Updates from the Annual ASH meeting, December 2018.
Surabhi Dangi-Garimella, PhD
A multicenter retrospective study that evaluated the efficacy and safety of chimeric antigen receptor (CAR) T-cell treatment, axicabtagene ciloleucel (axi-cel; Yescarta), in a real-world setting produced a similar response as well as toxicity compared with the ZUMA-1 clinical trial results. Predictors of response included low day 0 C-reactive protein (CRP) and high absolute lymphocyte count at leukapheresis. The results were presented December 1, 2018, during the 60th American Society of Hematology (ASH) Annual Meeting & Exposition, held in San Diego, California.1
The FDA approved axi-cel in October 2017,2 and long-term results for the treatment were presented at the annual meeting of the American Society of Clinical Oncology (ASCO) in June 2018.3 The results presented at ASCO showed the objective response rate (ORR) was 82% at 8.7 months, which had been maintained by the long-term 15.4-months follow-up time. The complete response (CR) rate was 58% at the long-term follow-up.
Caron A. Jacobson, MD, instructor in medicine, Department of Medical Oncology, Dana-Farber Cancer Institute, and first author of the study presented at ASH, said that in a real-world setting, eligibility criteria and patient management factors may be very distinct from a clinical trial setting. Therefore, bridging therapy may be needed between leukapheresis and treatment with the CAR T product.
Thus, the study examined patient and disease characteristics and biomarkers of response or toxicity following axi-cel treatment in a real-world setting. In this case, the settings were 6 US academic medical centers that used commercial axi-cel.
Among the 104 patients with lymphoma who were part of this retrospective study (median age was 63.8 years; range, 21-80), 94 (90%) had an European Cooperative Oncology Group (ECOG) performance status of 0-1 and 48 (46%) had a prelymphodepletion International Prognostic Index (IPI) score of ≥3. Twenty-eight (27%) patients had a prior autologous transplant and 3 (3%) had received a prior allogenic stem cell transplant. Forty-two patients (40%) each had bulky disease (tumor bulk >5 cm) and received bridging therapy following leukapheresis.
At a median follow-up of 5.6 months, 13 patients had their T cells collected but they did not receive CAR T-cell infusion: 6 patients had progressive disease, 2 patients had infections, 3 patients had technical issues with cell production, 1 patient had a CR to bridging therapy, and 1 patient had another malignancy diagnosed.
When evaluated in 95 patients, an overall response (OR) was observed in 67 (62%) patients. A CR was observed in 42 (44%) patients and a partial response (PR) in 25 (26%). Among 51 patients with a 6-month follow-up, an OR was observed in 22 (43%).
Half of the patients who had an initial PR and who were not being followed went on to have a CR, Jacobson said. The median duration of response was 4.9 months.
According to Jacobson, univariate analysis showed that ECOG performance status (P = .009), tumor bulk (P = .016), baseline CRP (P = .029), and prior ibrutinib (Imbruvica) treatment (P = .002) had a significant association with lack of response to treatment with axi-cel.
A majority (96%) of the treated patients experienced cytokine-release syndrome (CRS), a flare-up that is characterized by low/high fever and low blood pressure and can sometimes lead to capillary leak syndrome, according to Stephen J. Schuster, MD, of the Perelman School of Medicine.4
In 17 (16%) patients, CRS was grade 3 or higher; 2 patients (2%) died. A median time to onset was 1 day (range, 0-14), symptoms associated with the flare-up lasted a median of 6 days (range, 1-27). Fifty-eight (55.7%) patients experienced neurotoxicity following treatment with axi-cel, 29 (50%) of whom had grade 3 or higher neurotoxicity. There was 1 fatality associated with this toxic effect. Neurotoxic effects had a median onset time of 5 days (range, 0-34) and lasted for a median of 8 days (range, 1-52 days). Patients received tocilizumab (Actemra; n = 70) and steroids (n = 66) to counter the toxicity, and 30% required a stay in an intensive care unit. Six patients died following disease progression, and 5 died from toxicity. Univariate analysis that the researchers conducted for toxicity, mainly grade 3 or higher CRS or neurotoxicity, found no association with performance status, tumor bulk, IPI, prior treatment, bridging therapy, or eligibility for ZUMA-1.
Cytogenetic and immunohistochemistry staining found that 3 patients with programmed death ligand -1 postive (PD-L1) tumors were refractory to CAR T-cell therapy. Based on the staining results, a positive response could be associated with increased expression of PD-1, 41BB, ICOS, and Ki67, as well as CC3 indicating apoptosis, when CAR T-cell levels peaked by day 7. Subsequently, CAR T cells fell by day 14.
Jacobson said that the deviation from the observations in the ZUMA-1 trial may be due to the inclusion of sicker patients with a poorer performance status and possible different histologies in this patient population. Although rates of CRS and neurotoxicity were similar to ZUMA-1, toxicity was not associated with tumor bulk or response, but with higher peak inflammatory markers and absolute lymphocyte count, indicative of peak CAR T-cell levels.
“Unique combination approaches are necessary for specific patients/tumors,” she noted, adding that their trial results support the use of axi-cel outside of strict clinical trial criteria, although the outcomes may be slightly inferior.