Hemodialysis Catheters in Infants: A Retrospective Single-Center Cohort Study
Take away pointIn infants, HD-CVCs require technical modifications and multiple maintenance procedures. Occlusive thromboses are frequent. Non-tunneled HD-CVCs are associated with higher catheter-related infection and may have higher thrombosis rates compare to tunneled HD-CVCs.
Ma GMY, Ventura LM, Amiribadi A, et al. Hemodialysis Catheters in Infants: A Retrospective Single-Center Cohort Study. J Vasc Interv Radiol. 2020;31(5):778-786.
Click here for abstract
Single-center, retrospective cohort study
No reported funding
The Hospital for Sick Children, University of Toronto, Toronto, ON
Children suffering from end-stage renal disease (ESRD) require renal replacement therapy (RRT) or transplantation to survive. While renal transplantation offers the best prognosis, it is not offered to children under the age of 24 months or weight of 10 kg. Children under the age of 2 years can receive RRT via peritoneal dialysis (PD) or hemodialysis (HD), with PD being the most commonly used due to several known benefits over HD. HD must be resorted to, however, when PD fails or is contraindicated. This must be done with central venous catheters (CVCs) in this population due to problematic nature of arteriovenous fistulas in children under 10 kg. The placement and maintenance of CVCs for HD in infants has not been well-studied. This study aimed to evaluate outcomes and challenges in CVC insertion and maintenance in these patients.
In this retrospective cohort study, 29 patients who had undergone initial CVC placement for HD before the age of 1 year were identified over a 14-year period (2002-2016). Both tunneled and non-tunneled CVCs, as well as both upper and lower CVCs were included. Non-IR-placed CVCs as well as CVCs placed for indications other than HD were excluded. CVC-related adverse events and dialysis sessions were assessed through 1 year of life and overall outcomes were assessed until December 2017 or death. Measured procedural details included procedure outcome, modifications, and complications. Postprocedural details included number of dialysis cycles, HD sessions, device service intervals, total access site service interval, CVC maintenance procedures, adverse events, and status at final review.
Out of the 29 patients analyzed, 13 (45%) were neonates when RRT was initiated. 10 patients received initial RRT via PD before switching to HD while 19 started with HD. 49 CVCs were placed in these patients (15 non-tunneled and 34 tunneled). Modifications were used in 51% of all catheter placements (33% of non-tunneled and 59% of tunneled CVCs) to allow for proper tip placement. For upper venous system CVCs, these modifications included suturing the catheter to the patient’s scalp behind the ipsilateral ear (n=4) and suturing to the scalp with placement of CVC tip in the IVC (n=1), trimming the tip to reduce distance between staggered ends (n=17), and tip trimming with catheter pullback and burying the cuff in a caudal extension of the tunnel (AKA “T-incision,” n=3). Technical success of modifications was 94% with no statistical difference between catheter types.
Mean number of dialysis cycles was 2.3 (range 1-6), with means of 4.1 sessions/catheter for non-tunneled CVCs and 49 sessions/catheter for tunneled CVCs. 26 (53%) CVCs required IR maintenance procedures, including 1 non-tunneled and 25 tunneled CVCs in 13 patients. Procedures included 25 contrast evaluations, 6 re-suturings, 1 repositioning of a non-tunneled CVC, and 13 catheter exchanges. Exchanges were done due to fibrin sheath formation (n=3), CVC fracture (n=2), malposition/dislodgement (n=7), or CVC disfunction (n=2). Difference in IR maintenance rates/1000 catheter-days was not significant between catheter types (p=0.454). Non-tunneled catheters had significantly decreased mean number of sessions/catheter (p < 0.002), dwell time (p < 0.004), primary device service interval (p < 0.002), and total access site interval (p < 0.001).
A total of 6 intraprocedural adverse events occurred, including 3 episodes of prolonged bleeding, 2 patients requiring second puncture, and 1 episode of profound hypotension of uncertain etiology after CVC insertion. 5 catheter-related bloodstream infections (CRBSIs) occurred in 4 patients, including 4 tunneled and 1 non-tunneled CVC. Infection rate was 1.03 infections/1000 catheter-days, with non-tunneled CVCs having a rate of 9.25 infections/1000 catheter-days and tunneled catheters having a rate of 0.83/1000 catheter-days (p < 0.02). Symptomatic thrombosis was identified in 10/29 patients (34%) and 12/49 catheters (24%), with 8 jugular venous thromboses and 4 femoro-iliac venous thromboses. 8 were occlusive and 4 were non-occlusive. An additional 4 patients experienced 4 asymptomatic occlusions discovered incidentally during CVC placement. 2 patients with upper venous system thrombosis developed chylothorax after CVC placement, associated with 1 death at 2 years of age. There was no difference in fibrin sheath or thrombosis incidence between trimmed and non-trimmed catheters (p=0.422). Patients with tunnel extensions were associated with an increase in cuff exposures compared to non-extended tunnels (p = 0.011).
10 patient deaths occurred during infancy, of which none were due to catheter-related complications. 3 deaths were defined as late deaths, occurring at 16 months (renal failure), 2 years (chylothorax/plastic bronchitis/vascular stenting), and 10 years (transplant complications) of patient age. The chylothorax-related death was attributed to a delayed complication of catheter placement. 19 patients survived past infancy, with 8 surviving past organ transplant (53%), 2 remaining on HD, 3 receiving RRT via PD, and 1 fully recovering renal function. 2 patients were lost to follow-up.
This investigation sought to identify procedural particularities and catheter-related outcomes in patients undergoing CVC placement for HD prior to 1 year of age. The mortality, rate of renal transplantation, vascular occlusion rate, and CRBSI rates were all within expected ranges. Technical success in cases requiring modification was high and did not differ with catheter type. Interestingly, the practice of trimming the tip of the catheter to improve positioning was not associated with an increase in incidence of thrombus or fibrin sheath formation despite the theoretical risk of a rougher tip that may lead to turbulent venous flow. The suturing of non-tunneled catheters to a patient’s scalp for positioning was found to be uncomfortable for patients and may have resulted in loss of sutures, leading to change in tip position in some cases. Other factors that affected correct tip position included changes in patient size with rapid growth and decrease in edema due to dialysis. Non-tunneled CVCs had higher infection and thrombosis rates, but the difference in thrombosis rate was not statistically different compared to tunneled CVCs per 1000 catheter-days. Both catheter types had a statistically similar rate of IR maintenance procedures.
The retrospective nature and small sample size limited this study. A larger study population would have allowed for multivariate analysis and may have elucidated a possible difference in thrombosis rates between catheter types. The loss of follow-up of 2 patients may also have affected the results. A larger study featuring multiple centers and multivariate analysis may draw more concrete conclusions. In terms of CVC modifications, the authors presented compelling evidence for the need of new catheter options for this population, including more flexible non-tunneled catheter options to allow for thoracic versus cephalic attachment and catheters that eliminate the need to trim the tip to ensure proper tip position.
4th Year Medical Student
Oregon Health & Science University School of Medicine, Portland, OR