Prostate cancer has always been a strangely “silent” disease—at least on paper. We measure it with PSA, we trend it over time, and we tell ourselves that if the blood work doesn’t move, the cancer must not be moving either. Personally, I think that comforting logic is exactly where medicine sometimes trips: not because PSA is useless, but because cancer is more creative than our benchmarks.
This week’s attention to two major clinical trials adds fuel to a growing suspicion in oncology: imaging can show progression even when PSA looks stable, particularly in patients treated with newer, more potent androgen receptor inhibitors like enzalutamide. What makes this particularly fascinating is that it forces clinicians—and patients—to confront a mismatch between biology we can measure in the bloodstream and biology that’s playing out in tissues. And once you notice that mismatch, a deeper question emerges: are we truly monitoring the cancer, or are we monitoring the version of the cancer that best responds to our favorite test?
PSA isn’t the whole story
PSA has been the workhorse biomarker for decades, and it’s easy to see why. It’s inexpensive, routinely available, and it gives clinicians a quantifiable signal that can guide decisions quickly. Personally, I think PSA’s endurance in clinical practice is also psychological: it feels objective. If you can chart a number, you feel like you can control the uncertainty.
But the new analysis suggests that this “charting” can miss clinically meaningful events. In a subset of patients, radiographic progression—worsening disease visible on scans—occurred without PSA rising. From my perspective, this is less a technical nuance and more a philosophical warning sign: some aspects of cancer behavior are not obligated to follow the biomarker’s script.
What many people don’t realize is that PSA can behave like a downstream readout rather than a direct “thermometer” of every malignant change. Tumor cells can evolve in ways that reduce PSA secretion or decouple PSA expression from growth patterns. This raises a deeper question about how we interpret biomarkers: are we treating cancer signals, or are we treating the signals cancer has chosen to broadcast?
The broader implication is that reliance on PSA alone becomes riskier as therapies get stronger. Personally, I think the field often moves too quickly from “a biomarker is helpful” to “a biomarker is sufficient,” and cancer punishes that leap.
What the imaging signal is really saying
The study analyzed data from more than 2,500 patients across two large trials: one involving metastatic hormone-sensitive prostate cancer and another involving nonmetastatic castration-resistant disease. In both settings, imaging progression without PSA progression appeared more often in patients treated with enzalutamide plus androgen-deprivation therapy than in control groups.
One thing that immediately stands out is the consistency across different disease stages. When you see a pattern repeated across distinct populations, it stops looking like an isolated anomaly and starts looking like a real feature of how treatment pressure reshapes disease dynamics. In my opinion, that matters because oncology is full of “interesting” findings that never quite translate into clinical habits.
Personally, I also interpret the imaging results as a reminder that cancer is not one uniform problem. Radiographic progression can reflect changes in tumor burden, growth rate, or metastatic behavior—dimensions that PSA may not fully capture. If scans look worse, it suggests the disease is finding ways to persist or expand despite hormone pathway inhibition.
What this really suggests is that progression isn’t merely “what we can measure with blood.” It’s what the body demonstrates in anatomy, and imaging is the method we use to see that anatomy.
The illusion of stable biomarkers
Stable PSA can create a dangerous kind of optimism. Clinicians may de-escalate concern because the test looks reassuring; patients may feel they’re “in the clear” because numbers don’t rise. Personally, I think that emotional dimension is not trivial—health anxiety and hope both affect adherence, symptom reporting, and even willingness to pursue more invasive evaluations.
From my perspective, the key misunderstanding is equating “biomarker stability” with “biological stability.” Cancer under strong androgen receptor inhibition might still evolve, invade new sites, or increase in aggressiveness, while simultaneously altering PSA production. So the PSA trend becomes a partial narrative at best.
This is where editorially, I feel the conversation should shift from “PSA failed us” to “PSA has boundaries.” Boundaries are normal in medicine. The mistake is ignoring them and pretending one measurement can stand in for the whole organism.
Why this seems more common now
The study’s authors noted that radiographic progression without PSA progression is something increasingly seen in the era of more powerful androgen receptor inhibitors. Personally, I think that observation fits a broader oncology pattern: as we improve systemic therapy, we also increase the odds that tumors will adapt and “reroute” their measurable outputs.
A good way to think about it is strategy games. If you keep hitting the same coordinate on the map (PSA secretion pathways), adversaries may shift tactics. They don’t need to defeat your overall plan immediately; they just need to avoid the specific sensor you’re using. Over time, that sensor—your biomarker—can become less reliable as a lone indicator.
This raises a deeper question about how health systems design monitoring protocols. Are we updating them as treatment changes the underlying biology, or are we stubbornly keeping yesterday’s schedule because it’s familiar?
Monitoring is a balancing act
The takeaway isn’t that clinicians should abandon PSA. Instead, the study reinforces the need for a more nuanced approach that balances blood markers with imaging reality. Personally, I think this is the heart of modern oncology: “more data” isn’t the goal—“the right data at the right times” is.
Periodic imaging, the authors argue, may be necessary to ensure clinicians are evaluating the cancer rather than only chasing blood test trends. That word—periodic—matters. It acknowledges that monitoring is a continuous negotiation between benefit (catch progression early) and cost (radiation exposure, imaging burden, and downstream procedures).
If you take a step back and think about it, the real clinical challenge is not interpreting scans. It’s deciding when scans are required, and how to respond to discordant signals. Personally, I think many patients and families experience these moments as confusing: “The PSA is stable, so why are we doing more tests?” That question deserves empathy, because it’s not intuitive.
When scans look worse, confirm what you’re seeing
The article also points toward a practical principle: if imaging worsens in a way that doesn’t “fit” the expected PSA-driven disease pattern, biopsy may be critical—especially if new metastases appear in organs like the liver or lungs.
From my perspective, this recommendation is an insistence on epistemic humility. Imaging can be wrong; PSA can be misleading; treatments can distort biology; and cancer can behave unpredictably. Biopsy is the method that collapses uncertainty by obtaining direct tissue evidence.
What this implies is that discordance between PSA and imaging should not automatically trigger blind reassurance or blind panic. Instead, it should trigger careful clarification. Personally, I think that’s where the art of oncology meets the science: not just measuring, but interpreting.
Broader implications for patients and clinicians
This finding also carries a systems-level message. Personally, I think modern cancer care sometimes over-optimizes for what’s easy to measure rather than what’s clinically decisive. In practice, that can translate into protocols where bloodwork drives decisions and imaging becomes an afterthought.
But the study suggests that in certain treatment contexts—particularly with potent androgen receptor inhibitors—imaging can reveal disease progression that PSA fails to detect. That means monitoring protocols may need recalibration so that PSA and imaging are complementary rather than competing.
One trend I’m watching closely across oncology is the movement toward “multi-modal surveillance,” where labs, imaging, symptoms, and sometimes pathology all feed into a single coherent clinical judgment. This is harder than running PSA-only algorithms, but it’s also more honest.
Conclusion: the uncomfortable truth
If there’s one thing I’d want patients to remember, it’s this: stable PSA doesn’t always mean the cancer is standing still. Personally, I think the most important shift is cultural as much as clinical. We should treat PSA as a valuable tool with known limits, not as a single oracle.
And for clinicians, the provocative implication is clear: balancing the convenience of blood biomarkers with the anatomical truth of imaging isn’t optional anymore. It’s what responsible monitoring looks like when cancer stops obeying our preferred metrics.
Would you like me to tailor this article toward a patient audience (simpler language, more reassurance) or toward clinicians/investors (more policy and practice implications, tighter medical framing)?
