Effect of adding bemiparin and cefepime to routine treatment in cancer patients with SARS-CoV-2 infection

Dear Editor:

Oncology patients are more vulnerable to the 2019 coronavirus disease (COVID-19), with greater in-hospital mortality rates being recorded among them compared with other types of patients. The possible causes of this could be their increased plasma levels of tumor necrosis factor alpha (TNF-α) and interleukin 6 (IL-6) secondary to the neoplastic cachexia process and the prothrombotic state inherent to cancer.¹

In this study we present our experience with hospitalized oncology patients with COVID-19, focusing on factors associated with their survival and the effect of the different treatments administered. In this last regard, we would like to highlight the use of bemiparin at a dose of 5000 IU as a treatment used to counteract the prothrombotic neoplastic stage, and of cefepime at a dosage of 2 g/8 h due to the potential anti-bradykinin activity of this molecule.

Our patient cohort was formed by 33 patients with a median age of 60–70 years, of whom 75.76% were men. The majority (78.79%) of these patients had a stage IV tumor. The most prevalent type of neoplasm in our patient cohort was lung cancer (27.27%).

Mortality among hospitalized oncology patients with COVID-19 was 70%. This rate dropped to 55.6% among patients treated with monoclonal antibodies (tocilizumab or anakinra). A total of eight patients of our cohort were treated from the time of admission with bemiparin at a dose of 5000 IU and cefepime at a dosage of 2 g/8 h as empirical antibiotic therapy. The mortality rate dropped to 37.5% among these patients. The causes of death of the COVID-19 patients who were admitted to our care were respiratory distress secondary to parenchymal lung damage (57.14%), neurological presentations of the infection (14.29%), and acute pulmonary embolism and/or multiorgan failure in the rest of cases.

The data collected from our patient cohort was subjected to a statistical analysis using software SPSS 22, applying a multivariate analysis model to eliminate potential confounding factors. The factors evaluated were use of remdesivir, tocilizumab, or anakinra; pulses of methylprednisolone; and a combination of bemiparin 5000 IU and cefepime. The covariates considered were sex, age, type of tumor, and tumor stage. The mortality rate was considered a dependent variable.

Nagelkerke’s R² was 0.964, therefore indicating that 96.4% of the variability in the mortality rates could be explained by the variables included in the study. Application of the multivariate model demonstrated statistical significance (p < 0.05) for covariates age (higher mortality at older ages), sex (higher male mortality), and type of cancer (higher mortality with lung and gastro-biliary-pancreatic tumors), but no significance for the tumor stage. As for the interventions, the only two that yielded statistical significance were the use of a combination of bemiparin 5000 IU+ cefepime and the use of monoclonal antibodies (tocilizumab or anakinra). According to this model, the use of methylprednisolone pulses and remdesivir were not statistically significant.

The higher mortality associated with lung and gastro-biliary-pancreatic tumors is likely associated with the cachetizing nature of those tumors. Patients with these tumors have increased circulating levels of TNF-α, which could cause hyperinflammatory reactions to be more severe.²

The rationale of the use of cefepime in patients admitted for COVID-19 is based on the findings of several studies, such as that carried out by Roche et al.,³ who claimed that the bradykinin cascade plays a fundamental role in pulmonary angioedema induced by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the activation of several inflammatory mediators. According to the results of a virtual screening study performed by Rasaeifar et al.,⁴ cefepime could have bradykinin receptor antagonist activity, which is why we used it as an empirical antibiotic therapy instead of another cephalosporin. Further studies with a greater number of patients should be carried out to confirm these findings.