The near future of prostate cancer

And why I mostly avoid Twitter/X now

“Just what we need. A non-pathologist telling us the future of our own discipline. Thanks for the amazing insights of your post.”

Or, at least, words very similar. I can’t remember exactly and, interestingly, it’s been deleted now.

This was the reply by a pathologist to a post I made recently on X on the future of diagnostics and treatment in prostate cancer. To be fair, to some I was stating the bleeding obvious - that in medicine, specialties involved only in pattern recognition will be the first to fall (in a good way for patients) to AI - pathology and medical imaging being the prime candidates. (More on that below.)

But this was a guy replying to me as not only a professional colleague, but of course, as a fellow human being. The exchange just happened to be online. I’ve never met this person, let alone heard of him. But I can almost guarantee he’d never respond like that if it were in person. He’d be polite and respectful, like all of us normally are - in real life. I’d probably like him. He’d probably even like me! But the conditioning we’ve had from Twitter and now X produced a sarcastic remark from one stranger to another. In real life, this is just not how we roll. X, in other words, is not real life, but a grotesque distortion of it.

The other interpretation I took from it was, Stay in your lane. There are times when staying in your lane is entirely appropriate, sure. But not all the time. I’d argue that some of the most valuable insights and exchanges of ideas occur precisely when we do have the courage to veer out of our lane sometimes. Certainly, the cross-pollination I’ve had over the years with radiologists and pathologists who are superb in their expertise would attest to that. We even created an online training program (mripro.io) as a result.

It’s not news that social media and their diabolical algorithms (yes, media, like data, is plural and I am that irritating grammar pedant) bring out the lesser angels of our nature. It’s one of the reasons I mostly stay off X these days. Which is a shame. Because it can also be such a powerful vector for transmitting important new information and ideas.  

Another reason I steer clear is that I can’t stand its new name. But that might have less to to do with the letter X and more to it symbolising its new owner. Honestly, you couldn’t make up a better name for an evil genius than Elon Musk. But that’s not quite fair. I’ve no doubt about his genius. But I don’t think he’s evil. More, say, lacking in wisdom. I think it was the footage of him smoking a cigar in a convertible that did it... (Lacking wisdom seems a common enough trait of the genius that it’s always baffled me. If so smart, why not wise also?) But I digress.

The other reason I avoid X is that, while it can transmit useful advances in science, and in my field of prostate cancer, in particular, it seems to create a bubble for professionals in that discipline. This serves a purpose, sure. And I do enjoy learning from super-smart colleagues there. But I’m more interested in communicating direct to a much wider audience.

There will always be a necessary imbalance between experts in their field and the layperson. But this gap has been too wide in medicine for too long. I’m interested in figuring out how to close this gap, at least in my small corner of medical practice. I’m hoping that Substack might be the right place to do this.

Publishing in scientific journals is also crucial*. But it serves a different purpose altogether. It’s written by academic researchers and clinicians for academic researchers and clinicians. It’s certainly not pitched for the general public. Even clinical guidelines, which I’ve worked on, are not written for consumption by laypeople, but for clinicians. It’s well-known that science has a communication problem, to wit:

*Like any medium, most scientific literature is not worth the candle, but rather is published for less edifying reasons than advancing useful knowledge. To find the nuggets of high-level evidence gold, you have to filter a lot of dross. I’m sure I’m guilty of publishing some of the latter myself. But this is a whole other topic for another day.

So what is out there for those interested in gaining a broader understanding and context of the current state of affairs in prostate cancer? As it has been said, context is everything. Thankfully, there are resources, like the excellent information put out by pcf.org and its counterpart pcfa.org.au here in Australia. I’ve been fortunate enough to have articles featured in both (see links above).

But this platform we’re on here now has the potential to provide something different again. To engage directly with anyone interested in this field, regardless of your background knowledge, from a place that hopefully combines scientific, professional expertise with a more humanistic and honest approach. Where uncertainties about our knowledge, or our straight-out lack thereof, are not shied away from, but openly disclosed and questioned.

This is, after all, the bedrock of scientific thought. Question everything. Recognise our uncertainty about everything. (Quantum mechanics - the basis of what we call “reality” is itself still unsettled.) And be humble in the realisation of our ignorance.

This doesn’t come naturally or easily to those of us with expertise in any area. Deep knowledge of a narrow field can give you a sense of conviction. And it is true that experts or specialists do, by definition, know more than those outside our field. But it is still a long way off “the whole truth.” We should also remember that what we think of as fact changes all the time, as new and better evidence emerges.

So, back to those predictions. Which are more projections anyway, as they’re already starting to happen.

Pathology, or more specifically, histopathology (reading of tissue under a microscope) will be done by AI. Initially, while trust is developed in AI’s accuracy, it will be co-reported with humans. When it becomes obvious that AI routinely outperforms human pathologists, those with vested interests in the old ways will resist. The smarter ones will want to own the AI. Eventually, despite delays imposed by regulators, AI alone will report pathology and we will wonder how we could ever rely on just a single human brain, however god-like its abilities, to do this.

Taking a history falls into the same category. It’s pattern recognition, just not visual. So the apparent aura of mystique in making a diagnosis we doctors once enjoyed (somewhat justified in the old days before modern imaging) will continue to fade and soon be of historical interest only.

The same will occur for medical imaging. It will be read by AI. Again, it already is. But AI imaging may even be so accurate as to supplant the need for its diagnostic cousin, histopathology altogether. In other words, if AI imaging can give us the diagnosis reliably enough, we may no longer need to biopsy. (Some clinicians already dare to do this using PSMA PET to obviate a biopsy.) Arguments about whether the prostate should be biopsied via the rectum or the perineum will become esoteric, as biopsy itself becomes obsolete.

I posted that the last discipline within medicine to fall to AI would be surgery because, in addition to the pattern recognition of our diagnostic techniques, it requires doing as treatment. And surgery can be exceedingly complex “doing”. You may interject and say that robots are already doing surgery. Nope, they’re not. The robotic devices we’ve been using for around 20 years now are brilliant for allowing us to perform complex manoeuvres in very tight spaces of the body, but they’re actually master-slave devices. Which doesn’t have quite the same marketing ring to it, does it? We do the surgery via the device. It does nothing autonomously.

But there’s a twist to the future of prostate cancer here. Why should we assume that surgery, whether performed by humans or true robots, will be a part of prostate cancer treatment at all?

I suspect that it won’t, and that we will have highly specific non-surgical treatments instead. Theranostics, for example, which uses highly specific radioactive ligands that bind to and destroy only cancer cells, is yet in its infancy, and has already seen spectacular results in the metastatic setting.

Surgery, perhaps for all cancers, may one day be looked back on as barbarous - an apt term, perhaps, when considering the origin of surgery in barber-surgeons! It may be that the last remaining justifiable use of surgery is in the field of trauma. Surgery itself is a form of precisely controlled trauma, so it would make most sense to keep using it when the uncontrolled trauma of physical accidents occurs. But who knows, maybe even trauma will be treated by the rapid tissue healing therapies we see depicted in sci-fi movies.

In the meantime, as I’ve written elsewhere, focal therapy is emerging as a likely middle way for prostate cancers too risky for active surveillance, but not risky enough for radical treatment of the whole gland. But we still don’t know which cancer-killing energy source is best. Nor do we yet have long-term cancer outcomes to guide us. Many of us are generating data to answer these questions.

Remember, this is a surgeon writing here. And I love what I do. But I’d be a fool to ignore the tsunami of change coming our way. Just as in biological evolution, we adapt or we perish. And to keep ourselves sane, we get off X.