Reactions to "Please be dying, but not too quickly" and what it's like for patients

Comments and conversation.

If you haven’t read the three part series on the clinical trial process, start here:

Part 1 is here: A clinical trial story.

Part 2 is here: The patient’s perspective.

Part 3 is here: The end of the journey

By far, the most interesting commentary on the clinical trial essay / guide came from the Astral Codex Ten comments section—though I also got interesting emails and comments here. If you have responses you don't want public, email me at DrBStillman@gmail.com.

On Astral Codex Ten, sclmlw writes:

I used to believe a lot more firmly in the promise that a 'cure' or at least a long-term solution for an unfavorable cancer prognosis lay in getting into the right clinical trial. I think this has to be balanced with the real possibility that the 'right' clinical trial won't start enrolling until too far into the future to do you any good. In other words, it's unknowable whether the right trial will extend your life, or by how much. But it is knowable how the search impacts your life and the lives of those closest to you. While I think it's unfair to ask anyone to "give up", it is fair to set boundaries on how the search/treatment is allowed to interfere in your lives. If your wife is still searching for open slots while preparing funeral arrangements you both may look back on this period and wonder if you sacrificed too much quality for a little extra

I believe in avoiding futile end-of-life medical efforts, which only prolong suffering and expense without extending quality life. I’ve seen plenty of those in the ER. But some (not all!) oncologists we consulted expressed pervasive fatalism and pessimism regarding clinical trials, despite the fact that Jake is a 39-y.o. man with good baseline functional status, not an 80 y.o. with multiple comorbidities who has lived a complete life. Maybe, as Jake speculates, it’s less fatalism and more overwhelm; they think there’s not time to understand the clinical trial landscape. They don’t have the bandwidth. Which is a shame; an oncologist at UCLA recently told us that there’s a lot of action in R / M HNSCC treatments. Jake has an essay in the works on that, too.

Commenters Turtle and Scooby Von Do both agree that oncologists have too much going on, and that most early trials don’t work. Turtle writes:

We want our patients to get on to promising trials, but we're limited by bandwidth. It's impossible for one human being to know all the research efforts. I completely understand how doctors trying to do the best by Bess recommended her to reach out to the big institutions like MD Anderson. This is basically what I say to patients too. 

Scooby Von Do writes:

First, since most drugs don’t actually make it through the trials, the docs may feel like it would be fruitless to follow the early stage trials. I believe they do know a lot about the drugs that have made it to phase 3 and are excited about them. Second, interacting with the trial system is hard and frustrating and the incentives are to avoid it - as you’ve probably experienced.

But bandwidth limitations can be symptoms of other inefficiencies: There’s no reason to reinvent the wheel for each patient. Oncologists should understand the national clinical trial landscape, especially the regional offerings, within their sub-specialty; after an initial time investment, it’d be simple to check once a month on promising studies.

Jake wrote:

Keeping up with the better trials, even via clinicaltrials.gov, shouldn’t consume lots of time. It took Bess an unbelievable amount of effort and energy to get up to speed from nothing, but the trial situation is like riding a bike: starting from a stop is much tougher than maintaining speed.

[T]o someone working in the field, it shouldn’t take more than a few hours once every month or two or even three to keep abreast of what’s happening.

A commenter on Marginal Revolution, in response to my proposal that this be part of an oncologist’s workflow, asked, “Who is paying them for their time?” But all physicians, myself included (ER), are required to stay up to date on relevant research and treatments within their specialty, with most of that professional commitment not being billable. I can’t defend myself in court that my patient had poor neurologic outcomes because no one ever paid me to read up on post-cardiac arrest therapeutic hypothermia.

That said, doctors are frequently valued by employers for their “productivity,” which isn’t the same thing as doing right by patients, leaving doctors to self-educate. Any hospital employer reading: oncologists should be paid for a few hours a month of protected time during which they’re expected to stay up to date on clinical trials. Everybody wins! Well, everybody would win more if there were a real-time, searchable database that matches patients to studies using EMR data and allows docs to see what spots are available and where. A girl can dream.

Back to sclmlw’s comment about how he “used to believe a lot more firmly in the promise that a 'cure' or at least a long-term solution for an unfavorable cancer prognosis lay in getting into the right clinical trial.” I don’t think that “belief” in the “right” clinical trial is relevant to the process at all. There’s often data to support a “best potential” trial. Maybe the trial that seems like the best shot will work clinically, maybe it won’t. All standard-of-care meds were once trial drugs, so the next standard of care drug may already be in trial. It might not. In Jake’s case, I’d wager on petosemtamab being standard of care for HNSCC within a year. Right now, clinical trials are the only major way to access that “maybe” before a drug reaches market.

“In other words, it's unknowable whether the right trial will extend your life, or by how much. But it is knowable how the search impacts your life and the lives of those closest to you.”

One doctor I shared our Excel spreadsheet of top 15 clinical trials with, hoping to get his input, responded by saying he didn’t think traveling out of state for a clinical trial worthwhile, because “there’s something to be said for waking up in your own bed.” And that was the end of it. Jake and I think there’s something to be said for not dying. And his tumors shrunk due to petosemtamab. If we’d taken that oncologist’s bad advice, Jake would be dying even faster than he is. We’ve learned to double check everything.

There’s a strange double standard in oncology around what’s an acceptable amount of interference in a patient’s life during treatment. Once Jake had his second recurrence on July 21, 2023, everyone thought palliative chemo, which might only extend his life a few months and cause intense discomfort, reasonable. But there were a lot of opinions on the (un-) reasonableness of traveling for a targeted therapy that might cause less suffering and a potentially better response.

There’s a paternalistic streak in many oncologists. We felt that paternalism especially when Jake offered to stop chemo and allow some progression in order to be eligible for a new study. Instead, he was told that he “had to fail the new line of treatment” even though he’d only received one dose of chemo at the time and would have been subject to the same washout period as anyone else. The idea that he’d stop a treatment that was “working,” even though it had an average duration of response of four months, and instead choose a treatment that “might not work, but might work better” was not well received.

It’s not worth traveling for most phase 1a trials, unless there’s significant compelling scientific evidence to support it. But it’s important not to assume futility. I received a some criticism, mostly via e-mail, for what seemed like my “obsessiveness” in the trial search: the Excel spreadsheets, the hundreds of calls and e-mails, as if the extent of my search was a sign of my own pathological inability to accept Jake’s fate. The only way I could learn what we did and get Jake into the trial he’s in is via the arduous path we took. And it worked! He’s still alive and sitting next to me right now. If there was an easier way for patients to find available trials (and for doctors to stay abreast on trial spots outside their own institutions and for patients to establish care and consent for the trials), it would improve everyone’s life.

Jake also replies:

I do recognize that I'm statistically unlikely to be cured: R / M HNSCC is overwhelmingly fatal, And sclmlw is right that the right clinical trial probably won't enroll until it's too late for me. But right now, while my quality of life isn’t great, it isn’t unbearable either, and I don’t want Bess to be lonely.

There are also a lot of promising treatments coming along. The most notable is Moderna's mRNA-4157 personalized cancer vaccine.

One more note on the way trials impact patient’s lives. Sclmlw says :

In my experience, these things work best when everyone discovers they have aligned interests. From the perspective of the pharma company, the most pressing interest is in finding the right patients to enroll. We pay a lot of money to start up sites across the country, focusing most of our efforts on determining whether the PI will be able to "find" the patients who fit the trial criteria. The focus of the current system is on selecting a physician who can find the patients, not the other way around. Lots of physicians promise they can totally find many patients. 

If drug companies want patients to enroll, and doctors want patients to enroll, why aren’t pharma companies reimbursing travel / lodging expenses for patients to travel to a trial, increasing the catchment area while decreasing barriers to entry?

Ragged Clown writes:

I think all of these data problems are easy to fix if there was a little political will. Vice President Biden organised a ‘Cancer Moonshot’ and one of the projects we worked on was to standardise the recording of trial data but it went nowhere. I wrote a paper on standardisation and recommended a Domain Specific Language that could be read by people as well as by search engines (and AIs). I counted 98 different ways of saying “The patient must not be pregnant” for example. I think this kind of standardisation should be easy for the FDA or NCI but no one is interested in fixing it.

While I was working on the Moonshot thing, CT.gov hired an intern to fix the presentation of the data to make it easier to search but the project got cancelled because they had no money

The political will required would be the willingness to hire programmers with UX experience. I stand by my recommendation that the same folks who create filter checkboxes to find the right dress on H&M would be a good place to start. Clinicaltrials.gov is not designed to be used as a clinical trial search, it just happens to be a poorly searchable study registry.

Update from Ragged Clown in the comments of this post:

It's definitely not a UX problem. It's a data problem. We spent several years building a trial search engine with the kind of dress-buying UX that you describe. It didn't work because the data is no good. It would be easy to add a handful of fields to the data in ct.gov that make the data relevant to patients and easy to search. AI will help a little — but not as much as fixing the data and fixing the data will be easy.

1. A patient-friendly summary would be good.

2. Tidy up and standardize the names of drugs and conditions so you can search consistently.

3. Use a consistent format that's patient-friendly and machine-readable format for inclusion criteria.

A good reminder of the garbage in\garbage out problem. I’ll amend by saying there needs to be political will to want to change ClinicalTrials.gov into something searchable and useable by clinicians and patients. If fixing the data is easy—maybe easy from a programming perspective, getting new data uploaded by thousands of different sites would be its own challenge unless there was a federal deadline—then the problem isn’t how to fix it, it’s how to get those in charge to want to fix it. Thinking about the ways in which the data is entered can correlate with a better user experience would be a nice touch.

At Marginal Revolution: mthrbmpr writes

“My daughter works for Friends of Cancer Research, which works with all the stakeholders in these trials, so I sent her Seliger's piece. Thought it might be interesting to readers to hear her reply:

"His second point "Hospital center sites and/or drug companies don’t appear to do much outreach to community or even specialist oncologists" is in the midst of changing - since COVID made everyone realize it's possible to run clinical trials without as many in-person visits, shipping meds, using community sites for routine testing, etc. there's been a lot of convos about how to "decentralize" clinical trials and part of that is large academic med centers/sponsors being more intentional about partnering with community oncologists. Except at first decentralized trials are likely to be limited to MOAs that have well-understood safety profiles so there's more confidence a patient won't be administered the drug and need specialized care because of a side effect. CAR T cell therapies, for example, are crazy effective but couldn't be administered in a community setting and even few big med centers can admin them because they require specialized storage/manufacturing/training.

"and his first point about onc docs not being super hepful with clinical trials is very true but also being addressed - my colleague Brittany was working at ASCO prior to working at FOCR and she was helping develop a "button" that could be incorporated in the EHR that would help to identify clinical trials based on the patients history.”

This sounds great! Except for the obvious limitations that the EHR would be identifying trials within that specific hospital system, as there’s no national database to connect to. I think the Mayo Clinic in Rochester and Cedars Sinai in L.A. have systems to facilitate easy search within the hospital system.

Here’s my fantasy, if anyone wishes to build it: Have a national clinical trial database that EMRs can connect to, in the style of Epic’s Care Everywhere. A patient signs a release to share their information, like Care Everywhere allows hospital A to download a patient’s information from hospital B if the patient signs a release. Trials with open slots could mark themselves as “recruiting” and update in real time. A doc could press a button and identify potential open studies. A patient available for a trial could flag their EMR profile as actively searching, allowing clinical trial sites to browse patients in their region who are looking for a study. Access to that patient’s EMR would allow them to scan for eligibility. This would be pretty easy to do, computationally. OKCupid can do it. The tech exists.

Finding a clinical trial already feels a bit like online dating, except if you get the wrong match, you die.

I heard from a number of patients and their families, many who worked in the research and healthcare sphere, and whose trial stories were similar to Jake’s and mine, despite them being insiders.

K.T. wrote:

I'm a molecular biologist in a field directly related to cancer research. My both my best friend and my cousin died of CIC::DUX4, non-Ewing's sarcoma this year (in what is a freakish coincidence, like winning the worst lottery I can imagine).

I cried reading part 1 and 2, remembering the chaotic way I followed the same path as you, thinking naively that it must be a straightforward process and being absolutely beaten down by the reality that the system is impossible to understand, even with expertise like ours.

I can’t forget a story sent via email from a patient named Dakota (she said I could use her first name), who shared a cautionary tale about strict eligibility criteria, hospital error, and the need for constant vigilance. Are patients always ultimately responsible if there’s an error in protocol that leads to disqualification?

I'm twenty eight years old and I was diagnosed with Ph+ acute lymphoblastic leukemia (ALL) at the beginning of 2022. I was initially consented to be part of a study for newly diagnosed Ph+ ALL patients to receive blincyto with a TKI as the first line of treatment instead of chemo with a TKI which is the standard of care. Chemo regimens for acute leukemias are very aggressive and Blincyto, a BiTE therapy, carries with it far fewer dangers and side effects compared to the chemo, and had already been shown to blow other types of treatment out of the water when used as a secondary line of treatment in cases of relapsed ALL.

When I was first admitted to the hospital, I was given hydroxyurea to get my white blood cell count under control until they could determine a long-term plan. I was consented for the Blincyto study on day five of hydroxyurea. A team of doctors came into my room the next day, several hours after I had received the sixth dose, to tell me that I had been disqualified from the study because of an exclusion criteria that said participants could not have received more than five days of hydroxyurea, which had been overlooked by the floor attending. There were no possible exceptions. 

Instead of Blincyto, I received 8 modules of the chemo regimen Hyper CVAD, which gave me neutropenic fever leading to sepsis on six out of eight rounds, resulting in three ICU stays.

Having to go through eight rounds of Hyper CVAD chemo because of an error that was no fault of her own, after being consented is horrible. Are there any protections in place for a patient when something like this occurs? Or are they out of luck— a data point that has to be discarded. What about an error that disqualifies a patient mid-study? These are real questions: comment or email if you know.

Too many medical malpractice lawsuits are filed, but Dakota’s story cries out for one.

Richard Chin’s comment relates to the clinical trial question:

Clinical trials are by their nature dangerous and experimental. And the FDA allow companies to conduct trials under very circumscribed conditions, in order to protect patient safety and rights. So the companies are not allowed to stray from what was agreed. Good protocolmanship would write into the protocol enough leeway so that edge cases can be determined by the clinician, and if that leeway is written in, then there can be discretion for the clinician. Alternatively, leeway or procedures for exceptions to be granted by the medical director at the company can be written in.

In Dakota’s case, the circumscribed conditions of the study didn’t protect her or her rights, and her only option was to undergo a treatment protocol that was significantly more dangerous than the study.

I’m an ER doc and I ask patients in trials if any meds will interfere. But this is an uncommon practice, and patients may be unconscious or unable to tell me. A “break the glass” warning in an electronic medical record would be helpful, along with an alert.

So many unnecessary warnings and popups in EMRs, but this would be useful for patients at their home hospitals. A medical alert bracelet should be given to patients in a trial if certain medications would deem them ineligible. Anything that keeps me from disqualifying a patient on accident.

I’ve also had a lot of responses from AI companies, but my feelings remain the same: The product works or it doesn’t. If it works, nothing I say matters. If it doesn’t work, nothing they say matters. Right now, it doesn’t work. I would love for it to work. I hope they get there.

Many people know that the clinical trial system is far from optimal and could be improved through relatively modest changes. I’d love to connect people here who have compatible skill sets to tackle the problem. E-mail me at DrBStillman@gmail.com

If you’ve gotten this far, consider the Go Fund Me that’s funding my husband Jake’s ongoing cancer treatment.

---

---