Post-Treatment ANC is Predictive of Poorer Overall Survival and Disease-Free Survival in Esophageal Cancer Patients after Tri-Modality Therapy

Abstract

Although neutrophils serve important roles in inflammation and host defense, increasing evidence points towards their involvement in tumorigenesis and progression. Recent studies show that increased pre-treatment neutrophils associate with disease recurrence and survival rates in esophageal cancer. Here, we evaluate the association of post-treatment absolute neutrophil count (ANC) with survival outcomes in primary non-metastatic esophageal cancer patients who have received tri-modality therapy. We reviewed 219 consecutive patients with stage I-III esophageal cancer who completed tri-modality therapy at our center between 1/2007 and 12/2015. Of the 219 patients, 123 completed all treatments within 5 months of starting radiation and had available ANC at 6 months after start of radiation (range: +/- 1 month). 70 (57%) were > 60 years old, and 110 (89%) were male. 116 (94%) tumors were adenocarcinoma, 89 (72%) were stage III, and 120 (98%) were located distally. 120 (98%) received induction chemotherapy of which 77 (64%) received platinum-taxane (PT). All received concurrent chemotherapy of which 69 (56%) received PT. 90 (73%) had the same induction and concurrent agents. 118 (96%) received intensity modulated radiation therapy, with median dose of radiation being 5040cGy in 28 fractions (range: 4320-5600cGy). Kaplan-Meier survival, log-rank tests, and multivariate Cox regression were used to analyze the association of ANC with survival outcomes. The median follow up was 31 months (range: 5.7-119.9 months). Multivariate Cox regression using age, gender, KPS, stage, histology, tumor location, time to surgery, induction PT, concurrent PT, lung V20, mean heart dose, PTV size, and pre- and post-treatment ANC showed that increased 6-month ANC was predictive of worse overall survival/OS (p=0.0001, HR 1.3). The median ANC was 3.9 K/microliter (range: 0.9-30.4). When stratified by median ANC, patients with ANC > 3.9 had significantly worse OS compared to those with ANC ≤ 3.9 (5-year OS=35% vs 57%, p=0.049), and the disease-free survival (DFS) was also worse although this was not significant (5-year DFS=11% vs 24%, p=0.09). Subgroup analysis showed that the predictive value of ANC is influenced by choice of concurrent chemotherapy agent. In patients receiving concurrent non-PT, ANC > 3.9 predicted for worse 5-year OS (25% vs 57%, p=0.024) and 5-year DFS (4% vs 30%, p=0.022). This relationship was not seen for concurrent PT. Higher post-treatment ANC is predictive of poorer survival in primary non-metastatic esophageal cancer patients who have received tri-modality therapy, and this effect is influenced by choice of concurrent chemotherapy agent. Further investigation could consider individual neutrophil subpopulations, how they are affected by different treatment regimens, and their roles in cancer progression.