OPTIMUM at five years: a large separation, against a control that was not randomised

Design

Multicentre, externally controlled, Bayesian phase 2 trial with a molecularly matched control cohort from Myeloma XI

Population

107 analysed patients with newly diagnosed high-risk multiple myeloma or plasma cell leukaemia at 22 UK centres; 120 matched external controls

Primary outcome

18-month progression-free survival (reported previously); this analysis reports 5-year progression-free and overall survival

Effect

Progression-free survival not reached vs 24.4 months (HR 0.32, 95% CI 0.22 to 0.45); overall survival not reached vs 57.4 months (HR 0.43, 0.28 to 0.65)

OPTIMUM enrolled 108 patients at 22 UK centres with newly diagnosed high-risk myeloma — defined by two or more high-risk cytogenetic abnormalities, high-risk gene expression profiling, or plasma cell leukaemia — and treated them with daratumumab, cyclophosphamide, bortezomib, lenalidomide and dexamethasone before and after autologous transplant, melphalan-bortezomib conditioning, two parts of extended consolidation, then daratumumab-lenalidomide maintenance until progression. Activity was compared with 120 genetically matched patients from Myeloma XI. At a median follow-up of 71.1 months, progression-free survival was not reached against 24.4 months (HR 0.32, 95% CI 0.22 to 0.45), overall survival not reached against 57.4 months (HR 0.43, 0.28 to 0.65), and progression-free survival 2 not reached against 43.9 months (HR 0.26, 0.17 to 0.41).

The size of the separation is not in doubt; what it is attributable to partly is. This is a phase 2 trial with an external control, so the comparison carries every problem of a historical comparison — Myeloma XI patients were followed for a median of 117.8 months, meaning they were treated in an earlier era with different salvage available at relapse. Matching was molecular, not temporal. The honest description is that risk-stratified intensified therapy is associated with a large and sustained survival advantage over a matched historical cohort, and that no randomised trial has yet delivered the same answer.

Two things follow for practice. First, the result depends entirely on knowing who is high-risk, which means cytogenetics and, ideally, gene expression profiling at diagnosis — the diagnostic step is the part most units can act on now. Second, the benefit was consistent across subgroups except in patients with three or more high-risk cytogenetic abnormalities, where it was not. That is the group most likely to be offered intensification and least likely, on these data, to gain from it.

• Do cytogenetics at diagnosis in every newly diagnosed myeloma — the whole strategy rests on identifying high risk.
• Describe this to a patient as an externally controlled trial, not as a randomised comparison.
• Note that patients with three or more high-risk abnormalities did not show the same benefit.
• Extended consolidation and daratumumab-lenalidomide maintenance until progression is a long, costly commitment — check funding and access before proposing it.
• Where gene expression profiling is unavailable, cytogenetic risk alone still identifies most of this population.

The statistics, in plain English

A hazard ratio of 0.32 with a tight interval of 0.22 to 0.45 would be decisive in a randomised trial. Here the control arm came from a different study, so the interval expresses only sampling uncertainty — it says nothing about the bias introduced by comparing across eras, and that bias is not quantifiable from these data. The clue to its size is the follow-up: 71 months in OPTIMUM against 118 months in Myeloma XI. The mismatch in subgroups is also worth noting: a benefit that holds everywhere except in the highest-risk cytogenetic group is the pattern you would expect if some of the advantage came from the comparison rather than the treatment, though small numbers in that subgroup make it equally compatible with chance.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.