Dr Josette Pengloan
University of Tours, France
Central venous catheter (CVC) is a major and vital route for haemodialysis. It allows haemodialysis treatment to be started in the absence of an arteriovenous fistula. As a consequence of the ageing population and predialysis care policies, both the incidence and prevalence of CVC use for haemodialysis are growing worldwide.1 At least 23% of haemodialysis (HD) patients used a catheter in the UK, Belgium, Sweden, Canada and the USA in 2005 to 2007 and this proportion is increasing. This figure is higher than the recommended level of less than 10% CVC prevalence.2 In nearly half of the countries, 23% to 73% of patients initiated dialysis with a catheter.
The higher risks associated with catheter use are related to both severe infection and dysfunction rates. CVC-related infection is a life-threatening complication with severe manifestations: bacteraemia, distant abscesses and a chronic inflammatory state appear to contribute to increases in both overall and cardiovascular mortality.
CVC obstruction is more frequent than infection. In the Strategic HealthCare Programs National Database,3 totalling 50,470 patients and 2.83 million catheter days, the catheter dysfunction rate observed was four-fold higher than the bloodstream infection rate (0.83 and 0.19 per 1,000 catheter days respectively). In addition, thrombotic occlusion occurred in 28% of patients. The use of CVC in haemodialysis results in a decrease in dialysis efficiency, additional work for nurses and disruption of the dialysis schedules. In many cases the CVC has to be replaced, which increases the burden of the disease and the costs of haemodialysis.
According to the National Kidney Foundation/Disease Outcomes Quality Initiative, CVC dysfunction is defined as the “failure to attain and maintain an extracorporeal blood flow of 300ml/min or greater at a prepump arterial pressure more negative than -250mm Hg”.2 High venous pressure also limits the blood inflow. The other signs are difficulty in aspirating the contents of the lumen before connecting the lines, frequent arterial or venous pressure alarms during the session and inadequate dialysis circuit blood flow. When necessary, nurses try measures such as aspirating heparin and flushing with saline followed by gentle but firm manoeuvres to reverse the blockage. Reopening the lines can sometimes be impossible. In every case intraluminal fibrinolytics have to be used to attempt to obtain an adequate circuit blood flow.
Whatever the initial results of the manoeuvres, the cause of the dysfunction has to be carefully and actively investigated with a view to undertaking the appropriate treatment and prevent repetitive dysfunction. Kinking or misplacement of one of the lines (see Figure 1), can be diagnosed by chest X-ray and can be corrected by CVC replacement, if caused by kinking, or by repositioning in case of misplacement of the tip. The tips of the long-term tunnelled CVCs must be positioned in the right atrium and not in the vena cava or at the junction between vena cava and atrium. Sometimes, even with apparently well-positioned CVCs, imaging of the catheter by injection of contrast media through the ports can reveal other types of CVC misplacement (for instance, the tip lying against the wall of the atrium, near the valve or in the ventricle); this requires CVC repositioning. Thrombus accumulation (see Figure 2), fibrin sheath formation (see Figure 3) and stenosis of the adjacent vessel are the other frequent causes observed.
The fibrin sheath covering the CVC extremity can develop within a few days after CVC placement,4 and was present in more than 70% of refractory CVC dysfunction.5 Its composition changes over time. At the early stage it is composed of a mixed cellular and non-cellular covering consisting of smooth muscle cells, thrombus and areas with endothelial cell populations. Later on the sheath becomes less cellular and is composed of more collagen.4 Histological changes occur in the adjacent vein wall. Short-term catheters were associated with foci of local intimal injury with endothelial denudation and a layer of adherent thrombus. The second pattern, seen with long-term catheters, consisted of smooth muscle cell proliferation leading to vein wall thickening and focal areas of catheter attachment to the vein wall.6
Fibrinolytics can dissolve, at least partly, the fibrin sheath. Similar results have been reported with percutaneous and transcatheter high doses of urokinase (UK; 250,000U over four hours)7 or recombinant tissue plasminogen activator (rt-PA) (2.5mg rt-PA over a three-hour period).8
Nevertheless, with either method only half of treated CVCs will maintain function for a long period. CVC exchange, CVC exchange and balloon disruption, and CVC stripping by the femoral route, seem to be equivalent in terms of immediate technical success but disruption of sheaths by angioplasty balloon seems to result in more durable catheter patency.5,9
The progression over time of the composition of fibrin sheaths could explain these observations. Fibrinolytics could dissolve thrombus and proteins of ‘young’ fibrin sleeves, and the other procedures would be more effective in ‘old’ fibrin sleeves and could be used successively.
A close relationship between the presence of thrombus or fibrin sheath and CVC colonisation can be easily explained. Fibrin adhering to the CVC lumens or at the tips is a nutrient for bacteria and the risk of introduction of pathogens into the lumen of the CVC is increased during the numerous manoeuvres performed to clear a clotted CVC. It follows that a decrease in the incidence of thrombotic events should reduce the CVC-related infection rate. In an experimental study the incidence of catheter colonisation decreased when the amount of fibrin within the pericatheter sheath decreased.10 In a study comparing prophylactic heparin 100IU/ml and UK 5,000IU/ml every two weeks as a lock solution in children receiving chemotherapy through tunnelled catheters, UK administration resulted both in fewer occlusive events than heparin and a 1.4-fold decrease in the infection rate.11
Another explanation could be that sodium heparin, which is the most widely used catheter lock solution, stimulates Staphylococcus aureus biofilm while tissue plasminogen activator had little effect on S. aureus biofilm formation.12
Handling of lytic agents
Thrombolytic agents dissolve blood clots through activation of plasminogen. The currently available lytic agents are effective and safe. UK is a plasminogen activator, isolated from human urine or obtained from human neonatal kidney cells grown in tissue culture. It is indicated for occluded CVCs in Europe. Alteplase is the recombinant analogue of human tissue plasminogen activator (r-tPA) secreted from normal vascular endothelium. It is the only thrombolytic agent marketed and FDA-approved for restoration of function of occluded CVCs in the USA. Other tissue plasminogen activators can be used such as reteplase or tenecteplase.
Despite the fact that fibrinolytics have been used successfully for more than two decades to restore patency of an occluded CVC, there has been no standardisation of drugs, dose, concentration, regimen, modalities of injection (instillation or infusion) or duration of infusion. Many regimens with good immediate results using either UK or rt-PA have been published. For UK these include: lock solution; bolus of 5,000U; infusion of 20,000IU/hour over six hours; infusion of 125,000IU over two hours through one lumen followed by 125,000IU through the other lumen; and a high dose of 250,000IU of UK infused over three hours into the venous chamber during dialysis. This latter regimen was associated with an 81% success rate after the first infusion and 99% after the third infusion.13 In contrast, push-protocols (bolus)14 that have generally been recommended for rt-PA were usually proposed. A number of infusion protocols have also been reported for rt-PA including: 2mg/h over four hours; 2.5mg/h over three hours through each port; and 5mg/h over three hours through each port. However, it is difficult to evaluate the effectiveness of these protocols in the absence of criteria for the severity of occlusion which would allow meaningful comparisons. It is also difficult to compare the effectiveness of alteplase versus UK – 1mg of alteplase was described as “likely equivalent in thrombolytic potency to 36,000U of urokinase”, in one study.15 So it is not surprising that comparisons between low doses of UK 5,000IU versus alteplase 1mg show better results for alteplase.
We perform a simpler modality of infusion: 100,000U of UK after reconstitution is diluted with 20–30ml saline and infused using a syringe driver over 20–30min through each of two lumens before a dialysis session. At the end, nurses aspirate, flush with saline and connect for dialysis. If the presence of a fibrin sheath is suspected, the infusion is repeated before the next three sessions or more; if there is incomplete restoration, imaging is performed. This is a high dose, relatively rapid infusion that does not disrupt the organisation of dialysis.
In summary, CVC obstruction is a central issue in the field of nephrology, contributing to the high morbidity and mortality of dialysed patients. Careful attention to the tip position, better understanding of the mechanisms of obstruction, and prompt and aggressive investigation and treatment are the keys to reduce the CVC-related complication rate.
Figure 1. Misplacement of one line (arrow) of a tunnelled twin central venous catheter
1. Ethier J et al. Nephrol Dial Transplant 2008;23:3219–26.
2. 11 DOQI. Am J Kidney Dis 2006;48;S248–57.
3. Moureau N et al. J Vas Interv Radiol 2002;13:1009–1916.
4. Forauer A et al. Radiology 2006;240:427–34.
5. Oliver M et al. Clin J Am Soc Nephrol 2007;2: 1201–06.
6. Forauer A, Theoaris C. J Vasc Interv Radiol 2003;14:1163–68.
7. Gray R, Gupta A. J Vasc Interv Radiol 2000; 11:1121–29.
8. Savader S et al. J Vasc Interv Radiol 2000;11: 1131–36.
9. Jane d’Othée B et al. J Vasc Interv Radiol 2006;17:1011–15.
10. Keller J et al. Crit Care Med 2006;34:1450–55.
11. Dillon P et al. J Clin Oncol 2004;22:2718–23.
12. Shanks R et al. Nephrol Dial Transplant 2006;21:2247–55.
13. Twardowski Z. Nephrol Dial Transpl 1998;13:2203–06.
14. Beathard G. Semin Dial 2001;14:4415.
15. Clase C et al. J Thromb Thrombolysis 2001;11:127–36.