A propos of nothing in particular, a memory came back to me this morning. I was a first year — more precisely, first month — oncology fellow at the National Cancer Institute, and had the dubious honor of having bone marrow transplants as my first rotation. This was back when NCI did not require proof of legal residency to enter the campus, so we had quite a few referrals from all around the country for people who otherwise would never have been able to afford one. These were clinical trials of bone marrow transplants, but the risks of the research procedure — whether to use drug A or drug B, bot approved, to prevent graft versus host disease — were minor compared to the risks of a standard full-on allogeneic transplant, so before you think “but isn’t this undue influence?” let me reassure you that, for the most part, it was not.
Anyway, one of these patients was a young woman with Burkitt lymphoma not responding to any chemotherapy you threw at it. I saw her in clinic early on during the rotation and despite having this rapidly progressing cancer didn’t look all that bad. She had just finished the last round of Hail Mary chemotherapy meant to control the disease before the transplant so had to wait a couple of weeks before being eligible for the trial, and so she went home somewhere in the Midwest to be with her family during this “washout period”.
She did come back, with a portruding belly and a hollowed out face, both of those tells that the lymphoma did not care an iota about any hail marys her doctors threw at it. Her labs, from blood counts to kidney and liver function, looked great, but true enough her PET scan had more areas lighting up than not, lymphoma everywhere.
In most other places this would have prompted a discussion about hospice and end-of-life care, but not at the NIH. She was admitted to the largest research hospital in the world to get some more chemo followed by, let’s hope, the life-saving transplant.
So I got to see her and talk about her at rounds each morning, and each morning that I saw her, belly stil protruding, face still hollow, I’d paint an uggly picture and at first hint then overtly ask, what were we doing here and why wasn’t she already in hospice?
One of those mornings, the inpatient attending who was by the luck of the draw also the primary investigator of the clinical trial she was supposed to be in had enough. All of her organs were functioning. She didn’t have an infection. The cells will be ready in a few days. “What exactly do you think will be the mechanism of her demise?”, he asked, and the phrase struck me as so out of touch that I remembered it to this day. This was a scientist first, maybe a great scientist, asking the question, but a piss-poor doctor.
The point of much of medicine is that we have no clue what the exact mechanism of someone’s demise will be, but we have seen enough and read enough to know that a quick demise was the most likely outcome. There were many things that could have gone wrong at any second: tumor eroding into a large blood vessel causing her to exsanguinate; massive infection taking over after her bone marrow was completely wiped out by our chemotherapy and her lymphoma combined; a large blood clot clogging up her lung’s blood vessels… It was a long list and I wasn’t a fortuneteller, so I shrugged my shoulders and moved on. She died a few weeks after I left the inpatient BMT service.
So anyway, when people ask how exactly the AI bubble will burst, in a way that’s challenging the very idea that it is a bubble that may burst, I remember this patient, shrug, and move on.