A Qualifying PD-L1 Score, and a Liver That May Not Wait For It
A single patient whose PD-L1 score cleanly qualifies him for immunotherapy monotherapy. The disagreement is whether a liver trending toward organ-threatening tumor burden can afford to wait for a slower-onset response.
R.F., 70, spent 32 years walking the same postal route rain or shine and still walks his neighborhood daily, “out of habit more than exercise now.” He presented six weeks ago with unintentional 18-pound weight loss and right upper quadrant discomfort his primary care physician initially attributed to gallstones seen incidentally on ultrasound — until a CT ordered to look more closely at the gallbladder found something else entirely: a 5.2cm right lower lobe mass and multiple hepatic lesions, the largest 4cm, biopsy-confirmed as metastatic lung adenocarcinoma. Molecular testing returned negative for EGFR, ALK, ROS1, BRAF, MET, RET, and KRAS G12C — no actionable driver, which by itself settles this toward immunotherapy-based treatment rather than a targeted agent. His PD-L1 tumor proportion score came back at 55%, comfortably above the 50% cutoff KEYNOTE-024 used to enroll — the trial that made single-agent pembrolizumab standard first-line therapy for exactly this biomarker profile.
What the trial's own population looked less like is what showed up on his labs three days ago: his alkaline phosphatase and bilirubin have both risen meaningfully over the past two weeks, tracking the growth of his liver metastases and raising a real, if not yet certain, possibility that his liver is beginning to fail under the disease's own weight rather than merely hosting it. A patient whose hepatic metastases are threatening organ function before treatment has even started isn't the patient KEYNOTE-024 was built to represent, and its 44.8% response rate is also a 55% chance that the first thing he is given does nothing while his bilirubin climbs. His wife, who has driven him to every appointment since his diagnosis, has been pressing hardest for whichever option works fastest — a preference the team takes seriously without letting it substitute for the judgment call. He qualifies for the monotherapy trial on one of his numbers and belongs to the liver-metastasis subgroup of a different trial on another, and the two memberships point opposite ways. The tiebreaker is not which number is more flattering but which is more securely earned: the 55% is a prespecified entry criterion he meets exactly, while the liver-metastasis survival figure comes from an exploratory subgroup that was never powered to guide a decision like his.
A qualifying PD-L1 score, and a liver that may not have time to wait for it
Pembrolizumab monotherapy. KEYNOTE-024 is what made single-agent pembrolizumab first-line standard at a tumor proportion score of 50% or above, and at 55% he qualifies on the trial's own criterion rather than by extrapolation. He's already lost significant weight; avoiding added chemotherapy toxicity in a 70-year-old who qualifies for the simpler regimen is the guideline-concordant choice.
KEYNOTE-024's response rate was 44.8%, which also means a majority of patients on monotherapy did not respond, and immunotherapy responses characteristically take longer to manifest than chemotherapy's more immediate cytoreductive effect. Look instead at KEYNOTE-189's exploratory analysis of patients with liver metastases: median overall survival roughly doubled with the chemotherapy combination, 12.6 months against 6.6. That is the subgroup he is actually in.
You're right that his PD-L1 score is the cleaner guideline match on paper — but a biomarker qualifying him for monotherapy in general doesn't settle whether monotherapy is fast enough for him specifically, and I'd add that the liver-metastasis figure I just quoted is exploratory, so I'm arguing direction, not magnitude.
Before committing to either extreme, get a clearer read on how urgent this actually is — repeat liver function tests and a hepatology-informed read on trajectory over the next 48 hours. You're both reasoning from one draw. A mild, non-accelerating rise still leaves real room for monotherapy's toxicity advantage in an older, already-underweight patient; a doubling bilirubin over two days settles the question toward combination on its own, on firmer ground than a KEYNOTE-189 subgroup that was never powered to answer it.
Agreed: repeat hepatic function panel and hepatology consult obtained within 48 hours; chemo-immunotherapy combination started empirically now given the trend already visible, rather than waiting the full 48 hours to begin treatment.
Not agreed: whether monotherapy should be reconsidered as a switch if the repeat labs show the trajectory has genuinely stabilized. The first oncologist wants that option explicitly kept open; the second is skeptical a single stabilized data point would justify switching a working regimen.