Impact of G-CSF during neoadjuvant therapy on outcomes of operable pancreatic cancer.

Abstract

4126 Background: Pancreatic ductal adenocarcinoma (PDAC) is an aggressive disease characterized by chronic inflammation and a tolerogenic immune response. Neutropenia is a common side effect of cytotoxic chemotherapy, managed with administration of recombinant granulocyte-colony stimulating factor (G-CSF, Filgrastim). The interleukin 17 – G-CSF – neutrophil extracellular trap (NET) axis promotes oncogenesis and progression of PDAC, inhibiting adaptive immunity. We evaluated the impact of G-CSF administration during neoadjuvant therapy (NAT) on oncologic outcomes in patients with operable pancreatic cancer. Methods: A retrospective review of all patients with localized PDAC treated with NAT prior to pancreatic resection between 2014 – 2020 was completed at a single institution. G-CSF administration, type, and dose were collected from inpatient and outpatient medical records. Results: Of 351 patients treated, 138 (39%) received G-CSF during NAT with a median follow-up of 45.8 months. Patients who received G-CSF were younger (64.0 vs 66.7, p = 0.008), had lower BMI (26.5 vs 27.9, p = 0.021), and were more likely to receive 5-FU based chemotherapy (42% vs 28.2%, p < 0.0001), NAT dose reduction (40.6% vs 25.4%, p = 0.003), or experience febrile neutropenia (8.7% vs 3.3%, p = 0.029). No differences were observed in baseline or pathologic tumor staging. In patients who received G-CSF, 130 (94%) received Pegfilgrastim with a median cumulative dose of 12 mg (IQR 6-12). Patients who received G-CSF were more likely to have an elevated post-NAT neutrophil to lymphocyte ratio (45% vs 29.6%, p = 0.004) and systemic immune-inflammation index (39.5% vs 29.6%, p = 0.061). Receiving G-CSF was an independent predictor of perineural invasion (HR 2.4, 95 CI [1.08, 5.5], p = 0.031) and margin positive resection (HR 1.69, 95 CI [1.01, 2.83], p = 0.043). Patients who received G-CSF had decreased overall survival compared to patients who did not receive G-CSF (median OS: 29.2 vs 38.7 months, p = 0.0001). Receiving G-CSF during NAT was an independent negative predictor of progression free (HR 1.38, 95 CI [1.04, 1.83], p = 0.022) and overall survival (HR 2.02, 95 CI [1.45, 2.79], p < 0.0001). In a subset of patients with available pre- and post-NAT serum specimens (n = 28), G-CSF administration resulted in an increased number of citrullinated histone H3 complexes following NAT (+1378±1502 vs -300.7±1147 pg/ml, p = 0.007), indicative of enhanced peripheral NET formation. Conclusions: In patients with localized PDAC receiving NAT prior to surgical extirpation, G-CSF administration is associated with worse oncologic outcomes and should be administered with caution. Prospective randomized as well as confirmatory clinical studies are in order.