A Failing Lille Score: Stopping Steroids Without a Plan Behind It
A patient's day-7 Lille score confirms she isn't responding to corticosteroid therapy for severe alcohol-associated hepatitis. Stopping the drug isn't controversial. What happens in the same conversation — whether to open an early transplant evaluation that bypasses the traditional sobriety clock — is.
Priya A., a 41-year-old woman, has two teenagers at home and worked as a dental hygienist until this admission, her first for anything liver-related despite roughly a decade of escalating alcohol use that accelerated after a divorce three years ago. She was started on prednisolone six days ago for severe alcohol-associated hepatitis, with a Maddrey discriminant function of 44 and MELD of 29 at admission. Her bilirubin, which should be trending down by day 7 in a responder, has instead climbed from 16.8 to 19.4, and her calculated Lille score at day 7 comes back at 0.58 — above the 0.45 threshold that defines non-response and predicts roughly a 75 percent six-month mortality risk if nothing else changes.
That number is what actually reframes the conversation. A Lille score above threshold isn't a soft signal open to interpretation the way an early trend might be — it's the validated stopping rule Louvet and colleagues derived specifically to identify patients for whom continued steroids carry the infection risk of immunosuppression without the corresponding chance of benefit. What it doesn't answer, and what the team disagrees about, is what comes next: whether a mortality risk this high, three years into a decade-long pattern that started with one clearly identifiable life event, is severe enough to justify evaluating her for transplant now, well short of the six months of documented sobriety programs traditionally require.
Stopping steroids is not the hard part of this conversation
Stopping prednisolone today isn't in question for me — a Lille score of 0.58 is a validated futility signal, and continuing a drug that isn't working just adds infection risk on top of a bilirubin that's already climbing. Where I want us to be careful is not letting 'stop the drug' quietly become the whole plan. A 75 percent six-month mortality risk at her Lille score is not a number we get to sit with passively.
I'd open the evaluation today, not after a waiting period. The traditional six-month sobriety rule was built around older, more uniform outcome assumptions; structured early-liver-transplant protocols in steroid non-responders — the shape her case fits exactly — beginning with Mathurin's matched-cohort series and since tested prospectively in QuickTrans, have shown real post-transplant survival and, importantly, relapse rates that don't look categorically different from patients who cleared the traditional waiting period. Her six days of sobriety this admission isn't disqualifying under those protocols; it's simply the starting point of the assessment, not a pass/fail gate on its own.
Waiting to open the evaluation until she's demonstrated more sobriety time isn't caution, it's applying the older rule by default while telling ourselves we're keeping an open mind — the whole point of the structured protocol is that the assessment happens now, in parallel with her acute course, not after it.
I'm not objecting to early evaluation on principle. I'm objecting to doing it badly under time pressure. A real psychosocial assessment — her actual relapse risk, whether her sister's support is as solid as it sounds today, what happens to her kids during a transplant hospitalization and recovery — takes more than the two or three days a rapidly climbing MELD score leaves us. If the evaluation gets compressed to keep pace with her bilirubin, we're not doing the structured protocol Transplant Hepatology just described — we're doing a rushed version of it that predicts worse, not better, who actually does well afterward.
Prednisolone stopped. Transplant evaluation opened the same day, run in parallel with, not gated behind, the psychosocial assessment.
Not agreed: how much time the psychosocial assessment genuinely needs before a listing decision is made, given her climbing MELD score. The addiction medicine specialist asked directly that the timeline not be treated as fixed by the liver numbers alone, and that request was documented rather than resolved.