Diffuse large B-cell lymphoma, we’ve recognised over the past several years, is not a single entity but is comprised of multiple subtypes originally described as germinal centre and activated B-cell subtype. More recent advanced molecular profiling has demonstrated that within each of these groups there are high risk and low risk subsets. That risk stratification is described with frontline chemotherapy treatments but it hasn’t really been explored in the setting of chimeric antigen receptor T-cell therapy. So the purpose of the study was to determine the impact of these subtypes on the outcomes of patients treated with CAR T-cell treatment.

What was the design of the study?

We sought to collect biopsy specimens from patients with relapsed or refractory diffuse large B-cell lymphoma, all of whom were treated with commercial CAR T-cell therapy across 13 US academic medical centres. We started with over 100 cases and after applying genomic analysis, including gene expression and next generation sequencing, to these cases the final interpretation was applied to 96 biopsies. The subtypes, according to the two dominant modes of sub-classification, were assigned and then the treatment outcomes were described.

What were the key findings?

What we discovered is that actually the patients who are traditionally thought of as being at the highest risk, or some of the highest risk, groups, including those with so-called cluster 5 or MCD subtype, as well as those with what’s called the A53 subtype, which has a preponderance of p53 mutations, these are so-called cluster 2 patients, all of these patients in these groups had very similar outcomes to the other, lower risk, groups. The only signal that we saw was for a trend towards inferior progression free survival in the so-called cluster 3 patients with EZB subtype. This suggests that CAR T-cell therapy can overcome some of the high risk features that we see in the frontline setting.

What could be the clinical implications of this study?

We know that patients with relapsed/refractory large B-cell lymphoma still have a certain degree of failure with CAR T-cell therapy. Probably about 40% of patients achieve durable remission which means that up to 50-60% have failure to respond durably. We did look at some of the mutations in these biopsies that were associated with increased risk of treatment failure and we did find likely activating mutations in MYC and BCL-2 were associated with treatment failure as well as a few other mutations, including some cyclin dependent kinase inhibitor 2A and KLHL6.

So we will need more data to confirm these in larger prospective series or validation cohorts but as we seek to define the molecular features that may confer poor outcome with CAR T-cell therapy, there may be strategies going forward which would allow us to intervene perhaps earlier or at the same time as CAR T-cell treatment to improve the outcomes of these patients with high risk features.