Abstract 2656: Predictors of bloodstream infection associated with permanently implantable venous port in solid cancer patients

Abstract

Abstract Background There is no literature for stratification of permanent implantable venous ports (Port-A) related bloodstream infection (BSI) in solid cancer patients. We conduct this study to characterize the risk factors of BSI associated with the use of Port-A (PABSI) in solid cancer patients. Methods The study is done in a teaching hospital in Taiwan. Solid cancer patients who were implanted with a Port-A for systemic anti-cancer therapies were prospectively followed for the occurrence of PABSI, defined as BSI without other identifiable infection foci. Patients with PABSI during the follow-up period constitute the case cohort (N = 58) and 174 patients were randomly selected from those without BSI as the control cohort. PABSI risk score was developed by Cox proportional hazards model. Results 416 patients were registered (Table 1). Eighty-eight PABSI episodes occurred in 58 patients, with an incidence of 1.05 per 1000 catheter-days. All but one patient had stage 4 cancer. Independent predictors of PABSI occurrence included neutropenia (HR = 3.86, 95% CI = 1.75-8.50), total parental nutrition (TPN) (HR = 3.51, 95% CI = 1.86-6.63), chronic steroid use (HR = 7.01, 95% CI = 2.78-17.65), invasive procedure (HR = 3.00, 95% CI = 1.29-6.98), postoperative antibiotics (HR = 2.61, 95% CI = 1.43-4.75), and preoperative antibiotics (HR = 0.38, 95% CI = 0.17-0.83). A PABSI risk score, with a cut-off value of 0 (sensitivity 88.5%, specificity 64.3%), was defined for stage 4 cancer patients as follows: neutropenia, +1.350; TPN, +1.256; chronic steroid use, +1.947; preoperative antibiotics, –0.970; postoperative antibiotics, +0.959; and invasive procedure, +1.098. The median PABSI-free survival was 4.47 months for patients with score β0 but not reached for patients with score < 0 (p < 0.0001). Conclusions The PABSI risk score can help define high-risk groups in solid cancer patients and may help design future preventive strategies. Presence of any risk factor would render patients as high risk. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2656. doi:1538-7445.AM2012-2656