A systematic review and meta-analysis of the impact of radiation-related lymphopenia on outcomes in pancreatic cancer

Abstract

Background: Pancreatic cancer is a devastating disease with a 5-year survival rate of 5–10%. Radiation is commonly used in neoadjuvant and adjuvant settings to improve local control. Studies have shown that circulating lymphocyte count depletion after radiation has been associated with poor tumor control and inferior overall survival (OS) outcomes. Method: To better understand the impact of radiation-associated lymphopenia in pancreatic cancer, the authors undertook this systematic review and meta-analysis of clinical studies that have reported radiation-related lymphopenia in pancreatic cancer. Results: A systematic methodology search of PubMed, Embase and the Cochrane Library resulted in 2969 abstracts. Nine studies fulfilled the inclusion criteria. Six studies reported on outcomes in patients undergoing definitive chemoradiation and three studies comparing outcomes in stereotactic body radiotherapy versus definitive chemoradiation. The patients with severe lymphopenia were at increased risk of death with a pooled hazard ratio of 2.33 (95% CI: 1.79, 3.03; I2: 36%; p < 0.001) compared with patients with no severe lymphopenia. The odds of developing severe lymphopenia were 1.12 (95% CI: 0.45, 2.79; I2: 95%; p < 0.81). The pooled mean difference for OS was -6.80 months (95% CI: -10.35, -3.24; I2: 99%; p < 0.002), suggesting that patients who develop grade 3 or 4 lymphopenia have inferior median OS outcomes. Limiting the mean splenic dose to less than 9 Gy as well as various spleen dosimetric parameters such as visit (V)10 <32%, V15 <23% and V20 <15.4% can reduce the incidence of severe lymphopenia. Conclusion: Radiation-related lymphopenia is associated with an increased hazard of death and inferior median OS. Spleen dosimetric parameters correlate with the incidence of severe lymphopenia and with sub-optimal survival outcomes. There is a need to validate these findings in prospective studies.