitors; and at best half of those people will present to their doctor in time to be caught with stage 2 or 3 disease. Dostarlimab and other checkpoint inhibitors might be fantastic treatment options for people with certain cancer types, but only a small percentage of people have those types. It’s not a “cure for cancer,” much as the headlines might nudge us towards that reaction.
After Cousin Steve patiently answered a barrage of questions about the mechanism of action of checkpoint inhibitors and the Sloan Kettering study, I asked him to confirm my understanding that in no way does this study imply that checkpoint inhibitors would ever be highly effective against more than a small proportion of cancers. He agreed. However, he also reminded me of the huge number of other promising approaches to harnessing the immune system to treat cancers: vaccines, CAR-T and TCR-T therapies, bispecific antibodies. And that’s just immunotherapy; Steve’s own avenue of research is cancer-specific metabolism.
It seems plausible that while we might never drown cancer under a sudden tsunami of successful research, we might see a steadily rising tide, with high water marks like this rectal cancer study appearing with greater and greater frequency. As we get better at identifying the unique qualities of individual cancers, it stands to reason we’ll get better at locating their treatable weaknesses.
I asked Steve whether he thought there was a real chance that we would have broad breakthroughs in the coming decade, of the sort that we could effectively treat, say, 80% of cancer types. Long pause. I pressed; would it be a long shot to accomplish this, or a coin toss? Another pause. Then: “it’s a coin toss.”
I put a reminder into my iPhone for June 2032 to give him a follow-up call. I hope Cousin Steve ends up on the right side of medical history.
A Tiny Cancer Trial Generates Huge Headlines
I did med school in Germany 35 years ago and briefly practiced psychiatry thereafter. Long story, irrelevant to your post. I recall just enough to know how to read professional peer-reviewed papers and when to become suspect. I maintain a NEJM subscription.
I'm suspect of anything with a100% success rate, because medicine is the most primitive of the sciences. This does not make me a denier, it makes me someone who is not fully-convinced about most things, especially single small studies with astonishing results. I recall the definition of the scientific method containing the word "replicable" or repeatable. The referenced study is a data point. I want to wait until an expert institution in medical data analysis, such as Johns Hopkins University, has enough trials analyzed to offer an informed opinion. I know we're not there yet.