Dietary Therapy in Detail

3. Dietary Therapy in Detail

There are many diets available to assist in the management of CKD. These specifically formulated diets are different in several respects from standard diets formulated for feeding adult cats.

- When dietary changes are introduced in Stages II and III of CKD, the aim is mainly to address those factors which are likely to contribute to progressive renal injury and further loss of functioning nephrons. In this section the rationale will be reviewed for the manipulation of each element of the diet and published evidence for the efficacy of this treatment in slowing progression, presented.

- Once late stage III / IV has been reached, clinical signs of the uremic syndrome are evident and dietary treatment is designed more to improve the quality of life of the patient than to slow disease progression. Section 5 will deal with the approach and the use of renal care diets and supplements to address the problems of the uremic syndrome.

Phosphate Restriction and Management of Secondary Renal Hyperparathyroidism

Phosphate is freely filtered by the normal kidney but not actively secreted by the tubule. Thus, the amount of phosphate excreted from the body each day is highly dependent on the GFR. Re-absorption of phosphate occurs in the proximal convoluted tubule through a carrier-mediated process (cotransported with sodium ions). The maximum capacity of this system to re-absorb phosphate is influenced by parathyroid hormone (PTH) which reduces phosphate re-absorption and so increases the amount of phosphate excreted in the urine at a given plasma phosphate concentration and a given GFR.

As GFR falls, if dietary phosphate intake remains the same, the daily amount of phosphate excreted in the urine will not match with the daily phosphate intake. Thus, phosphate will start to accumulate in the body. Both intracellular stores and extracellular fluid concentration of phosphate increase. As the plasma phosphate concentration increases, so its rate of excretion will increase until a new steady state is reached at a higher plasma phosphate concentration and higher level of intracellular stores of phosphate. PTH plays a role in this process since increased PTH synthesis and secretion is triggered by a rise in both intracellular stores of phosphate and plasma phosphate concentration. Initially, this adaptive response is helpful as it enhances urinary excretion of phosphate compensating for the effect of the fall in GFR (Figure 5).

Figure 5. Effects of parathyroid hormone on calcium homeostasis and the involvement of its three target organs: bone, kidney and the gastrointestinal tract.

Unfortunately, the adaptive response of increasing PTH secretion to counterbalance the tendency for phosphate retention with loss of functioning nephrons and concomitant fall in GFR is limited by two factors:

1. At least 30% of the filtered load of phosphate has to be re-absorbed in the proximal tubule as part of the reabsorptive process for sodium and hence water
2. As increasing plasma concentrations of PTH are required, the actions of this hormone on bone bring more phosphate from bone stores into the extracellular fluid compartment adding to the problem of hyperphosphatemia.

As the CKD progresses and fewer functioning nephrons remain, secretion of PTH, driven by phosphate retention, becomes counter-productive and so mal-adaptive. Phosphate release from bone adds to the problem by inhibiting renal production of calcitriol and stimulating PTH synthesis and secretion and parathyroid gland growth. In the later stages of CKD (IRIS late stage III and stage IV), calcitriol deficiency (as a result of reduced kidney mass and the inhibitory effects of hyperphosphatemia on calcitriol synthesis) contributes to the problem of hyperparathyroidism in two ways:

• Calcitriol inhibits PTH synthesis and secretion by a direct action on the parathyroid gland. This hormone also prevents parathyroid gland hypertrophy
• With a lack of calcitriol, absorption of calcium from the intestine is reduced and hypocalcemia can occur (particularly low ionized calcium) in the more severe stages of CKD. With very high plasma phosphate concentrations, ionized calcium will also decrease due to complexing of calcium with phosphate and other small anions.

The above outline of the pathophysiology of secondary renal hyperparathyroidism is shown schematically in Figure 6. Scientific understanding of this process has changed the emphasis from the view that ionized calcium concentration decreases, which were once thought to drive PTH secretion, to recognize that phosphate retention is now central to this process.

Figure 6. Pathophysiology of secondary renal hyperparathyroidism.

That hyperphosphatemia and hyperparathyroidism are important in naturally occurring CKD is clearly evident from published studies (Barber & Elliott, 1998). Whether phosphate retention and/or increased parathyroid hormone synthesis and secretion are detrimental to the health and well being of the cat with CKD has been a topic of debate. Evidence from laboratory animal models and from human medicine suggests hyperphosphatemia and hyperparathyroidism are detrimental to the quality of life of the patient and may contribute to progressive renal injury. Direct evidence supporting these conclusions apply to cats is relatively sparse although some data from both an experimental model in the cat (Ross et al., 1982) and naturally occurring feline CKD support the conclusions that reduction of phosphate intake to control parathyroid hormone secretion results in:

• Reduced mineralization (Figure 7) and fibrosis in the remaining functioning kidney tissue (experimental model studies; Ross et al., 1982)
• A reduction in all cause mortality in cats with naturally occurring CKD (Elliott et al., 2000).

Figure 7. Renal calcification due to renal hyperparathyroidism in a cat. A band of calcification can be seen in the inner medulla which was confirmed on histological examinations (scale in mm).

The prospective diet study conducted by Elliott et al, (2000) was open label involving cats at stage II and III CKD where the aim was to use phosphate restriction by feeding a renal care diet to control plasma PTH and to study the effect this had on survival. The control group were permitted to continue eating their standard maintenance diet after their owners had rejected treatment with the renal care diet. Thus, scientifically the design of this study was not optimal since it was not masked and the control group were self-selecting.

A second study of a renal care diet has been published more recently where the design was a randomized controlled masked clinical trial (Ross et al., 2006) and the aim was to determine the benefit of the renal care diet on the time to uremic crises or renal death when fed to cats at stage II and III CKD. The renal care diet tested was compared to a standard maintenance diet that differed in protein, sodium, phosphate and lipid content. The renal care diet contained 0.5% phosphate on an as fed basis (1.2 g/1000 kcal for the dry renal diet, 1.0 g/1000 kcal for the wet renal diet) whereas the maintenance diet contained 0.9 or 1% phosphate on an as fed basis (1.8 g/1000 kcal for the dry maintenance diet, 2.3 g/1000 kcal for the wet maintenance diet). Feeding the renal care diet resulted in a lower plasma phosphate concentration at 12 and 24 months after introduction of the diets although plasma PTH concentrations did not differ significantly. The cats eating the renal care diet suffered significantly fewer uremic crises and there were significantly fewer renal deaths. In both of the above studies, because the renal care diets differed in a number of respects from standard maintenance diets, it is not possible to conclude whether phosphate restriction was responsible for the effect seen but it seems likely to have contributed.

Accumulation of phosphate and calcium in renal tissues will lead to nephrocalcinosis and may contribute to progressive renal injury and these processes are probably ongoing in IRIS stages II and III of CKD. Clearly, in the later stages of CKD (stage IV), extrarenal effects of hyperphosphatemia and hyperparathyroidism are evident with radiographic evidence of renal osteodystrophy and mineralization of soft tissues (Figure 8) accompanied by marked parathyroid gland hypertrophy. In human medicine, poor control of phosphate balance in the renal patient on dialysis leads to increased cardiovascular risk as calcium and phosphate accumulate in the vasculature (KDOQI, 2003).

Figure 8. Radiographic evidence of soft tissue calcification of the vasculature due to hyperparathyroidism in cats with chronic renal failure. A: Soft tissue calcification of the thoracic aorta in a 20 year old cat with chronic kidney disease (classified as uremic) (© Penney Barber). B: Soft tissue calcification of the abdominal aorta and some abdominal vasculature in a 19 year old cat with end stage chronic kidney disease. (© Penney Barber).

Management of Secondary Renal Hyperparathyroidism

From the above discussion of the pathophysiology of hyperphosphatemia and hyperparathyroidism secondary to CKD, the logical way to address these problems is to restrict dietary phosphate intake in the first instance. This can be done by restricting the amount of phosphate in the ration fed and /or adding phosphate binders to reduce phosphate bioavailability in the food that is fed.

Evidence that feeding a commercially formulated renal clinical diet reduces both plasma phosphate concentration and PTH concentrations when fed to cats with naturally occurring CKD has been published (Barber et al., 1999; Figure 9a and Figure 9b). The effect on plasma PTH concentration tends to be prolonged with plasma PTH concentrations falling with continued dietary phosphate restriction after the plasma phosphate concentration has stabilized (Figure 10). This probably results from depletion of intracellular stores of phosphate which influence PTH synthesis and secretion. In human medicine, the recommendations regarding control of plasma phosphate concentration have been published based on expert opinion and available clinical research evidence (KDOQI, 2003). These guidelines have been adapted by a group of veterinary nephrologists to apply to the cat and have been adopted by the IRIS group as recommendations according to the stage of CKD that is being treated.

Figure 9a. Effect of feeding a renal care diet to cats with ckd (stages II and III) on plasma phosphate concentration. Barber et al., 1999; From J Small Anim Pract (1999). Permission of reproduction granted by Blackwell Publishing.

Figure 9b. Effect of feeding a renal care diet to cats with ckd (stages II and III) on plasma PHT concentration. Barber et al., 1999; From J Small Anim Pract (1999). Permission of reproduction granted by Blackwell Publishing.

Figure 10. Effect of feeding a renal care diet on plasma phosphate (blue) and PHT (orange) concentrations in a cat with CKD. (Barber, 1999).

• For stage II CKD, the post-treatment plasma phosphate concentration should be below 1.45 mmol/L (4.5 mg/dL), but not <0.8 mmol/L (2.5 mg/dL). In our experience, cats where plasma phosphate can be maintained below 1.2 mmol/L (3.72 mg/dL) tend to remain very stable in stage II CKD for prolonged periods of time.
• For stage III CKD, the realistic post-treatment target is <1.61 mmol/L (5.0 mg/dL). Intestinal phosphate binders in addition to feeding a diet restricted in phosphate may be necessary to achieve this target in later stage III cases.
• For stage IV CKD, the realistic post-treatment plasma phosphate concentration target is 1.93 mmol/L (6.0 mg/dL) and this is unlikely to be achieved with dietary phosphate restriction alone.

Unpublished data from our research clinic shows that 55, 90 and 100% of cats presenting in stage II, III and IV CKD respectively have plasma phosphate concentrations above 1.45 mmol/L (4.5 mg/dL) at diagnosis. Re-analysis of the data from the prospective study of the effect of controlling plasma phosphate and PTH on survival of cats with stage II and III CKD (Elliott et al., 2000) demonstrated that:

- If the average plasma phosphate concentration was maintained at below 1.45 mmol/L (4.5 mg/dL) for the first half of their survival time (this was achieved in 18 of the 50 cats) their median survival time was 799 (interquartile range 569 - 1383) days

- For cats where the average plasma phosphate concentration exceeded 1.45 mmol/L (4.5 mg/dL) the median survival time was 283 (interquartile range 193 to 503) days (Figure 11).

Figure 11. Relationship between survival time and the mean plasma phosphate concentration achieved in the first half of the survival period. Data re-analyzed from Elliott et al. (2000).

These data are supportive of the extrapolation of the KDOQI (2003) recommendations on the control of plasma phosphate from human to feline medicine. Further prospective studies are necessary which are specifically designed to address the benefit of maintaining plasma phosphate concentrations below 1.45 mmol/L (4.5 mg/dL) in cats with CKD are still required, however, to verify this recommendation.

Adverse effects of restricting phosphate intake are rare. It is recommended that plasma phosphate and calcium (preferably ionized calcium) are measured routinely every 2 to 3 months in cats that have been stabilized on restricted phosphate diets and that hypophosphatemia (plasma phosphate concentration <0.8 mmol/L [2.5 mg/dL]) is avoided. Occasionally, hypercalcemia has been reported (Barber et al., 1998). This is a true hypercalcemia since ionized calcium as well as total calcium is outside of the reference range and plasma PTH is below the limit of detection. The underlying cause of the hypercalcemia in these cases is not understood but it appears to result from phosphate restriction since feeding more phosphate in the diet leads to the plasma calcium ion concentration returning to the reference range and the plasma PTH concentration increasing into the measurable range at the same time. Since PTH is important for normal bone turnover it does not seem appropriate to completely suppress PTH secretion in these cases, hence we would recommend feeding more phosphate to these cats. This small proportion of cases clearly do not need the degree of phosphate restriction provided by the commercial diets in order to control PTH and phosphate illustrating the point that any treatment should be tailored to the individual needs of the patient.

Dietary Sodium and Kidney Disease

Sodium is the major determinant of extracellular fluid (ECF) volume and blood pressure being the main ECF cation. Sodium ions are maintained at a stable concentration in ECF and plasma through the osmoreceptor and thirst mechanisms that regulate water balance. Plasma osmolality is maintained at a stable 280 to 290 mOsm/L.

Sodium is the major determinant of extracellular fluid (ECF) volume and blood pressure being the main ECF cation. Sodium ions are maintained at a stable concentration in ECF and plasma through the osmoreceptor and thirst mechanisms that regulate water balance. Plasma osmolality is maintained at a stable 280 to 290 mOsm/L.

Doppler blood pressure measuring device. (© Dr H. Syme).

Thus, in cats with normal kidney function, a wide range of sodium intakes can be tolerated without detrimental effects on arterial blood pressure. Indeed, one strategy adopted to reduce the tendency for the formation of uroliths in feline urine is to increase dietary sodium intake. This results in a larger volume of urine being produced and the cats will drink more water to compensate. Hence the urinary calcium and magnesium concentrations are reduced and the tendency for urolith formation also is decreased. Normal cats fed such diets show no tendency for their blood pressures to increase (Buranakarl et al., 2004; Luckschander et al., 2004) (Figure 12).

Figure 12. Influence of dietary sodium on blood pressure in healthy cats (Luckschander et al., 2004).

Formulated clinical renal diets tend to have a lower sodium content per calorie than foods designed for healthy cats. The content of the renal diets still provides more than 2 to 4 times the National Research Council (NRC) recommended daily intake of sodium (0.4 to 0.9 mmol/kg/day or 9.2 - 20.7 mg/kg/day) (Yu and Morris, 1999) at around 2 mmol/kg/day (46 mg/kg/day). Standard grocery diets provide between 4 to 6 mmol/kg/day (92 - 138 mg/kg/day) (Table 4). The rationale for this is that with loss of the number of functioning nephrons, ability to excrete sodium from the body is reduced. If dietary sodium intake remained the same then these cats would be at increased risk of developing hypertension associated with their CKD. There are no controlled studies in the published veterinary literature to demonstrate the benefit of reducing dietary sodium intake on blood pressure in cats with naturally occurring CKD.

In a cross-sectional study of cats presenting at different stages of CKD, we demonstrated that the fractional excretion of sodium increased with decreasing renal function (Figure 12; unpublished data taken from cats studied in Elliott et al., 2003a). The interpretation of urinary fractional excretion data from an individual case based on a spot urine sample should be made with caution since there appears to be significant intra-animal variability with time (Adams et al., 1991; Finco et al., 1997). A 24 hour urine collection would yield more reliable results but is impractical in feline clinical research. In addition, the pattern observed in the data presented in Figure 13 may well be blurred by the fact these cats were being fed heterogeneous diets. Nevertheless, despite these shortcomings there does appear to be a higher fractional excretion of sodium at the more severe stages of CKD, suggesting an adaptive change of the remaining functioning tubules ensuring more of the filtered load of sodium is excreted from the body. There was no difference in plasma sodium concentrations between the cats at the different stages of CKD in this cross-sectional study although the plasma chloride ion concentrations were lower in the severe stage (equivalent to IRIS stage IV; Elliott et al., 2003a), possibly associated with the development of metabolic acidosis (see section below).

Figure 13. Box and whisker plots showing fractional excretion of sodium in normal cats and cats with CKD at diagnosis. Data taken from cases included in Elliott et al., 2003a.

Some cats with naturally occurring CKD do present with severe hypertension. Figure 14 shows the distribution of blood pressure at initial diagnosis. These data are from 103 consecutive cases of naturally occurring CKD (Syme et al., 2002a). Categorization of these cats according to the IRIS staging system gives the following:

• Minimal risk (<150 mmHg) – 62/103 or 60%
• Mild risk (150 - 159 mmHg) – 10/103 or 10%
• Moderate risk (160 - 179 mmHg) – 15/103 or 14.5%
• Severe risk (>180 mmHg) – 16/103 or 15.5%

Figure 14. Distribution of systolic blood pressure measurements in 103 cats with CKD. From Syme et al (2002a).

This study was a cross-sectional one and did not address the question as to whether blood pressure rises with time in the feline patient with CKD. If sodium retention occurs over time in the CKD patient due to an inability to excrete the daily quantity of sodium taken in the diet, one might expect blood pressure to increase over time. However, Syme et al (2002a) found the plasma creatinine was not a risk factor for high blood pressure – in other words blood pressure did not appear to be higher in cats with more severe CKD. Indeed, the majority of the cats found in the high-risk blood pressure group were in stage II or early stage III CKD according to the IRIS classification system. However, these data are difficult to interpret since cases presenting in stage IV CKD may well have lower blood pressure due to dehydration.

Syme (2003) analyzed data from a population of cats with CKD followed longitudinally to determine whether blood pressure increased from diagnosis of CKD. The inclusion criteria of this retrospective study were carefully defined to avoid extraneous factors that might influence blood pressure other than chronicity of CKD. The study included 55 cats each followed for more than 3 months. Seven of the 55 cats showed an increase in blood pressure to a point where medical treatment was deemed necessary (systolic blood pressure persistently >175 mmHg). Of the 55 cats, 17 showed progression of their CKD (as evidenced by >20% increase in plasma creatinine concentration) over the period of follow-up and 38 cats were classified as non-progressive. The cumulative hazard rate for an increase in blood pressure to a level where treatment was necessary was not significantly different between the progressive and non-progressive groups. Taking the group as a whole, blood pressure increased significantly over time (0.38 [0.2 to 0.56] mmHg/month; P<0.001 by repeated measures linear mixed model approach). These data suggest that blood pressure increases gradually over time in cats with naturally occurring CKD. This phenomenon does not appear to be associated with a decline in kidney function as assessed by repeated measures of plasma creatinine concentrations, although more sensitive measures of kidney function over time (e.g., repeated assessment of glomerular filtration rate) would be necessary to be confident renal function has not changed over time in the non-progressive cases.

Similar findings were reported by Ross et al (2006) in their prospective study of the influence of diet on spontaneous CKD. Seven of the 45 cats entered into this study developed hypertension (systolic blood pressure >175 mmHg) and required medical treatment over the 2 year follow-up period despite having normal blood pressure at entry to the study. The overall effect of the renal care diet on the blood pressure of the cats involved in this study was not reported. Nevertheless, the renal care diet did not appear to limit the development of hypertension in this study since 5 of the 7 cats developing hypertension did so despite being fed the renal care diet. The numbers of cats developing hypertension in both these longitudinal studies are too small to conclude anything definitively.

From the above theoretical considerations it appears to be logical to restrict sodium intake in cats with naturally occurring CKD. Nevertheless, controlled studies are lacking to determine the benefit of such an intervention on blood pressure control or progressive deterioration in kidney function. Syme (2003) reported data on the effect of introduction of a renal care diet on blood pressure in cats with naturally occurring CKD. This was an uncontrolled study as all animals included were fed a standard renal care diet. In addition, this study did not involve cats deemed to be at high risk of end organ damage (Figure 15) resulting from high blood pressure as these cases were treated with drugs to control their blood pressure. Systolic blood pressure was measured twice before introduction of the diet and at two time points after the intervention (a minimum of 4 weeks and a maximum of 12 weeks post introduction of the diet) and blood pressure measurements were averaged at the two pre-treatment and two post-treatment time points. Compliance was demonstrated by a significant decline in plasma phosphate concentration (1.55 ± 0.53 mmol/L vs. 1.31 ± 0.32 mmol/L; 4.8 ± 1.64 mg/dl vs. 4.04 ± 0.99 mg/dL; n=28). No changes in plasma sodium or potassium ion concentrations were detected as a result of feeding the renal care diet. Systolic blood pressure did not change in response to introduction of the diet (139 ± 24 mmHg vs. 141 ± 32 mmHg; n=28). The power of the study to detect a 10 mmHg change in systolic blood pressure was calculated to be 90%. A sub-group of cats enrolled in this study had plasma aldosterone and plasma renin activity (PRA) measured before the introduction of the diet and whilst consuming the renal care diet. Plasma aldosterone concentration was higher when the cats were consuming the renal care diet (73 [43, 105] pg/mL vs. 123 [65, 191] pg/mL; pre-diet vs. whilst consuming diet respectively; n=22). Similar changes in PRA were detected following the introduction of the renal care diet (0.53 [0.17, 1.11] vs. 0.75 [0.21, 1.38] ng/mL/h). Both plasma aldosterone concentration and PRA remained in the reference range (derived from aged normal cats fed heterogeneous grocery diets formulated for adult cats) both before and during the renal care diet feeding period.

Figure 15. Secondary consequences of severe hypertension in cats with naturally occuring chronic kidney disease. (© A. Régnier-National Veterinary School of Toulouse: Ophtalmology unit. © Rebecca Elks, The Royal Veterinary College, London. © J. Elliott; Royal Veterinary College of London).

Results from a study involving the remnant kidney model in cats (Buranakarl et al., 2004) suggest that reduction of sodium intake may cause activation of the renin-angiotensin-aldosterone system (RAAS) resulting acutely in a fall in plasma potassium ion concentration and was without beneficial effect on arterial blood pressure. Three diets (with a respective sodium content of: 0.34%, 0.65% and 1.27%) were fed for 7 days sequentially to three groups of cats. The different sodium chloride intake were 50, 100 and 200 mg per kg of body weight (i.e: 0.5 g, 1.4 g and 2.8 g sodium for 1000 kcal), the lowest intake being equivalent to many renal care diets. The three groups of cats involved in this study were:
- Control cats with normal kidney function (young adults)
- Remnant kidney cats (11/12 nephrectomy model)
- Cats which had had a bilateral partial nephrectomy with one kidney wrapped in silk and cellophane (renal wrap model causing severe hypertension) (see Mathur et al., 2004). During the feeding trial, these cats received amlodipine besylate treatment to control their blood pressure and prevent development of hypertensive encephalopathy.

Both these models led to renal insufficiency accompanied by elevated arterial blood pressure of the similar order of magnitude as seen in naturally occurring CKD. However, activation of the RAAS (Figure 16) with elevated PRA (2 to 6 fold) compared to the control group and markedly elevated aldosterone (4 to 25 fold higher than control cats) was associated with both models (particularly marked in the renal wrap model). Cats with naturally occurring CKD and blood pressures placing them at minimal to moderate risk of end organ damage (up to 175 mmHg) tend to have either normal or suppressed PRAs compared to agematched control cats fed similar diets. Furthermore, plasma aldosterone concentrations also remain within the reference range and do not differ significantly from age-matched control cats (Syme et al., 2002b). Marked activation of the RAAS does occur in unstable naturally occurring CKD patients in stage IV (Syme, 2003). Thus, the remnant kidney and renal wrap models of hypertension appear to give rise to significant activation of the RAAS, a finding which is not relevant to naturally occurring CKD at stages II or III with no or only mild to moderate elevations in blood pressure. The relevance of these models to naturally occurring CKD in cats appears to be questionable.

Figure 16. Activation of the renin angiotensin aldosterone system (RAAS).

Cats with naturally occurring CKD with marked elevations in arterial blood pressure (systolic pressure >180 mmHg; high risk of target organ damage) tend to have normal or suppressed PRA associated with normal or marginally elevated plasma aldosterone concentrations (Jensen et al., 1997; Syme et al., 2002b). These cats also tend to have lower plasma potassium ion concentrations at diagnosis and are relatively resistant to the antihypertensive effects of standard doses of angiotensin converting enzyme inhibitors (Littman, 1994), both findings suggesting that hypertension in these cases is possibly the result of increased secretion and/or activity of aldosterone but not through activation of the RAAS. It is clear that restriction of sodium chloride intake in these severely hypertensive cats is not sufficient alone to manage their hypertension and pharmacological interventions are required to control blood pressure. Whether dietary sodium restriction helps to achieve control of blood pressure with drugs in these patients has not been studied in feline clinical patients. A clearer understanding of why some cats with naturally occurring CKD develop severe hypertension associated with a high risk of end organ damage remains to be established. Once the reason for this is understood, the role of sodium restriction in managing these patients may become clearer.

In summary, most renal care diets formulated for cats have reduced sodium content compared to standard adult maintenance foods. The logic behind this is that with reduced functional renal mass, maintenance of sodium homeostasis will prove more difficult to achieve and sodium retention could result in increased blood pressure. Hypertension could reduce the quality of life of cats with CKD and lead to further damage to the remaining functioning nephrons and so progressive renal injury. About 20% of cats with naturally occurring CKD do have arterial blood pressures at diagnosis which place them at severe risk of target organ damage (including renal damage) secondary to hypertension. Blood pressure does tend to increase gradually over time in the remaining 80% of cats with CKD where their blood pressure at initial diagnosis does not place them at high risk of target organ damage. However, certain observations have cast doubt on the routine restriction of dietary sodium in cats with naturally occurring CKD (Table 5).

Despite these observations, renal care diets, which restrict sodium intake, continue to be routinely used in cats with naturally occurring CKD. Their clinical use does not appear to be associated with worsening of hypokalemia (Elliott et al., 2000; Ross et al., 2006) or proteinuria (unpublished data), despite an increase in plasma aldosterone concentration within the physiological range (Syme, 2003). Whether reducing sodium intake is beneficial in limiting the chronic small increase in blood pressure detected over time in cats with naturally occurring CKD remains to be determined by future longitudinal studies, as does their potential benefit in managing severe hypertension in the cat in combination with antihypertensive drug therapy.

Potassium and Kidney Disease

The cat is somewhat unique in that there appears to be an association between CKD and hypokalemia. Loss of functioning nephrons puts the dog or human patient at increased risk of hyperkalemia. In cats, adaptive changes in the remaining functioning nephrons appear, in some 20 to 30% of cases of CKD, to over compensate and excess loss of potassium in the urine leads to hypokalemia (DiBartola et al., 1987; Elliott & Barber, 1998) unless they move into an oliguric stage as part of a uremic crisis. In the face of CKD, hypokalemia also appears to be associated with an increased risk of systemic hypertension (Syme et al., 2002a), possibly also due to the way the kidney responds to loss of functioning nephrons.

Potassium is the major intracellular cation and circulates in plasma at a concentration of around 4 mmol/L. This means measurement of plasma potassium concentration is an indirect assessment of whole body potassium status, particularly as potassium can change its distribution between the cells and ECF, for example in response to acid-base disturbances. Plasma potassium is freely filtered and most of the filtered load is returned to the plasma in the proximal convoluted tubule and the loop of Henle. The cortical connecting tubule is the site of potassium secretion into the tubular fluid (Figure 17). Fractional excretion of potassium will vary depending on various factors (Table 6).

Figure 17. Cortical collecting tubule cell.

Aldosterone acts on the cortical collecting tubule to increase potassium ion loss in the urine by increasing the number of potassium channels in the apical plasma membrane of the tubular cells through which potassium ions can diffuse. In effect, these potassium ions exchange for sodium ions diffusing from the tubular fluid into the tubular cell through epithelial sodium channels whose synthesis is also under the control of aldosterone. Intracellular potassium is maintained at a high concentration and intracellular sodium is maintained at a low concentration by the action of aldosterone causing the synthesis of basolateral membrane sodium potassium ATPase (pumps) (Figure 17).

Dow and Fettman (1992) hypothesized that potassium depletion may lead to a self-perpetuating cycle of renal damage and further potassium loss. This hypothesis is based on:

• The clinical observation of a strong statistical association between CKD and the occurrence of hypokalemia (Dow et al., 1989).
• Naturally occurring CKD was observed in association with feeding an acidifying diet that was marginally replete in potassium – renal function appeared to improve when the diet was changed and the hypokalemia was corrected (Dow et al., 1987).
• The experimental observation that feeding a diet that was deficient in potassium and supplemented with phosphoric acid (acidifying diet) led to severe hypokalemia and metabolic acidosis accompanied by a decline in glomerular filtration rate (Dow et al., 1990).

Support for this hypothesis was provided by demonstration that feeding a similarly formulated diet (marginally replete in potassium but high in protein and acidifying) led to the development of hypokalemia, clinical and laboratory evidence of renal dysfunction and renal lesions in cats over a twoyear study period (DiBartola et al., 1993). However, despite a number of small well-controlled and detailed studies, the causal relationship between whole body potassium deficit and progressive renal injury remains to be proven. Whilst it appears possible to induce renal injury by prolonged feeding of acidifying diets marginally replete in potassium, in general it seems most likely that hypokalemia associated with renal disease is mild and occurs as a result, rather than being a major cause of progressive renal disease.

Urinary excretion of potassium (calculated as fractional excretion) increases with increasing severity of renal dysfunction (Figure 18). In some cases, the fractional excretion of potassium can exceed 100%, indicating the capacity of the cortical connecting tubule to up-regulate potassium secretion in adaptation to nephron loss. Hypokalemia is found at all azotemic stages of CKD in our clinical patients with:
- 30% (6 of 20) in stage IV
- 25% (5/20) in stage III
- and 14.3% (3/21) in stage II (Elliott et al., 2003a).

Figure 18. Box and whisker plots showing fractional excretion of potassium in normal cats and cats with ckd at diagnosis. Data taken from cases included in Elliott et al., 2003a.

As stated previously, basing the assessment of whole body potassium status solely on plasma potassium concentration may under-estimate the prevalence (Theisen et al., 1997). The higher prevalence at the later stages of CKD is likely to be associated with metabolic acidosis which is more likely to occur at this stage of CKD. In our clinical case-load, the diets consumed by the cats presenting with hypokalemia tend to be standard adult maintenance formulations which are in no way limited in the amount of potassium they supply. Furthermore, hypokalemia in these cases is relatively mild (plasma potassium concentrations usually between 3.0 and 3.4 mmol/L; reference range 3.5 to 5.5 mmol/L) and is not generally associated with overt clinical signs (e.g., severe muscle weakness). Clinical improvements are seen in response to potassium supplementation in these cases, including increased appetite and improved level of activity. However, changes in renal function (as assessed by serial measurements of plasma creatinine concentration) are not seen in response to potassium supplementation alone.

In human medicine, observational studies have shown an inverse relationship between dietary potassium intake and blood pressure in some (Reed et al., 1985) but not all studies (Walsh et al., 2002). Randomized controlled clinical trials have shown that potassium supplementation reduces both diastolic and systolic blood pressure in human patients (Whelton et al., 1997). The observation that low plasma potassium concentration increased the risk of hypertension in cats with CKD led us to conduct a randomized controlled clinical trial to determine the effect of potassium supplementation on blood pressure in cats with naturally occurring CKD (Elliott & Syme, 2003). The trial was also designed to determine the benefits on general well-being (as assessed by body weight) and renal function (as assessed by serial plasma creatinine concentration). The supplement used was potassium gluconate at a dose of 2 mEq per cat twice daily as this formulation anecdotally is one of the best tolerated by cats. We chose to evaluate this supplement against corn starch rather than another salt of gluconate. Gluconate is a bicarbonate precursor and might assist in replenishing intracellular stores of potassium by addressing a sub-clinical metabolic acidosis exacerbating potassium loss from the body.

The trial was a prospective, randomized placebo controlled cross-over study with each phase lasting three months. Cases were selected that were in stages II or III CKD that had been on a stable diet for three months prior to enrolment. Cats treated for hypertension were excluded from the study, as were cats with plasma potassium concentrations <3.0 mmol/L. A total of 17 cats were evaluated in this protocol. The plasma potassium concentration (4.35 [4.21, 4.66] vs. 4.16 [3.92, 4.38] mmol/L) and the urine pH (6.08 [5.66, 6.51] vs. 5.63 [5.42, 5.96]) were significantly higher when the cats were taking the potassium supplement indicating at least partial compliance of the cases entered into the study. No beneficial effect of this level of potassium gluconate supplementation was detected on blood pressure or kidney function (as assessed by serial plasma creatinine concentrations and urine protein to creatinine ratio). This study was assessed on the basis of intention to treat. The major reason for owners withdrawing their cats from the study was that their cat would not eat the supplement (potassium gluconate or placebo).

In summary:

• Cats with CKD adapt to nephron loss by up-regulating potassium ion excretion. In some cases this can lead to excess urinary potassium loss and hypokalemia
• Hypokalemia occurs in about 20% of CKD cases and is found at all stages of this syndrome and clinical benefits are seen from correction of this electrolyte abnormality, particularly when the plasma potassium concentration is less than 3.0 mmol/L
• Severe hypokalemia can occur with feeding of acidifying diets which are marginally potassium replete and the feeding of these diets has been associated with the development of renal lesions – this appears to be a relatively uncommon cause of renal damage in cats in the UK
• Supplementation of dietary potassium by the addition of potassium gluconate (4 mEq /cat/day) for three months to cats with plasma potassium concentrations of >3.0 mmol/L did not result in any measurable clinical benefit on blood pressure or renal function in cats with naturally occurring stage II and stage III CKD
• Ensuring cats with CKD are fed rations which provide potassium in excess of requirements and which are not acidifying should avoid problems of hypokalemic nephropathy in cats. Routine additional supplementation with potassium (over that which is provided in renal care diets) does not appear to be necessary for the majority of cases.

Dietary Management of Proteinuria

The intact nephron hypothesis proposed by Hostetter et al, (1981) has shaped research into progression of CKD in the last 20 to 30 years. This hypothesis was based on observations involving experimental rats using surgical reduction of renal mass to mimic the loss of functioning nephrons that occurs in clinical kidney diseases. The observations that with surgical renal mass reduction results in the following adaptations in the remaining functioning nephrons form the basis of the intact nephron hypothesis.

These adaptations to loss of functioning nephrons appear to compensate for the reduction in the number of filtrating nephrons (Figure 3). Ultimately, these adaptations are thought of as being mal-adaptive since glomerular hypertension and proteinuria have been shown to lead to glomerulosclerosis and demise of the remaining functioning nephrons, particularly in rat nephrectomy models, where progression is rapid and closely related to the degree of proteinuria.

Intact Nephron Hypothesis

1. Hypertrophy – the remaining nephrons increase in size
2. Glomerular capillary hypertension – these nephrons function at a higher glomerular capillary pressure, increasing one of the forces for filtration
3. Hyperfiltration – as a result of the increased glomerular capillary pressure the filtration rate per individual nephron increases, partially compensating for the loss of the functional renal mass
4. Increased amounts of protein entering the glomerular filtrate and being excreted in the urine (proteinuria)
5. Increased protein entering the filtrate is indicative of glomerular hypertension but also overloads the tubular resorptive processes for protein. This stimulates tubular cells to secrete inflammatory and profibrotic mediators into the interstitial compartment, possibly stimulating interstitial fibrosis and inflammation and contributing to progressive renal damage.

Similar experimental models can be established using cats. Adaptive changes in feline nephrons following renal mass reduction include glomerular capillary hypertension, associated hyperfiltration and mild proteinuria (Brown and Brown, 1995). Functional progression of this feline model to severe end-stage kidney failure is much slower than the rat model and so interventions to slow that progression are more difficult to assess. Proteinuria has been used in the rat model as the hallmark of progressive renal injury, either as a marker of glomerular and/or tubular health or as a mediator of tubular damage.

Hyperfiltration and glomerular capillary hypertension appears to be driven, at least in part, in the surgical reduction models by local activation of the RAAS. In the face of afferent arteriolar vasodilation, this system leads to constriction of the efferent arteriole, glomerular capillary hypertension and exacerbates transglomerular passage of plasma proteins, the most abundant of which is albumin.

Leakage of protein into the glomerular filtrate has been implicated in causing renal pathology. Proteins that transfer across the glomerulus are normally taken back up by the proximal tubule through a process called pinocytosis, whereby the protein molecule is engulfed into a pinocytotic vesicle which buds off from the plasma membrane. This vesicle then fuses with a lysosome inside the cell, containing enzymes which break down the protein to its constituent amino acids which are returned to the plasma. Increasing the traffic through this uptake pathway seem to cause the proximal tubular cell to become overwhelmed with proteins taken up from the filtrate. This stimulates the cell to secrete a number of inflammatory cytokines from its basolateral surface, including endothelin-1, monocyte chemotractant protein-1 (MCP-1) and RANTES, leading to interstitial inflammation and fibrosis as a response to the proteinuria (Remuzzi & Bertani 1998) (Figure 19).

Figure 19. Pathogenesis of interstitial fibrosis. From Remuzzi and Bertani (1998).

Canine and human CKDs tend to be more proteinuric than feline CKDs. For example, in one pathological study, more than 50% of the dogs appear to have primary glomerular pathology (MacDougall et al., 1986). In cats, the pattern of pathology is predominantly interstitial inflammation and fibrosis with glomerulosclerosis occurring as a consequence of the CKD rather than as a primary disease process (Lucke, 1968). Loss of protein in the urine giving rise to urine protein to creatinine ratios greater than 2, usually indicative of primary glomerular pathology, is an uncommon finding in cats with CKD (Lees et al., 2005). Nevertheless, studies have underlined the importance of mild renal proteinuria in cats with CKD as a predictor of all cause mortality (King et al., 2006; Syme et al., 2006) and uremic crisis (Kuwahara et al., 2006).

Data from one of these studies (Syme et al., 2006) are presented in Figure 20. This study involved longitudinal follow-up of 94 cats from initial diagnosis of chronic kidney disease together with 28 aged-matched normal healthy cats and 14 aged cats with hypertension (systolic blood pressure >175 mmHg) but plasma creatinine concentrations within the laboratory reference range. The healthy aged normal cats used in this study defined a reference range for urine protein to creatinine ratio, the upper limit of which was 0.4. Multivariate regression analysis was used to identify risk factors at entry to the study that were associated with proteinuria. The variables identified were plasma creatinine concentration (the higher the creatinine the more likely the cats were to be proteinuric) and blood pressure. Survival analysis was undertaken using Cox’s regression analysis. Age, plasma creatinine and proteinuria (assessed by urine protein to creatinine ratio) were significant and independent risk factors associated with reduced survival time. No attempt was made in this study to determine cause of death as this is often difficult to define in aged cats with multiple problems.

Figure 20. Survival curves demonstrating the effect of urine protein to creatinine ratio and all cause mortality in cats with CKD. Reproduced with permission from Syme et al, (2006); From J Vet Intern Med 2006 Permission of reproduction granted by Blackwell Publishing.

The results of this study had been presented in abstract form prior to the full publication and were used to inform the American College of Veterinary Internal Medicine (ACVIM) Consensus Statement on proteinuria (Lees et al., 2005).

It is clear in cats with CKD as the number of functioning nephrons decreases (and plasma creatinine concentration increases) so the proteinuria worsens. This phenomenon has been confirmed by longitudinal studies of progressive CKD in feline patients (Hardman et al., 2004). The increase in UPC with progressive kidney disease probably underestimates the significance of the hyperfiltration that is occurring as progression occurs. This is because as the number of functioning nephrons decreases, so the surface area over which protein can be lost also decreases, tending to offset the amount of protein lost. The ACVIM Consensus statement on proteinuria recommends treatment for renal proteinuria should commence for azotemic cats when UPC exceeds 0.4. It should be accompanied by intensive investigation of factors that might cause or exacerbate proteinuria and extensive monitoring of the proteinuria to determine whether the prescribed treatments are effective.

Anti-proteinuric Therapy

As proteinuria seems to be a significant risk factor for reduced survival in cats with CKD it seems logical that treatments that reduce proteinuria should be prescribed when persistent proteinuria is identified in association with CKD. Specific treatment should be recommended when UPCs >0.4 are documented in an azotemic cat on 2 or more occasions in the absence of evidence of inflammation on urine sediment examination. The anti-proteinuric treatment for which there is greatest evidence of efficacy is ACE inhibitor therapy. Benazepril is authorized for use in the cat in Europe and has been shown to reduce glomerular capillary pressure in a renal reduction model in the cat (Brown et al., 2001) and to lower UPC in naturally occurring CKD in a randomized controlled masked clinical trial (King et al., 2006).

Dietary interventions designed to reduce proteinuria include:

• Feeding a reduced quantity of high quality protein
• Supplementing n-3 polyunsaturated fatty acids to produce a diet enriched in this component relative to n-6 polyunsaturated fatty acids

IRIS Staging System on Proteinuria for Cats

- UPCs <0.2 are considered normal,
- between 0.2 and 0.4 are considered borderline proteinuric
- >0.4 are considered to be proteinuric

Restriction of Protein Intake

Each time a meal of protein is consumed, renal hemodynamics are altered and glomerular filtration rate increases to an extent which depends on the quantity and nature of the protein fed. Restricting protein intake should limit these feeding related hyperfiltration responses. Much controversy surrounds the efficacy of reducing dietary protein intake as a means of managing proteinuria in both dogs and cats. In experimental models in rats, this approach proved highly successful in limiting proteinuria and slowing the rate of decline of renal function and progression of renal lesions in the renal mass reduction model (Brenner et al., 1982) and so was recommended for use in other species. Similar studies were conducted in cats, initially with results suggestive of a beneficial effect of protein restriction on glomerular lesion development in the remnant kidneys (Adams et al., 1993; 1994) although the cats fed a reduced amount of protein (2.7 g/kg/day) in these studies also consumed fewer calories (56 calories/kg/day) than the comparator group fed a higher quantity of protein (75 calories/kg/day and 6.8 g protein/kg/day). Furthermore, cats in the low protein group had evidence of protein malnutrition with reduced serum albumin concentration by the end of the study. In a subsequent study addressing the same question, the effect of calorie intake was distinguished from the effect of limiting protein intake and a markedly different pattern of renal lesions resulted with no evidence of a beneficial effect of restricting protein intake (Finco et al., 1998).

One problem with the model used in these two studies is that functional progression (progressive decline in GFR) is not evident over the 12 month post-surgery follow-up period, regardless of the diet that was fed. In the study reported by Finco et al (1998), surgical renal reduction caused cats to develop borderline proteinuria (UPC 0.24 to 0.27) whereas pre-surgery they were non-proteinuric (UPC 0.06 to 0.08). No significant difference in UPC was noted between any of the 4 groups of cats used in this study, thus diet had no effect on UPC. Renal histology of the remnant kidneys did, however, reveal a beneficial effect of reducing calorie (but not protein) intake on the severity of renal interstitial (but not glomerular) lesions. Cats in the low calorie intake groups consumed 55 and 58 calories/kg/day and those in the high calorie intake groups consumed 73 and 71 calories/kg/day. Protein consumption was 5.2 and 5.3 g/kg/day in the low protein diet groups and 9 g/kg/day in the high protein diet groups.

The differences between the results of these two studies are striking and are extensively discussed by Finco et al (1998), including:
- The source of protein (predominantly animal protein in the Adams et al, (1994) study whereas vegetable proteins made a major contribution to the diets fed in the Finco et al (1998) study)
- Dietary potassium which was lower in the Adams et al (1994) study (with cats developing hypokalemia when consuming the high protein diet)
- And dietary lipids which provided a higher proportion of the calories in the Adams et al (1994) study.

It is difficult to extrapolate from these two studies recommendations that can be confidently applied to stage II and III cats in terms of diets which will limit proteinuria and therefore possibly slow progressive renal injury by the mechanisms referred to above. Avoidance of diets which deliver a high quantity of animal protein would seem logical. Most renal diets formulated to limit phosphate intake will avoid excessive animal protein in their formulation. Restricting protein intake per se at these stages of kidney failure in the absence of other dietary modifications commonly encountered in renal care diets have not been investigated in the cat. By extrapolation from other species, the cases that are most likely to benefit from the potential renal hemodynamic modifying effects of dietary protein reduction are those with relatively marked proteinuria (UPCs >1.0).

Supplementation of N-3 Polyunsaturated Fatty Acids

Dietary lipids impact a variety of important parameters, including plasma cholesterol concentration and cell membrane structure. In people, hypercholesterolemia and hypertriglyceridemia are important risk factors for cardiovascular and renal disease. This does not appear to be the case in cats, at least partially because they possess only small amounts of low density lipoprotein (LDL) particles, which have been implicated, in their oxidized form, in human cardiovascular and renal disease progression.

However, there is potential in dogs, and possibly cats, for alterations in cell membrane structure through dietary lipid manipulations, specifically by altering the type of polyunsaturated fatty acid (PUFA) present in the diet. The manipulation that has been most well studied in dogs is alteration of the dietary ratio of n-6 PUFA (plant oils) to n-3 PUFA (fish oils). The n-6 and n-3 PUFA are incorporated into cell membrane phospholipids to serve as precursors for eicosanoids of importance in the renal vasculature, such as prostaglandin E2 and thromboxane A2. Altering the dietary n-6/n-3 ratio was hypothesized to be a nutritional method for altering renal hemodynamics in an effort to provide renoprotection, limiting the maladaptive hyperfiltration discussed above.

The cats that are most likely to benefit from the potential renal hemodynamic modifying effects of dietary protein reduction are those in stage II and II CKD with relatively marked proteinuria (UPCs >1.0).

Support for this hypothesis has been provided by studies in dogs using surgical renal reduction as a model of CKD. Feeding a diet markedly enriched in long chain n3-PUFAs lowered glomerular capillary pressure, reduced proteinuria and slowed progressive decline in GFR seen in this model ( [Brown et al., 1998].). By contrast, feeding a diet markedly enriched in n6-PUFAs raised glomerular capillary pressure, increased proteinuria and caused an accelerated rate of decline in GFR in the same renal reduction model ( [Brown et al., 2000].). These studies used extreme levels of PUFA supplementation but provide the proof of concept for the application of dietary manipulations adopted by some renal care diets where dietary lipids have been manipulated to provide a favorable n6:n3 PUFA ratio, generally achieved by the addition of fish oils. No such data are available in cats, which have somewhat unique PUFA metabolism. Providing long chain n-3 PUFA (eicosapentaenoic acid [EPA] and docosahexaenoic acid [DHA]) is probably even more important in cats than in dogs because the delta 6 desaturase is deficient in feline species. Renal care diets have been produced for cats where dietary lipids have been similarly manipulated.

One such diet has been used in a randomized controlled masked clinical trial in cats with naturally occurring CKD and was shown to be superior to a standardized maintenance diet when fed to cats in stage II and stage III CKD in preventing uremic crises and renal related deaths over a 2 year study (Ross et al., 2006). This positive beneficial effect was not associated with a detectable reduction in UPC in the group receiving the renal care diet. As discussed above, the renal care diet used in this study was also restricted in protein, phosphate and sodium as well as having a different lipid profile when compared to the standard maintenance diet against which it was compared. Further studies are necessary to determine whether n-3 PUFA supplementation to cats is effective in the management of proteinuria in cats and to determine what effect their use has on progression of CKD in the cat.

Other Dietary Manipulations Designed to Slow Progressive Renal Injury

The dietary manipulations discussed above are the main approaches used in the formulation of renal care diets for CKD in cats. There are, however, a number of newer approaches for which a rationale could be put forward based on extrapolation from data derived in other species. Much interest has centered on the phenomenon of endothelial cell dysfunction and the role this plays in progression of CKD in human patients. Endothelial cells line the entire cardiovascular system and they produce a plethora of mediators which in a healthy situation:

• Maintain a thromboresistant surface
• Produce tonic vasodilation of underlying smooth muscle cells to counter- balance vasoconstrictor mediators produced locally or present in the circulation
• Resist leucocyte adhesion and migration in the absence of major inflammatory stimuli
• Inhibit inappropriate smooth muscle and fibroblast proliferation

In some disease states, endothelial cell dysfunction is thought to contribute to the chronic and progressive nature of the disease (Figure 21). Examples include congestive heart failure, hypertension, cardiovascular complications that accompany diabetes mellitus and kidney diseases. In human patients and in some experimental models of CKD there is strong evidence to support the role of endothelial cell dysfunction in systemic hypertension, glomerular pathology, progressive proteinuria and tubular interstitial inflammation and fibrosis. In human patients, CKD is a major risk factor for cardiovascular disease and cardiovascular complications are a common cause of mortality.

Figure 21. Typical microscopic features of feline chronic tubulointerstitial nephritis. © Unité d’anatomie pathologique ; École Nationale Vétérinaire de Toulouse.

Endothelial cell dysfunction in renal disease may result from:

• Dyslipoproteinemia associated with disturbances in cholesterol metabolism
• Accumulation of inhibitors of endothelial nitric oxide synthase (principally asymmetric dimethylarginine [ADMA]) as a result of reduced renal excretion of ADMA and reduced catabolism by dimethylarginine dimethylamino-hydrolase as a result of oxidative stress (Baylis, 2006)
• Reduced renal synthesis of L-arginine, the amino acid substrate required for the synthesis of nitric oxide by the endothelium
• Increased oxidative stress which accompanies CKD and results in:

• Reduced bioavailability of nitric oxide released from the endothelium
• Stimulation of production of profibrotic, promitotic and vasoconstrictor mediators by the endothelium (e.g., endothelin-1, thromboxane A2 and hydrogen peroxide)

Although there is little published work relating to the relevance of these factors in progressive CKD in cats, some data have been published in abstract form supporting the problems of oxidative stress in naturally occurring feline CKD (Braun, 2000; personal communication) and the accumulation of ADMA in stages II, III and IV of CKD (Jepson et al., 2008), where the plasma concentration of ADMA correlated closely to the plasma creatinine concentration.

There are a number of dietary therapeutic approaches to correcting endothelial cell dysfunction associated with CKD. None of these approaches have been studied in the cat and their application to cats with CKD remains speculative at present. Possible approaches include:

• Supplementation of dietary L-arginine to boost the nitric oxide (NO) system, overcome inhibition induced by ADMA
• Dietary supplementation with flavanols (Figure 22) which have been shown to boost endothelial production of nitric oxide and improve endothelial cell health generally. By trapping free radicals, flavanols have a protective function in areas of necrosis that occur in the glomeruli following alternating ischemia-reperfusion arising from circulatory disorders that occur in CKD.
  The anti-hypertensive action of flavanols is due to several combined effects:

• Relaxation of smooth muscle fibers (Duarte et al., 1993; Huang et al., 1998). This property is beneficial in augmenting the filtration rate in surviving nephrons when functional renal tissue has decreased
• Stimulation of endogenous production of NO from arginine (Chevaux et al., 1999; Duarte et al., 2002). Nitric oxide is responsible for local vasodilation
• Inhibition of angiotensin converting enzyme (ACE), which has an important role in vasoconstriction (Hara et al., 1987; Cho et al., 1993)

• Use of diets enriched in antioxidants (e.g., vitamin E, vitamin C, taurine, lutein, lycopene, betacarotene etc.), adressing the balance between pro- and antioxidants and correcting the problem of oxidative stress in CKD.

Figure 22. Origin of flavanols.

Effective measures that address the problems of endothelial cell dysfunction are being actively sought for human medicine and some of the approaches listed above have shown promise. Endothelial cell dysfunction clearly complicates both the early stages of CKD as well as the end stage when renal replacement therapy is necessary and cardiovascular complications are a major cause of morbidity and mortality. Whether these measures will prove of benefit in cats with CKD and at what stage of the syndrome they are best applied remains to be determined.

Role of Fiber

Fermentable fiber is a recent addition to the nutritional management of CKD. It is hypothesized that the fermentable fiber provides a source of carbohydrate for gastrointestinal bacteria which consequently utilize blood urea as a source of nitrogen for growth. The increase in bacterial cell mass increases fecal nitrogen excretion and has been suggested to decrease the blood urea nitrogen concentration. However, unlike BUN, the classical uremic toxins (middle-molecules) are too large in molecular size to readily cross membrane barriers. As a consequence, it is highly unlikely that these toxins are reduced by bacterial utilization of ammonia. Fermentable fibers do have beneficial effects for modulating gastrointestinal health in patients with chronic kidney disease.

Summary

Section 4 of this chapter has dealt with the dietary manipulations commonly used in the production of a renal care diet and discussed them in relation to their application to stage II and early stage III CKD patients. The use of dietary therapy before obvious clinical signs of the uremic syndrome are evident has been somewhat controversial. The main treatment goal in this group of clinical patients is to slow the progression of CKD to stage IV and beyond. The rational basis for dietary modification by:

• Limiting phosphate intake
• Limiting sodium intake
• Supplementing potassium intake
• Limiting protein intake and modifying the lipid composition of the diet has been presented with the evidence for the efficacy of each of these dietary strategies in slowing progressive renal injury reviewed.

Evidence was presented from two prospective trials using renal care diets that clearly indicate that these diets can be beneficial in Stage II and Stage III CKD patients when assessed against the outcome of all cause mortality (Elliott et al., 2000) and time to uremic crisis or renal death (Ross et al., 2006). Although these two studies used diets which adopt a combination of the above dietary modifications and it is not possible to conclude precisely which provided the observed benefits, they do provide strong evidence for dietary intervention at stage II and III CKD in the cat.