BACKGROUND: Bortezomib-based induction followed by high-dose melphalan (200 mg/m) and autologous stem-cell transplantation (MEL200-ASCT) and maintenance treatment with lenalidomide alone is the current standard of care for young and fit patients with newly diagnosed multiple myeloma. We aimed to evaluate the efficacy and safety of different carfilzomib-based induction and consolidation approaches with or without transplantation and of maintenance treatment with carfilzomib plus lenalidomide versus lenalidomide alone in newly diagnosed multiple myeloma. METHODS: UNITO-MM-01/FORTE was a randomised, open-label, phase 2 trial done in 42 Italian academic and community practice centres. We enrolled transplant-eligible patients with newly diagnosed multiple myeloma aged 65 years or younger with a Karnofsky Performance Status of 60% or higher. Patients were stratified according to International Staging System stage (I vs II/III) and age (<60 years vs 60-65 years) and randomly assigned (1:1:1) to KRd plus ASCT (four 28-day induction cycles with carfilzomib plus lenalidomide plus dexamethasone [KRd], melphalan at 200 mg/m and autologous stem-cell transplantation [MEL200-ASCT], followed by four 28-day KRd consolidation cycles), KRd12 (12 28-day KRd cycles), or KCd plus ASCT (four 28-day induction cycles with carfilzomib plus cyclophosphamide plus dexamethasone [KCd], MEL200-ASCT, and four 28-day KCd consolidation cycles). Carfilzomib 36 mg/m was administered intravenously on days 1, 2, 8, 9, 15, and 16; lenalidomide 25 mg administered orally on days 1-21; cyclophosphamide 300 mg/m administered orally on days 1, 8, and 15; and dexamethasone 20 mg administered orally or intravenously on days 1, 2, 8, 9, 15, 16, 22, and 23. Thereafter, patients were stratified according to induction-consolidation treatment and randomly assigned (1:1) to maintenance treatment with carfilzomib plus lenalidomide or lenalidomide alone. Carfilzomib 36 mg/m was administered intravenously on days 1-2 and 15-16 every 28 days for up to 2 years; lenalidomide 10 mg was administered orally on days 1-21 every 28 days until progression or intolerance in both groups. The primary endpoints were the proportion of patients with at least a very good partial response after induction with KRd versus KCd and progression-free survival with carfilzomib plus lenalidomide versus lenalidomide alone as maintenance treatment, both assessed in the intention-to-treat population. This trial is registered with ClinicalTrials.gov, NCT02203643. Study recruitment is complete, and all patients are in the follow-up or maintenance phases. FINDINGS: Between Feb 23, 2015, and April 5, 2017, 474 patients were randomly assigned to one of the induction-intensification-consolidation groups (158 to KRd plus ASCT, 157 to KRd12, and 159 to KCd plus ASCT). The median duration of follow-up was 50·9 months (IQR 45·7-55·3) from the first randomisation. 222 (70%) of 315 patients in the KRd group and 84 (53%) of 159 patients in the KCd group had at least a very good partial response after induction (OR 2·14, 95% CI 1·44-3·19, p=0·0002). 356 patients were randomly assigned to maintenance treatment with carfilzomib plus lenalidomide (n=178) or lenalidomide alone (n=178). The median duration of follow-up was 37·3 months (IQR 32·9-41·9) from the second randomisation. 3-year progression-free survival was 75% (95% CI 68-82) with carfilzomib plus lenalidomide versus 65% (58-72) with lenalidomide alone (hazard ratio [HR] 0·64 [95% CI 0·44-0·94], p=0·023). During induction and consolidation, the most common grade 3-4 adverse events were neutropenia (21 [13%] of 158 patients in the KRd plus ASCT group vs 15 [10%] of 156 in the KRd12 group vs 18 [11%] of 159 in the KCd plus ASCT group); dermatological toxicity (nine [6%] vs 12 [8%] vs one [1%]); and hepatic toxicity (13 [8%] vs 12 [8%] vs none). Treatment-related serious adverse events were reported in 18 (11%) of 158 patients in the KRd-ASCT group, 29 (19%) of 156 in the KRd12 group, and 17 (11%) of 159 in the KCd plus ASCT group; the most common serious adverse event was pneumonia, in seven (4%) of 158, four (3%) of 156, and five (3%) of 159 patients. Treatment-emergent deaths were reported in two (1%) of 158 patients in the KRd plus ASCT group, two (1%) of 156 in the KRd12 group, and three (2%) of 159 in the KCd plus ASCT group. During maintenance, the most common grade 3-4 adverse events were neutropenia (35 [20%] of 173 patients on carfilzomib plus lenalidomide vs 41 [23%] of 177 patients on lenalidomide alone); infections (eight [5%] vs 13 [7%]); and vascular events (12 [7%] vs one [1%]). Treatment-related serious adverse events were reported in 24 (14%) of 173 patients on carfilzomib plus lenalidomide versus 15 (8%) of 177 on lenalidomide alone; the most common serious adverse event was pneumonia, in six (3%) of 173 versus five (3%) of 177 patients. One patient died of a treatment-emergent adverse event in the carfilzomib plus lenalidomide group. INTERPRETATION: Our data show that KRd plus ASCT showed superiority in terms of improved responses compared with the other two treatment approaches and support the prospective randomised evaluation of KRd plus ASCT versus standards of care (eg, daratumumab plus bortezomib plus thalidomide plus dexamethasone plus ASCT) in transplant-eligible patients with multiple myeloma. Carfilzomib plus lenalidomide as maintenance therapy also improved progression-free survival compared with the standard-of-care lenalidomide alone. FUNDING: Amgen, Celgene/Bristol Myers Squibb. TRANSLATION: For the Italian translation of the abstract see Supplementary Materials section.