Escalating Bone Pain, a Poker Game He Doesn't Want to Miss, and Three Different Levers
His bone pain has climbed steadily for three weeks, and there are three real drugs that could address it, each on a different clock. The disagreement isn't which one works — it's which one actually matches the timeline the pain is setting.
Neil B., 70, has driven himself across town to the same Thursday-night poker game with three retired coworkers from the machine shop for going on fifteen years, and the one thing he has said plainly to his oncology team is that he'd rather manage his pain than lose the ability to make that drive under his own power. He has metastatic castration-resistant prostate cancer with diffuse blastic bone metastases involving the lumbar spine and pelvis, currently on a fentanyl patch with oxycodone for breakthrough pain, needing an increased patch strength for the second time in three weeks as his pain has climbed steadily rather than settled.
His three coworkers-turned-poker-buddies have started taking turns calling him the morning after each game just to make sure he got home fine, a small ritual none of them have said out loud is really about something bigger. He is not currently on a bone-modifying agent, despite bone-only progressive disease that would ordinarily have prompted one earlier in his course. Fizazi et al.'s 2011 trial found denosumab reduced the risk of a first skeletal-related event more effectively than zoledronic acid in exactly this setting, with a benefit that compounds the longer it's given — a fact that argues for starting it now rather than treating his pain as a today-only problem, even though its analgesic effect, unlike an opioid dose increase, won't be felt for weeks. What nobody has yet added to his regimen is anything targeting the mechanism separate from opioid receptors entirely: metastatic bone pain has a real prostaglandin-mediated component, sensitizing nerve endings and promoting the same osteoclast activity driving his lytic disease, and the Cochrane review most often cited for adding an NSAID here, McNicol's, was withdrawn in 2015 as out of date; its 2017 replacement by Derry and colleagues found no high-quality evidence either way. That matters for what the adjunct can honestly be offered as — a mechanistically distinct attempt to spare him opioid dose, not a step with a trial behind it. He's already more sedated than he was a month ago, and each further opioid increase makes it harder for him to keep the one thing he's asked to keep. His most recent bone scan showed no new sites beyond the lumbar spine and pelvis he was already being followed for, which at least narrows tonight's question to how to manage pain from disease that's progressing in place rather than spreading somewhere new.
In clinic, choosing which lever to pull first
Increase the fentanyl patch again. His pain has climbed twice in three weeks on the current dose, and whatever else we start today, none of it works fast enough to be the actual answer to how he feels tonight. Denosumab's benefit is real, but it isn't felt for weeks — that's not a substitute for controlling pain now.
I'd push back on treating this as only a today problem. He has progressive bone-only disease and isn't on a bone-modifying agent yet — that's the actual gap driving where this pain is heading. Fizazi et al., 2011, found denosumab more effective than zoledronic acid at delaying the first skeletal-related event in exactly his disease setting, and that benefit compounds the longer he's on it.
Escalating opioids again treats the symptom this week without touching why it's still climbing — and it's climbing because the disease process generating it hasn't been addressed at all yet.
Both of you are solving real problems on real timelines, and neither is wrong. What I'd add is a third lever nobody's used yet: bone pain has a genuine prostaglandin-mediated component. I'll be straight about the evidence, though: the McNicol Cochrane review people still cite for this was withdrawn in 2015, and Derry's 2017 replacement found no high-quality evidence for or against NSAIDs added to opioids in cancer pain. I'm proposing it on mechanism and on what it might let us not do to his alertness, not on a trial result. He's already more sedated than he was a month ago, and he's told us plainly what that's costing him — a further opioid increase moves directly against the one thing he's asked us to protect.
Start denosumab today for the trajectory, add a short trial of a renally monitored NSAID rather than increasing the patch again tonight, and reserve further opioid escalation for if the combination genuinely doesn't hold him.
Agreed: start denosumab today, add a short renally monitored trial of naproxen rather than increasing his fentanyl patch tonight, and reassess his pain level and functional status in one week.
Not agreed: the pain medicine specialist remained uncomfortable holding the opioid increase given how consistently his pain has climbed on the current regimen these past three weeks, and wanted a lower threshold to escalate if the NSAID trial didn't show benefit within a few days rather than the full week. The geriatrician felt a full week was needed to fairly judge the NSAID's effect and worried a shorter trial would default back to opioid escalation before the adjunct had a real chance. Neither position was overruled; the one-week reassessment stood, with an explicit note to revisit sooner if his pain became unmanageable in the interim.