Metastatic Melanoma with Elevated LDH: Combination Immunotherapy Against Its Own Toxicity
A single patient, newly staged with widely metastatic melanoma and a lab value that reclassifies which trial subgroup he actually belongs to. The disagreement isn't about whether combination immunotherapy works — it's about whether his own life, away from reliable emergency care most weekends, can safely absorb what that combination costs.
D.K., a 52-year-old man, has run in every local half-marathon for the past nine years and was three weeks into training for his tenth when a mole on his upper back that his wife had been asking him to get checked for over a year finally got its dermatology appointment. Biopsy confirmed a 3.4mm-thick superficial spreading melanoma with ulceration; staging CT and brain MRI six weeks later showed more than a dozen bilateral pulmonary nodules and two asymptomatic subcentimeter cerebral metastases, neither with surrounding edema. He has no other real medical history beyond well-controlled hypertension on lisinopril, and has never needed a medication he had to think twice about taking.
His serum LDH, drawn at diagnosis, came back at 412 U/L against an upper limit of 245 — nearly double normal, and a number that by itself moves him into the subgroup CheckMate 067 found benefited most from adding ipilimumab to nivolumab rather than using anti-PD-1 alone. In that trial's own elevated-LDH stratum, the five-year survival gap between the combination and nivolumab monotherapy ran wide open; among patients with normal LDH, the same gap narrowed to something close to negligible. His disease burden compounds the same argument rather than complicating it — more than a dozen pulmonary lesions plus two brain metastases, however small and asymptomatic today, is exactly the volume the combination's added toxicity is supposed to be worth paying for.
That price is not abstract. Grade 3-4 immune-related adverse events occurred in roughly three of every five patients who received ipilimumab-nivolumab in CheckMate 067, against about one in five on nivolumab alone — colitis and hepatitis severe enough, in a real minority of cases, to require high-dose steroids or a biologic immunosuppressant and to end any further systemic therapy altogether. Median time to onset of a severe irAE in that trial clustered in the first eight to twelve weeks, the same window his coaching season runs through without a break. He spends most weekends several hours outside cell range, coaching a youth cross-country team from a folding chair on wooded trails, a detail the team keeps returning to when they talk through what a grade 3 diarrhea flare two hours from a hospital would actually mean for him.
His brain metastases do not cut both ways. The ABC trial (Long et al.) randomized exactly his situation — active, asymptomatic, untreated melanoma brain metastases — to the combination or to nivolumab alone, and found intracranial response in 46 percent against 20 percent, a gap its investigators read as reason to avoid single-agent anti-PD-1 here wherever possible. His two lesions are a second argument for the combination, not a hedge against it.
In clinic, weighing the combination against its own cost
Ipilimumab-nivolumab, not nivolumab alone. His LDH is nearly double the upper limit of normal, and CheckMate 067's own elevated-LDH subgroup is exactly where the combination's five-year survival curve pulls furthest away from monotherapy — in the normal-LDH majority, that gap narrows to something close to nothing. He isn't a generic stage IV patient we're extrapolating a population average onto; his labs put him specifically inside the subgroup the combination was shown to help.
If his LDH were normal and his disease burden were three or four small pulmonary nodules, I'd be making the opposite argument — nivolumab alone, save the toxicity. This is entirely about where his own numbers land, not a default preference for the combination.
I'm not disputing the subgroup data. I'm disputing whether the toxicity that buys it is survivable for this particular patient's life. Grade 3-4 irAEs ran in roughly three of five patients on the combination in CheckMate 067, against one in five on nivolumab alone — colitis severe enough to need high-dose steroids or infliximab, sometimes ending systemic therapy entirely. He spends most weekends without cell signal. A grade 3 diarrhea flare that starts Saturday morning on a trail isn't a statistic at that point, it's a real delay to treatment that the trial's own safety data never had to model.
And I'll concede the brain metastases don't help my case — ABC randomized this exact population and found intracranial response in 46 percent on the combination against 20 percent on nivolumab alone, so monotherapy is the weaker intracranial option, not an equivalent one. What I'm left arguing is narrower, and I think still right: the marginal survival gain from adding ipilimumab has to be weighed against a toxicity risk that, for him specifically, isn't equally distributed across a calendar week.
I don't think this is really combination-versus-monotherapy as an abstract choice. It's whether we build a plan around who he actually is. The LDH argument is real and I'm not overruling it — but neither is the fact that he's hours from an ER most Saturdays with a team of teenagers depending on him to be functional.
So: give him the combination the subgroup data support, but don't send him home with a prescription pad and a follow-up date. He carries an emergency oral steroid taper starting today, with an explicit written threshold for when to start it himself and call in rather than wait for Monday, and his coaching schedule gets a two-week buffer built around each infusion where someone else covers if anything develops. That's not a compromise on the pharmacology. It's the part CheckMate 067 was never going to tell us how to do.
Agreed within the visit: start ipilimumab-nivolumab per the elevated-LDH subgroup argument, with a patient-held emergency steroid taper and an explicit two-week post-infusion buffer built into his coaching schedule.