This site is intended for health professionals only!
Gavin Whelan, BPharm, Dip Clin Pharm
Principal Critical Care Pharmacist,
Guy’s and St Thomas’ NHS Foundation Trust,
Cathrine McKenzie, PhD, MRPharmS
Guy’s and St Thomas’ NHS Foundation Trust,
Continuous renal replacement therapy (CRRT) was first described in 1977 for the treatment of diureticunresponsive fluid overload in an intensive care unit. Today, CRRT is used widely within critical care and has completely replaced intermittent haemodialysis (IHD) in many centres.
CRRT is normally provided over a period of several days as opposed to the several hours that IHD is provided for. This places less demand upon the patient’s cardiovascular system and makes it a more suitable treatment option in haemodynamically unstable patients.1
CRRT is most commonly used as a supportive treatment in acute kidney injury (AKI). AKI is frequently encountered within critical care, with up to 20% of patients experiencing at least one episode during the course of their illness. It has a wide-ranging associated mortality that spans from 10% to 80% depending upon the population studied.2
The treatment of AKI will depend upon whether the cause is pre-renal, intrinsic or post-renal, but supportive therapy with CRRT is required by many patients within this group. CRRT has several other indications for use, both with and without the presence of AKI. CRRT should be considered in patients with:
There is also evidence to suggest that CRRT at high exchange rates can remove inflammatory mediators such as cytokines in patients with severe sepsis.3
Theory and types of CRRT
The umbrella term ‘continuous renal replacement therapy’ encompasses several different modalities of renal replacement. These are: continuous veno-venous haemodialysis (CVVHD), continuous veno-venous haemofiltration (CVVH) and continuous veno-venous diafiltration (CVVHDF). Knowledge of the different modalities is essential for the critical care pharmacist in order to assess drug clearance.
During CRRT, the patient is attached to the haemofilter via two catheters. Blood flows from the patient through one, passes through the haemofilter and is then returned to the patient’s circulation via the second catheter. Replacement fluid can be added before the blood reaches the filter (pre-dilution) or added after the filter (post-dilution). See Figure 1.
The filter on most haemofiltration machines consists of a column containing hollow semi-permeable fibres through which the patient’s blood passes. Pore size, the number of pores and the thickness of the membrane determine permeability of the filter fibres.
Larger pores or a higher number of pores will increase solute removal and thinner membranes decrease the resistance to solute removal by decreasing the distance the solute needs to travel. The surface area of the filter will also affect solute removal as larger filter columns with a higher surface area – due to an increase in the number of fibres they hold – allow greater removal.
Solute removal by the CRRT is dependant upon the solute size, so smaller molecules will pass freely through the filter membranes while larger molecules like proteins are retained within the blood as they are too large to pass through the pores.
The basic principles by which haemofilters remove fluid and impurities are convection or diffusion or a combination of both depending on the CRRT modality used.
How CVVHD works
In CVVHD, clearance is predominately achieved through diffusion. A concentration gradient occurs across the filter, as the patient’s blood will contain a higher concentration of dissolved solutes than the surrounding dialysis fluid. Solutes then diffuse across the filter membrane from the more concentrated blood into the less concentrated dialysis fluid.
As the concentration of solutes in the dialysate increases, the rate of diffusion decreases. In order to maintain the rate of clearance, the concentration gradient needs to be maintained. This can be achieved by increasing the rate of blood flow to maintain a high concentration on the blood side of the filter or increasing the flow rate of dialysis fluid to maintain a low concentration on the other side of the filter.
During CVVHD, the dialysis fluid runs through the filter in the opposite direction to blood flow, which also helps to maintain an adequate concentration gradient across the length of the filter.
Diffusion can also be a two-way process. For example, many dialysis fluids contain bicarbonate at a higher-thanplasma concentration and, as a result, bicarbonate will diffuse from the dialysis fluid into the plasma.4 In CVVHD, if an exchange rate of 2L/hr is selected, then two litres of dialysis fluid will flow into the column each hour and two litres will be removed. In CVVH, solute clearance is achieved through convection. A transmembrane pressure gradient is created across the filter. Blood flows through the filter and a pump creates negative pressure on the other side of the membrane. This pressure gradient forces water to move from the blood plasma across the membrane and, as it moves, solutes are drawn across with it.
This method allows removal of larger molecules that are too large to cross the membrane by diffusion alone. However, it can also lead to excessive clearance of smaller molecules such as sodium and potassium, so close monitoring of the patient’s electrolyte levels is essential.
As large amounts of water can be removed from the patient’s plasma during this process, additional fluid needs to be added into the blood supply to prevent dehydration. In CVVH, if an exchange rate of 2L/hr is selected, then no dialysis fluid runs through the column, but a pump will create sufficient negative pressure to draw two litres of ultrafiltrate from the blood plasma each hour.
CVVHDF is a combination of both CVVHD and CVVH and therefore achieves clearance through both diffusion and convection. If an exchange rate of 2L/hr is selected (1L/hr of CVVH and 1L/ hr of CVVHD), then one litre of dialysis fluid will run through the column each hour and a pump will create sufficient negative pressure to drawn an additional one litre of ultrafiltrate from the blood plasma each hour.5,6
Heparin and epoprostenol
In common with all extracorporeal circuits, CRRT circuits require anticoagulation. Inhibiting clot formation helps to extend the life of the circuit, which reduces costs and reduces the patient’s exposure to blood products, as the blood within the circuit is lost each time a circuit clots.
The amount lost will vary according to the circuit and filter size, but can be as much as a unit of blood. The aim is to anticoagulate the circuit while minimising the amount of anticoagulant reaching the patient’s systemic circulation.
Unfractionated heparin remains the anticoagulant of choice in most situations. A continuous infusion of low-dose heparin is infused into the circuit, some of which will be removed by the filtration process – so, in theory, very little should be present in the blood returning to the patient. Despite this, activated partial thromboplastin time (APTT) monitoring is still required and can be used to adjust infusion rates if the patient starts to exhibit systemic anticoagulation. Infusions of prostacyclin analogues such as epoprostenol can be used on their own to inhibit platelet activation or added into circuits that begin to clot despite heparin.
Due to their short half-life, prostacyclin analogues should have limited systemic effects and therefore may be preferred to heparin in patients who are at risk of or are actively bleeding. However, the costs of prostacyclin infusions are significantly more than unfractionated heparin and the anticoagulant effect is probably much less effective.
Citrate anticoagulation is a relatively new technique that has the benefit of only anticoagulating the circuit and not the patient. Sodium citrate is infused into the blood as it leaves the patient, this chelates any ionised calcium in the blood, which inhibits coagulation, and then calcium is infused back into the blood as it returns to the patient, restoring the clotting cascade.
This makes it the ideal choice in patients where avoidance of systemic anticoagulation is essential. However, citrate anticoagulation is still very expensive and not all the products required are commercially available. The costs may be offset against longer circuit life and reduced patient blood transfusion.
Caution should be used in patients with hepatic failure, as they may be unable to metabolise the citrate, leading to accumulation and toxicity. In patients who have a coagulopathy or are systemically anticoagulated for another indication, additional anticoagulation of the haemofiltration circuit is not normally required.
In addition, patients who are receiving drotrecogin alfa-activated (DAA) do not normally require CRRT anticoagulation.
Many factors need to be taken into account when assessing drug clearance via CRRT. Unfortunately, this remains an area where the evidence base is limited and often requires an understanding of the basic chemistry, pharmacokinetics and dynamics of the drugs involved.
Although most drug removal will occur through convection and/or diffusion, some drug removal will result from membrane interactions that are membrane- and drug-dependent. Many filters are made of synthetic hydrophobic materials with a high adsorptive capacity, such as sulfonated polyacrylonitrile, which have a tendency to bind negativelycharged proteins.
As the blood passes through the filter, polycationic drugs (e.g. aminoglycosides) can bind to these negatively-charged proteins, removing them from the circulation. The clinical significance of this is unknown, but it is likely that this process becomes rapidly saturated and therefore should have minimal impact upon drug clearance over the life of the filter.7
During CVVH, drug clearance will be mainly through convection. Only unbound drug will pass through the membrane pores, so protein binding is an important factor, but it should be noted that protein binding changes during critical illness. Therefore, drugs that are highly protein-bound – for example, amiodarone and midazolam – are less likely to be removed by CVVH.
The sieving coefficient
CVVH usually utilises highly permeable membranes that have a high cut-off value of between 20,000 and 50,000 Daltons. As most drugs are below this size, molecular weight has little impact upon clearance. As such, drug clearance is relatively easy to calculate and this is expressed mathematically in the sieving coefficient (S):
Replacement fluid can also affect drug clearance during haemofiltration. If replacement fluids are added to the blood before passing through the filter (predilution), then the blood is more dilute as it passes through the filter and drug clearance will be lower than that seen when replacement fluid is added after the blood has passed through the filter (postdilution).
Drug clearance during CVVHD occurs mainly through diffusion. This process is dependent upon molecular weight, concentration gradient, membrane porosity and surface area of the filter. In CVVHD, the countercurrent flow of dialysate is always much smaller than blood flow and, as a result, complete equilibrium occurs between the dialysate and blood serum so dialysate fluid leaving the filter will be 100% saturated with small easily diffusible solutes (<500 Daltons). Therefore, clearance of small easily-diffusible drugs will be equal to dialysate flow rate. Increasing molecular weight decreases the speed of diffusion, so the concentration of larger molecules in the dialysate will not reach equilibrium with the plasma concentration. If dialysate flow rate is increased, the time available for diffusion to take place is shorter and less diffusion will occur, resulting in less clearance.7
From the number of variables influencing drug clearance mentioned above, the difficulties in undertaking research in this area can be appreciated and it should be remembered that, during CVVHDF, all of these factors need to be considered, making estimates of drug clearance even more complex.
Key points for predicting drug clearance in CRRT
CRRT is a rapidly-changing area of therapy within critical care and the critical care pharmacist needs to maintain a good working knowledge of the principles involved to understand the factors affecting drug clearance and ensure therapy is both safe and efficacious for the patient.
1. Prowle JR et al. Nat Rev Nephrol Sept 2010;6:521- 29
2. Lewington A et al. The Renal Association Clinical Practice Guidelines. Module 5: Acute Kidney Injury 2010, www.renal.org
3. Sylvester W. Am J Kidney Dis. Nov 1997;30(5 Suppl 4):S38-43
4. Hilton PJ et al. Q J Med 1998;91:279-83
5. Bellomo R et al. Crit Care 2000;4:339-45
6. Forni LG et al. N Engl J Med 1997;336:1303-9
7. Ronco C, Bellomo R, Kellum JA (Eds). Critical Care Nephrology. Canada: Saunders Elsevier; 2009