AUTHORS
Roni Shouval Jr, MD, PhD, Teng Fei, PhD2, Luca Paruzzo, MD3, Maria Bromberg, MS4, Sophia Stock, MD5, Sigrun Einarsdottir, MD, PhD6, Abraham Avigdor, MD7, Matar Ben Yehuda8, Ori Ben Valid9, Arnon Nagler I, MD10, Ofrat Beyar-Katz, MD, PhD11, Jabour Halloun, MD12, Anat Stern13, Netta Glaubach, MD14, Giulia Gabrielli, MD15, Federico Stella, MD16, Kayleen Shi17, Guido Ghilardi18, Antonio Imparato, MD19, Shan Liu, PhD20, Genovefa Papanicolaou21, Jae H Park, MD22, Miguel-Angel Perales, MD23, Susan K. Seo24, Gunjan L. Shah, MD, MS25, Keimya Sadeghi, MS26, Chetan Kancharla27, John Slingerland28, Marion Subklewe I, MD29, Marcel R.M. van den Brink Jr, M.D., Ph.D.30, Zainab Shahid, MBBS31, Kai Rejeski, MD32, Karamjeet Singh Sandhu, MD33, Marco Ruella, MD34
AFFILIATIONS
1 Memorial Sloan Kettering Cancer Center, New York, NY, United States
2 Memorial Sloan Kettering Cancer Center, New York, NY, United States
3 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
4 Memorial Sloan Kettering Cancer Center, New York, NY, United States
5 LMU University Hospital, LMU Munich, Munich, Germany
6 Memorial Sloan Kettering Cancer Center, New York, NY, United States
7 Medical Center Dr. h Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel
8 Medical Center Dr. h Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel
9 Medical Center Dr. h Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel
10 Medical Center Dr. h Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel
11 Rambam Health Care Campus, Haifa, Israel
12 Rambam Health Care Campus, Haifa, Israel
13 Rambam Health Care Campus, Haifa, Israel
14 Rambam Health Care Campus, Haifa, Israel
15 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
16 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
17 LMU University Hospital, LMU Munich, Munich, Germany
18 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
19 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
20 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
21 Memorial Sloan Kettering Cancer Center, New York, NY, United States
22 Memorial Sloan Kettering Cancer Center, New York, NY, United States
23 Memorial Sloan Kettering Cancer Center, New York, NY, United States
24 Memorial Sloan Kettering Cancer Center, New York, NY, United States
25 Memorial Sloan Kettering Cancer Center, New York, NY, United States
26 City of Hope, Duarte, California, United States
27 City of Hope, Duarte, California, United States
28 City of Hope, Duarte, California, United States
29 LMU University Hospital, LMU Munich, Munich, Germany
30 City of Hope, Duarte, California, United States
31 Memorial Sloan Kettering Cancer Center, New York, NY, United States
32 LMU University Hospital, LMU Munich, Munich, Germany
33 City of Hope, Duarte, California, United States
34 Lymphoma Program, Abramson Cancer Center, University of Pennsylvania, Philadelphia, Pennsylvania, United States
ABSTRACT
Background: Broad-spectrum antibiotics (BSAs) disrupt the gut microbiome and have been associated with inferior outcomes after cancer immunotherapy. In CAR T-cell therapy, pre-infusion BSA exposure has been linked to worse progression-free survival (PFS) and overall survival (OS). BSAs are commonly initiated after infusion, primarily for fever accompanying cytokine release syndrome (CRS), and the impact of specific post-infusion agents remains unclear. Establishing the impact of post-infusion BSA exposure on CAR-T outcomes is essential to guide their appropriate use.
Aim: To evaluate the association between individual BSAs administered between day 0 to 30 after CD19-CAR-T infusion and survival in over 1,000 patients with large B-cell lymphoma (LBCL) treated worldwide.
Methods: Retrospective study of large B-cell lymphoma (LBCL) patients treated with axi-cel, tisa-cel, or liso-cel at 5 international tertiary centers (United States, Germany, Israel). Antibiotic exposure was evaluated from day -22 to day 30 relative to infusion. Study antibiotics included piperacillin-tazobactam (PTZ), cefepime, carbapenems, and intravenous (IV) vancomycin. Patients with pre-infusion bacteremia or pre-infusion study-antibiotic exposure were excluded from outcome analyses.
Results: We studied 1,093 patients with LBCL (median age 65 years) treated with axi-cel (45%), tisa-cel (32%), or liso-cel (22%) across five centers (68% US, 32% ex-US). With a median follow-up of 25.2 months, median OS and PFS were 39 and 9 months, respectively.
BSAs were given far more often after than before infusion: 125 patients (11%) pre-infusion versus 682 (62%) post-infusion. Post-infusion exposure tracked closely with CRS (81% of patients with CRS versus 14% without CRS; p<0.001). Bacteremia occurred in 15 patients (1.4%) before infusion and 112 (10%) during days 0-30. Median post-infusion antibiotic duration was 5 days for cefepime and 7 days for PTZ and carbapenems.
A total of 821 patients, who did not experience pre-infusion bacteremia or pre-infusion exposure to a study antibiotic, were considered for further analyses and categorized into four mutually exclusive groups defined by expected intensity: PTZ with or without cefepime but without carbapenem or IV vancomycin (n=261; 32%); cefepime monotherapy (n=60; 7%); carbapenem with or without other BSAs or IV vancomycin (n=100; 12%); or no exposure to any of the four study antibiotics (n=400; 49%). The remaining 147 patients had combined exposure to multiple study antibiotics and were not further analyzed.
In day-30 landmark Cox models stratified by center and adjusted for age, product, pre-lymphodepletion LDH, bridging, CRS, and day 0-30 bacteremia, using PTZ as the reference, cefepime monotherapy was not associated with different PFS (HR 0.88, 95% CI 0.53-1.44) or OS (HR 0.64, 0.34-1.23), and lack of BSA exposure was not associated with better outcomes than PTZ (PFS HR 1.34, 0.95-1.89; OS HR 1.22, 0.83-1.80). Any carbapenem was associated with inferior OS versus PTZ (HR 1.66, 1.10-2.49; p=0.015) and numerically inferior PFS (HR 1.45, 0.99-2.12; p=0.059).
In separate analyses comparing exposure duration, >=7 versus <7 days of PTZ or cefepime was not associated with worse outcomes within exposed patients, whereas longer carbapenem exposure trended toward worse PFS (HR 1.97, 0.91-4.27; p=0.085).
To address whether there is a preferable BSA post-infusion, we directly compared the PTZ and cefepime groups described above using 1:1 propensity-score matching (104 patients total), balancing age, sex, performance status, product, pre-lymphodepletion LDH, bridging therapy, and antibiotic exposure duration. PFS was comparable between agents (cefepime vs PTZ HR 0.93, 95% CI 0.57-1.52; p=0.8), while OS numerically favored cefepime without reaching significance (HR 0.60, 0.33-1.09; p=0.092).
Summary: In an international multicenter analysis of post-infusion antibiotic exposure, the two most common empiric beta-lactams, PTZ and cefepime, were not associated with inferior survival and were comparable to each other. Carbapenem exposure was associated with inferior OS relative to PTZ; confounding by indication cannot be excluded. These findings are reassuring for commonly used empiric beta-lactams after infusion and support reserving carbapenems for clear indications.
DISCLOSURES
Roni Shouval Jr, MD, PhD: CorMedix - Consultancy (Includes expert testimony); AROG Pharmaceutical - Consultancy (Includes expert testimony)