This is part II of “The emotional trial of clinical trials,” but it can be read independently.

The saga of switching from the petosemtamab clinical trial to the PDL1V trial is crazy upon crazy, but while the logistical nightmare hints at our personal frustration, the story didn’t express the  emotional and psychological struggles of trying to stay alive. With so much of the system feeding into the craziness, I started to wonder: is this what finally makes me lose my mind? Or what finally terminates my will to persist?

My grip on sanity in an insane medical bureaucracy came from other bastions of reason amidst the chaos, and from the aid of many others. The vast majority of people Bess and I have interacted with, including oncologists, trial nurses, admin staff, and others, have been supremely helpful. With only a small number of exceptions, we’ve overwhelmingly encountered people who’ve gone above and beyond duty to help us! I want to emphasize that, and to highlight that I’ve not lost all perspective, or a sense of gratitude. Most individuals have been fantastic, even if the system sometimes contorts those efforts into appearing like roadblocks.

The opaqueness of information causes new, unexpected problems, even as someone is trying to solve a different problem for us: many games of telephone, in which Bess or I ask something, then someone else asks an oncologist, then an oncologist asks the trial sponsor, then the sponsor gets back to the oncologist, then the oncologist tells someone at the site, then someone at the site tells us, with all the latency this implies. Sometimes we then have to retell that information to someone else at Mayo, or UCSD, or another trial site. Sure, I get why pharma companies don’t want the unwashed masses pinging them all the time, but the net result of information opacity is a lot of drag, during which small delays compound into larger, life-threatening delays. Things could be worse! There’s also an alternate world in which this is much harder than it’s been. But, more importantly, there’s also one in which it’s easier and the trial system more humane. Information is digital, so why are we constantly calling people to relay on it?

So much of my cancer treatment experience consists of “if only the doctor were in town.” I got delayed starting Keytruda in May due to oncologist absence at Mayo, without anyone really covering for him. The May 25 surgery was almost delayed into June due to a doctor vacation. In the period between Mar. 13 and Apr. 15, the missing information from MDA—the doctor was on vacation— caused tremendous struggle. I obviously don’t begrudge doctors getting vacation, and I have been told that doctors are somewhat human too.[1] But the lack of adequate, specialty-appropriate backup, has meant that anything short of an acute medical emergency that requires the ER, or pre-scheduled chemo, has to wait until the “main doc” returns. Which is to say, everything grinds to a halt, no matter how time-sensitive, for at least a couple of weeks.  

In another world, someone at MDA covering for the vacationing doc in mid-March would’ve looked at my chart in response to my message, seen that I’d consulted with the main doc a few months before to discuss the trail and said: “Yeah we have PDL1V, and we’ll get you in.” I would’ve flown out, qualified, and circumvented much of what followed. Coverage isn’t really coverage if the standard of care is reduced to the barest minimum, or simply instructions to wait. The people making these staffing decisions must have never been in a situation where two weeks might be the difference between life or death for themselves or someone they love. Maybe in some specialties, a delay isn’t optimal but won’t lead to death. In oncology, days matter when fighting an aggressive tumor.

Despite describing what my life—our lives—has been like, I know that in some ways I’m lucky relative to other people in my position. I may be exhausted and hurting, but I’m still able to travel. Without Bess, I would’ve walked the opioid road unapologetically long ago (and as still be my fate). In The Fellowship of the Ring, as the Fellowship is about to depart Rivendell, Elrond says: “You will meet many foes, some open, and some disguised; and you may find friends upon your way when you least look for it.” I’ve found many friends and many people who have helped me when I least looked for it. As I wrote above, an incredible number of doctors, nurses, administrators, and others have made sure I always had somewhere to turn while thousands of people (!), most not friends or family, have donated to the Go Fund Me (GFM) my brother set up. I’ve had many struggles, but the GFM has meant that money mostly hasn’t been so far, despite numerous flights, five-figure genetic testing bills, and all the other money struggles that debilitating illness foments. Many, many people have proffered help and support—you know who you are, and I thank you, though not here by name, and think of you.

Despite these real advantages, which I don’t want to be churlish about—Bess and I appreciate the many people who’ve helped us—there’s been too much to deal with. I’ve let emails drop when I shouldn’t have, and every spare, exhausted neuron fires its energy towards the clinical trial process. At night, when we decide we can’t do any more, I try to read and find reading too difficult and unsatisfying because my mind has been hosed by a combination of fighting cancer and working all day to survive. I’d call Netflix’s adaptation of 3 Body Problem a soporific, but exhaustion itself instead functions that way. We’re good at keeping track of information, yet despite being good, we’re paranoid: what if we misplace or forget something essential? What should we know that we don’t know we should know? And how would we know that we don’t know it?

We’re like detectives from cop novels, forever wondering if the most important datum will be one we don’t notice. We’re continuously double- and triple-checking things, to minimize the risk of missing a vital message. We can tell that we sometimes annoy doctors, nurses, and admins with our checking (and double and triple checking) into matters, but we’ve also found that no one else will do what we do. Finding out that Seagen was having meetings to update their trial rules a few days after getting clearance to do a pre-trial round of chemo, and then double checking with two different sites after the meeting, prevented me from being disqualified. If we hadn’t followed meeting dates, known that rules might change, and then double and triple checked the answer (which had changed), I might still be hunting for a trial. It might be too late. I’m still alive because UCSD had the petosemtamab trial, but a secondary reason is that we were diligent enough to establish care and make sure I could get into that trial. We want to relax, but when inattention means death, it’s hard to.

The non-design of what we call the “clinical trial system” is apparent in information transmission. So little information that could easily be exposed to the public on websites is exposed. I put quotes around “clinical trial system” because it’s not really a system—it appears to have evolved as a set of kludges, and those kludges frustrate. All of the trial information is in databases. It would not be hard to call the database from a website. “PDL1V—5 possible slots.” Someone gets a slot? Update the database to say that “Doe, John” is now in the trial. And then have trial.slots == 4 display on the website. I’m not even a programmer and could likely figure out how to make this work.[2]

A system like this is far more efficient than having prospective trial participants call and email to find out whether slots are open and when they might be open. The Internet is quite old! “Create, read, update and delete” processes are old. Instead of querying a database and showing the results on a website, Bess and I have to call trial sites to query administrative persons by voice. This is not efficient. The meta lesson may be that in many fields, the application of the simple, decades-old technologies remain un- or underapplied. One thesis driving Y Combinator and many venture capitalists is that the application of simple technologies can yield lopsided returns.

A number of “insiders” in the clinical trial system reached out to Bess after she posted her essay about our first trial-search saga, and when she posed the question of why it’s so hard to find trial slots, many admitted that drug companies don’t want people to know. Apparently, speed in successfully filling trials, and how many patients are currently in a trial, tells competitors where a company likely is on their trial timeline. This information can affect where competitors put their money or their time. They protect that information even as it seems to make their own studies harder to run and fill.  

If Bess ever moves to offering clinical-trial guidance services full time, part of what she might do is put all the information she gathers privately on a public website. Does someone tell her the PDL1V kickoff meeting is the week of Apr. 1, and first dosing for the “c” arm is Apr. 15? Put it online! VaccinateCA would be the model. She just read this paragraph and excitedly whispered, “I’ll be the clinical trial vigilante,” with a gleam in her eye usually reserved for unstoppable pet projects.

My brother read a version of essay and observes:

The lack of information about these trials seems like cruel and unusual punishment. Problems like these have been solved by the private technology industry. If Yelp and Google can figure out how to manage petabytes of data from local businesses, then surely the richest country in the history of the world can shore up its technological IP like ClinicalTrials.gov. There is no excuse for an important government website to have a rotten foundation of garbage data.

The things Bess and I most want, maybe even need, to do are getting deferred endlessly by hunting and pecking for information that should be readily available on websites, but isn’t. This seems to waste the extremely valuable time of oncologists, too—and trial managers, trial nurses, and so forth. How many people are like me, except that they give up much sooner? Or do they die while waiting, or experience too much disease progression?

Throughout writing about the clinical-trial process, between July 2023 and now, I’ve tried to give readers a sense of hope balanced by a sense of what the process is like. “I’m going to beat it,” was my big thought when Bess and I first embarked.

That was before my hands and feet hurt as much as they do now.

That was also when I thought I’d have to go through the elaborate trial-learning process once, not at least three times.

That was before I realized just how miserable flying is for someone like me, who needs to wear an n95 mask to avoid getting sick but who also needs to constantly spit, because I don’t have a tongue and my whole nasal-throat-mouth apparatus doesn’t work properly.

Just now, I had to pause writing to hack up a bunch of mucus—the same sort of mucus plug that I have to hack up dozens of times a day, and the type that has previously inspired thoughts of auto-termination because of physical misery. That was before I realized too that the attitude and culture of some of the biggest institutions is: “Do it our way, and suffer.” It seems like a lot of people don’t realize how hard it is to travel when sick, when I can’t eat without blending food, when every step hurts. The most meaningful time I have right now is spent with Bess, family, and friends—none of which happens when I’m on flights or at appointments.

Don’t get me wrong: I’m happy to do hard things that are necessarily hard. It’s the pointlessly hard things that feel intolerable. There are intrinsically hard things: writing well, startups, discovering nature’s secrets. Then there are incidentally hard things, like badly designed products or bureaucracy, and the latter are almost infinitely more frustrating in part because they are pointless, or nonsensical, or casually cruel. People respond to Kafka’s The Trial because it depicts this world of random misery. The first kind of hard thing doesn’t bother me, or I think most people, but the second is maddening because of the way it destroys human flourishing and wastes resources. Pointless waste is awful and that’s what I’ve been experiencing for the last few weeks. And there are people who defend this awful system. They say it “protects” me, from myself I guess.

These systems have evolved without patient input. Any temporary success leads to more hard. Even what should be relatively small problems ripple through the whole system. Not being able to do the things that I’m supposed to be here to do endlessly frustrates. The fight against the pain and fatigue in my body and mind are bad. All of us should be trying to enhance human progress and human flourishing, in however imperfect and cockeyed a way. The last five weeks have not been that for me.

Being alive at all right now is a surprise and blessing, I’m well aware. Petosemtamab bought five months of relatively low worry, and, from mid-December to mid-March, relatively good days. “Relatively” is doing a lot of work there, because compared to my pre-cancer universe, days were bad. But compared to being dead or utterly incapacitated, they were pretty good, and you can see from the writing I’ve done that I managed to engage in some generative activities.^[3]

I hate that these are probably the last months of my life and I’m spending them fighting bureaucracy. Relentless, crushing bureaucracy. Which studies? What am I eligible for? Which are open? Can we do telemedicine appointments? The slog feels more like a fight against the mud. “You’re eligible—just kidding!” Trying to decide what might be good. Blockages. People on vacation. Faxes. Demands that I fly places for ten-minute conversations. Moloch. Consuming all our time and attention. The truth is that what we’re doing is unlikely to work, but we’re trying to maximize the probability that it does. Meticulous record keeping. Every call, every message, when it happened, who it was with, what the person said. Example: MDA promise to do video calls. Now ignored. Demanding huge numbers of records up front. Epic partially solves this! Use it! Moving target. Never the same. Not what I want to spend what’re realistically my last months doing. Alternative is to take something locally and from there hope. This is why some oncologists say trials aren’t worth it; at first I didn’t believe them, and yet now I must concede they have a point. But I shouldn’t have to choose between frustration in the attempt to get more time, or a quick, inevitable decline. Everything takes a week or two. Endless emails, calls, everything. Medical bills strewn across my home. Being forced away from Bess, my siblings, friends. Care Everywhere exists! No one uses it! Why not? Easier ways exist.

Every time I’m like: we’re almost at the end! I’m not. There’s fractally more bullshit. Nothing is reliable. It didn’t have to be like this. We didn’t have to work ourselves ragged to get here. We didn’t have to start establishing care again. Record transfer should be easy. Finding clinical trials is like someone throwing pebbles at my head: the first isn’t so bad but by the hundredth you start to wonder if it’s worth it. The opioid road begins to look more appealing again. Balance it against the decent probability that none of this works anyway.

I see why people give up—take whatever’s closest and hope for the best. It’s like being a mortal fighting Morgoth in The Silmarillion: you can keep fighting, but probably there’s no winning condition. The end result of fighting is…more fighting (arguably this essay is part of the fight). Against nameless and faceless medical ethicists. Against whoever decides how many lines of systematic therapy a person can have. Against the FDA edifice and its preening satisfied bureaucrats.

The latency involved means we have to pursue many options at once. If you wait until you have definitive answers, it’s too late. We must pursue multiple concurrent possibilities, because we don’t know what’ll work out. This is itself a resource drain, but, since we can’t get information about what’s available, we’re stuck looping through appointments to try and understand the trial landscape.

Even trying to explain this is hard. “Why didn’t you do x?” people ask, and inevitably we have, requiring further explanation as I try to move them through the idea maze we’ve been traversing. And people within the system don’t recognize how seemingly small problems or requests cascade into new problems. I admit to my resolve wavering. I wrote to friends and family on Apr. 6:

I’ve been thinking about whether to end treatment, or scale it back to whatever is available locally in Phoenix, which is much the same thing, given that only phase 1a dose-finding trials are available. Treatment mainly produces more treatment: more flying, more struggle, more exhaustion, more needle pokes. What I choose may not matter, though, because I did manage to get new CT scans, and the results in the neck are bad. Really bad. I may have chosen the next step wrong; I probably should’ve gotten chemo and let go of Seagen’s PDL1V. I may not even be eligible for PDL1V because of the requirement that my life expectancy be a minimum of three months. 

That was a nadir of frustration, and I was wrestling with some of the challenges I wrote about in “Will things get better? Suicide and the possibility of waiting to find out.” But on April 15 I got the first dose of PDL1V.

There are lessons in this for others going through trials, and their friends and family who may be helping them. One important lesson is, I think, “You are not alone,” and that, in many cases, “You are not the problem.” The system is the problem. I’d love to get it improved.

Completing this essay feels miraculous. On top of the general exhaustion from growing tumors and poor sleep, the PDL1V’s main side effect is fatigue. I didn’t feel it on Apr. 15, when it was infused. The night of Apr. 17, though, I got home around 7:00 p.m. local time and could barely move. Fortunately, Bess had ordered some Ethiopian food and blended it for me. The next day, I did get out of bed, but not much else happened. Brain fog set in. This essay, which I’ve been intermittently working on for weeks, took on the character of climbing Mt. Denali, when in fact it’s just a guy typing at a keyboard. On Saturday, I began replying in earnest to emails that needed attention days prior, and apologizing for unreturned texts. The excuses are real but still excuses.

I’m reading Freeman Dyson’s book Disturbing the Universe (the first half is good; the second, not so much), and found this, about Dyson’s early relationship with Richard Feynman (Surely You’re Joking, Mr. Feynman! is also highly recommended):

In that little [hotel] room, with the rain drumming on the dirty window panes, we talked the night through. Dick talked of his dead wife, of the joy he had had in nursing her and making her last days tolerable, of the tricks they had played together on the Los Alamos security people, of her jokes and her courage. He talked of death with an easy familiarity which can come only to one who has lived with spirit unbroken through the worst that death can do. (59)

“With spirit unbroken:” that is my hope for those who exceed me.

If you’ve gotten this far, consider the Go Fund Me that’s funding ongoing care. In addition, this essay is really for everyone who has encouraged me to keep going, and for everyone who is wrangling, fighting, and struggling with the clinical trial system: you are not alone, and, as I wrote in the body of the essay, the problem is probably not you. If you are looking for more, see also “You Do Not Own This.”

---

[1] Verification pending.

[2] As my brother observes, two-way APIs have been a thing for decades. The problems I’m writing about are policy problems. The base data is not accurate, and there are no penalties for non-compliance, and no apparent rules around the updating of information for potential patients in clinical trials. Clinicaltrials.gov is surely a product of the same disaster artists behind the original healthcare.gov.

In software, we refer to the backend of an application as the “infrastructure.” The U.S.’s physical infrastructure is crumbling; the same is unfortunately true for its software. We should prioritize solving digital challenges by asking for help from the brilliant minds behind user-centric software. This is a classic problem in data science.

[3] I also edited Bess’s work, and I sometimes wrote down funny, interesting things she’d say, and then give her a potential outline for an essay. Bess is an amazing writer who often struggles to get started or to see the way forward from a blank page. She’s getting better at it than she used to be, but external boosts still help her.