Outcomes
We analyzed overall survival, progression-free survival (PFS), response rate, treatment-related mortality (TRM), quality of life (QoL), grade 3 and 4 adverse events, and incidences of acute and chronic graft-versus-host disease (aGVHD and cGVHD, respectively).
Risk of bias: We assessed the risk of bias (RoB) in the included studies using Cochrane's RoB 2 tool for RCTs and the Cochrane Childhood Cancer Group's assessment criteria tool for observational studies. We rated the certainty of evidence using the GRADE approach for all previously stated outcomes.
Synthesis methods: We followed standard Cochrane methodology, and two independent review authors performed all steps regarding study selection and risk of bias assessment in duplicate.
Included studies: We included one RCT of alloSCT as first-line treatment for MTCL (103 participants), and two prospectively planned trials of alloSCT as salvage treatment (119 participants): one Mendelian-randomized controlled trial for cutaneous T-cell lymphomas (CTCL) (102 participants) and one prospective observational trial for non-cutaneous MTCL (17 participants). We identified three ongoing trials. All of them are non-randomized, phase-II studies. They will mostly include relapsed or refractory lymphomas. Two other studies are awaiting classification. Overall, we rated both included RCTs to have some concerns of bias due to bias arising from the randomization process, and missing outcome data. The RoB for the prospective observational study was rated serious due to the study design, small number of participants, and reporting and detection bias.
Synthesis of results: As we did not identify enough RCTs, we did not conduct a meta-analysis. We rated the evidence for all outcomes to be of low to very low certainty, due to the very small number of events and limitations in study design. AlloSCT compared to autoSCT for adults with MTCL (first-line) We included one RCT with 103 participants. AlloSCT compared to autoSCT may result in little to no difference in two-year overall survival (hazard ratio (HR) 1.3 (95% confidence interval (CI) 0.7 to 2.4; low-certainty evidence) or PFS HR 0.9 (95% CI 0.5 to 1.5; low-certainty evidence). AlloSCT compared to autoSCT may result in a large increase in TRM (RR 18.70, 95% CI 1.11 to 315.72; low-certainty evidence), but may have no effect on overall response rate (RR 1.03, 95% CI 0.73 to 1.45; low-certainty evidence). AlloSCT compared to autoSCT may result in little to no difference in the number of adverse events (RR 1.10, 95% CI 0.69 to 1.75). 84% of all participants receiving alloSCT had aGVHD and 47% had cGVHD (very low-certainty evidence). AlloSCT for adults with non-cutaneous MTCL (relapsed or refractory setting) We included one phase-II study with 17 participants. Overall survival and PFS at two years were presented graphically in the included study, and were approximately 80% and 64%, respectively (very low-certainty evidence for both). The overall response rate was 76% after one month (70% in complete remission and 6% in partial remission; very low-certainty evidence). The two-year TRM rate was presented graphically in the included study, without making reference to absolute numbers, and was approximately 6% (very low-certainty evidence). 35% of all participants had aGVHD and 41% had cGVHD (very low-certainty evidence). Adverse events were not reported. AlloSCT compared to other treatment for adults with CTCL (relapsed or refractory setting) We included one RCT with 102 participants. AlloSCT compared to any other treatment may make little or no difference to overall survival (HR 0.54, 95% CI 0.23 to 1.29; very low-certainty evidence), but may increase PFS (HR 0.38, 95% CI 0.21 to 0.69; very low-certainty evidence). AlloSCT may make little or no difference to the number of adverse events (RR 1.16, 95% CI 0.90 to 1.50; very low-certainty evidence). The one-year cumulative incidence of cGVHD was 31.4% (very low-certainty evidence). The overall response rate and TRM were not reported.
Authors' conclusions: In this systematic review, we identified one RCT of alloSCT as first-line treatment for MTCL, and one Mendelian RCT and one prospectively planned phase-II trial of alloSCT for relapsed or refractory MTCL. We encountered difficulties while extracting data from these studies, because there were major differences in study designs, intervention and control groups. We identified low- to very low-certainty evidence on the effectiveness and safety of alloSCT for adults with MTCL. We had some concerns about the risk of bias in the two RCTs and serious concerns about bias in the phase-II study. As there are three ongoing studies evaluating alloSCT, mostly for relapsed or refractory MTCL, it is possible that an update of this review will be published in the near future. However, none of the ongoing studies are of an RCT design.
