Recombinant tissue plasminogen activator reduces incidence of catheter malfunction and bacteremia

Hemmelgarn B, Moist L, Lok C, et al. Prevention of dialysis catheter malfunction with recombinant tissue plasminogen activator. N Engl J Med. 2011;364:303-312.

Malfunctioning and infected hemodialysis catheters increase morbidity, mortality, and the cost of care of hemodialysis patients. More than 50% to 70% of hemodialysis catheters will be expected to fail within the first year, with half to two-thirds of these failures attributable to catheter thrombosis.1 Heparin has been shown to be superior to saline as a locking solution in preventing catheter malfunction. Only one study, however, has previously evaluated recombinant tissue plasminogen activator (rt-PA) as a locking solution for primary prevention of catheter failure rather than as a treatment of existing/suspected thrombosis. In addition, line-related sepsis constitutes up to 75% of deaths from infection among this population,2 and infection risk increases in the presence of partial or complete thrombosis.3

In this month's literature watch, we review a study by Hemmelgarn and colleagues designed to evaluate whether rt-PA administered once weekly as locking solution in substitution for the scheduled dose of heparin is superior to the current customary care of locking catheters with heparin only after each dialysis session, in preventing catheter malfunction (primary outcome) and infection (secondary outcome).

The ideal study design to investigate whether one intervention is superior to another is a randomized clinical trial (RCT) with masked allocation and determination of outcomes. The study by Hemmelgarn and colleagues represents a well-designed RCT with random allocation of the patients and full masking (blinding) of group assignment, in which the investigators evaluate whether a protocol including 1 mg of rt-PA as a locking solution used once a week in place of 5,000 U of heparin is superior to heparin alone in preventing catheter malfunction and bacteremia.

Participants were recruited from 11 Canadian hemodialysis centers within 2 weeks of a newly inserted hemodialysis catheter. Patients were excluded from the study if their catheters at baseline failed to function adequately, defined as a blood flow of at least 300 mL/min during the dialysis sessions immediately prior to randomization. Masked random treatment allocation in a 1:1 ratio was performed centrally with the use of a permuted-block design stratified according to center and catheter status (first ever catheter for patient or not). Catheter malfunction and catheter-related bacteremia episodes by pre-defined criteria were documented. Analysis of outcomes was conducted on an intention-to-treat basis. Additionally, a cost assessment of maintaining future catheter patency was conducted.

Although the study was originally designed for an enrollment of 340 patients to ensure 80% power to detect approximately a 34% reduction in the incidence of catheter malfunction with rt-PA once weekly, due to lower than expected enrollment, only 225 patients were enrolled. Despite a smaller size, the study did show a significant reduction in the primary outcome because of a higher than expected baseline risk and higher than expected risk reduction with treatment. The strengths of the study included masked allocation, pre-defined inclusion and exclusion criteria, and clinically relevant pre-defined primary and secondary outcomes. Additionally, there were no significant differences in baseline characteristics of both groups, indicating the success of the blocked stratified randomization schema used. Finally, the study had excellent follow-up with robust ascertainment of the outcomes. The single limitation of this study (smaller than expected study size) does not invalidate the overall findings but does reduce the precision of the final estimates of effect size or benefit.

The current study by Hemmelgarn and colleagues demonstrates that the use of rt-PA as a locking solution for central venous catheters in place of heparin for one out of every three hemodialysis sessions decreases significantly the incidence of hemodialysis catheter malfunction and bacteremia. Hemmelgarn and colleagues demonstrated a reduction in events of hemodialysis catheter malfunction of 14.8%. If this reflects the true absolute risk reduction, than a number needed to treat (NNT) of seven patients to prevent one catheter malfunction episode can be expected. Similarly, locking hemodialysis catheters once weekly with rt-PA reduced episodes of catheter-related bacteremia by 0.97 episodes per 1,000 patient days, resulting in an NNT of approximately three patients treated for 1 year to prevent one episode of bacteremia.

Several clinically important questions are not addressed by this robust RCT. Most importantly, this study does not determine whether additional benefit or risk can be obtained with larger or more frequent rt-PA doses.4 If infection risk is reduced, a reduction in mortality might also be anticipated. The current study was insufficiently powered and of insufficient duration to determine whether the reduced incidence of the primary or secondary outcomes resulted in improved survival. There were five deaths among patients receiving customary care and three in the experimental rt-PA group. There was no difference between the groups in total or in serious adverse events. Adverse events were experienced by 70% versus 68.7% of patients in the rt-PA and heparin-only groups, respectively. Significant intracranial bleeding episodes were observed in the customary care (heparin-only) group only.

The authors also explored to some degree the cost differences of the two strategies and noted that for an incremental cost for rt-PA of $13,956 (Canadian dollars) one would expect to prevent one episode of catheter-related bacteremia. This apparent benefit did not include considerations of substantial potential improvements in patient quality of life, and it is likely that when expressed in terms of patient-centered outcomes such as QOL, the intervention might be even more clearly cost effective. This study does not address whether similar benefits might be obtained with other catheter locking solutions.5

As long as a substantial number of patients begins hemodialysis with a catheter for hemodialysis access, complications attributable to the catheters will remain significant. The current study provides compelling evidence that a strategy using once weekly rt-PA in place of heparin as a locking solution, which should reduce partial and/or total catheter thrombosis, will result in improved catheter function and reduced risk of infection. Although this strategy is likely to result in a favorable cost effectiveness, the costs of rt-PA will largely be borne by the dialysis units, the benefits will accrue almost entirely to the patient, and therefore, it should be anticipated that there will be additional disincentives for the adoption of this potentially beneficial therapy in an era of bundled payments for dialysis.