12105 Background: Infection by SARS-CoV-2 can turn into an acute respiratory infection. Approximately 15% of patients will develop a distress syndrome responsible in most cases of mortality. A host hyperinflammatory response induced by a cytokine storm, is the main cause of this severe complication. Chemotherapy myelosuppression is associated with higher risk of infections and mortality in cancer patients. There have been no previous reports about the clinical management of patients with neutropenia and concomitant COVID-19. Herein, we present a multicenter experience in several hospitals during COVID-19 outbreak in neutropenic cancer patients infected by SARS-Cov-2. Methods: Retrospective clinical data were collected from clinical reports. Protocol was approved by a Clinical Research Ethics Committee (HULP: PI-4194). Inclusion criteria were cancer patients with neutropenia (<1500 cells/mm3) and concomitant COVID-19 infection. Comorbidities, tumor type and stage, treatment, neutropenia severity, filgrastim (G-CSF), COVID-19 parameters and mortality were analyzed. Exploratory analysis included a description of all data collected and bivariate analyses among different pairs of variables, including their impact in mortality in this cohort. In addition, multivariable logistic regression was used to predict respiratory failure and death as a function of multiple variables. Results: Among 943 patients with cancer screened in 14 hospitals in Spain, eighty-three patients (8%) had a febrile neutropenia and COVID-19 infection. Lung (26%), breast (22%), colorectal (13%) and digestive non-colorectal (17%) cancers were the main locations and most patients had advanced disease (67%). Fifty-three (63%) of patients included died because respiratory failure. Neumonia was presented in 76% of patients, bilateral in 47% and 12% of all patients had thrombotic events. The median of neutrophils was 650cls/mm3 and 49% received G-CSF with a median of days on treatment around 4,5 days. Among all variables related with mortality in neutropenic cancer patients with COVID-19 infection, we found that the number of days with G-CSF showed a significant trend toward worse outcome and higher mortality. In particular, a logistic regression model was developed to predict respiratory failure, as a function of the number of days of G-CSF treatment. As adjusting covariates, sex, age, treatment purpose (palliative vs curative, to adjust for patient status), tumor type, and the lowest level of neutrophils in the patient (to adjust for neutropenic status) were used. A significant effect was obtained for the days of G-CSF treatment (OR = 1.4, 95% CI [1.03, 1.92], p-value = 0.01). Conclusions: Our findings suggest that a prolonged G-CSF treatment could be disadvantageous for these cancer patients with COVID-19, with a higher probability of worse outcome.
