Introduction

Acute kidney injury (AKI) is a debilitating condition, associated with high morbidity and mortality in both veterinary and human patients.^1,2 Metabolic disturbances, fluid imbalance, electrolyte abnormalities, hypertension, and involvement of multiple organ systems are commonly present in dogs with severe AKI.³ Diagnostic investigation is aimed to identify the inciting cause and to eliminate it when possible as well as addressing complications.

Severe renal dysfunction leads to accumulation of uremic toxins affecting all body organs, thus providing only a narrow window of opportunity for therapeutic intervention. Conventional management is limited and often insufficient, and animals might succumb to the consequences of AKI before kidney function has improved. In such cases, dialytic intervention is indicated, extending the window of opportunity for renal recovery.

There are several modalities for delivering renal replacement therapy, including, continuous renal replacement therapy (CRRT), intermittent hemodialysis (IHD) and peritoneal dialysis.⁴ IHD is a very effective renal replacement modality, in which solutes are removed mostly by diffusion and excess fluids are removed by ultrafiltration, typically over 4 to 5 hours.⁵ However, IHD is currently available at a limited number of specialty centers equipped with an IHD unit, and requires expensive equipment and maintenance (eg, water purification system) as well as high level of expertise.

As implied by its name, CRRT is delivered continuously, resulting in a slow and gradual removal of uremic toxins. This modality enables the utilization of both diffusive and convective principles and allows several treatment modalities: slow continuous ultrafiltration (SCUF; used primarily for removing excess body fluid), continuous venovenous hemofiltration (CVVH; clearance is achieved by convection), continuous venovenous hemodialysis (CVVHD; clearance is achieved by diffusion), and continuous venovenous hemodiafiltration (CVVHDF; combines both diffusive and convective clearance). SCUF is aimed at removing excess body fluid (ultrafiltrate) and is not used for removal of uremic toxins. CVVH is a purely convective modality in which the filtered fluids, along with the dissolved uremic toxins, are replaced with a replacement solution, either before or after the filter (pre- or post-filter configuration, respectively). CVVHD is a diffusive therapy, similar to IHD, where solutes diffuse based on their concentration gradient between the blood and the dialysate. Solutes present in high concentrations in the blood, such as uremic toxins, diffuse across the semipermeable membrane to the dialysate, and solutes present in a higher concentration in the dialysate (eg, bicarbonate) diffuse into the blood. CVVHDF combines both diffusive and convective properties. Although CRRT is used for treating a variety of life-threatening conditions, the main indication for CRRT in human and animal patients is severe AKI.⁶ One of the advantages of CRRT over IHD is a substantially longer treatment time allowing for a more gradual solute and fluid removal, thereby decreasing the risk for dialysis disequilibrium syndrome (DDS) and intravascular contraction, while the patient is closely monitored in an intensive care setting. Despite the growing use of CRRT platforms in veterinary practices worldwide, very limited number of studies, mostly case reports, book chapter, and review papers, described the use of this treatment modality in dogs and cats with naturally occurring disease.^7–9 More importantly, unlike in human medicine where guidelines for CRRT have been established,^10,11 currently, there are no standardized, outcome based CRRT prescription protocols in veterinary medicine.

In humans, CRRT is typically delivered over days, until kidney function improves or the patient is transitioned to IHD. This prolonged CRRT treatment duration impose a huge financial burden and is extremely labor intensive, therefore this approach is not applicable in many veterinary clinical settings and shorter treatment times are often delivered.

We hypothesize that CRRT is a safe and effective therapeutic modality for dogs with severe AKI. The aims of this study were to describe a relatively large cohort of dogs managed by CRRT, to assess CRRT safety and to investigate the relationship between a prescribed CRRT target dose and the urea reduction as well as dose delivered in dogs with naturally occurring AKI.

Materials and Methods

Results

Treatment dose, adequacy and efficacy

Of the 48 CRRT treatments, 2 (4%) were delivered as CVVHD and the other 46 (96%) as CVVHDF. Extracorporeal circuits with priming volumes of 97 mL, 60 mL, and 193 mL were used in 31 (64%), 13 (28%) and 2 (4%) of the treatments, respectively. In 2 treatments (4%) of dogs with severe inflammatory response (1 dog with heat stroke and another dog with severe pancreatitis), a 193 mL extra corporeal circuit, designed for both CVVHDF and cytokine removal was used. In 3 (6%) treatments the extracorporeal circuit volume was > 20% of the dog’s blood volume, therefore synthetic colloid (2 dogs) or packed red blood cells (1 dog) were used for priming the circuit. Pre or post filter replacement configurations were used in 42 (93%) and 3 (7%) of the treatments, respectively. The median blood flow rate during CRRT was 63 mL/min (IQR, 30 mL/min; range, 24 to 111 mL/min), corresponding to a median of 3 mL/kg/min (IQR, 2 mL/kg/min; range, 1 to 14 mL/kg/min). Ultrafiltration was prescribed in 23 (48%) of the treatments with a median ultrafiltration rate of 2.6 mL/kg/h (IQR 1.3 mL/kg/h; range, 0.6 to 6.6 mL/kg/h), over a median duration of 11 hours (IQR, 11 hours; range, 2 to 22 hours). Systemic heparinization and RCA were used in 24 treatments each (50%). Calcium gluconate 10% was used when RCA was selected as the anticoagulation method with an initial rate of 0.5 mL/kg/h (IQR, 0.3 mL/kg/h; range, 0.2 to 1.2 mL/kg/h). The median target dose for the entire treatment was 1 mL/kg/min (IQR, 0.4 mL/kg/min; range, 0.3 to 2.5 mL/kg/min). The median dose of the 1st (48 treatments), 2nd (33 treatments), and 3rd (29 treatments) segments was 0.9 mL/kg/min (IQR, 0.26 mL/kg/min; range, 0.3 to 1.26 mL/kg/min), 1.0 mL/kg/min (IQR, 0.32 mL/kg/min; range, 0.6 to 1.5 mL/kg/min) and 1.4 mL/kg/min (IQR, 0.48 mL/kg/min; range, 0.7 to 2.5 mL/kg/min), respectively, with a significant difference among the treatment segments (P < .001; Figure 1).

View Full Size

Target dose (A) and hourly urea reduction ratio (urea reduction ratio; B) among the treatment segments in dogs with acute kidney injury managed by continuous renal replacement therapy. Data are presented as box and whiskers. The box represents the 2nd and 3rd quartiles. The horizontal line within the box represents the median. The whiskers represent the range, and the circles indicate statistical outliers.

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.07.0294

• Download Figure
• Download figure as PowerPoint slide

Urea concentration and URR (at the early stages of the treatment) were the main indicators for treatment adequacy and were measured and calculated, respectively, by measuring urea before, during, and after (immediately prior to CRRT termination) the treatment. Urea concentration significantly (P < .001) decreased among segments (Figure 2). Median serum urea concentration at the end of CRRT was 67 mg/dL (IQR, 66 mg/dL; range, 21 to 206 mg/dL) (Table 1, Figure 2). The overall median URR was 76% (IQR, 30%; range, 11% to 92%) over a median treatment time of 21 hours (IQR, 9 hours; range, 3 to 32 hours). Twenty-four (54%) treatments had an overall URR of > 75%, and 10 (21%) treatments had an overall URR of > 85%. The median (linearly extrapolated) hourly URR for the entire treatment was 4% (IQR, 1%; range, 2% to 12%). The median Kt/V of the entire treatment was 2.34 (IQR, 1.9; range, 0.24 to 7.02), and was > 1.2 and > 2.0 in in 37 treatments (77%) and 29 treatments (60%), respectively. When excluding prematurely terminated treatments, 34 (100%) and 28 (85%) of the treatments had a Kt/V of > 1.2 and > 2.0, respectively.

View Full Size

Urea concentration among the treatment segment in dogs with acute kidney injury managed by continuous renal replacement therapy. Data are presented as box and whiskers. The box represents the 2nd and 3rd quartiles. The horizontal line within the box represents the median. The whiskers represent the range, and the circles indicate statistical outliers.

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.07.0294

• Download Figure
• Download figure as PowerPoint slide

The median hourly URRs of the 1st, 2nd, and 3rd treatment segments were 6.5% (IQR, 2.4%; range, 3.2% to 18.8%), 6.0% (IQR, 4.0%; range, 0.5% to 18.1%) and 5.5% (IQR, 3.0%; range, 1.5% to 7.7%), respectively, with a significant difference among the segments (P = .05) (Figure 1). Five dogs (all of which with urea < 600 mg/dL) had hourly URRs of > 10% at the first segment (range, 10% to 16%). There was a significant (P = .026) positive correlation between the CRRT target dose and URR (Figure 3). Additionally, the initial urea concentration was significantly (P = .018) albeit weakly (r = .26) correlated with the URR.

View Full Size

Scatter plot depicts the association between the dose and the hourly urea reduction ratio among the different segments. Note that the slope of the segments is decreasing in the second and third segment.

Citation: Journal of the American Veterinary Medical Association 261, 1; 10.2460/javma.22.07.0294

• Download Figure
• Download figure as PowerPoint slide

Discussion

The results of this study demonstrate that CRRT is safe and effective for management of severe AKI in dogs. The dose prescribed resulted in gradual reduction in urea concentration, well within the current veterinary guidelines.

In recent years, CRRT and hybrid therapies providing renal replacement therapy over an extended period of time, gained popularity in veterinary practice and their availability is increasing globally, particularly in the intensive care unit setting. In humans, CRRT is often delivered over several days, until the patient regains kidney function, or at least for 24 hours, before the patient is transitioned to IHD once stable enough.^16,17

Rapid solute removal during hemodialysis is a major concern, as it predisposes the patient to non-physiologic osmotic changes, resulting in fluid shift from the intravascular to the extravascular compartment, potentially leading to neurological signs, cumulatively known as DDS.¹⁸ In IHD, rapid osmotic changes are more likely, due to the relatively short treatment duration compared with CRRT and the risk for DDS, a potentially fatal complication, is considered higher.^19,20 Yet, when adhering to the published veterinary guidelines, IHD can also be delivered safely.¹⁸ CRRT is relatively expensive, labor intensive and requires prolong treatment duration compared with IHD. One of its advantages is slow and safe removal of solutes, thereby decreasing the risk for DDS. CRRT is also considered to have superior hemodynamic stability in human patients for several reasons.^21–23 Continuous, prolonged treatment allows slower fluid removal, enabling better patient tolerance to fluid removal while maintaining the intravascular volume, which is the main contributing factor for hemodynamic stability during hemodialysis. The slower and more gradual decrease in solutes concentration also negates the fluid shifts from the intravascular space, further decreasing the risk for hypovolemia. Based on the evidence from human patients, there is no specific modality that shows a survival benefit over the other²⁴; however CRRT is indicated in hemodynamically compromised patients (even though it does not improve the survival rate compared with other modalities).²⁵ Hemodynamic stability and the risk for DDS are very important considerations in veterinary patients, as small patients are more prone to these complications, and veterinary patients are often hemodynamically unstable prior to the initial treatment. This notion is supported by our results, indicating development of hemodynamic instability during CRRT in only 2 dogs (4%).

Veterinary guidelines for the delivery of safe IHD treatments were published decades ago,¹⁸ however, recommendations for CRRT prescription for veterinary patients are scarce.⁹ We therefore aimed to decrease urea concentration slowly, not exceeding the recommended maximal hourly URR. Safety and adequacy were assured by periodically measuring urea concentration and calculating the hourly URR. The initial median target dose for patients in this study was 0.9 mL/kg/min (range, 0.3 to 1.26 mL/kg/min) and was individualized based on the severity of uremia and the risk for DDS. In most dogs, the initial target dose was higher than the initial recommended dose for human patients (0.3 to 0.4 mL/kg/min) since the treatment duration prescribed herein was shorter (due to practical and financial considerations), thus we have aimed to maximize efficiency while providing a safe treatment.²⁶ Indeed, in only 5 treatments the hourly URR was slightly greater than the recommended in the veterinary IHD guidelines, and none presented clinical signs of DDS. Therefore, a median target dose of approximately 1 mL/kg/min might be considered appropriate for veterinary patients; however, each patient was considered individually for the risk of complications, and in highly uremic animals and small dogs, lower CRRT target dose is initially recommended to decrease the risk for DDS and hemodynamic instability. Consistent with the human guidelines,^26,27 frequent assessment of the prescribed target dose is recommended to assure that the URR is within the target, not only to assure safe urea reduction, but also to assess treatment adequacy. The treatment target time is also a factor that should be accounted for when planning the prescription. In this study, we aimed to deliver the treatment over about 24 hours, with an overall URR of > 75%, Kt/V of > 1.2, and a urea concentration that will minimize the risk for uremic complications until the next treatment, typically 24 to 48 hours later, if needed. This scenario is often different compared with human patients managed by CRRT, where treatments are usually longer, lasting several days^28,29 and so urea reduction can be done more gradually, thus an initial lower dose is adequate. This is an innovative approach which combines CRRT technique and the experience accumulated in veterinary IHD to provide a safe treatment, while considering both financial and technical limitations of most veterinary clinical settings that offer bedside extracorporeal treatments.

Solute removal in dialysis follows a first order kinetics, namely for a given clearance, the amount of solutes removed depends on the initial urea concentration. While the median target dose was increased between the 1st to the 3rd segment, from 0.9 to 1.4 mL/kg/min, the median hourly URRs decreased from 6.5% to 5.5% (Figure 1). A likely reason for this observation is inherent in the principles of urea kinetics; at a constant flow, the clearance of a given solute will remain constant (assuming there is no clotting of the filter), however, its removal rate will be proportional to its initial concentration. For example, a urea clearance of 20 dL/h in a patient with an initial urea concentration of 100 mg/dL will result in urea removal rate of 2000 mg/h (20 dL/h X 100 mg/dL), but it will increase to 4000 mg/h, for the same clearance, if the initial urea concentration is 200 mg/dL. Hence, to sustain a relatively constant URR and to meet the CRRT target, the dose needs to be increased proportionally to the gradual decrease in urea concentration. This is supported by the positive effect of initial urea concentration on the URR; the higher the initial urea is, the greater the URR will be. This approach of gradually increasing the treatment dose over the treatment course, resulted in a relatively constant URR throughout the treatment.

Other reasons for the apparent reduction in URR toward the last stages of the treatment relates to filter clotting resulting in decreased dialyzer efficiency, increasing pre-filter replacement fluid rate (hence increasing pre-dilution), and the narrowing difference between the amount of urea removed compared to the amount of urea generated. In health, urea elimination rate equals urea generation rate.³⁰ During hemodialysis, serum urea concentration is determined by the urea generation rate and its removal rate, which depends on the inlet urea concentration and urea clearance. As urea concentration is decreasing, its removal rate decreases as well, and at a certain point, it equals to the urea generation rate, resulting in an apparently ineffective treatment (based on URR and urea concentration), namely urea concentration does not change in face of treatment delivery (URR approaches 0%). Therefore, as the treatment progresses URR becomes an inappropriate measure of treatment adequacy and thus the Kt/V should be evaluated as well.

Toward the end of the treatment, consideration was also given to the added benefit of increasing the dose over financial costs and labor of the technical staff. Especially in large dogs, increasing the CRRT dose substantially increases treatment costs and labor. In such circumstances, one has to balance between a satisfactory urea reduction and treatment costs and labor.

The overall median URR in this study was 76%, and in approximately a quarter of the dogs it was higher than 85%. Currently available veterinary IHD recommendations suggest the URR for the first and second IHD treatment should not exceed 50% and 70%, respectively, to decrease the risk for DDS.¹⁸ Although the delivered treatments herein had higher total URRs for a single treatment, the prolonged treatment duration allowed their safe delivery, well within the recommended hourly URRs. In that sense, one CRRT treatment was equivalent to 2 to 3 IHD treatments. This might justify, at least partially, the extra financial cost of 1 CRRT treatment compared with 1 IHD treatment.

Urea kinetic modeling led to the development of several measures of dialysis adequacy, including single and double pool Kt/V, Equilibrated Kt/V, Kt/resting energy expenditure, Kt/ body surface area and Kt/ high metabolic rate organ.^31–34 Each of these measures of adequacy has its limitations and there is no consensus regarding the ideal method of assessing dialysis adequacy in general and in CRRT is specific. An accepted method to assess treatment dose in IHD is by calculating the Kt/V. This measure of dose adequacy is calculated by the urea dialyzer clearance multiplied by dialysis time and normalized for urea volume of distribution (Appendix). In this study, the median Kt/V of the entire treatment was 2.34 with a wide range of 0.24 to 7.02. This extensive range is a result of differences in treatment duration (including premature treatment termination). Practice guideline for humans with AKI recommends delivering a weekly Kt/V of 3.9 (Kt/V 1.2 to 1.4, 3 times weekly), when using an IHD platform.²⁶ According to these guidelines, some of the dogs in our study had a Kt/V greater than the weekly recommended dose, while others did not reach the desired target for a single intermittent treatment, mostly due to complications and premature treatment cessation. Current human literature suggest that Kt/V is not a suitable measure for CRRT adequacy, and other methods, however limited as well, should be applied instead. The main arguments against the use of Kt/V for dose assessment are its lack of reliability in patients with AKI (as oppose to CKD), and the inaccuracy of clearance estimation in CRRT.^35–38 These arguments will need to be validated in veterinary patients as well; however, it seems prudent to take into consideration the human experience and avoid using the Kt/V as a sole measure of treatment adequacy in dogs with AKI managed by CRRT. On the other hand, until other quantification methods of CRRT adequacy are validated, Kt/V is still being used as the main indicator for clearance in CRRT and is a simple and cost- effective method for assessment of treatment efficacy, as long as its limitations are taken into account.³⁷ In this study both URR and Kt/V were used to assess treatment efficacy and overall adequacy. When prescribing the treatments herein, the IHD veterinary experience was combined with the human guidelines for CRRT while adjusting these to the clinical setting. During the first hours of treatment URR was measured periodically to assure a safe reduction in urea concentration (as the same prescribed dose for a given urea concentration might result in different URRs among patients), and to evaluate the treatment efficacy. As the treatment progresses the URR becomes less relevant as a marker of treatment efficacy, because urea removal rate approaches urea generation rate and the URR approaches 0% despite continuous urea clearance. Therefore, Kt/V is a more appropriate measure of the overall treatment efficacy. The limitation of URR as a measure of treatment efficacy in CRRT is a reason for the relatively low proportion of treatments with an overall URR of > 75% and the substantially higher proportion of the treatments with a Kt/V of > 1.2

Complications of CRRT are considered common in human patients and include electrolyte and acid-base disturbances (eg, hypocalcemia, hypernatremia, metabolic acidosis and alkalosis), hypotension, vascular access related complications, nutrition losses, altered drug pharmacokinetics, inadvertent fluid imbalances and extra-corporeal circuit complications (eg, filter clotting).³⁹ Our results demonstrate a relatively low incidence of life threatening complications, and fewer complications in comparison to human reports, most likely due to the relatively longer duration of CRRT in human patients, increasing the risk for the aforementioned complications.^39–41 Clotting of the extra-corporeal circuit was one of the major complications in our study, occurring in 13% of the treatments, and accounted for the termination of almost a third of the prematurely terminated treatments. These results are comparable to the human literature, where the most common cause for circuit downtime is clotting.^40,41 Despite this complication, treatment time of prematurely terminated treatments due to clotting was rather long (median, 16 hours). In accordance with human studies, there was no difference in clotting rates for heparin and citrate anticoagulated circuits, however this needs to be further validated in a larger scale study.⁴¹

Bleeding was an infrequent (2%), however fatal complication in one dog in our study that was systemically heparinized. Bleeding tendency is common in patients with AKI,⁴² and the decision regarding anticoagulation method should be thoroughly considered. Bleeding is also uncommon but is potentially life-threatening in human patients managed by CRRT.^41,43 Similar to the dog that presumably died from uncontrolled bleeding, most bleeding complications in humans are associated with heparin anticoagulation. Despite its low occurrence, this catastrophic complication emphasizes the need for rigorous bed-side ACT monitoring and provision of adequate and quick response in any case of bleeding or excessive anticoagulation. Ionized hypocalcemia was common in dogs anticoagulated by RCA, consistent with the most common electrolyte disorder in human patients receiving citrate anticoagulation.⁴⁴ Although this complication was documented frequently in our study, clinical evidence of hypocalcemia was rarely seen. Additionally, pre-existing hypocalcemia was documented in some of the dogs in this study, as commonly documented in AKI patients.^45,46

Seven treatments were prematurely terminated due to technical problems. Four treatments were ended prematurely due to excessive patient fluid loss/gain, which is an integrated feature of the machine monitoring system, designed to prevent unintended fluid imbalance. This complication might occur, for example, if a bag is being replaced without an appropriate indication, leading to an unexplained weight imbalance. The latter is attributed by the machine to the patient, resulting in treatment discontinuation. Such errors are expected to decrease with experience and training of the technical staff.

Survival to discharge in this study was 53%, consistent with previous studies of dogs managed by hemodialysis.^47,48 This survival rate is comparable to the survival rate of dogs with AKI managed medically,³ however it is likely masked by disease severity of animals managed by hemodialysis compared with those managed medically. Twenty of the discharged dogs (83%) were non-dialysis dependent, while 2 dogs (8%) continued to be managed with IHD after discharge.

There are several limitations to this study. Due to the retrospective nature of this study, medical records were occasionally incomplete, precluding evaluation of certain parameters. Additionally, the target (prescribed) dose might overestimate the actual delivered dose due to variables such as machine alarms and filter clotting.⁴⁹ Direct measures of solute clearance reflect more accurately the CRRT dose,⁵⁰ and should be considered in future studies. Finally, changes to urea volume of distribution due to alternations in total body water were not accounted for in the Kt/V calculations (Appendix).

In conclusion, CRRT was safe and effective for the initial management of dogs with severe AKI. These results demonstrate that the prescribed CRRT target dose, led to hourly URRs which are consistent with the current veterinary guidelines and human recommendations for an adequate and safe treatment. Additionally, we have shown that maintaining the hourly URR is achieved by periodic dose increments. Results of this study may lay the foundation for development of veterinary recommendations for CRRT.