CAR T-cell therapy may overcome poor prognosis tied to conventional tiNHL therapies

A real-world study presented at ASH 2024 shows that CD19 chimeric antigen receptor (CAR) T-cell therapy is highly effective and has an acceptable toxicity profile for individuals with transformed indolent non-Hodgkin lymphoma (tiNHL).

“About 10–30 percent of iNHL patients undergo transformation to aggressive LBCL* (aLBCL). Prior studies have shown that tiNHL is associated with chemoresistance, worse disease control, and inferior overall survival (OS),” said study co-first author Dr Swetha Thiruvengadam from the City of Hope National Medical Center, Duarte, California, US.

There is no standard-of-care (SoC) treatment for this high-risk patient subset, and although CD19 CAR T has revolutionized the treatment landscape of aLBCL, tiNHL patients were largely underrepresented in prior pivotal trials, she noted.

The investigators retrospectively evaluated adults with relapsed/refractory aLBCL (DLBCL** [89 percent] and HGBCL***) – both de novo and transformed – who received SoC CD19 CAR T-cell therapy between 1 December 2017 and 31 October 2022. Of the 1,182 patients included, 884 (71 percent) had de novo aLBCL, while 338 (29 percent) had tiNHL. Most tiNHL cases transformed from follicular lymphoma (FL; 84 percent), followed by marginal zone lymphoma (12 percent) and Waldenstrom’s macroglobulinaemia (4 percent).

Overall, median age at infusion was 64 years and 36 percent were women. At the time of CAR T-cell therapy, 80 percent of patients had advanced disease (III–IV), 56 percent had elevated LDH^#, 40 percent had >1 site of extranodal disease, and 12 percent had bulky disease (≥10 cm). The most common CAR T used was axicabtagene ciloleucel (78 percent), followed by tisagenlecleucel (14 percent) and lisocabtagene maraleucel (9 percent).

Overall response rates were similar between the tiNHL and de novo arms (83 percent vs 81 percent; p=0.3), but complete response (CR) rate was significantly higher in the former vs the latter (67 percent vs 59 percent; p=0.017).

Median duration of response was comparable in both tiNHL and de novo arms (23.6 vs 20 months; p=0.28), as was median duration of CR (48.3 vs 36.3 months; p=0.81). [ASH 2024, abstract 524]

After a median follow-up of 13 months post-CAR among all patients, the tiNHL and de novo arms had similar 24-month progression-free survival (PFS; 40.6 percent vs 38 percent; p=0.16) and OS (57.6 percent vs 52 percent; p=0.15).

The cumulative incidence of non-relapse mortality was also comparable between the tiNHL and de novo arms at 12 (7.8 percent vs 7.5 percent) and 24 months (10 percent vs 11 percent) after CAR T infusion.

A trend towards a lower risk of disease progression, relapse, or death post-CAR was observed in the tiNHL vs de novo subgroup after adjusting for key baseline characteristics on multivariate analyses, but this failed to achieve statistical significance (adjusted hazard ratio [aHR], 0.84; p=0.07). However, in a separate post hoc analysis, the risk of disease progression or death was significantly lower in patients with transformed FL vs those with de novo disease (aHR, 0.81; p=0.049).

The regression model suggested that other factors tied to worse PFS are advanced disease stage prior to CAR T (aHR, 1.4), elevated LDH (aHR, 1.7), ≥3 prior lines of therapy (aHR, 1.2), bendamustine use within 12 months prior to CAR (aHR, 1.5), central nervous system involvement of aLBCL prior to CAR (aHR, 1.3), and receipt of bridging therapy (aHR, 1.4).

Compared with the de novo arm, the tiNHL arm had similar incidences of grade ≥3 CRS^## (7 percent vs 8 percent; p=0.6) and CAR T toxicity-related ICU stay within the first 30 days (12 percent vs 14 percent; p=0.2), and lower rates of grade ≥3 ICANS^### (21 percent vs 27 percent; p=0.024), tocilizumab use for CRS (50 percent vs 61 percent; p<0.001), and glucocorticoid use for ICANS (36 percent vs 44 percent; p=0.01).

“Our study suggests that CD19 CAR T is highly effective and safe for tiNHL with potentially higher CR rates, lower progression risk, comparable CRS, and lower incidence of ICANS compared with de novo aLBCL,” Thiruvengadam said. “[The findings imply that] CAR may be able to overcome the historically poor prognosis associated with conventional therapies for tiNHL.”