**4. Evolution of the controversy over the existence and prevalence of RSW**

The derivation of the controversy regarding the existence and relative prevalence of RSW and SIADH can be appreciated by a brief review of salt balance in normal subjects. Studies in Yanomamo Indians, the "no salt society", support the notion that we require virtually no salt in our diet to maintain normal ECV. (Hollenberg, 1980; Oliver et al, 1975) In Yanomamo Indians, the mean sodium excretion is 1 mmol/day, mean serum sodium 140 mmol/L, mean urine volume 1 L/day and mean blood pressure 102/62 mmHg. (Oliver et al, 1975) These studies suggest that we require little or no salt in our diets to maintain normal ECV.

Normal kidneys appear to have an innate sense of what is a normal ECV for that individual and adjust to any fluctuations in sodium intake to maintain ECV within narrow limits. (Hollenberg, 1980) The adjustments, however, are not instantaneous as sodium excretion will exceed input for up to 5 days before reaching equilibrium after an acute reduction in

Complexity of Differentiating Cerebral-Renal Salt Wasting

wasting. (Schwartz et al, 1957)

(Hoorn et al, 2005)

concentration below UNa. (Schwartz et al, 1957)

from SIADH, Emerging Importance of Determining Fractional Urate Excretion 47

In 1957 Schwartz et al published their seminal report on SIADH that captured the fancy of physiologists and clinicians by reproducing the data in studies of vasopressin injections in healthy subjects by Leaf et al to propose the inappropriate secretion of ADH without the benefit of measuring plasma ADH levels. (Leaf et al, 1953; Schwartz et al, 1957) They proved convincingly that a hyponatremic patient, who presents with a concentrated urine, high UNa and increased ECV, as determined by radiosulfate measurements of ECV, must be due to an inappropriate secretion of ADH. ADH did not respond to the usual volume or osmolar stimuli and thus termed it inappropriate. The hyponatremia that was associated with a high UNa of 70 mmol/L and euvolemia or hypervolemia strengthened by determination of radiosulfate space, defined a syndrome that was not consistent with cerebral or renal salt

There are several characteristics of SIADH that are worth reviewing as they relate to hyponatremia. These patients go into a period of negative sodium balance followed by an equilibrated state when sodium intake matches output. (Janenike & Waterhouse,1961) There is an increased blood volume as determined by sulfate space and by radioiodinated serum albumin and 51Cr labeled red blood cells. (Bitew et al, 2009; Schwartz et al, 1957) The hypervolemia reduces plasma renin and aldosterone and increase plasma A/BNP. (Bitew et al, 2009; Fichman et al, 1974) GFR increases and urine osmolality is invariably concentrated. (Beck, 1979) Urine osmolality can, however, be dilute under circumstances of "ADH escape". Dilute urines have been noted after rapid infusion of saline at 2 L over a 2 hr period and after reducing sodium intake. (Jaenike & Waterhouse, 1961; Schwartz et al, 1957) Several possible explanations for this interesting phenomenon include a down regulation of V2 receptor or increased urine flow rates cannot equilibrate with the hypertonic medulla.

An unappreciated observation is an increase in serum sodium despite high fluid intake in SIADH. A balance study reported an increase in serum sodium from 105 to 135mmol/L over an 8 day period, when the mean fluid intake was 2648 ml/day and mean daily sodium intake of 315 mmol/day. The mean sodium concentration of 124.4 mmol/L in the input fluid was higher than the mean UNa of 86.8 mmol/l over the 8 day period. The sodium concentration in the intake fluid exceeded UNa on every day of the study, suggesting that serum sodium can increase even in SIADH as long as sodium concentration in the intake fluid exceeds UNa, regardless of the intake volume. (Schwartz et al, 1957) This reasoning can be applied to desalination when saline infusion decreases serum sodium. (Steele et al,1997). In the first case of SIADH, serum sodium decreased from 121 to 114 mmol/L after saline infusion when UNa was 70 mmol/L and to 103 mmol/L after hypertonic saline before undergoing a metabolic study. It is unlikely that serum sodium decreased while receiving saline, with a sodium concentration of 155 mmol/L, when UNa was 70 mmol/L. This is not consistent with desalination. (Steele et al, 1997) The best explanation for this phenomenon is an unrecorded intake of water that decreased the input sodium

These elegantly designed studies in the initial report of SIADH proved that a hyponatremic patient can have high UNa without invoking RSW, largely because the volume status was shown to be increased by credible methods of determining ECV and not by tenuous clinical criteria as in the original report. (Peters et al, 1950) The existence of cerebral salt wasting was appropriately questioned. Since the assessment of ECV is critical in differentiating SIADH from RSW, it would be appropriate to review the various methods by which we assess ECV.

sodium intake. (Valtin, 1997) When a normal subject is placed in negative sodium balance by increasing urinary sodium excretion by diuretics or increased sweating, urinary sodium excretion decreases to as low as 1 mmol/day. (McCance, 1936; Strauss et al, 1958) Sodium excretion does not increase until the sodium losses have been replenished. (McCance, 1936; Strauss et al, 1958) This important role of normal kidneys to conserve sodium, when in a state of negative sodium balance is, in retrospect, the basis for the birth of the term cerebral salt wasting syndrome in 1950. (Peters et al, 1950) Peters et al reported 3 subjects with cerebral disease, acute encephalitis, subarachnoid hemorrhage and bulbar poliomyelitis. They concluded unconvincingly that these patients presented evidence of salt wasting, which was characterized by nitrogen retention, low blood pressure and correction of their hyponatremia by large salt intake. Nitrogen retention occurs in a volume depleted patient, referred to as prerenal azotemia, with retention of urea or nonprotein nitrogen (NPN), approximately double the BUN. (Abuelo, 2007) This is a reasonable assumption because urea excretion increases with any increase in urine output, even in RSW during volume repletion. (Shannon, 1936) NPN decreased from a baseline 44 to 25 mg/dL after receiving large amounts of salt in the first case. Only one NPN determination was reported in the second case with SAH and none was reported in the third case with bulbar poliomyelitis. The blood pressure in the first patient was 110/70 mmHg when the NPN was 44 mg/dL with preceding blood pressures of 120/80 to 130/88 mmHg without testing for postural changes in blood pressure or pulse. The second patient with SAH had one blood pressure reading of 220/110 mmHg and none was reported for the third case. The hyponatremia in all 3 patients did not respond to long periods of increased salt intake. The salt balance study that lasted 39 hours revealed the patient to be in negative sodium balance after salt intake was acutely reduced from 15 g/day to no salt intake. This delay in reaching equilibrium on the third day after an acute reduction in salt intake was construed as salt wasting, but is actually consistent with observations made in normal subjects. (Valtin, 1997) The negative sodium balance for 39 hours after an acute reduction in sodium intake does not justify the diagnosis of salt wasting. The first "dehydrated"case, however, could have had salt wasting. McCance and Strauss et al reported that a volume depleted subjects would avidly conserve sodium until their sodium losses were replaced. (McCance, 1936; Strauss et al, 1958) This "dehydrated" or assumed volume depleted patient had a urine chloride of 61.6 mmol/L, which can be explained by RSW. The inability to assess clinically the state of ECV has been the basis for doubting the existence of RSW. The same shortcomings were repeated in another report by the same authors on salt wasting. (Welt et al, 1952)

Four years later, Cort reported a hyponatremic patient with astrocytoma and papilledema, who had signs of dehydration. (Cort, 1954) Sodium intake of 15 g/day for many days failed to correct the hyponatremia. In a nine-day balance study, sodium intake was acutely reduced to 142.5 mg/day. The patient received corticotrophin on days 4, 5 and 6, deoxycorone on days 7, 8 and 9 and restarted on 15 mg/day salt on day 10. The patient went into negative sodium balance of 100 mmol on the first day and 60-70 mmol/day for the next 8 days. Sodium balance was unaffected by corticotrophin or deoxycortone. This study was compared to a similar study by McCance, who found normal subjects to go into sodium balance by the 5th day. (Cort, 1954) Determinations of daily chloride space revealed a 1.4 L reduction on the first day and 690 ml on the 9th day. Resumption of 15 g/day salt intake increased the chloride space by 1 L and her serum sodium "restored toward normal". (Cort, 1954) The reduction in chloride space and prolonged negative sodium balance prove the existence of RSW.

sodium intake. (Valtin, 1997) When a normal subject is placed in negative sodium balance by increasing urinary sodium excretion by diuretics or increased sweating, urinary sodium excretion decreases to as low as 1 mmol/day. (McCance, 1936; Strauss et al, 1958) Sodium excretion does not increase until the sodium losses have been replenished. (McCance, 1936; Strauss et al, 1958) This important role of normal kidneys to conserve sodium, when in a state of negative sodium balance is, in retrospect, the basis for the birth of the term cerebral salt wasting syndrome in 1950. (Peters et al, 1950) Peters et al reported 3 subjects with cerebral disease, acute encephalitis, subarachnoid hemorrhage and bulbar poliomyelitis. They concluded unconvincingly that these patients presented evidence of salt wasting, which was characterized by nitrogen retention, low blood pressure and correction of their hyponatremia by large salt intake. Nitrogen retention occurs in a volume depleted patient, referred to as prerenal azotemia, with retention of urea or nonprotein nitrogen (NPN), approximately double the BUN. (Abuelo, 2007) This is a reasonable assumption because urea excretion increases with any increase in urine output, even in RSW during volume repletion. (Shannon, 1936) NPN decreased from a baseline 44 to 25 mg/dL after receiving large amounts of salt in the first case. Only one NPN determination was reported in the second case with SAH and none was reported in the third case with bulbar poliomyelitis. The blood pressure in the first patient was 110/70 mmHg when the NPN was 44 mg/dL with preceding blood pressures of 120/80 to 130/88 mmHg without testing for postural changes in blood pressure or pulse. The second patient with SAH had one blood pressure reading of 220/110 mmHg and none was reported for the third case. The hyponatremia in all 3 patients did not respond to long periods of increased salt intake. The salt balance study that lasted 39 hours revealed the patient to be in negative sodium balance after salt intake was acutely reduced from 15 g/day to no salt intake. This delay in reaching equilibrium on the third day after an acute reduction in salt intake was construed as salt wasting, but is actually consistent with observations made in normal subjects. (Valtin, 1997) The negative sodium balance for 39 hours after an acute reduction in sodium intake does not justify the diagnosis of salt wasting. The first "dehydrated"case, however, could have had salt wasting. McCance and Strauss et al reported that a volume depleted subjects would avidly conserve sodium until their sodium losses were replaced. (McCance, 1936; Strauss et al, 1958) This "dehydrated" or assumed volume depleted patient had a urine chloride of 61.6 mmol/L, which can be explained by RSW. The inability to assess clinically the state of ECV has been the basis for doubting the existence of RSW. The same shortcomings were repeated in

another report by the same authors on salt wasting. (Welt et al, 1952)

existence of RSW.

Four years later, Cort reported a hyponatremic patient with astrocytoma and papilledema, who had signs of dehydration. (Cort, 1954) Sodium intake of 15 g/day for many days failed to correct the hyponatremia. In a nine-day balance study, sodium intake was acutely reduced to 142.5 mg/day. The patient received corticotrophin on days 4, 5 and 6, deoxycorone on days 7, 8 and 9 and restarted on 15 mg/day salt on day 10. The patient went into negative sodium balance of 100 mmol on the first day and 60-70 mmol/day for the next 8 days. Sodium balance was unaffected by corticotrophin or deoxycortone. This study was compared to a similar study by McCance, who found normal subjects to go into sodium balance by the 5th day. (Cort, 1954) Determinations of daily chloride space revealed a 1.4 L reduction on the first day and 690 ml on the 9th day. Resumption of 15 g/day salt intake increased the chloride space by 1 L and her serum sodium "restored toward normal". (Cort, 1954) The reduction in chloride space and prolonged negative sodium balance prove the In 1957 Schwartz et al published their seminal report on SIADH that captured the fancy of physiologists and clinicians by reproducing the data in studies of vasopressin injections in healthy subjects by Leaf et al to propose the inappropriate secretion of ADH without the benefit of measuring plasma ADH levels. (Leaf et al, 1953; Schwartz et al, 1957) They proved convincingly that a hyponatremic patient, who presents with a concentrated urine, high UNa and increased ECV, as determined by radiosulfate measurements of ECV, must be due to an inappropriate secretion of ADH. ADH did not respond to the usual volume or osmolar stimuli and thus termed it inappropriate. The hyponatremia that was associated with a high UNa of 70 mmol/L and euvolemia or hypervolemia strengthened by determination of radiosulfate space, defined a syndrome that was not consistent with cerebral or renal salt wasting. (Schwartz et al, 1957)

There are several characteristics of SIADH that are worth reviewing as they relate to hyponatremia. These patients go into a period of negative sodium balance followed by an equilibrated state when sodium intake matches output. (Janenike & Waterhouse,1961) There is an increased blood volume as determined by sulfate space and by radioiodinated serum albumin and 51Cr labeled red blood cells. (Bitew et al, 2009; Schwartz et al, 1957) The hypervolemia reduces plasma renin and aldosterone and increase plasma A/BNP. (Bitew et al, 2009; Fichman et al, 1974) GFR increases and urine osmolality is invariably concentrated. (Beck, 1979) Urine osmolality can, however, be dilute under circumstances of "ADH escape". Dilute urines have been noted after rapid infusion of saline at 2 L over a 2 hr period and after reducing sodium intake. (Jaenike & Waterhouse, 1961; Schwartz et al, 1957) Several possible explanations for this interesting phenomenon include a down regulation of V2 receptor or increased urine flow rates cannot equilibrate with the hypertonic medulla. (Hoorn et al, 2005)

An unappreciated observation is an increase in serum sodium despite high fluid intake in SIADH. A balance study reported an increase in serum sodium from 105 to 135mmol/L over an 8 day period, when the mean fluid intake was 2648 ml/day and mean daily sodium intake of 315 mmol/day. The mean sodium concentration of 124.4 mmol/L in the input fluid was higher than the mean UNa of 86.8 mmol/l over the 8 day period. The sodium concentration in the intake fluid exceeded UNa on every day of the study, suggesting that serum sodium can increase even in SIADH as long as sodium concentration in the intake fluid exceeds UNa, regardless of the intake volume. (Schwartz et al, 1957) This reasoning can be applied to desalination when saline infusion decreases serum sodium. (Steele et al,1997). In the first case of SIADH, serum sodium decreased from 121 to 114 mmol/L after saline infusion when UNa was 70 mmol/L and to 103 mmol/L after hypertonic saline before undergoing a metabolic study. It is unlikely that serum sodium decreased while receiving saline, with a sodium concentration of 155 mmol/L, when UNa was 70 mmol/L. This is not consistent with desalination. (Steele et al, 1997) The best explanation for this phenomenon is an unrecorded intake of water that decreased the input sodium concentration below UNa. (Schwartz et al, 1957)

These elegantly designed studies in the initial report of SIADH proved that a hyponatremic patient can have high UNa without invoking RSW, largely because the volume status was shown to be increased by credible methods of determining ECV and not by tenuous clinical criteria as in the original report. (Peters et al, 1950) The existence of cerebral salt wasting was appropriately questioned. Since the assessment of ECV is critical in differentiating SIADH from RSW, it would be appropriate to review the various methods by which we assess ECV.

Complexity of Differentiating Cerebral-Renal Salt Wasting

circulation. (Wijdicks et al, 1985)

in neurosurgical patients, especially SAH.

proper "control" and hyponatremic groups.

diagnosis of RSW. (Cusano et al, 1990; Maesaka et al, 1990)

from SIADH, Emerging Importance of Determining Fractional Urate Excretion 49

determined by radioiodinated serum albumin in 21 patients on the first day of admission within 48 hours after SAH and on the 6th day after SAH. Comparisons between the first and second volume determination revealed blood volume to decrease in 8 of 9 hyponatremic patients, suggesting that 88.9% had RSW and 11.1% had SIADH. UNa values were not reported. (Wijdicks et al, 1985) Interestingly, plasma volume was decreased in 8, 66.7%, and increased in 4, 33.3%, of 12 nonhyponatremic patients with SAH, suggesting that RSW can occur in nonhyponatremic patients. Moreover, all 8 hyponatremic and 8 of 12 nonhyponatremic patients with decreased blood volume were in negative sodium balance. The increased blood volume in the 4 patients with normonatremia raises the question of whether or not SIADH can occur with normonatremia. In a separate study, Water restriction in volume depleted patients with SAH increased morbidity and mortality, probably due to decreased perfusion of brain and extension of ischemia in an already compromised

Another study used 51Cr tagged red cells and CVP measurements in 18 hyponatremic patients of various etiologies with UNa of 43-210 mmol/L. (Sivakumar et al, 1994) Seventeen of 18 patients had decreased blood volumes, 18 of 18 patients had decreased CVP and all 18 patients corrected their hyponatremia within 72 hours after initiating saline infusion. The high UNa, decreased blood volume and correction of hyponatremia within 72 hours after initiating saline therapy argue strongly for RSW. We demonstrated a similar correction of hyponatremia within 48 hours after initiation of saline therapy in two patients with RSW and failure of saline therapy to correct the hyponatremia in two patients with SIADH. (Bitew et al, 2009; Maesaka et al, 2007) The three neurosurgical studies demonstrate by acceptable methods of determining ECV, that RSW is much more common than SIADH

A study in neurosurgical patients determined blood volume by 51Cr labeled red cells in 20 hyponatremic and 20 nonhyponatremic "control" patients. Patients with evidence of "dehydration or hypovolemia" were excluded. All met criteria for SIADH. (Brimioulle et al, 2008) The exclusion of dehydrated or hypovolemic patients might have excluded patients with RSW and meeting the criteria for SIADH would include patients with RSW. Blood volumes were found to be comparable in the "control" and hyponatremic groups, suggesting that the hyponatremic group was entirely SIADH. Interestingly both the "control" and experimental groups were hypouricemic, mean serum urate 2.7 and 1.3 mg/dL and had high mean FEurate of 19% and 32% (normal 5-10%), respectively. There is ample evidence to suggest that the high FEurate in the hyponatremic group was consistent with both SIADH and RSW, while the nonhyponatremic group with increased FEurate would be consistent with RSW or normal controls depending on whether the FEurate was high or normal, respectively, figures 3, 4, table 1. (Maesaka et al, 1996, 1998, 1999, 2009) Based on these diagnostic possibilities, this study suffers first from a protocol-based elimination of patients with evidence of volume depletion as in RSW and failure to select the

Ten patients with acquired immune deficiency syndrome with saline-responsive postural hypotension had CVP of 0 cm water, increased renin and aldosterone, hyponatremia, hypouricemia, elevated FEurate and UNa >40 mmol/l, which collectively support the
