Skip to content

Multiple Myeloma Trial Design Patterns

Definition

Multiple myeloma (MM) is a plasma cell malignancy characterized by clonal proliferation of plasma cells secreting monoclonal immunoglobulin. MM is a chronic, relapsing disease with patients typically receiving 4–7 lines of therapy over their lifetime. Trial design is heavily stratified by: transplant eligibility (transplant-eligible [TE] vs. transplant-ineligible [TNE] newly diagnosed), line of therapy (1L NDMM, 2L+, relapsed/refractory [RRMM]), cytogenetic risk (high-risk: del17p, t(4;14), t(14;16) vs. standard-risk), and response depth (MRD negativity).

CTG Phase 3 myeloma dataset (256 trials): PFS primary 99/256 (39%), ORR/CR primary ~40/256 (16%), OS primary ~15/256 (6%). Median enrollment 307 patients. Predominantly open-label.

Standard response criteria: IMWG (International Myeloma Working Group) 2016 consensus criteria. Key categories: sCR (stringent CR — CR + normal FLC ratio + no clonal PCs by IHC/IFC), CR (no M-protein by SPEP/UPEP, <5% plasma cells on BM biopsy), VGPR (≥90% reduction in serum M-protein), PR (≥50% reduction), MR (minimal response, 25–49% reduction), SD, PD.

Regulatory Position

PFS (dominant primary for regular approval): PFS is the standard primary endpoint for myeloma regular approval across all treatment lines. The IMWG definition — "time from randomization to progressive disease per IMWG criteria or death from any cause, whichever first" — implements the composite variable strategy natively (death is a PFS component).

FDA myeloma-specific position: PFS HRs of 0.3–0.6 (typical for novel combinations) represent large, clinically meaningful effects. PFS differences of 10–20+ months in median PFS are typical for modern daratumumab-based combinations. These magnitudes support PFS as the primary endpoint without requiring OS maturity.

MRD-negative CR (emerging primary, January 2026 FDA draft): FDA issued a draft guidance (January 2026) proposing MRD-negative CR rate at ≥10⁻⁵ sensitivity as a primary endpoint for regular approval in NDMM TE settings. MRD-negative CR rate would need to be large and durable with supporting OS trend. See Emerging Endpoints in Oncology Trials.

Accelerated approval: CR rate and VGPR+ rate used for accelerated approval in RRMM when no effective standard therapy available and ORR is large and durable (e.g., belantamab mafodotin initial approval based on ORR 31%, CR rate 19%).

Status: FDA Cancer Endpoints 2018 = Final; FDA MRD/MM draft guidance = January 2026, Draft

When to Use Each Endpoint

By Setting and Treatment Line

Setting Primary Endpoint Key Secondary Notes
1L NDMM transplant-ineligible (TNE) PFS OS, ORR, MRD-neg rate MAIA (dara+Rd, PFS HR 0.56); ALCYONE (dara+VMP, HR 0.50)
1L NDMM transplant-eligible (TE) induction sCR/CR rate MRD-neg rate, PFS GRIFFIN (dara+VRd, sCR 42% vs 32%)
1L NDMM TE post-ASCT maintenance PFS OS Lenalidomide maintenance CALGB 100104; CASSINI
2L RRMM (1–3 prior lines) PFS OS, ORR POLLUX (dara+Rd, HR 0.37); CASTOR (dara+VD, HR 0.39)
2L+ RRMM heavily pretreated PFS or ORR OS KarMMa-3 (ide-cel CAR-T, PFS HR 0.49); CARTITUDE-4
RRMM accelerated approval ORR (CR/VGPR rate) DOR, OS Belantamab, selinexor initial approvals
High-risk smoldering myeloma PFS or time to progression OS Prevention design; no approved drug yet
MRD-driven adaptive therapy MRD-negative CR rate PFS, OS Emerging design per Jan 2026 FDA draft

Endpoint Hierarchy Pattern (Standard Phase 3 RRMM)

Primary:     PFS by IRC using IMWG criteria (α = 0.05) → if significant →
Secondary 1: OS (α = 0.05) — tested at pre-specified maturity threshold
Secondary 2: ORR (CR + VGPR + PR rate) → MRD-negative rate
Secondary 3: DOR (among responders)
Secondary 4: PRO (QoL, symptom burden)

Design Considerations

IMWG Response Criteria Implementation

PFS assessment in myeloma requires:

  • M-protein quantification: Serum protein electrophoresis (SPEP) + immunofixation (IFE) at each assessment; 24-hour urine protein electrophoresis (UPEP) for Bence-Jones myeloma
  • Free light chain (FLC) ratio: Required for sCR assessment and for non-secretory myeloma monitoring
  • Bone marrow biopsy: Required only for CR confirmation; not at every assessment (unlike AML)
  • Imaging: WBLDCT or PET-CT at baseline; repeat only for suspected PD (new lesions, extramedullary disease)

Assessment schedule: Every cycle during induction (q3w or q4w schedule); every 2–3 cycles during maintenance. Response confirmed by repeat assessment ≥4 weeks later for CR/VGPR+.

IRC Requirements in Myeloma

For open-label myeloma trials with PFS as primary:

  • IRC reviews M-protein data (central lab) and imaging blinded to treatment
  • Central laboratory (not local) M-protein measurement is required for IRC-assessed PFS
  • IRC is less universally required in myeloma vs. solid tumors (some FDA-approved trials used investigator assessment) but increasingly expected for Phase 3 NDA submissions

MRD Assessment Design

When MRD-negative rate is a primary or key secondary endpoint:

  • Assay: Next-generation flow cytometry (NGF, EuroFlow) at ≥10⁻⁵ sensitivity (FDA Jan 2026 draft requires ≥10⁻⁵, not 10⁻⁴)
  • Timing: At the time of best response (CR confirmation) and after consolidation/maintenance at pre-specified time points (12, 18, 24 months)
  • Bone marrow biopsy sample adequacy: ≥5 × 10⁶ total cells assessed required to achieve 10⁻⁵ sensitivity
  • Central vs. local MRD: Central assessment required for regulatory primary endpoint; local may be used for treatment decisions if central is impractical
  • Treatment discontinuation before scheduled MRD assessment: Treatment policy (patient counted as MRD-positive); hypothetical sensitivity (what rate if all patients completed therapy)

Cytogenetic Risk Stratification

High-risk cytogenetics (del17p, t(4;14), t(14;16), t(14;20), amp1q21) are mandatory stratification factors. FDA expects:

  • Pre-specified subgroup analysis in high-risk vs. standard-risk cytogenetic groups
  • Protocol to specify FISH assay thresholds for del17p (≥20% of plasma cells)
  • Sensitivity analysis demonstrating effect is not driven entirely by one cytogenetic subgroup

ASCT as Intercurrent Event (Transplant-Eligible NDMM)

In TE NDMM trials, ASCT occurs between induction and maintenance phases. If the trial randomizes at diagnosis (not post-ASCT), ASCT is an IE:

Strategy: Treatment policy — ASCT outcomes included in OS and PFS. ASCT eligibility is itself a drug benefit in induction trials. Rationale mirrors FDA AML 2022 HSCT position.

Design option: Randomize post-ASCT (maintenance trial) — avoids ASCT as IE; limits generalizability to TE population completing ASCT.

Alpha Allocation for PFS → OS Testing

Most myeloma Phase 3 trials use hierarchical testing:

  • OS tested at pre-specified maturity (e.g., 50% events = ~8–12 years follow-up for 1L NDMM)
  • At PFS primary analysis, OS is typically 10–20% mature — insufficient for formal OS testing
  • OS boundary at PFS primary analysis: very small alpha (0.0001) — O'Brien-Fleming boundary preserves most alpha for final OS analysis
  • Intermediate OS analysis (when ~30% mature): small alpha spending per O'Brien-Fleming function

Intercurrent Events

Subsequent Antimyeloma Therapy

Myeloma patients receive multiple lines of IMiDs, PIs, daratumumab, and novel agents:

  • PFS (primary): Treatment policy in most modern Phase 3 (POLLUX, CASTOR, MAIA) — progression counted regardless of subsequent therapy
  • Rationale for treatment policy (not hypothetical) in myeloma: Subsequent therapy is an expected part of myeloma management; the clinical question is "how does this regimen compare in real-world patients who will receive further lines?"
  • OS: Treatment policy universally

Historical shift: Early myeloma trials used hypothetical strategy (censored at new therapy) influenced by FDA 2007 PFS guidance. Post-2015, treatment policy has become standard following EMA guidance and FDA acceptance of both approaches.

In induction trials, patients stop study drug to undergo ASCT:

  • Strategy: Treatment policy — treatment interruption for ASCT is part of the overall treatment plan
  • Study drug may restart post-ASCT (e.g., maintenance phase of the same trial)
  • Protocol must specify treatment interruption window (typically ≤3 months) distinguishing planned ASCT interruption from treatment discontinuation

Death Without Documented Progression (IMWG-Specific)

IMWG PFS definition already includes death as a component event (composite strategy). The standard IMWG definition — PD or death, whichever first — handles this:

SAP language: "Progression-free survival is defined as the time from randomization to first documented progressive disease per IMWG 2016 criteria or death from any cause, whichever occurs first, implementing a composite variable strategy. Patients without a PFS event will be censored at their last adequate response assessment."

Missing M-Protein Assessments

When M-protein assessment is missed (patient too ill, logistical), this is a missing data problem (not IE):

  • NRI approach: treat missed assessment at a protocol-specified window as PD — most conservative
  • Alternative: censor at last adequate assessment before the missing window
  • FDA prefers the approach that avoids PFS inflation; SAP must pre-specify handling

Regulatory Precedent

Trial Drug Setting Primary Endpoint HR / Result Approval
POLLUX Daratumumab + Rd 2L+ RRMM PFS by IRC HR 0.37 Regular (2016)
CASTOR Daratumumab + VD 2L+ RRMM PFS by IRC HR 0.39 Regular (2016)
MAIA Daratumumab + Rd 1L TNE PFS HR 0.56 Regular (2019)
ALCYONE Daratumumab + VMP 1L TNE PFS by IRC HR 0.50 Regular (2018)
ELOQUENT-2 Elotuzumab + Rd 2L+ RRMM PFS HR 0.70 Regular (2015)
KarMMa-3 Ide-cel (CAR-T) 2–4 prior lines PFS HR 0.49 Regular (2023)
CARTITUDE-4 Cilta-cel (CAR-T) 1–3 prior lines PFS HR 0.26 Regular (2024)
IMROZ Isatuximab + VRd 1L TE PFS HR 0.40 Regular (2024)
GRIFFIN Daratumumab + VRd 1L TE sCR rate (Phase 2 primary) sCR 42% vs 32% Supported supplemental approval

Limitations and Pitfalls

Long OS follow-up required: For 1L NDMM, OS follow-up of 8–12 years is needed for 50% event maturity. PFS (median 3–5 years in modern regimens) drives approval; OS is confirmatory after many years.

PFS-OS disconnect risk: Several myeloma drugs with PFS benefit have not shown OS benefit in follow-up (bortezomib-based regimens in some settings). Subsequent therapy effectiveness may dilute OS signal even when PFS benefit is large.

MRD assay standardization: Multiple NGF and NGS MRD platforms with different sensitivities and definitions. FDA Jan 2026 draft requires NGF at ≥10⁻⁵ but does not mandate a specific platform. Cross-trial MRD comparisons remain unreliable until standardization is complete.

ASCT as confounding factor in 1L TE trials: If ASCT rates differ between arms (e.g., experimental arm achieves deeper response → higher ASCT eligibility), the ASCT rate difference itself may confound OS estimates. Treatment policy avoids censoring but OS interpretation must acknowledge this.

Approval in one RRMM line does not guarantee all lines: Daratumumab approval in 3L did not automatically extend to 2L or 1L — separate trials required for each line-specific approval.

Source: FDA Cancer Endpoints 2018 (Final); FDA MRD/CR in Multiple Myeloma Draft Guidance (January 2026, Draft) Status: Cancer Endpoints 2018 = Final; MRD/MM draft = Draft (January 2026) Compiled from FDA guidance + CTG Phase 3 myeloma dataset (256 trials) + estimand_framework_oncology_review.md §3.3