Abstract
OBJECTIVE To compare dialysate sodium concentration and patient plasma sodium concentration of dogs during intermittent hemodialysis treatments.
SAMPLE 211 intermittent hemodialysis treatments performed on 40 client-owned dogs for the management of dialysis-dependent uremia.
PROCEDURES Medical records were reviewed to determine the plasma sodium concentration of each dog before and after routine hemodialysis treatments. Associations between detected changes in plasma sodium concentration and dialysate sodium concentration were evaluated by use of Spearman rank correlations and linear regression analysis.
RESULTS Significant linear correlations were found between the dialysate sodium concentration and patient sodium concentration. The starting dialysate-to-patient sodium gradient was associated with the strongest correlation to the change in patient sodium concentration at the end of the dialysis session. Modest correlations existed between the dialysate sodium concentration and postdialysis patient sodium concentration as well as between the predialysis dialysate-to-patient sodium gradient and postdialysis dialysate-to-patient sodium gradient.
CONCLUSIONS AND CLINICAL RELEVANCE The dialysate sodium concentration was correlated with the patient sodium concentration in dogs, and the dialysate-to-patient sodium gradient could be used to further refine this association to predict the postdialysis patient sodium concentration and potentially manage dysnatremia during hemodialysis. Prospective studies should be performed to determine how these associations can be used to correct aberrations as well as to avoid unwanted alterations in patient sodium concentrations.
Derangements of sodium and water balance are commonly reported consequences of dialysis-dependent kidney failure and must be considered when performing hemodialysis treatments. Common consequences of disordered renal sodium and water homeostasis include dysnatremia, dehydration, and hypervolemia.1,2 During hemodialysis, sodium concentrations will equilibrate between the patient's plasma and the dialysate, which provides a mechanism for therapeutic adjustment of abnormalities in the PNa.3 However, a positive value for the DNa-PNa has been found to influence interdialytic thirst and water accumulation, predialysis blood pressure, hyperkalemia, and intradialytic hypotension and hypertension in humans.4–6 A positive DNa-PNa also has been associated with patient morbidity and mortality rates in humans with end-stage renal disease.7
Goals for the use of a DNa are to achieve or maintain a PNa within the reference range at the end of dialysis, facilitate appropriate removal of excessive fluid through ultrafiltration, and protect against the development of dialysis disequilibrium. Currently, the effect of the DNa on the PNa of dogs undergoing intermittent hemodialysis is not known.
Therefore, the purpose of the retrospective study reported here was to compare the DNa with the predialysis and postdialysis plasma sodium concentrations in dogs undergoing intermittent hemodialysis. We intended to identify factors that could be used to improve the use of DNa to protect against the development of alterations in plasma sodium concentration.
Materials and Methods
Sample
Medical records from the School of Veterinary Medicine at the University of Pennsylvania and the William R. Pritchard Veterinary Medical Teaching Hospital at the University of California-Davis from October 1, 2009, to April 1, 2014, were evaluated retrospectively to identify dogs that underwent routine intermittent hemodialysis by use of a commercially available system.a The DNa was predicted on the basis of the programmed dialysate conductivity. The PNa concentration before and immediately at the termination of each dialysis treatment was measured by use of indirect and direct potentiometry.b-d Dogs were excluded if other analytic methods were used to measure predialysis and postdialysis plasma sodium concentrations, IV fluids were administered concurrently during dialysis, or dialysate sodium profiling (variable DNas throughout a single treatment) was performed during dialysis.
Statistical analysis
Correlations between DNa and PNa were evaluated by use of Spearman rank correlation analysis. Linear regression analysis was used to determine correlations between DNa and PNa. A statistical software programe was used for all analyses. Values were considered significant at P ≤ 0.05.
Results
A total of 364 dialysis treatments was identified for possible inclusion in the study. Of these, 153 were excluded because of a lack of PNa measurements, other methods used to measure the PNa, concurrent administration of IV fluids, or dialysate sodium profiling. Thus, 211 dialysis treatments performed on 40 dogs were included for analysis. The median number of dialysis treatments for each dog was 4 (range, 1 to 51).
Significant (P < 0.001) correlations were found between the DNa-PNa and ΔPNa (R2, 0.73; 95% confidence interval, 0.67 to 0.79), the DNa and post-dialysis PNa (R2, 0.76; 95% confidence interval, 0.66 to 0.86), and the predialysis DNa-PNa and postdialysis DNa-PNa (R2, 0.27; 95% confidence interval, 0.21 to 0.33). The predialysis DNa-PNa had the strongest correlation with the observed ΔPNa (R2, 0.72). The change in plasma sodium concentration at the end of the dialysis treatment was predicted by use of the following equation: ΔPNa = 0.73 × (DNa-PNa - 5.83); Figure 1). The DNa was moderately correlated with the postdialysis PNa (R2, 0.50; Figure 2). Predialysis DNa-PNa had the weakest correlation with the postdialysis DNa-PNa (R2, 0.27; Figure 3).
Graph of the association between the predialysis DNa-PNa and the ΔPNa (ie, postdialysis PNa - predialysis PNa) for 211 intermittent hemodialysis treatments performed on 40 client-owned dogs. Each symbol represents results for 1 dialysis treatment, and the line of best fit is indicated.
Citation: American Journal of Veterinary Research 79, 4; 10.2460/ajvr.79.4.450
Graph of the association between DNa and post-dialysis PNa for 211 intermittent hemodialysis treatments performed on 40 client-owned dogs. See Figure 1 for key.
Citation: American Journal of Veterinary Research 79, 4; 10.2460/ajvr.79.4.450
Graph of the association between the predialysis DNa-PNa and postdialysis DNa-PNa for 211 intermittent hemodialysis treatments performed on 40 client-owned dogs. See Figure 1 for key.
Citation: American Journal of Veterinary Research 79, 4; 10.2460/ajvr.79.4.450
Discussion
Results of the study reported here indicated significant linear correlations between the DNa and changes in plasma sodium concentrations during hemodialysis treatments in uremic dogs treated with intermittent hemodialysis. The strongest predictive correlation was the magnitude of the predialysis DNa-PNa and the ΔPNa at the end of a dialysis treatment. Liner regression of this association predicted that a DNa-PNa of 1 mmol/L would result in a change in the postdialysis PNa of 0.73 mmol/L. Analysis of the graph revealed that to yield no change in the plasma sodium concentration (ie, isonatricity; ΔPNa = 0) during dialysis treatment, the DNa should be approximately 7.9 mmol/L greater than the predialysis PNa.
These findings are in contrast to results for human patients whereby isonatricity was maintained with a mean DNa-PNa of −6.8 mmol/L.8 The general recommendation to achieve isonatric dialysis in humans is to individually customize the DNa so that it is slightly less than that for the starting PNa.9 The reason that dogs of the cohort reported here required a higher DNa-PNa to maintain isonatricity is not known. Mannitol was administered during approximately 15% of these treatments, although the timing of mannitol administration relative to collection of samples used to measure the sodium concentration was not always recorded. Owing to its hypertonicity, mannitol results in a flux of intracellular water to the intravascular space. This effect is only temporary (generally < 30 to 40 minutes, as determined from unpublished data of one of the authors [LDC]) because mannitol is removed via diffusion across the dialyzer. However, if mannitol were administered near the end of a dialysis treatment, it could cause a decrease in the postdialysis PNa. This could result in a dog having a postdialysis PNa lower than the DNa. Unfortunately, the retrospective nature of the study reported here prevented statistical evaluation of the effects mannitol may have had on the postdialysis PNa.
The DNa was correlated directly with the postdialysis PNa; however, the value for goodness of fit was lower than for other correlations. For example, dogs treated by use of a DNa of 151 mmol/L had a postdialysis PNa that ranged from 134 to 149 mmol/L, which suggested that the DNa was not the most important determinant of postdialysis PNa. The predialysis DNa-PNa had a positive linear correlation with the postdialysis DNa-PNa; however, this was the weakest correlation detected.
Studies10,11 of human patients support the recommendation that the DNa should be individually customized to each patient's predialysis sodium concentration; however, this is not standard practice in human dialysis centers, where the patient's predialysis sodium concentration is typically assessed only on a monthly biochemical panel and a batch dialysate with a fixed sodium concentration is delivered to multiple patients simultaneously. Additionally, the recommended DNa for dialysis of humans has not been definitively established, with the suggested DNa ranging from 0.5 to 8.8 mmol/L less than the predialysis PNa.12,13 Interestingly, analysis of data for the study reported here suggested that the DNa should be almost 8 mmol/L higher than the predialysis PNa to cause no change in PNa during a dialysis treatment. Evaluation of a similar DNa-PNa in humans revealed that a mean DNa-PNa of 8.4 mmol/L greater than the mean predialysis PNa resulted in a mean change in PNa of 2.5 mmol/L during dialysis.10 A common recommendation for individually customized DNa of humans is to use the DNa that is equal to the predialysis PNa.10,14 Analysis of data for the present study suggested that this practice in this cohort of dogs would have resulted in a decrease in the postdialysis sodium concentration (ie, ΔPNa) of 5.8 mmol/L. A potential difference between results for human and canine patients likely reflects differences in patient populations. The reference range for plasma sodium concentration in humans (134 to 145 mmol/L) is lower than that in dogs (144 to 160 mmol/L). A similar difference exists for plasma albumin concentration, for which it is higher in humans (3.5 to 5.0 mg/dL) than in dogs (2.2 to 3.9 mg/dL). It is unknown whether these differences would lead to interspecies variability in Gibbs-Donnan equilibrium potentials, which, if present, could have a role in the contrasting DNa-PNas required to maintain isonatricity in each species.
Information for human dialysis treatments was derived from patients who underwent chronic maintenance hemodialysis and established steady-state PNas for prolonged periods. In contrast, the cohort of dogs in the present study comprised primarily dogs that underwent dialysis support for acute kidney injury, and most received dialysis treatment for only several weeks. It is unlikely that most of these dogs achieved steady-state conditions for sodium homeostasis during the course of dialysis management. It would be interesting to compare these results with those for a group of dogs undergoing chronic maintenance hemodialysis for > 3 or 4 months, although such canine patients are not common.
A positive DNa-PNa causes sodium accumulation in patients and movement of water from the intracellular space to offset the osmotic gradient that develops. This shift of water across body compartments can cause or contribute to some of the complications observed in patients with kidney disease. A high DNa can contribute to interdialytic weight gain, an increase in blood pressure, hyperkalemia, and ultrafiltration requirements at subsequent dialysis treatments as consequences of sodium accumulation.15 Therefore, there is a need to recognize the consequences of an unwanted positive DNa-PNa as well as to develop techniques to determine an appropriate DNa to decrease the risk of these complications.
The ΔPNa after dialysis does not predict the magnitude of sodium flux that occurs during dialysis. A hypernatric dialysate causes sodium to accumulate in the extracellular matrix in a concentration-dependent, nonosmotic manner.16 In this situation, sodium ions diffuse across the dialyzer membrane into the patient's plasma water; subsequently, the sodium content of the mucopolysaccharides within the extracellular matrix also will increase. As each ion of sodium complexes with the extracellular matrix, it leaves the osmotic and diffusive pools, lowers the plasma sodium concentration, and restores the DNa-PNa. Sodium molecules within the mucopolysaccharide complexes are osmotically inactive, and it has been determined that the quantity of osmotically inactive sodium molecules modulates the plasma water sodium concentration.17 Sodium will continue to diffuse into patients until the matrix is saturated; however, there will not be a substantial increase in the plasma osmolality of patients. Thus, the net transfer of sodium into a patient may be much more than can be predicted simply on the basis of the difference between the predialysis and postdialysis PNa, and PNA may correlate poorly with the total amount of exchangeable sodium.18 Plasma sodium activity is reduced via the Gibbs-Donnan effect, in which non-diffusable plasma proteins trap sodium molecules and thus prevent them from crossing the dialyzer membrane. Differences in composition and size of a dialyzer membrane as well as plasma protein concentration of a patient result in the fact that the Gibbs-Donnan effect becomes unpredictable and can lead to differences in sodium transport, despite similar DNa-PNas.19
Multifrequency bioimpedance spectroscopy is a noninvasive tool that can be used to measure changes in intracellular, extracellular, and total body water during hemodialysis.20–22 Use of this modality for dogs has been reported.23 Results of multifrequency bioimpedance spectroscopy could provide additional data on sodium and water flux during dialysis that cannot be predicted on the basis of changes in PNa and body weight alone.
The present study had several limitations. First, dogs often received food and water through an esophagostomy tube during a dialysis treatment. There was great variation in the volume and sodium concentration of the diets administered, and these could not be controlled within this study population. These variables are a further reflection of the differences in study populations between humans and dogs. Second, bioimpedance spectroscopy data were not available for most treatments. These data could have been extremely helpful for evaluating changes in interstitial, intracellular, and total body water flux that have the potential to influence PNa throughout dialysis treatments. Third, the DNa was not measured; instead, it was predicted on the basis of the programmed dialysate conductivity. However, dialysate conductivity is an accurate predictor of the sodium concentration within dialysate.24,25 Finally, each dialysis treatment was evaluated separately, which is consistent with the methods used in similar studies15,26,27 for people undergoing dialysis. It is possible that there may have been individual factors that affected the sodium flux of dogs during dialysis. However, significant associations were detected when the treatments were considered as separate events.
The DNa was correlated with the PNa, and the DNa-PNa can be used to further refine this association. Prospective studies should be performed to validate these descriptive associations and their use in predicting the postdialysis PNa for various hemodialysis treatments. Each patient should have the DNa individually customized on the basis of the predialysis PNa to avoid unwanted changes in PNa as well as plasma osmolality. Prospective studies that involve the use of bioimpedance spectroscopy may be helpful in providing additional information about sodium and water flux during intermittent hemodialysis.
Acknowledgments
There were no sources of funding for this study. The authors declare that there were no conflicts of interest.
Presented in abstract form at the 2015 American College of Veterinary Internal Medicine Forum, Indianapolis, June 2015.
ABBREVIATIONS
| ΔPNa | Change in patient plasma sodium concentration |
| DNa | Dialysate sodium concentration |
| DNa-PNa | Dialysate sodium concentration-to-patient plasma sodium concentration gradient |
| PNa | Patient plasma sodium concentration |
Footnotes
Phoenix intermittent hemodialysis platform, Baxter International Inc, Deerfield, Ill.
Vitros 350, Ortho Clinical Diagnostics, Rochester, NY.
Nova pHOx, Nova Biomedical, Waltham, Mass.
Cobas 6000 c501, Roche Diagnostics, Indianapolis, Ind.
Stata for Mac, version 12.0, Stata Corp, College Station, Tex.
References
1. Guillaumin J, DiBartola SP. Disorders of sodium and water homeostasis. Vet Clin North Am Small Anim Pract 2017;47:293–312.
2. Langston C. Managing fluid and electrolyte disorders in kidney disease. Vet Clin North Am Small Anim Pract 2017;47:471–490.
3. Funck-Brentano JL, Man NK. Optimization of Na content of dialysis fluid. Nephron 1984;36:197–200.
4. Flanigan MJ, Khairullah QT, Lim VS. Dialysate sodium delivery can alter chronic blood pressure management. Am J Kidney Dis 1997;29:383–391.
5. De Nicola L, Bellizzi V, Minutolo R, et al. Effect of dialysate sodium concentration on interdialytic increase of potassium. J Am Soc Nephrol 2000;11:2337–2343.
6. Movilli E, Camerini C, Gaggia P, et al. Role of dialysis sodium gradient on intradialytic hypertension: an observational study. Am J Nephrol 2013;38:413–419.
7. Thomson BKA, Huang S-HS, Leitch RE, et al. Pre to post-dialysis plasma sodium change better predicts clinical outcomes than dialysate to plasma sodium gradient in quotidian hemodialysis. Hemodial Int 2013;17:548–556.
8. Pedrini LA, Ponti R, Faranna P, et al. Sodium modeling in hemodiafiltration. Kidney Int 1991;40:525–532.
9. Flanigan MJ. Role of sodium in hemodialysis. Kidney Int Suppl 2000;76:S72–S78.
10. Song JH, Lee SW, Suh C-K, et al. Time-averaged concentration of dialysate sodium relates with sodium load and interdialytic weight gain during sodium-profiling hemodialysis. Am J Kidney Dis 2002;40:291–301.
11. Santos SFF, Peixoto AJ. Revisiting the dialysate sodium prescription as a tool for better blood pressure and interdialytic weight gain management in hemodialysis patients. Clin J Am Soc Nephrol 2008;3:522–530.
12. Redaelli B, Sforzini S, Bonoldi G, et al. Hemodialysis with “adequate” sodium concentration in dialysate. Int J Artif Organs 1979;2:133–140.
13. Kim M-J, Song JH, Kim GA, et al. Optimization of dialysate sodium in sodium profiling haemodialysis. Nephrology (Carlton) 2003;8(suppl):S16–S22.
14. de Paula FM, Peixoto AJ, Pinto LV, et al. Clinical consequences of an individualized dialysate sodium prescription in hemodialysis patients. Kidney Int 2004;66:1232–1238.
15. Hecking M, Karaboyas A, Saran R, et al. Dialysate sodium concentration and the association with interdialytic weight gain, hospitalization, and mortality. Clin J Am Soc Nephrol 2012;7:92–100.
16. Farber SJ, Schubert M, Schuster N. The binding of cations by chondroitin sulfate. J Clin Invest 1957;36:1715–1722.
17. Nguyen MK. Quantitative interrelationship between Gibbs-Donnan equilibrium, osmolality of body fluid compartments, and plasma water sodium concentration. J Appl Physiol 2006;100:1293–1300.
18. Edelman IS, Leibman J, O'Meara MP, et al. Interrelations between serum sodium concentration, serum osmolarity and total exchangeable sodium, total exchangeable potassium and total body water. J Clin Invest 1958;37:1236–1256.
19. Flythe JE, McCausland FR. Dialysate sodium: rationale for evolution over time. Semin Dial 2017;30:99–111.
20. Jaffrin MY, Fenech M, de Fremont J-F, et al. Continuous monitoring of plasma, interstitial, and intracellular fluid volumes in dialyzed patients by bioimpedance and hematocrit measurements. ASAIO J 2002;48:326–333.
21. Sarkar SR, Wystrychowski G, Zhu F, et al. Fluid dynamics during hemodialysis in relationship to sodium gradient between dialysate and plasma. ASAIO J 2007;53:339–342.
22. Moissl U, Arias-Guillén M, Wabel P, et al. Bioimpedance-guided fluid management in hemodialysis patients. Clin J Am Soc Nephrol 2013;8:1575–1582.
23. Scheltinga MR, Helton WS, Rounds J, et al. Impedance electrodes positioned on proximal portions of limbs quantify fluid compartments in dogs. J Appl Physiol 1991;70:2039–2044.
24. Polaschegg HD. Automatic, noninvasive intradialytic clearance measurement. Int J Artif Organs 1993;16:185–191.
25. Locatelli F, Di Filippo S, Manzoni C, et al. Monitoring sodium removal and delivered dialysis by conductivity. Int J Artif Organs 1995;18:716–721.
26. Kumar S, Khosravi M, Massart A, et al. Are serum to dialysate sodium gradient and segmental bioimpedance volumes associated with the fall in blood pressure with hemodialysis? Int J Artif Organs 2014;37:21–28.
27. Davenport A. Negative dialysate to sodium gradient does not lead to intracellular volume expansion post hemodialysis. Int J Artif Organs 2010;33:700–705.
