Prospective Cohort Study of the Incidence, Risk Factors and Long-Term Sequelae of Symptomatic Catheter-Related Thrombosis in Adults with Cancer.

Long-term indwelling central venous catheters (CVCs) are used for delivering chemotherapy, parenteral nutrition, antibiotics, and blood products, as well as for facilitating blood drawing, in many patients with malignancy. Although the important supportive role of CVCs is unquestioned, there is uncertainty regarding the prevention and treatment of catheter-related thrombosis (CRT) because there is a lack of prospective and contemporary data on the natural history of this complication. As a first step towards improving CRT management, we conducted a prospective cohort study to examine the incidence, clinical risk factors, and the long-term sequelae of symptomatic CRT in adults with cancer. Consecutive patients undergoing insertion of a CVC at a tertiary care center were enrolled and followed for the duration of their catheter-dwell time plus 4 weeks or a maximum of 52 weeks, whichever comes first. Scheduled assessments were done at weeks 1, 2, 4, 8, 12, 24, 36 and 52 weeks after insertion. Patients with symptomatic CRT were treated with anticoagulants and were followed for an additional 52 weeks from the date of CRT diagnosis. Baseline information and follow-up data regarding catheter patency, thromboprophylaxis, clinical symptoms, and thrombotic events was collected. Standardized regional guidelines for catheter care were followed and symptomatic CRT was diagnosed based on objective testing and satisfaction of prespecified criteria. Between March 2002 and July 2003, 444 patients underwent 500 catheter insertions. The mean patient age was 56 y (range 18–91 y) and 55% of patients were female. Catheters inserted included PICCs (65%), ports (18%), pheresis (11%), and Hickman catheters (6%). As of July 22, 2004, 442 patients had completed follow-up. The total catheter-dwell time was 59,959 d (median 88 d), while the total follow-up was 73,654 pt-d (median 151 d). Colorectal was the most common tumor type in 18% of patients and 41% of all patients at enrolment had metastatic solid tumor. Overall, there were 19 episodes of symptomatic CRT, representing an incidence of 4.3% (95% CI 2.6–6.6%) of patients or 0.3 CRTs per 1000 catheter-dwell days (95% CI 0.2–0.5 per 1000 d). The mean time to CRT was 53 d (range 6–162 d). Development of CRT was not associated with age, ECOG performance status, cancer treatment, catheter type, side of insertion, thromboprophylaxis, infection, or previous history of thrombosis. The only significant risk factor was ovarian cancer (P=0.02). In patients with symptomatic CRT, 89% (17/19) of CRTs were treated with anticoagulant therapy alone, 5.3% (1/19) had the catheter removed, and 5.3% (1/19) were treated with both; none had symptomatic pulmonary embolism or post-thrombotic syndrome during follow-up. In summary, the incidence of symptomatic CRT in adults with cancer is low and treatment with anticoagulant therapy alone was not associated with any serious long-term sequelae. Due to the small number of CRTs observed, larger studies are required to further evaluate risk factors and identify the optimal therapeutic approach for CRTs.