Outcomes of infectious workup and antibiotic use during tarlatamab initiation in patients with advanced neuroendocrine carcinomas.

Abstract

e23307 Background: Tarlatamab has improved outcomes in extensive-stage small cell lung cancer (ES-SCLC) and is under investigation in other neuroendocrine carcinomas (NEC). Patients are monitored for cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) during tarlatamab. As it can be difficult to discern CRS/CANS from sepsis, we evaluated culture acquisition patterns and empiric antibiotic use during tarlatamab initiation. Methods: We conducted a retrospective chart review of patients with advanced NEC treated with tarlatamab at Yale Cancer Center (2023–2025) through an IRB-approved protocol. Data cutoff was December 20, 2025. Antibiotics appropriateness was determined by culture positivity and clinical review. Results: Thirty-four patients were identified. Smoking history varied (20 former, 10 current, 4 never). Histologies included SCLC (n = 24), large cell neuroendocrine carcinoma (LCNEC; n = 2), mixed SCLC/LCNEC (n = 1), mixed LCNEC/adenocarcinoma (n = 1), prostate NEC (n = 2), nasal cavity NEC (n = 1), and small bladder cell carcinoma (n = 1). 30/34 patients received ≥1 dose of tarlatamab. At C1D1, 33 patients were observed inpatient and 1 outpatient. Thirty continued to C1D8 (29 inpatient, 1 outpatient), and 27 to C1D15 (24 outpatient, 3 inpatient). At C1D1 and C1D8, CRS and/or ICANS were noted in 19/34 and 9/30 patients respectively (Table 1). Cultures were obtained for 20/64 encounters: 18 blood, 12 urine, 1 cerebrospinal fluid, 2 sputum, and 1 pleural fluid. Documented reasons for culture were fever (n = 17), hypotension (n = 4), hypoxia (n = 3), altered mental status (n = 5), and dysuria on admission (n = 1). Most cultures (88.2%) were negative after 5 days of incubation. Four urine cultures were positive: Klebsiella pneumoniae (n = 2), Staphylococcus aureus (n = 1), and Serratia and Lactobacillus species (n = 1). Nine patients received antibiotics, with 88% intravenous. No cultures or antibiotics were documented during C1D15. Retrospective adjudication of clinical details and culture data deemed antibiotics unnecessary in all cases, as symptoms improved with CRS/ICANS-directed therapy, or an organism was deemed a colonizer. Conclusions: During tarlatamab initiation, documented culture growth was rare and antibiotics were deemed unwarranted after clinical review. Routine infectious workup and empiric antibiotic initiation during tarlatamab observation may be unnecessary and should be reserved for clear evidence of infection or clinical deterioration requiring escalation of care despite treatment for CRS/ICANS. These results support protocol changes to reduce unnecessary testing, antibiotic exposure, and patient discomfort. Patient characteristics. C1D1 (N=34) C1D8 (N=30) Fever (≥38°C) 14 5 CRS G1-2 14 7 CRS G3-4 3 1 ICANS G1-2 6 4 ICANS G3 2 0 Patients cultured 13 7 Positive cultures (urine) 1 3 Antibiotics given 6 3