QJM vol. 97 no. 10 © Association of Physicians 2004; all rights reserved.
Masterclasses in medicine |
Acidosis in a patient with cholera: a need to redefine concepts
From the 1Division of Nephrology, Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada, 2McGill University, Montreal, Quebec, Canada, 3Nephrology Unit and Department of Internal Medicine, Stellenbosch University, Cape Town, South Africa, and 4Division of Nephrology, St Michael's Hospital, University of Toronto, Toronto, Canada
 |
Summary |
|---|
 Top Summary IntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
A patient presented with cholera and a severe degree of ECF volume contraction. Despite large losses of bicarbonate (
)-containing diarrhoeal fluid, laboratory acid-base values were remarkably close to normal. A detailed analysis emphasizing principles of physiology and a quantitative approach provided new insights and eventually better definitions of metabolic and respiratory acidosis. A shift in focus from 
concentration to content in the extracellular fluid (ECF) compartment revealed the presence of metabolic acidosis. Central to this analysis was an emphasis on the haematocrit to enable a more accurate estimate of the degree of ECF volume contraction. The latter also revealed ‘contraction’ metabolic alkalosis, which masked the underlying metabolic acidosis. The presence of a respiratory acidosis of the tissue type was evident from the raised venous PCO2, which was not surprising once the magnitude of the ECF contraction had been appreciated. ‘Bad buffering‘, as defined by Professor McCance, was the immediate danger and prompted swift action to restore an effective circulation. The haematocrit and the venous PCO2 also contribute valuable information to monitor the response to therapy. Nevertheless, there were still dangers to be discovered when an in-depth analysis suggested that the administration of isotonic saline would introduce an unanticipated danger for the patient. |
Introduction |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
In this case discussion, once again the central figure is Professor McCance, our imaginary consultant based on the historical figure who practiced medicine
70 years ago. He deals with data from actual clinical cases and applies principles of integrative physiology at the bedside. On this occasion, our Professor will require information that was not available during his medical career—data concerning electrolyte losses in diarrhoeal fluid during a cholera epidemic.1 However, we provide him with assistance from ‘acid-base heaven’ in the form of Dr R.A. Phillips, a prominent clinical investigator who studied patients suffering from cholera.2 As usual, the emphasis will be on concepts that depend on an understanding of physiology that crosses subspecialty boundaries. To avoid overwhelming the reader with details, only key facts are provided, and only when necessary. The overall objective is to demonstrate how the application of simple principles of integrative physiology at the bedside can be extremely helpful in making the correct clinical diagnosis and planning optimal therapy. |
The consultation |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
To describe our patient with cholera, we have used a collage of data drawn from four of Dr Phillip's patients.2 Because there were insufficient data reported in these four cases, we also added data from an additional patient with a similar clinical picture.3 All the patients were healthy prior to this illness, and the medical histories were similar. The initial event was the ingestion of contaminated food and/or water, followed shortly thereafter by the development of profuse watery diarrhoea with a consistency that was like rice water. In very early descriptions, the emphasis was on how markedly contracted the blood volume was.4 On physical examination, our patient appeared to be very ill. There were findings suggestive of a contracted extracellular fluid (ECF) volume—blood pressure was 90/60 mmHg and the heart rate was 110 bpm. There were no other abnormalities reported in the remainder of the examination.
Laboratory investigations (Table 1) did not reveal the expected metabolic acidosis resulting from the loss of NaHCO3—plasma bicarbonate () concentration (PHCO3) was 21 mmol/l, arterial pH 7.36, and arterial PCO2 38 mmHg. The total plasma protein level was 132 g/l and the haematocrit was 60%. The glomerular filtration rate (GFR) was markedly reduced, as evidenced by the high plasma creatinine and urea concentrations.
|
Preliminary impression
The ‘normal-appearing’ arterial blood gas results were troublesome, because a severe degree of metabolic acidosis was anticipated. To explain these findings, the ER consultant suggested that the diarrhoeal fluid might not contain an appreciable amount of NaHCO3 and therefore had not induced an acid-base abnormality. Because the housestaff were not satisfied with this simple interpretation, they called upon Professors McCance and Phillips to help resolve the discrepancy between the clinical findings and the laboratory data. The team would soon recognize that the patient had both metabolic acidosis and respiratory acidosis despite these laboratory data. Of even greater importance, there would be lessons to learn about the dangers of infusing isotonic saline in severely ill patients with cholera.
|
Issues concerning metabolic acidosis |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
Relishing the prospect of tackling another fascinating case, Professor McCance began the discussion by asking a deceptively simple question: ‘What is the best way to diagnose metabolic acidosis?’
Question 1. What is the best way to diagnose metabolic acidosis?
Physiology principle 1: Metabolic acidosis is usually defined as a process that lowers the PHCO3 and blood pH (5).
Return to the bedside: When this definition was applied to our patient, he did not appear to have an appreciable degree of metabolic acidosis due to a deficit of NaHCO3, because of the near-normal values for the PHCO3 and the plasma pH (see Appendix for a discussion of the elevated value for the anion gap in plasma on admission). Everyone found this quite disconcerting—the patient had the typical history and also appeared to have ECF volume contraction. Professor McCance thought aloud: ‘I suspect that either metabolic acidosis is not present or that there is another abnormality masking its presence.’ And so he asked: ‘Could another process be present that leads to the addition of new ?’
Question 2: Could another process be present that leads to the addition of new ?
Physiology principle 2: There are two tools that should be applied to evaluate whether 
was added: first, there must be electroneutrality when the anion is added to the body; second, one must examine balance data to understand why the PHCO3 rose.6
Return to the bedside: There are two ways to add to the ECF compartment while maintaining electroneutrality. First, can be added with a cation other than H+ or ammonium ions (
)—usually sodium (Na+). If NaHCO3 was ingested and retained, the ECF volume would increase. Second, can be added without a gain of Na+ or potassium (K+) while maintaining electroneutrality, providing that there is a simultaneous loss of chloride anions (Cl–) in the form of hydrochloric acid (HCl) and/or ammonium chloride (NH4Cl) (Figure 1). If only this occurred, the ECF volume would be close to normal,7 but a contracted ECF volume could occur, however, if there was also a deficit of sodium chloride (NaCl) and/or potassium chloride (KCl).
|
Professor McCance felt that he could almost certainly rule out alkalosis due to the addition of new
to the ECF compartment. He dealt with this issue as follows: (i) Raise the PHCO3 due to ingestion and absorption of NaHCO3. Given the lack of intake of NaHCO3, the large diarrhoeal losses, the raised plasma creatinine concentration (Table 1), and the prompt fall in this creatinine concentration with an infusion of isotonic saline, the ECF volume is likely to be very contracted. Therefore metabolic alkalosis could not be due to a simple addition of NaHCO3, which would be expected to expand the ECF volume.
(ii) Raise the PHCO3 due to a deficit of HCl. Electroneutrality is maintained in the ECF and intracellular fluid (ICF) compartments when HCl is lost (Figure 1). Because vomiting or nasogastric suction were not prominent features, gastric HCl loss does not seem to play an important role in this story. Although the Zollinger-Ellison syndrome could augment the loss of HCl,8 it is a rare condition and thus an extremely unlikely basis for the metabolic alkalosis in patients with cholera.
(iii) Raise the PHCO3 due to the renal excretion of a large amount of NH4Cl. The high plasma creatinine concentration indicates that the glomerular filtration rate (GFR) is low enough to compromise the kidney's ability to produce and excrete (and generate new ).9 Thus, our patient would be expected to develop metabolic acidosis with a smaller amount of loss in the diarrhoeal fluid, compared to a patient with normal renal function. Also, the absence of chronic metabolic acidosis10 makes it even more unlikely that there is a high rate of NH4Cl excretion.
Professor McCance was not finished with his analysis of the high PHCO3. He asked: ‘Is there another way to raise the PHCO3 other than adding to the ECF compartment?’
Question 3: Is there another way to raise the PHCO3 other than adding to the ECF compartment?
Physiology principle 3: Concentration terms have both numerators and denominators (equation 1)—each must be examined independently.
 | (1) |
Return to the bedside: The PHCO3 will rise when the ECF volume falls. It is a common belief that a significant degree of contraction type of metabolic alkalosis11 is not clinically important unless the patient has the combination of an expanded ECF volume and a normal PHCO3, and is then given a diuretic to induce a marked decline in ECF volume. Not being a follower of common beliefs, Professor McCance set about demonstrating that this mechanism was indeed operative in this patient. He started by asking: ‘What is the most accurate way to assess the ECF volume in these patients?’
Question 4: What is the most accurate way to assess the ECF volume?
Physiology principle 4: There is one easy way to obtain a quantitative estimate of the ECF volume in a patient—measure the concentration of a constituent that is restricted to the ECF compartment and whose content is not altered when the ECF volume changes.
Return to the bedside: Because there is a lack of precision in our clinical ability to make a quantitative assessment of the ECF volume,12 Professor McCance stated that he preferred to use the rise in the haematocrit to assess changes in the ECF volume. The intern asked Professor McCance to explain how he could do this because red blood cells (RBC) were located in a small subdivision of the ECF compartment, the plasma volume. ‘This,‘ said Professor McCance, ‘is an excellent question, and its answer will illustrate another important principle of physiology that may help guide therapy in these patients’.
Question 5: How can the haematocrit be used to assess the ECF volume if RBCs are confined to the circulating volume?
To understand the answer, it is essential to understand physiology principles 5 and 6.
Physiology principle 5: The haematocrit (equation 2) is the ratio of the RBC volume to the blood volume (RBC + plasma).
 | (2) |
Return to the bedside: The normal value for the haematocrit is 40% (0.40) in a healthy young male. A change in haematocrit is the most reliable, simple way to assess changes in vascular volume, and thereby the degree of ECF volume contraction. Although there were no values for the haematocrit before therapy, it was 60% on admission and declined to 40% in the first 2 h after infusing saline—moreover, it remained 40% after therapy.2 As shown in equation 3, when the haematocrit was 60%, the plasma volume had decreased by 56% (1.33 vs 3.0 l), a value similar to that calculated using the total protein concentration.13
 | (3) |
Therefore plasma volume=1.33 l
Question 6: How does one relate the change in the plasma volume to the ECF volume?
Physiology principle 6: After determining the plasma volume, one must infer the change in the interstitial fluid volume by examining the Starling forces across capillary membranes (Figure 2).
Return to the bedside: The major outward-oriented Starling force is the capillary blood pressure. Because this should be low in our patient, there is a diminished tendency for albumin-free fluid to exit these tiny blood vessels. The major inward-oriented Starling force is the colloid osmotic pressure (COP). Because the plasma total protein level was very elevated in our patient (Table 1), there is a larger inward force (the plasma albumin concentration should be elevated by more than two-fold, because the plasma volume was reduced by >50%), yet the albumin concentration in plasma rose by somewhat less than two-fold. Therefore the patient has lost albumin from his ECF compartment. Because diarrhoeal fluid in cholera does not contain an appreciable amount of albumin and little urine was produced, our patient might have redistributed albumin from the plasma to the interstitial compartment for the following reasons. First, albumin crosses capillary membranes through fenestra. This permits the delivery of water-insoluble fatty acids to muscle cells, where they are oxidized.14 Second, albumin returns to vascular compartment via the lymphatics (Figure 3). The driving force for lymphatic flow is the tissue hydrostatic pressure, and valves in lymph vessels ensure that flow proceeds towards the large veins. It is possible that the very low interstitial volume and reduced skeletal muscle activity that would be anticipated in this disease state could be responsible for slower flow in the lymphatic vessels and thereby the redistribution of albumin into the interstitial compartment (Figure 3).
|
|
A medical student asked: ‘Do these changes in Starling forces help explain why the GFR was so low?’
Question 7. Do these changes in Starling forces help explain why the GFR was so low?
Physiology principle 7. To filter albumin-free plasma across glomerular capillaries, one needs a much larger capillary hydrostatic pressure than in other capillary beds (40 vs. 25 mmHg), because 25% of the renal plasma is filtered at the glomerulus.
Return to the bedside: The very high haematocrit leads to a lower hydrostatic pressure in capillaries, because the blood is so viscous.15 Therefore although there is a higher systemic blood pressure (the blood pressure was not a reliable sign of the degree of contraction of the ECF volume in our patient), there is a much larger pressure drop before the renal glomerular capillary bed due to the high viscosity of blood. This may help explain why pre-renal failure is so prevalent in these patients.16 Nevertheless, renal blood flow is likely to be sufficient to deliver enough oxygen to cells of the nephron, because of their low metabolic work (reabsorbing filtered Na+). Hence we must distinguish between a low GFR (pre-renal failure) and tubular cell necrosis (the latter is unlikely to occur at earlier stages of the disease).
Question 8. What should his PHCO3 be if the only influence was the decrease in his ECF volume?
Revisit physiology principle 4. To raise the PHCO3 solely on the basis of a contracted ECF volume, one needs to relate the % decline in the ECF volume to the expected rise in the PHCO3.
Return to the bedside: Because our patient had close to a 56% reduction in ECF volume, his expected PHCO3 should have been >50 mmol/l (>2 x 25 mmol/l) if this were the only influence. Therefore to have a PHCO3 of 21 mmol/l, the content of in the ECF compartment had to decline by >50% due to the loss of NaHCO3 in the diarrhoeal fluid (Figure 4). A similar degree of contraction of the ECF volume in a patient with diabetic ketoacidosis and a normal PHCO3 was recently described.3
|
Dr Phillips then related his experience with patients suffering with cholera, and confirmed that the stool from these patients usually contained
45 mmol per litre, with its concentration being remarkably similar among patients (Table 2). In his series of 11 patients who passed more than 3 l of stool on the first hospital day, seven had metabolic acidosis (average PHCO3 of 18 mmol/l), but four had a PHCO3 that was at the lower limit of the normal range. Dr Phillips made one other important point. He had observed that the stool volume rose appreciably once therapy had begun—to as much as 21 l/day! This volume is >2-fold larger than the entire ECF volume, and implies that there was a very poor circulation to the intestine prior to therapy that limited the secretion of Cl– and Na+ and thereby the diarrhoeal volume. This is similar to the basis for the low GFR in these patients.
|
In summary, our patients with cholera had two simultaneous acid-base disorders: contraction metabolic alkalosis and metabolic acidosis due to the loss of NaHCO3 in diarrhoeal fluid, resulting in a PHCO3 in the near-normal range (Figure 4). Thus definitions of metabolic acidosis should include an assessment of both the concentration and the content of
in the ECF compartment. |
Issues concerning the definition of respiratory acidosis |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
While the housestaff now felt confident in their ability to recognize metabolic acidosis, they were quite puzzled when Professor McCance asked, ‘Might the patient have respiratory acidosis?’ After all, the PHCO3, blood pH, and arterial PCO2 were very close to normal values.
Question 9. Might the patient have respiratory acidosis?
Physiology principle 8. The traditional definition of respiratory acidosis relies on an assessment of changes in ventilation to alter the arterial PCO2 as well as an assessment of the stimulus for ventilation by the H+ concentration (a low PHCO3 and a low blood pH).
Return to the bedside: Because the arterial blood pH and PHCO3 were in the normal range, the appropriate arterial PCO2 should also be in the normal range—it was 38 mmHg (Table 1). There seemed to be no evidence for respiratory acidosis. Professor McCance nodded in agreement but immediately added the following caveat: ‘I wish to emphasize that this patient does not have a ventilatory form of respiratory acidosis’ (Table 3). ‘In metabolic acidosis, however,’ he added, ‘the objective is to lower the PCO2 inside cells to ensure that H+ are buffered by rather than non-bicarbonate buffers’ (Figure 5).17 Professor McCance stressed that it was a disadvantage to buffer H+ by non-bicarbonate buffers, calling this ‘bad buffering’.
|
|
Question 10. What is the difference between ‘good’ and ‘bad’ buffering?
Physiology principle 9. Good buffering means that the tissue PCO2 can fall enough to ensure that most of the added H+ will be removed by binding to
and therefore they will not bind to proteins, altering their charge, shape, and possibly their function.Return to the bedside: The PCO2 in tissues is controlled by factors in addition to the arterial PCO2. These factors are the rate of production of CO2 in cells and the rate of blood flow to remove CO2 from cells (Figure 5). In patients with a significant degree of ECF volume contraction, tissue perfusion may be compromised, impairing the removal of CO2. Accordingly, CO2 accumulation may lead to intracellular respiratory acidosis (tissue form of respiratory acidosis, Table 3).
The following facts suggested that the tissue PCO2 was too high in our patient. CO2 diffuses from cells to the capillary down its concentration difference and therefore the venous PCO2 sets the minimum value for the PCO2 in tissues. Because our patient had a contracted ECF volume and very viscous blood, a very slow capillary blood flow rate should be anticipated. When Professor McCance was shown data from a similar patient with a haematocrit of 60%,3 he pointed out that her venous PCO2 was much higher than in arterial blood (PvCO2 69 mmHg, PaCO2 43 mmHg). Therefore he believed that our patient had a tissue form of respiratory acidosis, despite the normal arterial PCO2 value.
To summarize, there are two types of respiratory acidosis. The first occurs on the arterial side of the circulation. Hypoventilation, whatever its cause, leads to an arterial PCO2 that is inappropriately elevated, results in a decreased arterial pH, and forces H+ to be buffered by non-bicarbonate buffers in all cells. The second, or tissue, form of respiratory acidosis, is suspected on clinical grounds and is confirmed by measuring the PCO2 on the venous side of the circulation. Therefore it affects those tissues with a low blood flow rate. Early on, the brain might be spared because of its autoregulated blood flow.
|
Therapy |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
Three issues are emphasized in therapy, the volume of isotonic infusions, their anionic composition, and the amount of K+ to infuse. One must also bear in mind benefits and potential risks of these therapies. Dr Phillips took the lead in this discussion because of his extensive experience with these patients.2
(i) Assess the volume of isotonic saline required
Question 11. What are the goals of therapy and what are the risks when treating the ECF volume contraction?
Physiology principle 10. To re-expand the ECF volume towards normal, enough of an isotonic solution must be given to partially restore the deficit plus replace the anticipated ongoing losses.
Return to the bedside: The ECF volume was contracted by 56% (>5 l) and the ongoing losses were 3 l/day before therapy.13 Once therapy begins, however, these diarrhoeal losses could rise enormously2—up to 200% of the ECF volume. Therefore, one must be prepared to infuse as much as 25 l/day. Dr Phillips emphasized three reasons why the initial infusion should be rapid. First, the GFR was very low (pre-renal failure); second, venous thrombosis might develop due to the high blood viscosity and slow blood flow rate; third, the high PCO2 in blood drawn from the brachial vein confirms a tissue form of respiratory acidosis in cells drained by this venous system.3 On the other hand, one must also think about potential dangers of this aggressive therapy. Dr. Phillips told the group of his first attempts at aggressive therapy that had disastrous results.18 Of the first 40 patients he treated with intravenous isotonic saline plus an oral rehydration solution (ORS) containing NaCl and glucose, five died of pulmonary oedema. He concluded that while aggressive ECF volume re-expansion is warranted for haemodynamically compromised patients, they must be followed closely, not only for signs of adequate volume re-expansion but also for signs of over-expansion of the central blood volume. Apart from careful and frequent clinical examination, it is useful to follow the fall in the haematocrit, the fall in the venous PCO2 and the rise in urine output.3 This may be particularly important in elderly patients and those with cardiac disease who may be more likely to develop pulmonary oedema.
Route of therapy: Our bias is that with a severe degree of ECF volume contraction, one should give the Na+-containing solution intravenously. Nevertheless, this is not always essential if one can ‘encourage’ the intestinal tract to absorb enough Na+ and Cl–.19 To do this, glucose is added to the ORS, because there is a luminal Na+-linked glucose transporter (SLGT) that uses 2 Na+/glucose—glucose ‘drives’ the electrogenic absorption of Na+ (Figure 6). By giving 1 l of ORS per hour for 6 h, adequate volume therapy can be provided.19
|
(ii) Assess the need to infuse alkali
The content of
in the ECF compartment was reduced by >50% due to loss of in diarrhoeal fluid (Figure 4, Table 2).2,20 If the ECF volume were re-expanded with isotonic saline, the PHCO3 would fall from 21 to 10.5 mmol/l. While a PHCO3 of 10 mmol/l is unlikely to be life-threatening, the same 50% fall might have untoward consequences if the initial PHCO3 had been 10 mmol/l, because we would now be dealing with a very severe degree of metabolic acidosis once the ECF volume is re-expanded (PHCO3 5 mmol/l). In fact, the PHCO3 might be even lower, since this calculation does not account for the ongoing loss of NaHCO3 in diarrhoeal fluid. The pH in cells could be even lower if there was a high venous PCO2 (a tissue-form of respiratory acidosis).
(iii) Assess the need to infuse K+
Professor McCance pointed out that, like the PHCO3, the plasma K+ concentration (PK) was also within normal limits. He asked the team to consider what they thought of the K+ status in this patient.
Question 12. Might this patient have a deficit of K+?
Physiology principle 11: K + will move from the ICF to the ECF compartment if the magnitude of the negative voltage in cells is diminished while there are open K+ channels in cell membranes.
Return to the bedside: The majority of K+ in the body is inside cells, thus the PK provides a tiny and inadequate window on the total body K+ stores. Patients with a total body K+ deficit may have a PK that is within normal limits, because most of the deficit of K+ reflects changes inside cells. This was almost certainly the case with these patients with cholera, because the average K+ concentration in their diarrhoeal fluid is 15 mmol/l (Table 2), and with the loss of many litres of diarrhoeal fluid, there is a very large K+ loss, exceeding the likely intake of K+. Measured deficits of K+ averaged 158 mmol in 24 h, with a range of 50–245 mmol/24 h.21 These authors of this paper estimated that the total faecal loss of K+ could be as high as 400–700 mmol of K+ for the entire period of diarrhoea (10–15% of total exchangeable K+).
The voltage across cell membranes and the open probability of their K+ channels determines the distribution of K+ between the ICF and ECF compartments. This voltage is created by the Na-K-ATPase, which pumps 3 Na+ ions out of the cell in exchange for 2 K+ ions, creating a negative voltage inside cells.22 The activity of the Na-K-ATPase is affected by the concentration of Na+ in the ICF compartment. The major electroneutral route of entry of Na+ into the ICF is via the Na+/H+ exchanger (NHE).23 This activity of NHE is stimulated by insulin, which raises the concentration of Na+ in the ICF and thereby augments the exit of positive voltage via the Na-K-ATPase—this makes the voltage inside cells more negative. This then causes a shift of K+ from the ECF to the ICF, down its electrical gradient. On the other hand, the magnitude of the negative voltage inside cells is likely to diminish in patients with cholera. Accordingly, the direction of K+ movement is from the ICF to the ECF compartment, provided that the K+ channels in the cell membrane remain open. As to mechanism, the severe degree of ECF volume contraction raises the level of 
-adrenergic hormones, and they inhibit the release of insulin.24 This outward shift of K+ can maintain the PK in the normal or mildly elevated range despite a large intracellular K+ deficit, akin to events in diabetic ketoacidosis.25
Dealing with the K+ deficit is another important part of therapy for these patients. One should anticipate a fall in the PK once the ECF volume returns to normal and the acidosis is largely corrected. Therefore it is advisable to add at least 10 mmol of K+ to each litre of infusate, and to follow the PK closely. There is one other point to emphasize. Should hypokalaemia develop, the patient might develop paralytic ileus (called cholera sicca21). The danger here is that fluid is accumulating in the lumen of the GI tract without diarrhoea and may cause the attending physicians to slow down the intravenous fluid administration, based on the observed decrease in diarrhoeal losses. This again emphasizes the need for clinical observation, including listening for bowel sounds and following the more reliable signs of ECF volume contraction (haematocrit, total plasma proteins, venous PCO2, and the urine output3).
After this thorough exploration of their patient's metabolic mysteries, and satisfied that they were now able to implement a logical and accurate plan of management, our team adjourned. At the request of Professors McCance and Phillips, they agreed to meet over tea the following day to brief their learned professors on recent advances with respect to cholera and its therapy. The nephrology fellow was given the task of burning the midnight oil, and the next morning, he took centre stage.
|
After the adjournment |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
The nephrology fellow summarized the findings from several interesting articles he had read, highlighting three points for Professors McCance and Phillips to comment on.1,16,26–28
(i) Urine output. When compared directly, patients given isotonic saline had a much larger urine output in the first several hours of therapy compared to those receiving a 
-containing infusion.16 These data suggested that saline was superior to the -containing intravenous infusion, and should therefore be the recommended replacement solution, suggested our budding nephrologist.
(ii) Blood volume. Prior to therapy, all patients had a very low total blood volume (60% reduction); however, of great interest, the central blood volume was only modestly diminished,28 raising interesting questions about the regulation of the central vs. the peripheral blood volume.
(iii) The effects of using on the likelihood of developing pulmonary oedema. Although the data were not robust, the renal fellow made the following comments. When these severely ill patients were given isotonic saline, they were more likely to develop pulmonary oedema, even though the positive balance of saline did not re-expand their ECF volume to its normal state.26 In contrast, pulmonary oedema was not reported in patients who received the -containing intravenous fluid. Most surprising of all, those patients treated with isotonic saline who developed pulmonary oedema could be rescued by infusing isotonic NaHCO3 at fairly brisk rates. This was totally unexpected, said the nephrology fellow.
All eyes turned to Professor McCance, who appeared deep in thought. After long moments of silence, our Professor was ready to begin his analysis and dispense some pearls of wisdom. ‘To deal with these three unanticipated observations,' he said, ‘I would like to begin with the data that deal with the cardiac output prior to therapy. Since three-quarters of the blood volume is in the venous side, and because there is a reasonably well-maintained central blood volume in these patients with a low total blood volume, this strongly suggests that constriction of the peripheral venous capacitance vessels is a very prominent part of the picture (Figure 7). While I do not anticipate unique effects due to cholera on this vascular bed, venoconstriction should occur with a very contracted ECF volume, perhaps due to adrenergic actions. Let me now address the specific issues.’
|
(i) The improvement in GFR with intravenous saline. When saline was infused in this setting, much of the volume probably remained in the central blood volume. This would account for the marked fall in the level of the haemotocrit and total proteins in plasma in <2 h, as observed by Dr. Phillips.2 Now, with a lower viscosity, blood flow to the kidneys should increase and the GFR should rise. Professor McCance added another point. Isotonic saline led to a better initial restoration of the central blood volume, but it did not increase flow dramatically in the peripheral circulation.
(ii) NaHCO3-containing intravenous solutions were less efficient as expanders of the central blood volume. The major difference in net effect of infusing NaCl vs. NaHCO3 is on the PHCO3. Because isotonic infusions are -free, they will result in a fall in the PHCO3 by expanding the ECF volume. There is no special role for Cl– in this story—rather, the important factor is the lack of a source of in the isotonic saline.
Professor McCance also commented on a speculation to account for the lesser effect on central blood volume of infusing alkali.28 These authors suggested that acidaemia might lead to more pronounced constriction of peripheral veins in these patients (Figure 7). This was present in most dramatic form when the infusion of saline led to a fall in the PHCO3, and it would be less obvious with a -containing infusion.
(iii) Explain why an infusion of isotonic NaHCO3 might ameliorate pulmonary oedema. When viewed in isolation, an infusion of an isotonic Na+-containing solution should be contraindicated in patients with pulmonary oedema unless it improves myocardial function (there were no data on the latter point). Another possibility is that local H+ could augment peripheral venoconstriction, although no mechanism for this is known. Now raising this PHCO3 could cause venodilatation and a redistribution of blood from central venous to peripheral venous compartments. The volume delivered to the peripheral circulation with transudation of saline to the interstitial space could have led to a net decline in the central blood volume, even while the patient was receiving isotonic NaHCO3.
The consultant physician pointed out that ß–adrenergic hormones bind less well to their receptors when the H+ concentration is elevated.29 ‘Does this make your hypothesis less robust?’ he asked. The response of Professor McCance was typical of his way of thinking. He first examined the ‘big picture’ and reminded everyone that a maximum adrenergic response is a component of the flight or fight response, a time when adrenergic levels are very high and L-lactic acid is being produced as we sprint from danger. It would be an imperfect design to amputate the desired vascular effects of adrenaline in this setting, so perhaps there are other factors for the diminished adrenergic response in these in vitro studies.
Professor McCance summarized by saying: ‘The take-home message here is that I will now always use an intravenous solution that has an appreciable concentration of or an anion that is readily converted to (lactate or acetate anions) in treating severely ill patients with cholera.30 Because can lead to a fall in the ionized Ca2+ concentration, there might be a theoretical advantage in using the lactate anion-containing solution. It is imperative to be sure the lactate anions are being metabolized (follow the anion gap in plasma initially) and to examine the patient frequently for symptoms and signs of pulmonary oedema and the development of hypokalaemia.’
The discussion now turned to the molecular advances in cholera, and again it was the renal fellow who provided a brief update.
Question 13. What is the expected duration of the toxicity due to the cholera toxin?
Physiology principle 12. The cholera toxin binds tightly and irreversibly to intestinal cells and has a permanent effect.
Return to the bedside: The duration of the toxicity can be predicted to be as long as the affected cells live—this is 7 days.31 The secretion of electrolytes is stimulated in cells near the base of the crypts in the intestinal villus after the cholera toxin binds to their luminal membrane—the bacteria do not enter the cells or the blood stream. The intact toxin consists of dimeric A chains and five identical B subunits. When B units bind irreversibly to a glycosaminoglycan, GM1, on the target cells, the A chain gains access to the cell interior and generates the signal molecule, cyclic AMP. As a result, a special Cl– channel is inserted into the luminal membrane of these cells (cystic fibrosis transmembrane regulator, CFTR) (Figure 8). Two features are needed for Cl– secretion—the driving force to secrete Cl– must be electrical because the concentration of Cl– is much lower in cells than in the intestinal lumen. This driving force—negative voltage in cells—must not be dissipated when Cl– anions are secreted. This is achieved by having cyclic AMP lead to a persistently open K+ channel in the basolateral membrane of these cells—thus there is Cl– secretion with little change in ICF voltage.
|
From the above, it is clear that we need a way for K+ to enter cells and a source of Na+ to enter the lumen of the intestinal tract to have a net secretion of electroneutral Na+ + Cl–. The pathway for K+ and Cl– entry into cells is via a secretory form of the Na+, K+, 2-Cl– cotransporter (NKCC-1) in the basolateral membrane of enterocytes (Figure 8).
The route for Na+ entry into the lumen of the intestinal tract is somewhat circuitous. Na+ enter enterocytes on the NKCC-1 and are pumped out of these cells by an electrogenic Na-K-ATPase in the lateral membrane. Finally, Na+ ions diffuse passively between enterocytes into the lumen of the intestinal tract, completing the process of the electroneutral entry of Na+ and Cl– into its lumen.
Question 14: Can these new insights lead to new therapies for cholera?
Physiology principle 13. If one could develop a drug that blocks CFTR, the intestinal Cl– channel, one might reduce fluid losses dramatically.
Return to the chemistry lab: Thiazide natriuretics inhibit Na+ and Cl– transport in the distal convoluted tubule by binding to the Cl– binding site on the Na+, Cl– co-transporter. Recently a modified thiazide derivative was synthesized and it inhibits the CFTR in enterocytes—its name is thiazolidinone32 and it may be very valuable in the future for intractable cases of cholera.
|
Conclusions |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
The physiology principles that were central to the analysis are summarized in Table 4. This case vividly illustrates the limitations of the traditional definitions of metabolic and respiratory acidosis. If we were to rely simply on the concentration of
, the PCO2, pH, and the plasma anion gap, we would not have made a diagnosis of metabolic acidosis, nor of respiratory acidosis. Nevertheless, on the basis of the history and also from the data obtained by Dr Phillips, we can predict that our patient lost a very large amount of NaHCO3 in his diarrhoeal fluid. Because this loss exceeds all input, metabolic acidosis must be present—the challenge was to marry the clinical story and the laboratory data. Once again, the application of simple principles of physiology at the bedside and a quantitative approach offered a potential solution to this problem and in so doing, arrived at a new understanding of acidosis. By examining the mass balance for 
and focusing on its content, rather than just the concentration, it was clear that a large deficit of (and hence metabolic acidosis) was present. To quantify the deficit of , an estimate of ECF volume was required—this was calculated from the change in the haematocrit, because clinical signs are not reliable in quantitative terms. The usual clues to diagnose metabolic acidosis had therefore been masked by a contraction alkalosis. No method that relies only on concentrations will permit one to make an accurate assessment of all acid-base disorders—this includes the approach of Stewart.33 ‘An assessment of ECF volume should be part of the diagnostic workup, and I would recommend using the change in the haematocrit as a simple tool to improve the accuracy of this assessment' said Professor McLance.
|
An appreciation of the significance of the severe degree of ECF volume contraction and the high venous PCO2 also led to a diagnosis of respiratory acidosis of the tissue-form, despite the normal arterial PCO2. It pointed to poor tissue perfusion and CO2 removal with a higher capillary, and tissue, PCO2. ‘Bad buffering’ was the danger in this situation and measures to restore an effective circulation were therefore promptly implemented.
A danger could be created by not dealing with the metabolic acidosis, because a high H+ concentration may restrict infused saline to the central blood volume. As a result, it is possible to develop pulmonary oedema. To minimize this risk, a form of alkali should be added to the intravenous fluids.
|
Additional comment from the housestaff |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
The medical registrar looked concerned and Professor McCance asked if she had a question. She stated that she grasped all the concepts, but there was still one detail that was not clear. ‘I just did a quantitative analysis and I cannot understand why my patient presented with a PHCO3 of 21 mmol/l. Before the illness, I agree that he had
250 mmol of HCO3 in his ECF compartment (10 l x 25 mmol/l). When he lost 5 l of diarrhoea fluid with a 40 mmol/l concentration of HCO3, his deficit of HCO3 should have been 200 mmol, leaving him with only 50 mmol of HCO3 in his ECF compartment. On admission with his 50% reduction in the ECF volume, should he not have had a PHCO3 of 10 mmol/l (50 mmol in 5 l)?’ Professor McCance agreed and asked her to continue. ‘Well,‘ she said, ‘I cannot find a source for the extra HCO3 in the ECF compartment, because there was no appreciable loss of HCl (vomiting), excretion of NH4Cl (the lack of a chronic stimulus, the lag period, the low GFR, and the patient was anuric), and there was no exogenous input of NaHCO3’.
Professor McCance was happy that she had not completed her analysis. ‘I believe that the most logical source of this extra HCO3 was from the ICF compartment,’ she said. ‘To make new HCO3 ions and maintain electroneutrality, these HCO3 ions must be made with the cations, H+ or , and the former is far more likely for the following reason. Because of his very severe degree of ECF volume contraction, there is a markedly reduced blood flow rate to tissues. This will cause more extraction of O2 from each litre of blood flowing to tissues. As a result, there will be a larger amount of CO2 added to each litre of blood. This in turn will raise the venous Pco2 to values that greatly exceed the arterial PCO2, and the venous PCO2 will virtually equal that in cells drained by this vein, because CO2 moves via diffusion from cells to the capillary. The high tissue PCO2 should drive the bicarbonate buffer reaction back to H+ and HCO3 ions (Figure 9). Moreover, if the PHCO3 rises, these H+ must be retained in cells; the high H+ concentration should lead to more H+ binding to ICF proteins. What I am not sure about,‘ she concluded, ‘is how HCO3 ions move from the ICF to the ECF compartment’.
|
|
At this point, the nephrology consultant took the lead. He had read a paper that addressed this very issue.34 Once the Cl/HCO3 ion exchanger becomes active, Cl– ions will enter the cell while HCO3 ions exit in a 1:1 stoichiometry on this electroneutral transporter because of their concentrations in the ECF and ICF compartments. Moreover, most cells have Cl– ion channels with a high open probability.35 Accordingly, Cl– ions will be pushed out of cells by the negative voltage—this process will diminish the net negative voltage in cells (Figure 10). As a result, there will be a shift of K+ ions out of cells. The net result is the export of K+ and
ions.
‘Thank you both’ said an excited Professor McCance. ‘Now I can also understand why the PK was not low in this patient, who should have lost more K than he had in his ECF compartment. The loss of K in diarrhoea should be 65 mmol (5 l x 15 mmol/l, Table 3) and his ECF K content should be 40 mmol (10 l x 4 mmol/l). Nevertheless, I can foresee a new problem that will develop during therapy: a much more severe degree of metabolic acidosis.’ What could Professor McCance be thinking?
Question 15. What problem will develop when isotonic saline is used to partially re-expand the ECF volume?
Physiology principle 3, restated: The concentration of HCO3 will be the net result of changes in the content of HCO3 and the volume of the ECF compartment.
Return to the bedside: Everyone quickly identified several reasons why the PHCO3 should fall when isotonic saline was infused. First, replacing diarrhoeal fluid with its high HCO3 concentration with an equal volume of HCO3-free infusate will result in a larger deficit of HCO3. Second, as the blood pressure rises, the splanchnic circulation will improve and this might increase the production of diarrhoeal fluid that contains HCO3, a point well-recognized by Dr Phillips who treated these patients.2 Third, re-expanding the ECF volume without adding HCO3 will lower the PHCO3 by raising the denominator of the HCO3 concentration.
Professor McCance was impressed with these answers, but he had one more mechanism in mind. As soon as blood flow to cells rises, there will be the same total extraction of O2, but less extraction of O2 per litre of blood flow, less addition of CO2 per litre of blood flow, and a fall in the venous and tissue PCO2. This should reverse the events described in Figure 9, and cause both K and HCO3 ions to enter cells. As a result, the degree of metabolic acidosis will become much more severe. The implications of this rapid fall in the PHCO3 on the microcirculation can contribute to the development of pulmonary oedema as discussed above.
Professor McCance concluded that he never would have thought of all these interesting mechanisms had it not been for the curiosity of his young teachers, the medical team, and the updated information provided by the nephrology consultant. ‘Learning must be an ongoing process, and it's a rather exciting process!’ he exclaimed. ‘I have also realized that you cannot understand an acid-base disorder if you rely solely on the pH, PCO2, the concentration of HCO3 and the anion gap in blood tests. There is an even larger error if you do not factor in the ECF volume, rely solely on the arterial values, and do not assess or measure the venous PCO2. In short, there is no substitute for the application of integrative physiology at the bedside!’
|
Appendix |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
High value for the anion gap in plasma (Table 1). One might expect to find a high value for the anion gap in plasma due to the addition of L-lactic acid if there was an inadequate delivery of oxygen to tissues to meet their demand for aerobic generation of ATP.36 The blood L-lactate was in the 2–4 mmol/l range in other patients with cholera30 and the anion gap was minimally elevated 2 h after infusing isotonic saline, yet the PHCO3 had decreased.2 Reasons to accumulate large amounts of the other usual acids, ß-hydroxybutyric acid (prolonged starvation37) and D-lactic acid (antibiotics, GI stasis, and a supply of carbohydrate in the lumen of the GI tract38) were not evident from the history. Perhaps the major reason for an increased anion gap in plasma was a high concentration of albumin in plasma, the consequence of the severe ECF volume contraction. In addition, the net charge on albumin appears to rise in settings with a contracted ECF volume.39
|
Footnotes |
|---|
Address correspondence to Professor M.L. Halperin, University of Toronto, St Michael's Hospital Annex, Lab #1, Research Wing, 38 Shuter Street, Toronto, Ontario, M5B 1A6, Canada. e-mail: mitchell.halperin{at}utoronto.ca
|
References |
|---|
|
TopSummaryIntroductionThe consultationIssues concerning metabolic...Issues concerning the definition...TherapyAfter the adjournmentConclusionsAdditional comment from the...AppendixReferences |
|---|
1. Greenough WB, III. The human, societal, and scientific legacy of cholera. J Clin Invest 2004; 113:334–9.[CrossRef][Web of Science][Medline]
2. Watten RH, Morgan FM, Songkhla YN, Vanikiati B, Phillips RA. Water and electrolyte studies in cholera. J Clin Invest 1959; 38:1879–89.[Web of Science][Medline]
3. Napolova O, Urbach S, Davids MR, Halperin ML. How to assess the degree of extracellular fluid volume contraction in a patient with a severe degree of hyperglycemia. Nephrol Dial Trans 2003; 18:2674–7.
4. Latta T. Cases of malignant cholera in which 480 ounces of fluid were injected into the veins with success. Lancet 1831–32; 2:366–71.[CrossRef]
5. Halperin ML, Goldstein MB. Fluid, Electrolyte and Acid-Base physiology; a problem-based approach. Philadelphia, WB Saunders, 1998.
6. Carlotti APCP, Bohn D, Mallie J-P, Halperin ML. Tonicity balance and not electrolyte-free water calculations more accurately guide therapy for acute changes in natremia. Intensive Care Med 2001; 27:921–4.[CrossRef][Web of Science][Medline]
7. Scheich A, Donnelly S, Cheema-Dhadli S, Schweigert M, Vasuvattakul S, Halperin ML. Does saline ‘correct’ the abnormal mass balance in metabolic alkalosis associated with chloride-depletion in the rat. Clin Invest Med 1994; 17:448–60.[Web of Science][Medline]
8. Rambaud J-C, Modigliani R, Emonts P, Matuchansky C, Vidon N, Besterman H, et al. Fluid secretion in the duodenum and intestinal handling of water and electrolytes in Zollinger-Ellison syndrome. Digestive Dis 1978; 23:1089–97.[CrossRef]
9. Halperin ML, Jungas RL, Pichette C, Goldstein MB. A quantitative analysis of renal ammoniagenesis and energy balance: a theoretical approach. Can J Physiol Pharmacol 1982; 60:1431–5.[Web of Science][Medline]
10. Brosnan JT, Lowry M, Vinay P, Gougoux A, Halperin ML. Renal ammonium production—une vue Canadienne. Can J Physiol Pharmacol 1987; 65:489–98.[Web of Science][Medline]
11. Cannon PJ, Heinemann HO, Albert MS, Laragh JH, Winters RW. ‘Contraction’ alkalosis after diuresis of edematous patients with ethacrynic acid. Ann Int Med 1965; 62:979–90.
12. Chung HM, Kluge R, Schrier RW. Clinical assessment of extracellular fluid volume in hyponatremia. Am J Med 1987; 83:905–8.[CrossRef][Web of Science][Medline]
13. Love AHG, Phillips RA. Measurement of dehydration in cholera. J Inf Disease 1969; 119:39–42.
14. Halperin ML, Rolleston FS. Clinical Detective Stories: A Problem-Based Approach to Clinical Cases in Energy and Acid-Base Metabolism, 1st edn. London, Portland Press, 1993.
15. O'Shaughnessy WB. The blue epidemic cholera. Lancet 1832; 2:366–71.[CrossRef]
16. Cash RA, Lehman JS, Jr., Hare RS, Nalin DR. Renal function in acute asiatic cholera. Trans Roy Soc Trop Med Hyg 1973; 67:217–24.[CrossRef][Web of Science][Medline]
17. Vasuvattakul S, Warner LC, Halperin ML. Quantitative role of the intracellular bicarbonate buffer system in response to an acute acid load. Am J Physiol 1992; 262:R305–9.
18. Phillips RA. Cholera in the perspective of 1966. Ann Int Med 1966; 65:922–30.
19. Cash RA, Nalin DR, Forrest JN, Abrutyn E. Rapid correction of acidosis and dehydration of cholera with oral electrolyte and glucose solution. Lancet 1970; ii:549–50.
20. Beisel WR, Watten RH, Blackwell RQ, Benyajati C, Phillips RA. The role of bicarbonate pathophysiology and therapy in asiatic cholera. Am J Med 1963; 35:58–66.[CrossRef][Web of Science][Medline]
21. Watten RH, Phillips RA. Potassium in the treatment of cholera. Lancet 1960; 2:999–1001.[CrossRef][Web of Science][Medline]
22. Russell JM. Sodium-potassium-chloride cotransport. Physiol Rev 2000; 80:211–76.
23. Soleimani M, Singh G. Physiologic and molecular aspects of the Na+/H+ exchangers in health and disease processes. J Invest Med 1995; 43:419–30.[Web of Science][Medline]
24. Porte DJ. Sympathetic regulation of insulin secretion. Arch Intern Med 1969; 123:252–60.
25. Beigelman P. Potassium in severe diabetic ketoacidosis. Am J Med 1973; 54:419–20.[CrossRef][Web of Science][Medline]
26. Greenough WB, III., Hirschhorn N, Gordon RS, Jr., Lindenbaum J, Ally KM. Pulmonary oedema associated with acidosis in patients with cholera. Trop Georg Med 1976; 28:86–90.
27. Harvey RM, Enson Y, Lewis ML, Greenough WB, Ally KM, Panno RA. Hemodynamic effects of dehydration and metabolic acidosis in asiatic cholera. Trans Assoc Am Phys 1966; 79:177–86.[Medline]
28. Harvey RM, Enson Y, Lewis ML, Greenough WB, Ally KM, Panno RA. Hemodynamic studies on cholera: Effects of hypovolemia and acidosis. Circulation 1968; 37:709–28.
29. Nahas GG, Zagury D, Milhaud A, Manger WM, Pappas GD. Acidemia and catecholamine output of the isolated canine adrenal gland. Am J Physiol 1967; 213:1186–92.
30. Watten RH, Gutman RA. Comparison of acetate, lactate, and bicarbonate in treating the acidosis of cholera. Lancet 1969; ii:512–14.[CrossRef]
31. Field M. Intestinal transport and the pathophysiology of diarrhea. J Clin Invest 2003; 111:931–43.[CrossRef][Web of Science][Medline]
32. Ma T, Thiagarajah JR, Yang H, Sonawane ND, Folli C, Galietta LJV, et al. Thiazolidinone CFTR inhibitor identified by high-throughput screening blocks cholera toxin-induced intestinal fluid secretion. J Clin Invest 2002; 110:1651–8.[CrossRef][Web of Science][Medline]
33. Corey HE. Stewart and beyond: New models of acid-base balance. Kidney Int 2003; 64:777–87.[CrossRef][Web of Science][Medline]
34. DeMars C, Hollister K, Tomassoni A, Himmelfarb J, Halperin ML. Citric acidosis: A life-threatening cause of metabolic acidosis. Ann Emerg Med 2001; 38:588–91.[CrossRef][Web of Science][Medline]
35. Waldegger S, Jentsch TJ. From tonus to tonicity: physiology of CIC chloride channels. J Am Soc Nephrol 2000; 11:1331–9.
36. Cohen RD, Woods HF. Clinical and Biochemical Aspects of Lactic Acidosis. Oxford, Blackwell Scientific Publications, 1976.
37. Kamel KS, Lin S-H, Cheema-Dhadli S, Marliss EB, Halperin ML. Prolonged total fasting: a feast for the integrative physiologist. Kidney Int 1998; 53:531–9.[CrossRef][Web of Science][Medline]
38. Halperin ML, Kamel KS. Turning sugar into acids in the gastrointestinal tract. Kidney Int 1996; 49:1–8.[Web of Science][Medline]
39. Kamel KS, Cheema-Dhadli S, Halperin FA, Vasudevan S, Halperin ML. Anion gap: Do the anions restricted to the intravascular space have modifications in their valence? Nephron 1996; 73:382–9.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. G. Narins, Participants:, M. Halperin, G. Danovitch, R. Falk, and J. Bargman The Nephrology Quiz and Questionnaire: 2005 Clin. J. Am. Soc. Nephrol., May 1, 2006; 1(3): 592 - 608. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||