decisions – getting closer and UCSF second opinion

How much uncertainty is tolerable when making a life and death decision? And what are the forcing-functions on that decision? I really like what Jeff Bezos says in one of his annual letters about one-way versus two way doors: 2-way doors ought to be made with a bias towards speed over perfection (you can always go back or iterate the decision). One-way doors in contrast are non-reversible and ought to be made with great care and deliberation. This second-line decision is clearly a one-way door…and I’m getting kicked through it from behind!

Schematic of decisioning second line treatment (pardon my quick scrawling). Y-axis is relevant decision info vs. time. There is no perfect information and much that will never be known. The majority of the relevant information must be gathered within 2 weeks (big shaded area), with incremental utility coming after that (small shaded area). Not shown: there is risk to dragging a decision out further than week 2 as risk of cancer progression.

As with any decision there is a key distinction to be made between the decision itself and the outcome: sound decisions can turn out badly, poor decisions sometimes result in success. But lucky outcomes over time tend to favor the better decisioning process, while any single decision may not. So the mental framework I’ve adopted here is to saturate my information gathering space such that I’m in the best possible place to make an informed decision under a hard constraint of time (see schematic). In this case that’s roughly 3 weeks from the biopsy (affording for context-specifics like exploring any clinical trial options).

Thus an important question I’m asking myself is which option minimizes regret in the case of a poor outcome. For example, if I did a trial for CAR-T and it failed (though on paper the odds might look better) I would likely have regret for not having tried a standard of care treatment with a reasonably decent chance of success first (see table below). That’s a strong argument against CAR-T as a 2nd line option. I’m about 85%+ convinced that going the standard 2nd line treatment is the way to go. I could see reversing this viewpoint if either I get new information from my conversation with Stanford specific to 2nd line outcomes, AND/OR if there’s some ‘gotcha’ about 3rd line CAR-T options that changes the landscape (say unavailability of the drug which I’ve heard can happen, payment issues, outcome changes given that 2nd line failed, etc). Before I start treatment I want to ensure I fully understand how exactly CAR-T would be an option (see questions below, e.g., where would it happen, which manufacturer, etc.) for me, should I need to move onto 3rd line treatment.

For all the talk about valuing independent thinking in our society – and especially in silicon valley – it’s a challenge, and perhaps of questionable merit when it comes to making complex medical decisions. Going against standard treatment options or recommendations is a form of mild heresy. But as situations become increasingly desperate and odds of success dwindle there is probably some threshold whereby it becomes sensible to start thinking more laterally, more independently. I’m not saying this applies to me at this point but it is something I think about as a person that likes to reason for himself. It’s worth asking questions and ruling them out for ones-self versus not considering them in the first place. For example, I’ve been pushing to start treatment IMMEDIATELY. But if I think more holistically I can ask the question: what if I sit on my hands and do nothing for 3-6 months here? Part of the motivation might be to see if yet more treatments come online as there’s much in the current FDA pipeline. No medical professional would recommend it. It’s like the bioinformatics joke my friend Mike tells about protein folding problems… Protein folding problems are hard under any circumstance. The PhD students in his group would sometimes ask whether they would solve the problem more quickly by hitting their head against the wall NOW, versus putting the problem aside, playing hookie for 6 months, then coming back to it then. Counter to intuition, and assuming Moore’s law holds (computational power improves markedly over that period), the latter approach might sometimes yield a faster result. It’s not meant to be laugh out loud funny but it illustrates the point well.

You might be this very moment thinking that the chemo has compromised my ability to think straight for even going there. But there are documented cases – at least one – of spontaneous regression/remission in aggressive B-cell NHL; life is certainly weird. And a quick skim of the literature suggests these rates (>=10%, at least in the more indolent types) warrants further investigation. Which means we don’t know enough about it to make any kind of substantive claims one way or another…which means it’s not a topic of conversation and hence, for practical purposes doesn’t exist in the possibility space of options. Don’t worry – that’s not what I’m going to do. While it may take a bit longer my bias is to consider it and explicitly rule it out versus not thinking at all about it all. I really take this seriously and want to consider every reasonable possibility, if only for peace of mind.

I’ll do my best to summarize where things stand after taking in a LOT of information. You’ll have to bear with me if this is dense as I’m trying to synthesize a ton of information and not sugar coat anything. To paraphrase (French mathematician) Blaise Pascale “I’d have written you a shorter letter if I had more time”.

UCSF second opinion:

Tuesday I met with a UCSF oncologist. I wasn’t expecting to get much from this apart from validation about what (I think) I understand. I’m paying out of pocket for all these other opinions. And it’s money well spent. The goal here is to maximize the useful information I get toward making a decision within a very contracted timeframe. This was supposed to be a one hour conversation, which seemed like more than long enough. We ended up meeting for nearly 2.5 hours. We didn’t start until around 4:45pm and ended at 7:15. I first met with one of the fellows. “I’ve read a lot about your case” she said. “I can’t believe you exercised through R-CHOP. I haven’t seen anybody do that!“. She then inventoried the questions I had for the doctor before bringing him in. I rattled off some of the questions I’ve been pondering. “So how did you get so medically literate?” she asked. “I probably just take more of an interest than most patients“.

First off we went through my information on EPIC. Because I took the time to read through the pathology report, have been studying this quite a lot lately, and have been having many conversations with knowledgeable folks, we were able to go VERY deep on various issues. Not only was I incredibly impressed with the level of answers (at one point we were sketching outcomes on paper), but his willingness to go there WAY beyond anything I’ve seen thus far; I expected him to kick me out after an hour but we kept on going until he was sure I was satisfied. Extremely impressive. I walked away feeling like I took a step-function jump in my understanding and comfort level in terms of proceeding with standard line therapy. I’ve now got nearly sufficient information to press ahead. After the meeting the fellow actually came in and thanked me, “Thank you for going so deep and probing so hard, I learned a TON in this conversation”. I thanked her but said that of course I was the one now better informed. UCSF must understand the placebo effect and must be banking on those non-trivial effects by boosting patient egos:). I’ll take it! Here’s a summary of the discussion and where things currently stand.


Kaiser told me the genetic study wouldn’t come back until Monday (per my last post, when you’ll recall I lost my cool). Turns out the FISH study was in the system after all! So I got my results from UCSF before I got it from Kaiser. That’s the second time in as many weeks. Ugh Kaiser! Anyway the news is good there that I don’t have any genetic mutations, all negative, which is great. That information could really only have complicated an already complex situation but I’ll take the small win where I can get it!

For Non-Hodgkin’s lymphoma, at least my type, there’s a continuum from follicular to large b-cell lymphoma (DLBCL), with the former being slower growing/less aggressive and the latter being more aggressive. Follicular is manageable whereas DLBCL is curable. I’m somewhere in the middle. For that reason I always thought that, given a choice, DLBCL is better assuming you catch it. But turns out the odds of curing it if first line fails are not so high and therefore options get limited with each subsequent treatment – basically Russian roulette. The doctor basically rebutted the argument saying effectively: would you prefer to live a normal life managing cancer (if follicular), or roll the dice on curing it and possibly not succeeding (DLBCL)? This was a VERY hard conversation. 

Standard treatment

There is still some disagreement about which protocol to use (see table below). Specifically, the recommendation from Kaiser is to use R-GDP as the chemo protocol. UCSF explicitly recommended not doing that one, opting instead for D-HAP or R-ICE due to my ‘ABC’ subtype. Note these are 3-day inpatient treatments, not outpatient which reflects that they are ratcheting up the toxicity. The second line chemo has a 50% response rate.

Triangulating various professional opinions on second-line chemo regimen. I’m considering 2 dimensions: trust vs. strength of recommendation. R-ICE scores highest across 3 opinions, though I’m still waiting on Stanford recommendations.

It’s important to understand that this is only part 1 of a 2-part success story. If it’s successful I then move to high-dose chemo and autologous transplant. The success rate of that (combined) procedure is 60%, meaning there is no recurrence (I think this is measured on a 5 year timeframe but not totally clear on that piece). Anyway, taken together this gives a 30-35% overall success rate for second-line chemo + high dose chemo and transplant. That’s really the number that I think matters. I ensure everyone is clear in their descriptions because numbers are thrown around without definitions, which is maddening. There’s a bunch of fuzzy parts to those rates (age, attitude, strength, my specific mix of cells; 3b, 3a, not sure about gender, etc), that I’ve not been able to segment out. But triangulating all these opinions the overall rate is likely somewhere in the 30-40% range, maybe 50% if I put on rose-colored goggles. Turns out that a mere 3 years ago second line treatment was the last standard care option and if you fail that (at least with a DLBCL diagnosis) the outcomes are poor. Only recently has Car-t therapy started to offer a (very expensive, $500K) third-line option. I’ve been told that should be encouraging news and take some of the pressure off. Easier to say than to actually feel.

I want to get to a place where the autogenic stem cell transplant is an option as that provides a curative outcome with long term data (30+ years). Again, it’s not the transplant that’s curative but rather the high dose chemo that precedes it (think: rebooting your computer hard drive). The transplant is a means to repopulate my stem cells after they’ve been destroyed by chemo. By the way the reason they don’t jump right into the high-dose chemo (I asked) is that it’s not likely to succeed if the second line didn’t show response. Also, they wouldn’t use radiation to zap the tumor here either (I asked), as that would be a local solution to what’s a systemic issue. Makes sense.  In case you’re curious this will be assessed with CT/PET imaging and then a system called Deauville score.

Car-T & side effects:

It seems there is not much data on follicular grade ‘3b’ for CAR-T therapy, the current somewhat questionable diagnosis based on first pathology report. The response rates for DLBCL 3rd line are about 35% (I’m hearing different numbers here so I’ll verify with Stanford). Also it turns out there are 3 drug makers of CAR-T  treatment that all differ in terms of how they are engineered and the side effect profiles. A Gilead/Kite version, called Yescarta is used by Stanford and apparently hits the cancer hard and fast, with a high initial toxicity profile. On the other hand the Novartis and Juno versions are used by UCSF and are slower acting, over 6 months I think, with lower initial toxicity (there’s something called cytokine storm which doesn’t sound pleasant that seems the main issue, along with some neural toxicity). Essentially the t-cells get infused and stick around longer. I gather both are equally effective and all hit the same target (CD19).

Side effects of chemo:

Me: So what other long term effects might I expect from the high-dose chemo and stem cell transplant, assuming I get there?

UCSF Doctor: “Are you done having kids”?

Turns out there’s a high risk of sterility from the high dose chemo. I haven’t thought deeply about this piece but something to consider. If there’s anything I’m thankful for on a daily basis it’s that I’ve already had 3 beautiful kids. While we do talk about a 4th – at least we have before this (we want 6 more Lua’s!) – I’m happy. This would be even more agonizing if I didn’t feel satisfied with that part and was contemplating treatments that result in effective castration. Of course the normal stuff you would expect is also going to be an issue: weakness, possible nausea, neutropenia, etc., but more severe than first line chemo. Sounds fun.

Nutrition/Integrative medicine:

I asked about this topic. There’s really not that the medical establishment will say about it. No strong opinions on any of it. He told me to see folks at Osher center for more detailed stuff.

Next Steps/summary:

  • I spoke with Stanford coordinator and finally things are moving w/r/t getting that tissue from kaiser and processing the pathology report. I’ll hear from them end of this week or early next week to setup an appointment to come in and review with the doctor.
  • I finally got that referral and am now getting updates from Kaiser after my mini tirade on the phone the other day. For all my complaints they do seem responsive to (very strong) patient feedback. Let’s see if it lasts…
  • As discussed above I’m close to convinced that standard second line chemo is my best next move. I think the only remaining piece is if I get some new information from Stanford regarding the possibility of a CAR-T option. Hoping this happens my early next week.
  • Validate pathology report w/stanford/Kaiser — align on diagnosis specifics and assess if that changes anything.
  • Finalize CAR-T options for 3rd line treatment. Before proceeding I want to know what this looks like in the pessimistic scenario that second line fails. I need to understand several key pieces (outlined below).
  • Decide on which specific 2nd line chemo treatment to go with.
  • Once I speak with stanford I may circle back with UCSF/Dana-Farber to validate everything, assuming any open questions.
  • Align with my oncologist at Kaiser about open questions and fill him in on the
    information from various other opinions. This will happen this evening.

Here’s a list of open questions I’ve been mulling meantime:

–diagnosis validation–
a) second pathology report @kaiser & what’s the status? Is it 3b vs. 3a?
b) Second pathology report @stanford: (1) what’s it say? (2) does anything change w/r/t 2nd line? 3rd line?
c) FISH results are negative – does that change anything regarding odds?

–Referrals & financials–
a) Is referral to UCSF an option for CAR-T (different drug specs: Novartis vs. Gilead/Kite) should we need to go that path?
b) Does Kaiser pay for CAR-T as 3rd line as standard of care or is it only supported in the context of studies ($0.5M)? Reason for asking is based on conversation with UCSF where they suggested Kaiser only would support in research context.

–2nd Line: standard treatment–
a) Is referral to UCSF an option for the autologous transplant? All else equal (are they?) this would be logistically easier.
b) Is referral to UCSF an option for CAR-T, or just Stanford?
c) which chemo? D-HAP vs. R-ICE vs. R-GDP? UCSF says R-GDP is not good for ‘ABC’ subtype as second line…
d) odds of sterility post high-dose chemo? How long to recover if it does?
e) Risks/other things to think about here?

–2nd Line option: CAR-T–
a) odds analysis
b) Scenarios:
– are outcomes worse if chemo fails THEN Car-T? What’s the benefit of CAR-T? pros/cons. Seems it’s mostly about the quality of life during treatment itself (and long term effects, e.g., sterility) more than about the outcomes?
– If go straight to CAR-T are we simply missing a ‘shot on goal’ with chemo? Supposing CAR-T failed would chemo be an option at all or is that option simply bypassed?
c) Is there any concern about actually GETTING CAR-T if 2nd line fails/as 3rd line (waiting list etc)?
d) Are there other trials that I should consider or is the one at Stanford the most relevant (follow up LLS on this piece)?
e) What do we know about my specific pathology (follicular 3b) and CAR-T?

–3rd Line Treatment–
a) where I would get this? Is a trial necessary?
b) would Kaiser refer/pay for this?
c) is there a waitlist or other ‘gotcha!’ to be aware of?,
d) which specific product would I get (kite vs. novartis),
e) validate my eligibility given the specifics of the pathology report.

–Other thoughts/peripheral questions–
a) China studies? I’ve heard they may actually be farther along than US for this?
b) Thinking about last resorts and allogeneic transplant if 3rd line fails…do siblings need to see if they’re a match? Should I start checking databases/if so when to do that?
c) What if I adopt a wait-and-see approach and sit on my hands for 3-6 months? Let’s talk about specific issues with this:
*Is it that the tumor grows quickly and the tumor burden is higher when I do treat?
*That it spreads (will this type spread too?) and does nasty things?
*might the cancer morph, say from less favorable to more favorable type?
*something else?