This article is more than 5 years old. Information may no longer be current.
Patient characteristics correlate with toxicity, response to CAR T-cell therapy in diffuse large B-cell lymphoma
Click Here to Manage Email Alerts
ATLANTA — The presence of inflammation and high tumor burden prior to chimeric antigen receptor T-cell treatment indicated increased risk for toxicity among patients with diffuse large B-cell lymphoma, according to safety and efficacy data from the TRANSCEND NHL study presented at the ASH Annual Meeting and Exposition.
The analysis of patients’ clinical and laboratory biomarkers before treatment with JCAR017 (Juno Therapeutics) may identify patients at risk for toxicity and reduced durability of response, the research showed.
“JCAR017, a defined composition chimeric antigen receptor [CAR] T-cell product administered at precise doses of CD4-positive and CD8-positive CAR T cells, better enables the identification of potential patient factors associated with clinical outcomes due to lack of product variability,” Tanya Siddiqi, MD, assistant clinical professor in the department of hematology and hematopoietic cell transplantation at City of Hope, said during her presentation.
JCAR017 — CD19-directed CAR T-cell product — demonstrated high complete response rates and low incidence of cytokine release syndrome and neurotoxicity.
Patients received lymphodepletion with fludarabine and cyclophosphamide followed by a single dose of JCAR017 at 5 × 107 cells or 1 × 108 cells. A separate, smaller cohort received a two-dose schedule of JCAR017 at 5 × 107 cells. Researchers collected and evaluated patient blood samples at various time points for pharmacokinetic, pharmacodynamic and cytokine levels.
Researchers sought to determine potential correlations of clinical endpoints with patient characteristics, baseline clinical lab values and cytokine levels, and expansion and exposure to CAR T cells.
Best overall response, durable response at 3 months, any-grade neurotoxicity, grade 3 to grade 4 neurotoxicity, and cytokine release syndrome served as the clinical endpoints.
Ninety-one patients in the DLBCL cohort were evaluable for safety. Among these, 35% developed any-grade cytokine release syndrome, including 34% with grade 1 to 2, 0% with grade 3 and 1% with grade 4. Nineteen percent of patients also experienced neurotoxicity: 7% had grade 1 to 2, and 12% had grade 3 to 4.
Dose level of JCAR017 did not correlate with cytokine release syndrome or neurotoxicity.
Higher baseline lactate dehydrogenase (LDH) and higher baseline disease burden — measured by the sum of the products of diameters (SPD) of lymph nodes on imaging — (P < .001 each) appeared to be correlated with any-grade cytokine release syndrome and neurotoxicity. In fact, the risk for cytokine release syndrome and neurotoxicity was eight-fold higher if LDH was 500 or higher at baseline and SPD was 50 cm2 or higher.
Pre-CAR T-cell clinical labs and cytokines that were associated with the occurrence of any-grade neurotoxicity included trends for higher levels of ferritin, C-reactive protein, D dimer, plasma interleukin 6, interleukin 15, TNF- and MIP-1-alpha (P < .05 for each).
Researchers observed an association between any-grade cytokine release syndrome and higher trends of pre-CAR T-cell infusion ferritin, C-reactive protein, interleukin 10, plasma interleukin 16, interleukin 15, TNF-alpha and MIP-1 beta (P < .05 for all except ferritin and C-reactive protein).
Among 88 patients evaluable for efficacy, 74% achieved best overall response of complete or partial remission. Of these, 53% remained in complete or partial remission at 3 months.
For best overall response, researchers observed an association between pre-CAR T-cell analytes and lower levels of ferritin, LDH, C-X-C motif chemokine 10, granulocyte-colony stimulating factor and interleukin 10.
Those associated with durable response at 3 months included lower ferritin, C-reactive protein, LDH, C-X-C motif chemokine 10, interleukin 8, interleukin 10, interleukin 15, monocyte chemoattractant protein-1, MIP-1, TNF-, and higher pre-CAR T-cell hemoglobin and albumin (P < .05 for each).
Tumor burden (SPD and LDH), baseline markers of inflammation and inflammatory cytokines trended lower among patients with durable responses at the 3-month mark, whereas higher baseline tumor burden and cytokines appeared associated with higher CAR T-cell expansion in patients that may lead to early T-cell exhaustion and potentially lower durability of response.
“Preliminary modeling data suggest a therapeutic window may exist for CAR T-cell expansion that could limit toxicity and optimized efficacy,” Siddiqi said. “[We] may be able to drive all patients into this window by identifying patients at risk for low or high CAR T-cell expansion and evaluating strategies, like combination therapy, to enhance or limit CAR T-cell expansion.” – by Melinda Stevens
Reference s :
Siddiqi T, et al. Abstract 193. Presented at: ASH Annual Meeting and Exposition; Dec. 9-12, 2017; Atlanta.
Abramson JS, et al. Abstract 581. Presented at: ASH Annual Meeting and Exposition; Dec. 9-12, 2017; Atlanta.
Disclosures: Siddiqi reports speaker roles with Juno and Seattle Genetics; membership on the steering committees for ibrutinib (Imbruvica; Pharmacyclics, Janssen) and JCAR017; and a financial relationship with Pharmacyclics. Please see the abstract for all other authors’ relevant financial disclosures.
Perspective
Back to Top
Caron Jacobson, MD
CAR T cells have had remarkable success in refractory and relapsed B-cell malignancies. This has resulted in the FDA approval of tisagenlecleucel (Kymriah, Novartis) for pediatric and young adult B-cell acute lymphoblastic leukemia and axicabtagene ciloleucel (Yescarta, Kite Pharma/Gilead) for adult DLBCL. Response rates following anti-CD19-directed CAR T-cell therapy in these diseases have been 80% to 90%, with long-term responses seen in a proportion of patients. There are patients who do not respond, however, and among the patients who do respond, some will relapse. Mechanisms and predictors of resistance and relapse are being investigated. In addition, these therapies are associated with unique and high-risk side effects of cytokine release syndrome and neurotoxicity. Predictors of these side effects, and mediators of neurotoxicity, also comprise an area of active investigation. The shared goal is to improve efficacy while improving safety.
The results of the TRANSCEND study of JCAR017 in DLBCL are similar in terms of efficacy to that seen with other anti-CD19 CAR T cells in this disease. The toxicity profile, however, compares favorably with that seen on other studies, with only 2% of patients experiencing grade 3 and grade 4 cytokine release syndrome; rates of neurotoxicity are more similar across studies.
This study retrospectively analyzed several patient and disease characteristics, as well as serum biomarkers of response and toxicity. Following JCAR017, patients with worse performance status, with disease transformation from marginal zone lymphoma or chronic lymphocytic lymphoma, and with serum markers of increased tumor burden and inflammation had an inferior outcome. Additionally, higher disease burden and worse performance status correlated with both cytokine release syndrome and neurotoxicity, and higher serum markers of inflammation and disease burden correlated with neurotoxicity, specifically. The authors interpret this to suggest earlier treatment in one’s disease course may yield improved outcomes and safety.
Although these results are intriguing, they are not wholly consistent with those seen on other studies, which demonstrate no difference in efficacy by disease stage, risk score, burden or number of prior therapies. However, disease burden and baseline inflammatory markers more consistently correlate with toxicity. The differences in association with response may be that the different CAR T-cell constructs result in different postinfusion kinetics and, thus, the factors predictive of response with one therapy may not be the same for another. It may also be that we have still treated only a relatively small number of patients, and data from more patients will be necessary to identify a robust predictive model.
Rather than use these data to argue for treating patients earlier in their disease course with CAR T cells, a more compelling argument, perhaps, is that CAR T-cell therapy benefits patients regardless of chemotherapy sensitivity. Therefore, identifying patients who are least likely to respond to further chemotherapy and treating them instead with CAR T cells may spare patients unnecessary and unsuccessful treatments down the road. The ongoing and planned studies of these therapies in high-risk patients with DLBCL — who either do not respond to or who relapse within a year of upfront chemotherapy — randomly assigned to CAR T cells or standard of care, including salvage chemotherapy and autologous stem cell transplantation, are expected to address this issue and answer this question.
Click Here to Manage Email Alerts