Incidence of central port failure among patients receiving bevacizumab.

Abstract

e19579 Background: Central venous port (CVP) failure is a common problem in cancer patients (pts) during chemotherapy. Risk factors include increasing age, male gender, and hematogenous malignancy. The incidence of CVP failure in adult cancer pts previously reported in the literature ranges from 3.4–10%, with infection as the most common cause. Pts receiving Bevacizumab (bev) showed an increase in wound dehiscence, as well as increase thromboembolic event. We conducted a study to determine the frequency of port failure among pts receiving bev. Methods: A multi-center chart review of cancer pts receiving ≥ 2 doses of bev from Cleveland Clinic Cancer Center (Moll Pavilion) and Beaumont Hospitals was conducted. All pts had a CVP placed between July 2006 and August 2009. Pts who had wound adhesions, CVP failure due to infection, without proper monthly heparinization were excluded. The incidence of CVP failure was calculated. CVP failure was defined as difficulty to inject and blood draw requiring port removal. Results: We identified 164 pts (61 males; 103 females; age range 36-88). Twenty eight pts (17 %) experienced CVP failure (9 males, age 42-79). There were 18 breast, 5 colorectal and 5 non-squamous cell lung cancer pts. Median follow up was 142 days (range 30-258). The CVP was placed 59% in the left side, and 41% in the right side of the chest. Using Venography to assess the case of malfunction; thrombus or fibrin sheath was the most common cause of failure (16 Pts) 57%, other causes (disconnection, cath rupture or fracture) in (7 pts) 25%, and the venograph was normal in (5 pts) 18%. The proportion of CVP failure in pts receiving bev in this study is significantly higher than previously reported. Since we excluded infection as a cause of the CVP failure, it is likely that the observed CVP failure was underestimated. Conclusions: Bev is associated with a higher rate (17%) of CVP failure compared with previous reports even after excluding infection as a cause. The thrombus as the main cause may be secondary to bev increasing thromboembolic event. Since the number of pts is small, a prospective cohort study is needed to verify our results. No significant financial relationships to disclose.