I’m following ASCO 2026 from home this year. Three kids under nine means a week in Chicago is more complicated than it used to be, and so I’m doing what a lot of community oncologists do: reading abstracts in the early morning, catching select sessions when I can, and trying to figure out what, if anything, will actually change when I walk back into clinic.
That last part is the thing I keep returning to.
What “practice-changing” actually means in a community clinic
Most data presented at a meeting like ASCO is not practice-changing in the way the word implies. Some of it is genuinely new. Some is confirmatory. And a fair portion, particularly the long-term follow-up readouts in settings where we already know what we’re doing, sits somewhere in between. We see that the survival curves are holding. We note the updated hazard ratios. We come away with a reinforced sense that the decisions we were already making are the right ones.
I want to be honest about something, though. Whether the data is new or confirmatory, most of it doesn’t change the actual conversations I have with patients. Not because the data isn’t useful, but because of how I end up using it, which is qualitatively. Does this treatment work? Yes. Is it appropriate for this kind of patient? Yes. Okay, let’s use it. That’s the framework most of us are operating in, and the new data feeds into that framework at a fairly high level. What it almost never does is get translated into something a specific patient can actually hold onto. I don’t say: given the updated five-year follow-up data, and given your specific characteristics, here is what your expected trajectory looks like. I say: this is the best option for you and here is why we feel good about it.
Those are very different conversations. The second one is the one I actually have. And patients, in my experience, are not asking whether the drug works in a trial. They are asking what they personally should expect. That question mostly goes unanswered.
The CROWN trial as a specific example
This Friday at ASCO, the seven-year CROWN update is being presented. Lorlatinib versus crizotinib in first-line ALK-positive non-small cell lung cancer. I want to walk through what happens to this data between the podium and my clinic, because I think it illustrates the problem precisely.
The hazard ratio for progression-free survival is 0.19. That is the most information-rich number in the dataset. An HR of 0.19 is not just a ranking. It is a continuous, precise estimate of the relative risk of progression at any given point in time. It tells you something about the shape of the benefit, its consistency across the follow-up period, the magnitude of the separation between arms. The intracranial HR is 0.06. Those two numbers, taken together, carry an enormous amount of clinical signal.
By the time that data gets summarized for a broad audience, it becomes landmark rates: 55% progression-free survival at seven years for lorlatinib, versus 3% for crizotinib. Those numbers are still quantitative, still meaningful, but they are a reduction. A single time point rather than the full curve. You have already lost something.
By the time it reaches my clinic, it becomes: yes, lorlatinib works, use it.
And then, for the question I actually need to answer, which is what a patient should expect from lorlatinib compared to alectinib, it becomes silence. The ALEX final OS data was presented at ESMO last year: alectinib also convincingly beats crizotinib, median OS of 81 months versus 54 months. Two drugs, two trials, the same comparator, both comprehensively answered. The question of lorlatinib versus alectinib for the specific patient in front of me has no randomized head-to-head answer, and the CROWN seven-year update does not bring us any closer to one.
So I sit with a 44-year-old woman, ALK-positive, good performance status, no CNS involvement at baseline. She wants to understand her options. She wants to know what each one means for her, not for the trial population, but for her specifically, given her age, her CNS picture, her life. I have an HR of 0.19 and seven years of follow-up data on one of those options, and what I can tell her is: both drugs are excellent, here is why I lean toward one, here is what we watch for. The precision of that hazard ratio never reaches her. It informed my confidence. It did not inform her understanding of what she is actually choosing between, and it did not give her the kind of individualized picture that genuine shared decision-making requires.
What the data is actually worth
A phase III trial like CROWN represents hundreds of millions of dollars and the better part of a decade. The seven-year follow-up alone, the ongoing data collection, the statistical work, the regulatory submissions, the ASCO presentation, is itself a substantial additional commitment on top of that. The science behind an HR of 0.19 did not come cheaply, and neither did the ALEX OS data, or the MONARCH long-term follow-up, or the extended CheckMate and KEYNOTE analyses. The investment across all of these programs is enormous.
And the downstream effect on clinical practice is largely that oncologists feel more confident in decisions they were already making. The data does not reach patients in any meaningful form. My 44-year-old patient does not get a more precise picture of her expected trajectory. She does not leave the appointment with a genuine understanding of what she is choosing between. She gets my best clinical judgment, which is careful and considered, but which is not the same as a conversation grounded in the actual precision the data is capable of providing. That gap is not good for patients, and I would argue it is not good for the return on the investment that generated the data in the first place.
The thing that would change this isn’t more data. The data is already there. What is missing is investment in the last mile: translating what is already known into something a clinician can use at the point of a decision. Specifically, a tool that takes the trial data, takes the characteristics of the patient in front of me, and returns an individualized expected benefit estimate. Not the population median. Not the headline hazard ratio. Something I can bring into the room and say, for someone like you, here is what we would expect, here is how your CNS picture factors in, here is what this means for your situation. That conversation changes things for patients in a way that another confirmatory trial update does not.
The ROI case for that investment is straightforward. The marginal cost of the translation layer is small relative to what has already been spent generating the data. But the return is asymmetric: more patients having the conversations that actually reflect the trial benefit, better uptake among the patients who would genuinely benefit most, and real shared decision-making rather than a confident nod at the end of a busy appointment. The companies that funded the CROWN trial and the ALEX trial have already done the hardest part. The investment that remains is putting that data to work at the bedside. That is a much smaller ask, and it is the one that has not yet been met.
I find that a genuinely interesting problem to sit with. If you’re working on something adjacent to it, I’d like to hear about it. Contact.
Dr. Henry Conter is a Medical Oncologist and Hematologist at William Osler Health System and the founder of Kesis & Sisters. He trained in Medical Oncology at MD Anderson Cancer Center and spent six years at Hoffmann-La Roche in progressively senior roles spanning oncology clinical development, portfolio strategy, and medical and regulatory affairs, across both national and global functions.