Uncovering the basis of a severe degree of acidemia in a patient with diabetic ketoacidosis
From the 1Division of Pediatric Nephrology, Stollery Children's Hospital, University of Alberta, Edmonton, Canada, 2Hospital das Clinicas, Faculty of Medicine of Ribeirão Preto, University of São Paulo, Ribeirão Preto, Brazil, 3Department of Critical Care Medicine, Hospital for Sick Children and Departments of Anaesthesia and Medicine, University of Toronto, Toronto, Canada, 4Division of Nephrology, St Michael's Hospital, University of Toronto, Toronto, Canada, and 5Division of Nephrology and Department of Medicine, Stellenbosch University, Cape Town, South Africa
Address correspondence to Professor M.L. Halperin, Emeritus Professor of Medicine, University of Toronto, St Michael's Hospital Annex, Room 5078, 30 Bond Street Toronto, Ontario, M5B 1W8, Canada. email: mitchell.halperin{at}utoronto.ca
 |
Summary |
|---|
 Top Summary IntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
In this teaching exercise, the goal is to demonstrate how an application of principles of physiology can reveal the basis for a severe degree of acidaemia (pH 6.81, bicarbonate <3 mmol/l (PHCO3), PCO2 8 mmHg), why it was tolerated for a long period of time, and the issues for its therapy in an 8-year-old female with diabetic ketoacidosis. The relatively low value for the anion gap in plasma (19 mEq/l) suggested that its cause was both a direct and an indirect loss of NaHCO3. Professor McCance suggested that ileus due to hypokalaemia might cause this direct loss of NaHCO3, and that an excessive excretion of ketoacid anions without
in the urine accounted for the indirect loss of NaHCO3. In addition, he suspected that another factor also contributing to the severity of the acidaemia was a low input of alkali. He was also able to explain why there was a 16-h delay before there was a rise in the PHCO3 once therapy began. The missing links in this interesting story, including a possible basis for the hypokalaemia, emerge during the discussion between the medical team and Professor McCance. |
Introduction |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
In this clinical teaching exercise, the focus is on the approach to the patient who has a more severe degree of acidaemia than is usually observed in patients with diabetic ketoacidosis (DKA). Once again, the central figure in this clinical detective story is the imaginary consultant Professor McCance, who practiced medicine close to 80 years ago. Readers will recognize that his overall objective is to demonstrate that an application of principles of integrative physiology at the bedside and a quantitative analysis are essential to reach a more accurate clinical diagnosis, to reveal the underlying pathophysiology, and to help design better therapy.
|
The consultation |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
There was a lively discussion at Admission Rounds, when an explanation was sought for a very severe degree of acidaemia (arterial blood pH 6.81, bicarbonate (
) concentration in plasma (PHCO3) <3 mmol/l, arterial PCO2 8 mmHg), and only a modest increase in the anion gap in plasma (PAnion gap)) in an 8-year-old female who presented with her first episode of DKA (see Appendix 1). She had only a moderate degree of hyperglycaemia (19.5 mmol/l, 351 mg/dl), however, and her GFR was only minimally reduced. Of significant importance, hypokalaemia (3.3 mmol/l) was present, and this also caused a vigorous debate concerning its pathophysiology and its implications for therapy. The house staff thought that the low concentration of potassium (K+) in plasma (PK) might be ‘tainted’ because the patient's initial presentation was at a walk-in clinic, with chief complaints of intense ‘air-hunger’, cough and difficulty breathing. The diagnosis made at that time was asthma, and the patient was treated with a bronchodilator and an antibiotic, despite the lack of findings of airway obstruction on physical examination. At this point, all agreed that a consultation with Professor McCance was the obvious way to improve their understanding of these issues. As always, he was delighted to come by for a brief initial discussion, but he would have to return later in the day to complete the rounds. He asked the team to prepare a synopsis of the laboratory data prior to therapy, which they were most eager to provide (Table 1).
|
|
Morning rounds |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
Professor McCance thanked the team for sharing their very interesting case with him. He too had wondered why some patients have more and others have less severe degrees of acidaemia, and why there were large differences in the PAnion gap on presentation for DKA. He realized that the discussion of this case would provide a rich learning experience: ‘There are many interesting aspects of your case and all of them need to be integrated into a comprehensive picture‘, he said.
The intern volunteered to summarize what the team had agreed upon in their discussion of the case on admission. Seeing the nod of approval from Professor McCance, he began.
‘First, we are quite confident that we are dealing with DKA, and have not missed another cause for metabolic acidosis and an increased PAnion gap in this patient. We have also ruled out those causes of metabolic acidosis with little increase in the PAnion gap corrected for the concentration of albumin in plasma (PAlbumin), including diarrhoea1 and glue sniffing.2 Second, although all of our findings are consistent with a diagnosis of DKA in qualitative terms (hyperglycaemia, ECF volume contraction, the odour of acetone on her breath, and metabolic acidosis), they are less convincing when viewed from a quantitative perspective’. He then asked Professor McCance to focus on two aspects of the acid-base disorder, the severity of the acidaemia and the small increase in the PAnion gap.
The registrar wished to add some background to focus the discussion and asked whether she could provide a brief summary of their understanding of ketoacidosis. Sensing Professor McCance's approval, she proceeded.3
‘As per your teaching, we performed a function/control analysis to understand the pathophysiology of ketoacidosis.4 The most important function of this metabolic process during prolonged fasting is to provide the brain with a fat-derived, water-soluble fuel when its usual fuel, glucose, is in short supply. The starting point is triglycerides stored in adipose tissue and the end points are regeneration of ATP that was consumed when cerebral and renal work was performed (and also the excretion of ß-hydroxybutyrate anions (ß-HB–) and 
in the urine. The latter ensures that a sufficient number of ‘effective’ osmoles in the urine to avoid a profound degree of oliguria).5 Many of the pathways in this metabolic process occur in different organs (e.g. adipose tissue, liver, brain, kidneys) (Figure 1). In each organ, control will be exerted to ensure that the rate of formation of ketoacids is high enough to supply the brain with all the ATP it needs, but there should not be a much higher rate of ketogenesis, as these extra ketoacids will accumulate and lead to a much more severe degree of acidaemia’.
|
‘In quantitative terms, the rate of production of ketoacids in the liver in prolonged fasting is
1 mmol/min;3,6 half of these ketoacids appear to be oxidized in the brain,7 and one-fourth are removed by the kidney. In the kidney, of which 150 mmol/day are excreted in the urine, and 250 mmol/day are oxidized to regenerate renal ATP.6 Of the remaining ketoacids 150 mmol are removed by conversion to acetone, which was detected on the breath of our patient. In addition, a small quantity of ketoacids could be oxidized in the intestinal tract to provide the ATP needed to absorb nutrients if they were ingested. Other organs, particularly skeletal muscle, do not oxidize an appreciable quantity of ketoacids’. And thus she put their first question to Professor McCance, ‘Why did such a severe degree of metabolic acidosis develop in our patient with DKA, whereas this does not occur in prolonged fasting? ’
Question 1. Why did such a severe degree of metabolic acidosis develop in our patient with DKA, whereas this does not occur in prolonged fasting?
Physiology principle 1. Metabolic acidosis is a process that leads to a higher concentration of H+ in plasma along with a decrease in the PHCO3. In the case of ketoacidosis, it develops when ketoacids are formed faster than their H+ can be removed by metabolism or via the excretion of ketoacid anions with in the urine (Figure 1).
Return to the bedside. Ketoacids are produced in the liver when insulin levels are very low and levels of hormones with actions that oppose those of insulin are elevated.8 The maximum capacity to synthesize ketoacids is set by the delivery of fatty acids to the liver, and this pathway is actually regulated by the rate of provision of the cofactors, ADP and NAD+, which are produced when the liver performs work (Figure 2).9,10 The major components of liver work are the biosynthesis of proteins and electrical work due to Na+ pumping by the Na-K-ATPase.11
|
‘Of great importance, there is not an appreciable amount of biosynthetic or electrical work in the liver during prolonged fasting.8 Accordingly, I would expect that the rate of ketoacid production should also be low, but that is not what occurs,’ said Professor McCance.
In a previous session, and commented on a possible resolution to this conundrum—partial oxidation of fatty acids may still occur in the liver if control by availability of ADP and NAD+ is bypassed—i.e. there is ‘uncoupling’ of oxidative phosphorylation (Figure 2). This may be due either to the presence of more uncoupler proteins in the mitochondrial membrane, increased entry of H+ through these H+ channels due to the generation of a greater proton-motive force when fatty acids are oxidized (large negative intra-mitochondrial voltage),12 and/or the presence of anions with a dispersed charge (e.g. iodide) or that behave as if they are ‘lipid-soluble’ (e.g. salicylate, dinitrophenol) that can bind protons and carry them through the inner mitochondrial membrane, while their non-protonated form can readily exit from the mitochondrion.13
I now wish to relate this biochemistry to the current patient: ‘She may have a higher rate of production of ketoacids because of a higher rate of uncoupling of oxidative phosphorylation than in other patients with DKA, but I was not informed that she ingested salicylates, which may act as an uncoupler of oxidative phosphorylation. Nevertheless, I have a more important question for the moment: Can ketoacids accumulate even if there is no further increase in ketoacid formation? ’
Question 2. Can ketoacids accumulate even if there is no further increase in ketoacid formation?
Physiology principle 2. To determine why there is a high concentration of a metabolite, one must examine both its rate of input and its rate of output (Figure 3).
|
Return to the bedside. To answer this question, we must examine ketoacid output to see if a low rate of removal of ketoacids could be responsible for the greater accumulation of H+. The major organs that remove ketoacids are the brain and the kidneys.8 Therefore depression of the central nervous system is one setting or a relate decreased rate of removal of ketoacids. In the kidney should cause a more severe degree of ketoaciosis. There are two renal pathways for ketoacid anion plus H+ removal. First, ketoacids can be oxidized to regenerate the ATP that is consumed when filtered Na+ is reabsorbed.14 Second, there is a need to excrete ketoacid anions plus ammonium (
) in prolonged fasting to ensure that there is a safe, minimum urine flow rate to decrease the risk of forming kidney stones (5) (Figure 1); this is the fate of close to 20% of the filtered load of ketoacid anions in prolonged fasting.15 There is also a limit on the renal rate of production of by the availability of ADP (i.e. renal work to reabsorb filtered Na+). Since there was no reason for diminished brain work due to sedation and the GFR was not very low, there was no evidence for a slower removal of ketoacids as an explanation for the severe degree of acidaemia in this patient. At this point, the registrar expressed her concern that the PAnion gap was only moderately elevated, yet the PHCO3 was extremely low. ‘Does this mean that a loss of NaHCO3 was an important cause of the metabolic acidosis? ‘, she asked.
Question 3. Was a large loss of NaHCO3 an important cause of the metabolic acidosis?
Physiology principle 3. There are two ways to lose NaHCO3: a direct way and an indirect way.16 In the direct loss, NaHCO3 is excreted via the GI tract (e.g. diarrhoea1) or in the urine (e.g. early in proximal renal tubular acidosis17). In contrast, in the indirect loss of NaHCO3, the process begins with the production of H+ and ß-HB– anions (Figure 4). Next, the ions are titrated by H+, and the resulting CO2 is exhaled by the lungs. In the final step, Na+ is lost in the urine with ß-HB–, because the rate of excretion of ß-HB– exceeds the capacity of the kidney to excrete .16
|
Return to the bedside. ‘First, there is a possible ‘occult’ direct loss of NaHCO3 in the GI tract‘, said Professor McCance. ‘Since the patient presented with a PK of 3.3 mmol/l, this hypokalaemia may cause GI motility to decline and, as a result, NaHCO3 that was secreted into the lumen of the early small intestine may remain unabsorbed in this location. If this were true, there may be a ‘surprising’ rise in the PHCO3 after therapy with exogenous KCl, and this will become important when we discuss therapy in this patient’.
It is also possible that the patient had a large indirect loss of NaHCO3, which is characterized by the excretion of appreciably more Na+ + K+ than Cl– in the urine. Said another way, the urine will contain more ß-HB– than (Figure 4). This does not mean that the urine must contain little ammonium (); rather it means that indirect loss of NaHCO3 only due to the extra portion of urine ß-HB– that is not accompanied by .
'Therefore, I have two questions,' continued Professor McCance. ‘First, did you measure the concentrations of Na+, K+, and Cl– in the urine? Second, was measured or was its concentration estimated in the urine?‘
The registrar had anticipated these questions, but unfortunately urine had not been sent for measurements of electrolytes, osmolality, and creatinine prior to therapy. Notwithstanding, she had reviewed the laboratory results in several cases of DKA previously seen in the past year and found several patients with initial urine data. In all of them, the concentration of Na+ + K+ was 30 to 80 mEq/l higher than that of Cl–.
‘Aha,’ said Professor McCance, ‘just as I expected! The urine electrolytes suggest that these patients are very likely to have had an indirect loss of NaHCO3.’
He continued, and pointed out that there is a predictable, quantitative relationship between the fall in the content of in the ECF compartment and the PHCO3 in patients with metabolic acidosis due to a loss of NaHCO3 if the ECF volume is not contracted, but there is not an obligatory association with acidaemia in this setting as reflected in the following three hypothetical examples where there are different concentrations of in the fluid that was lost. In the first one, the concentration of in the fluid that is lost is <25 mmol/l. Accordingly, the plasma pH and the PHCO3 will rise, obscuring the presence of the metabolic acidosis. In the second example, if the concentration of in the fluid that was lost was 25 mmol/l, there would be no change in the pH or the PHCO3, again obscuring the presence of the metabolic acidosis. In the third example, if the concentration of in the fluid that was lost was >25 mmol/l, there would be a fall in both the pH and the PHCO3 (Table 2). Notwithstanding, if therapy is simply to re-expand the ECF volume with isotonic saline, all three examples will present acidaemia.
|
In patients with DKA, the concentration of Na+ + K+ was 30 to 80 mEq/l higher than that of Cl–, the calculated concentration of potential
(ß-HB–) in the urine was >25 mEq/l, and this is why acidaemia would be present; when there is a higher total excretion of potential , the degree of acidaemia will be more profound (Table 2). ‘This is why I was so excited to finally understand this indirect loss of NaHCO3, and why there is metabolic acidosis, with or without acidaemia, in virtually every patient presenting with DKA18,19 or profound diarrhoea‘,1 said McCance.
The registrar agreed that re-expanding the ECF volume with isotonic NaCl would unmask the process which created the deficit of NaHCO3 and thus metabolic acidosis would be converted from an ‘occult’ form to an overt one. She continued: ‘This same scenario is interpreted by those who believe in the strong ion difference as an effect of adding Cl– without Na+.’20,21
‘I understand your points, ’responded Professor McCance. ‘One must expand the definition of metabolic acidosis to a process that would lower the PHCO3, but this could be from a high value to a less high value rather than just to a value that must be less than the normal value for the PHCO3.1 The comment about adding Cl– without Na+ does not make ‘electrical sense’—it is shocking! Any chemist would insist on the presence of electroneutrality. The strong ion difference proponents must define the composition of every solution, and recognize that concentrations have numerators and denominators. Hence they will lower the PHCO3 by dilution by adding a solution with equal concentrations of Cl– and Na+, and acidaemia would be the result, but this was not caused simply by the ‘addition’ of Cl–. On the other hand, if they added less Cl– than Na+, and if the other anion were either or potential , they would prevent a fall or even create a raise the PHCO3, but this again was not an effect due directly to the absence of Cl–—it is that simple, I am afraid’.
Returning to the patient, Professor McCance recommended the use of the urine osmolal gap to estimate the rate of excretion of (see Appendix 1, section 2).22 Patients with DKA excrete abundant quantities of (150 mmol/day) if they do not have a very low GFR,14 and the principal anion in the urine will be ß-HB–.5 This leads to another question, ‘Why was so much ß-HB– excreted in this patient? ’
Question 4. Why was so much ß-HB– excreted in the urine in this patient?
Physiology principle 4. The rate of excretion of substances such as ketoacid anions that are not secreted by the proximal convoluted tubule depends on its filtered load, minus the amount of ß-HB– reabsorbed in this nephron site.
Return to the bedside. As this patient had a relatively high GFR, reflected by her near normal PCreatinine, she would have a high filtered load of ß-HB even though there was only a modest elevation in its concentration in plasma (high GFRxmodest elevation of ß-HB–). It is also possible that she might have a reduced rate of reabsorption of ß-HB– by the proximal convoluted tubule if a drug such as salicylic acid or compounds such as L-lactate were elevated in concentration in plasma as they diminish the renal reabsorption of ß-HB–;23 the net result would be fewer ß-HB– anions in plasma and more ß-HB– anions in the urine. Unfortunately, her PL-Lactate was not measured. Although uncoupling of oxidative phosphorylation, hyperventilation, and hypokalaemia are seen with salicylate intoxication,24 again there were no data to assess a possible contribution of salicylate to the pathophysiology in this case. It will be important to establish whether the parents gave their daughter aspirin during her illness.
Notwithstanding, because the above do not permit us to deduce why the acidaemia was so severe, we must ask, ‘What other possible mechanisms might explain why her degree of acidaemia was so severe? ’
Question 5. What other possible mechanisms might explain why her degree of acidaemia was so severe?
Physiology principle 2 revisited. To understand why the acidaemia was so severe, one must think in terms of the input and output of . Parenthetically, it is important to detect the process that caused the deficit of (metabolic acidosis) as well as whether acidaemia was present.
Return to the bedside. While we do not have a reason to invoke a reduced input of NaHCO3 as a cause for the severe degree of acidaemia in this patient, nevertheless, there are some clues for us to think about. These are the hypokalaemia, the absence of a severe degree of hyperglycaemia, the near-normal GFR, and the degree of contraction of the ECF volume, which was not excessive. ‘I think I have an idea but I need to develop it more thoroughly’, said McCance. 'When I return this afternoon, we shall explore these issues and comment on why she could tolerate such a severe degree of acidaemia for such a long period of time. There are therapy issues in this patient that we need to think carefully about; this really is a challenging one‘, he added as he departed, deep in thought. The team thanked him for his insights, which as always left them with much to think about.
|
The afternoon session |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
Prompted by the suggestion that there could be clues from an assessment of the degree of hyperglycaemia, hypokalaemia and the degree of ECF volume contraction that might help to reveal why the PHCO3 was so low, the medical team began their analysis. ‘I am sure we shall learn something we have observed, but have not seen‘, said the registrar. ‘Let us ask Professor McCance: why was the PGlucose lower than expected? ’
Question 6. Why was the PGlucose lower than expected?
Physiology principle 5. Glucose sources are either exogenous or endogenous. In the endogenous category, we must consider glycogen, glycerol and protein. Catabolism of proteins is the most important, as it has a major price to pay: loss of lean body mass.
Return to the bedside. ‘The production of glucose from endogenous sources is relatively small,4 and I believe we can dismiss a lack of catabolism as a major reason for the lower than expected degree of hyperglycaemia in this patient‘, said Professor McCance.
When protein is converted to glucose, there will be both weight loss and the appearance of urea.4 In this process, 572 mmol of urea are formed for every 100 g of protein converted to 60 g of glucose (333 mmol), as 60% of the weight of protein can be converted to glucose.25 Although the patient presumably lost between 2 and 3 kg of weight, there was close to a 1.5 l deficit in her ECF volume (weight 1.5 kg). This suggests that there was little catabolism of lean body mass if these weights are accurate. Since we do not know how much urea was excreted, we cannot assess the degree of catabolism.
The head nurse rose and provided the following information. ‘Weights were not measured on admission in this patient. We asked the parents about their initial weight, so these are not necessarily accurate values. In addition, there is an unknown weight of fluid in the gastrointestinal tract that will raise the body weight, so please take the values, even if they were measured, with a grain of salt‘, she said with a smile. Professor McCance thanked her for these comments and added, ‘Using weight to estimate the ECF volume must similarly not be relied on’.
Professor McCance concluded that if we were to exclude a lower rate of protein catabolism as a cause for the lower than expected PGlucose, he would guess that the patient did not consume a large amount of glucose-containing drinks, but he was especially interested in her intake of fruit juice. Hence he asked, ‘Does anyone know what kind of fluid she drank?’ As this question had not been asked, the intern went off to the waiting area in search of the parents. Meanwhile, a final year medical student asked, ‘Is there a way to deduce how much glucose the patient drank?’
Question 7. Is there a way to deduce how much glucose the patient drank?
Physiology principle 6. The major fate of ingested glucose is its excretion in the urine when the PGlucose is >15 mmol/l.26 During glucosuria, there is an obligatory loss of Na+ and Cl– in the urine.
Return to the bedside. Since we already know that the PCreatinine was only minimally elevated (Table 1), our next step will be to assess her ECF volume on admission. Since the physical examination cannot provide a quantitative assessment of the ECF volume,27,28 the registrar had performed a quantitative estimate using the haematocrit (0.48) and the haemoglobin concentration (166 g/l) (Table 1).1,18 An illustration of the relationship between the degree of rise in the haematocrit or haemoglobin concentration and the decline in plasma volume is shown in Table 3. She calculated that the plasma volume was reduced by 27% before therapy began and hence the ECF volume. ‘Since there was no value for the PAlbumin on admission,’ she said, ‘I cannot verify the degree of contraction of the ECF volume using a calculation based on the albumin concentration’.
|
Professor McCance thanked her for this information. He commented that the value for the PCreatinine, which reflected a minor decrease in the GFR, was consistent with this conclusion. Nevertheless, there was one fact that might cast some doubt on this conclusion. Depending on the method used, PCreatinine may be falsely low in the presence of a severe degree of hyperglycaemia29 and falsely high when there is a high concentration of acetoacetate.30 They all agreed that confirmation of the modest degree of ECF volume contraction was needed. This information would also be very useful to interpret the clinical impact of her acidaemia.
Question 8. What additional supporting evidence is there to imply that there was a modest degree of contraction of her ‘effective’ arterial blood volume?
Physiology principle 7. Before the degree of contraction of the ‘effective’ arterial blood volume becomes more severe, the blood flow to vital organs such as the brain and heart is maintained. In contrast, blood flow to the splanchnic area, skin and resting skeletal muscle declines. ‘Therefore,’ said McCance, ‘I would like to know what the PCO2 was in the arterial blood and from blood drawn simultaneously from the brachial vein.’
Return to the bedside. The medical intern said that both these measurements were made early in therapy, and that the arterial PCO2 was 8 mmHg while the PCO2 was 11 mmHg in brachial venous blood. This provided supporting evidence for a well-maintained blood flow to skeletal muscle in the forearm.31 Professor McCance added that a similar conclusion could be drawn from PO2 measurements, after adjusting the PO2 for the content of O2 in the brachial venous and arterial blood. Before leaving the issue of the ECF volume contraction, Professor McCance had one more question: ‘How large is the deficit of Na+ in this patient? ’
Question 9. How large is the deficit of Na+ in this patient?
Physiology principle 8. The major physiological role for Na+ is that this cation and its attending anions, Cl– and ‘force’ water to stay out of cells and hence they determine the ECF volume.
Return to the bedside. The quantity of Na+ in the ECF compartment is the product of the ECF volume and the PNa. When normal, an 8-year-old, 27 kg child has a blood volume of 2 l (75 ml/kg). With a normal haematocrit of 0.40, her red blood cell (RBC) volume would be 0.8 l, her plasma volume 1.2 l (haematocrit = RBC volume/total blood volume). In contrast, with a haematocrit of 0.48 and the same RBC volume, her blood volume would be 1.67 l. Thus, with the same 0.8 l RBC volume, her plasma volume would be 0.87 l, reduced by 27%. Ignoring changes in Starling forces for simplicity, the ECF volume should have declined by approximately 27% of its normal value: from 5.4 l (20% weight) to 3.9 l.
When her ECF volume is normal, the total content of Na+ in her ECF compartment would be 756 mmol (140 mmol/l x 5.4 l). With an ECF volume of 3.9 l on admission, her Na+ content would be 519 mmol (PNa 133 mmol/l x 3.9 l). Therefore, her deficit of Na+ would be 237 mmol (756 minus 519), which is 9 mmol/kg. To have a loss of 237 mmol of Na+, the patient would need to have lost 5 l of urine and thereby excreted 1500 mmol of glucose in the urine to drive this osmotic diuresis (300 mmol/l of glucose-induced osmotic diuresis, with 50 mmol of Na+ per litre). While clearly some of this glucose would come from endogenous sources and dietary carbohydrate, the remainder would come from ingested fruit juice or soda pop. In most fruit juices and soda pop, there is very little Na+. The concentration of hexoses in fruit juice after digestion of sucrose is close to 750 mmol/l.26 Therefore a reasonable estimate for the maximum volume of fruit juice ingested in the 2-week period would be close to 2 l (i.e. 1500 mmol/750 mmol/l). The registrar then continued with the second clue and asked, ‘What could be the reason for the lower-than-expected PK in this patient?’
Question 10. What could be the reason for the lower than expected PK in this patient?
Physiology principle 9. A low PK could be due to low intake of K+, a shift of K+ into cells, and/or a high rate of excretion of K+.
Return to the bedside. Professor McCance began his analysis by considering the intake of K+. A major source of dietary K+ is from the ingestion of the ‘intracellular fluid’ of fruit, vegetables and meat. Linking the response to the previous questions, perhaps this patient had a low intake of fruit juice.
With respect to a shift of K+ into cells, ß2-adrenergics activate the Na-K-ATPase in the cell membrane of skeletal muscle, and this action could be responsible for the shift of K+ into these cells.11 Our patient was given bronchodilators prior to coming to hospital, and one of their actions is to cause K+ to shift into cells,32 which could be one reason why the PK was 3.3 mmol/l. Nevertheless, the drug was administered only briefly, and Professor McCance was not sure that it would still have an effect on admission. However, the air hunger due to the severe acidaemia would be an important adrenergic drive and this may have increased the shift of K+ into muscle cells.
With respect to the excretion of K+, we do have data on the usual concentration of K+ in the urine in other patients with DKA (25 mmol/l).26 In addition, from the calculation of the content of Na+ in her ECF compartment performed above, we have a suggestion that her urine volume was not very high throughout her illness. If these estimates based on a low intake of sugar were valid, a low urine volume would imply that a large renal loss of K+ would be an unlikely basis for the very low PK in this setting.
Just then, they were interrupted by the intern who had returned to report that only a small fraction of the patient's fluid intake was from fruit juice. Professor McCance became visibly excited—he had clearly had a sudden insight. ‘It appears that we are converging on a common denominator. After our rounds this morning, I checked in the library and discovered that fruit juice contains 50 mEq of K+ per litre,’26 he said. ‘Now can anyone identify a cause of a low input of that we have not yet considered? ’
Question 11. Can anyone identify a cause of a low input of 
that we have not yet considered?
Physiology principle 10. Electroneutrality must be present in fruit juice that contains 50 mEq/l of K+.
Return to the bedside. The major anions in fruit juice are organic (e.g. citrate); and they too should provide 50 mEq of this potential after metabolism, which will result in the formation of 50 mEq of new .33 Accordingly, this provides ‘therapy’ for acidaemia. If this patient did not ingest much fruit juice, there would be a small input of K+ and potential , compared to other patients with DKA, and this might be an additional factor contributing to the more severe degree acidaemia and the presence of hypokalaemia.
At this point there was a heated discussion among the medical team concerning the therapeutic implications of these new insights. On the one hand, there was a group who felt that the acidaemia was so severe that there was a need to give NaHCO3. This view was vigorously opposed by another group, who emphasized the danger of causing a shift of K+ into cells and the dangers of giving NaHCO3 to this population of patients, as it is reported to be a risk factor for the development of cerebral oedema34,35, but this relationship was not observed in a more recent, carefully controlled study.36 There was yet another suggestion by a lone member of the team. The intern suggested that they delay giving the insulin for an hour, until they had replaced some of the deficit of K+ and raised the PK to a less dangerous level. Professor McCance did not want to interrupt this interchange of ideas, as each of the proposals had theoretical merit. Once he sensed that all opinions had been expressed, he said, ‘Let us first analyse the mechanisms that permitted this patient to survive with such a prolonged and serious degree of acidaemia with a pH <7 and a PHCO3 <3 before deciding on the course of action. We can begin by asking why acidaemia might have a more severe impact on the brain in patients with DKA.’
Question 12. What makes acidaemia have a more severe impact on the brain in patients with DKA?
Physiology principle 11. The danger of acidaemia is to have H+ bind to proteins in cells, which changes their ideal charge (Protein° to H•Protein+), shape, and possibly functions (equation 1).31,36,37 Forcing H+ to react with can prevent the binding of H+ to proteins. In more detail, the bicarbonate buffer system can ‘out-compete’ the protein system for H+ removal, mainly because acidaemia stimulates ventilation, which lowers the PCO2 (Figure 5, equation 2).
|
| (1) |
|
| (2) |
|
Return to the bedside. The acidaemia will become more severe if the majority of the H+ load is not buffered by
in skeletal muscle. To understand why this occurs, one must ‘dissect’ the bicarbonate buffer system’. The best way to assess this buffer system is to measure the venous PCO2, as this reflects the PCO2 in capillaries draining individual organs. This PCO2 is in equilibrium with the bicarbonate buffer system in the ECF and ICF compartments in individual organs. The process of lowering the capillary PCO2 begins with stimulation of the respiratory centre in the brain. This is an ideal response, as the PCO2 in the ECF and ICF compartments of the brain is directly proportional to the arterial PCO2, because the rates of oxygen and glucose consumption in the brain are almost constant because of its auto-regulated rate of blood flow. Therefore, by having this ‘ideal’ PCO2, there may only be minimal binding of H+ to intracellular proteins in the brain during acidaemia, which decreases possible detrimental effects on neuronal function.
The arterial PCO2, however, is not sufficient to ensure that the BBS will function optimally in other organs because the PCO2 in their capillaries differs from the PCO2 in arterial blood, and the venous PCO2 is not constant in these organs as they consume a variable quantity of oxygen and add a variable quantity of CO2 to their capillary blood (Table 4). Since CO2 diffuses rapidly, distances are short, and time is not a limiting factor, the PCO2 in capillaries is virtually identical to the PCO2 in cells; this is also true for the interstitial compartment of the ECF in this region. Therefore the arterial PCO2 does not reveal whether the BBS has operated efficiently in the vast majority of the ICF and ECF compartments, but the arterial PCO2 does set the lower limit for the PCO2 in capillaries. In contrast, the venous PCO2 provides information about the effectiveness of this ‘good form of buffering’—i.e. the removal of H+—without requiring a higher H+ concentration. There is, however, an important caveat: if an appreciable quantity of blood shunts from the arterial to the venous circulation and bypasses cells, this venous PCO2 will not reflect the PCO2 in cells in this drainage bed. If there is a markedly contracted ECF volume, as is usually the case in DKA, there will be ineffective buffering in skeletal muscle and therefore more H+ will be available to bind to proteins in these tissues as well as to be carried by the circulation to the brain.
|
Professor McCance drew the discussion to a close with his final question, ‘Should this patient be treated with KHCO3 because of the severe degree of acidaemia, the low PK, and the fact that there are few circulating ß-HB anions? ’
Question 13. Should this patient be treated with KHCO3 because of the severe degree of acidaemia, the low PK, and the fact that there are few circulating ß-HB anions?
Physiology principle 12. In the absence of ketoacid anions in the circulation that can be metabolized to , this patient will be dependent on her rate of excretion of to raise her PHCO3, and the rate of excretion yields only a small quantity of per day. For example, in a 50 kg person who is excreting 180 mmol of per day, the hourly addition of due to the excretion of would only be 7.5 mmol/h (180 mmol/24 h), and this will cause only a tiny rise in the PHCO3 when acidaemia is severe in degree.
Return to the bedside. The endocrine consultant had a look of alarm on his face as he said, ‘Professor McCance, the studies we rely on to define risk factors for cerebral oedema in young patients with DKA have identified the administration of NaHCO3 as an important risk factor for the development of this dreaded complication during the recommended therapy of DKA.34,39 Therefore,’ he concluded firmly, ‘I would definitely not give any !’
Professor McCance thanked the consultant about bringing him up to date on the potential association of the danger of therapy with NaHCO3. ‘Nevertheless‘, he continued, ‘we are dealing with your patient, who in this context is perhaps very different from the vast majority of other patients with the same diagnosis’. What was unique was the severity of the acidaemia, the paucity of precursors to regenerate , and the lack of patient-induced pre-treatment with the K+ and derived from fruit juice. Perhaps this unique patient did indeed have a need for KHCO3. To illustrate this point, he asked the registrar for one other bit of information. ‘How long did it take for this patient to have a significant rise in her PHCO3 once therapy began?’
The registrar was delighted with this question. She just about to raise the same issue, and provided Figure 6 for Professor McCance. Over 16 h of therapy, which consisted of the administration of isotonic saline, KCl, and insulin, there was virtually no rise in the PHCO3. ‘Having such a severe degree of acidaemia with pH <7 was very alarming to us, but we were afraid to administer NaHCO3 in part to avoid cerebral oedema, and in part because of the low PK,' she explained. 'Should we have withheld insulin? Should we have administered KCl plus an equivalent amount of NaHCO3 so as to effectively replace KHCO3?’ The group waited for Professor McCance to comment, but everyone in the room was silent, including their mentor, who was clearly deep in thought.
|
|
Concluding remarks |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
At Professor McCance's request, the registrar drew the rounds to a close with the following summary. The principles of physiology are summarized in Table 5. While the typical presentation of DKA is for metabolic acidosis to occur with a high PAnion gap, this patient did not have the usual levels of circulating ketoacid anions. In addition, the acidosis was severe and sustained. By good detective work, using the clues pointing to a small intake of fruit juice with its K+ and organic anion input, a possible explanation for the severity of the hypokalaemia and the severe degree of acidaemia became evident. Added to this, we had the effects of the exogenous administration of an adrenergic agent. The role of filtration of ketoacid anions and their rate of re-absorption, as well as their metabolic clearance, all contribute to whether the new anions remain in the body or are excreted in the urine. All of these factors have to be taken into consideration when assessing the severity of acidaemia. Finally, the value of administering NaHCO3 (with or without KCl) should be determined on an individual case basis, rather than by looking at a large population of patients, as the latter can only identify associations, rather than cause and effect relationships. It is possible that the association of risk of cerebral oedema with therapy using NaHCO3 simply reflects the duration or seriousness of the illness in that particular population.36
|
Professor McCance reminded the team to be careful about the administration of NaHCO3. If it were given in bolus form, there could be a large rise in the rate of production of CO2 per litre of venous blood, and if all of this new CO2 were not removed in a single pass through the lungs, there could be a significant rise in the arterial PCO2, with potentially devastating effects in the brain. Accordingly he asked, ‘Did the studies on large populations of patients with DKA who developed cerebral oedema control for whether
was given as a bolus vs. slow and steady addition? If they did not, one should be hesitant to apply that information to every patient with DKA, especially if there is a good indication to administer NaHCO3’. The message from their mentor was clear—Professor McCance did not want to provide a standard recipe for the use (or not) of NaHCO3 in DKA. Rather, he was encouraging them to think like physicians, examine and weigh the data, then make their own decisions in individual cases. Professor McCance had been very impressed with the skills of this medical team, and he congratulated them on a very complete and well thought through analysis of the derangements in their patient. Once again, he said in conclusion, he had learned more about this fascinating subject at the bedside, from both them and their patient, than he had from reading textbooks or articles.
|
Appendix I |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
1. Case summary
An 8-year-old, previously healthy female child presented to hospital with a history of shortness of breath and cough for the past two days. When she was seen by her family physician, the initial diagnosis was otitis media, and she was started on an oral antibiotic (cefprozil). Upon arrival in the emergency department, in addition to the symptoms related to the cough and cold, she complained of polyuria, polydipsia, and of being tired; there was also weight loss of 2 to 3 kg over the past 2 weeks. Physical examination revealed heart rate 123 bpm, respiratory rate 34/min, O2 saturation 99% in room air, blood pressure 113/73 mmHg, tympanic temperature 36.6°C, Glasgow coma scale 12/15, and weight 27 kg. The patient was noted to be using her accessory muscles for breathing, and was felt to have asthma, thus received two doses of salbutamol in rapid succession by nebulizer and one dose of 40 mg oral prednisone.
Approximately 3 h after admission, she was noted to be drowsy and lethargic. Her heart rate, respiratory rate, temperature and blood pressure remained the same. She received a bolus of 250 ml 0.9% saline (10 ml/kg) and a Foley urinary catheter was inserted. The 0.9% saline infusion was continued at 125 ml/h after the bolus. Urinalysis revealed pH 5.5, specific gravity >1.030, glucose 2+, and ketones 3+. The patient became less drowsy, and started talking to her mother. Because the diagnosis of DKA had been established, an intravenous infusion of insulin (0.1 U/kg/h) was before, and the patient was transferred to the referral centre.
2. Estimating the rate of excretion of 
The first step in estimating the rate of excretion of is to estimate its concentration by using the urine osmolal gap (equations 3a and 3b).22 The second step is to divide this value by the concentration of creatinine in the urine. The normal renal response in a patient with chronic metabolic acidosis is the excretion of 15–20 mmol of per mmol creatinine.
|
| (3a) |
|
| (3b) |
|
References |
|---|
|
TopSummaryIntroductionThe consultationMorning roundsThe afternoon sessionConcluding remarksAppendix IReferences |
|---|
1. Zalunardo N, Lemaire M, Davids MR, Halperin ML. Acidosis in a patient with cholera: A need to redefine concepts. Q J Med (2004) 97:681–96.[Web of Science]
2. Carlisle EJF, Donnelly SM, Vasuvattakul S, Kamel KS, Tobe S, Halperin ML. Glue-sniffing and distal renal tubular acidosis: sticking to the facts. J Am Soc Nephrol (1991) 1:1019–27.[Abstract]
3. Cahill GFJ. Starvation in man. N Engl J Med (1970) 282:668–75.[Medline]
4. Halperin ML, Rolleston FS. Clinical Detective Stories: A Problem-Based Approach to Clinical Cases in Energy and Acid-Base Metabolism (1993) 1. London, England: Portland Press.
5. Kamel KS, Lin S-H, Cheema-Dhadli S, Marliss EB, Halperin ML. Prolonged total fasting: a feast for the integrative physiologist. Kid Internat (1998) 53:531–9.[CrossRef]
6. Owen OE, Felig P, Morgan AP, Wahren J, Cahill GF Jr. Liver and kidney metabolism during prolonged starvation. J Clin Invest (1969) 48:574–83.[Web of Science][Medline]
7. Owen OE, Morgan AP, Kemp HG, Sullivan JM, Herrera MG, Cahill GFJ. Brain metabolism during fasting. J Clin Invest (1967) 46:1589–95.[Web of Science][Medline]
8. Halperin ML, Cherney DZI, Kamel KS. Ketoacidosis. In: Acid-Base and Electrolyte Disorders: a companion to Brenner and Rector's, The Kidney—DuBose TD, Jr, Hamm LL, eds. (2002) Philadelphia: WB Saunders. 67–82.
9. Flatt JP. Conversion of carbohydrate to fat in adipose tissue, an energy yielding and therefore self-limiting process. J Lipid Res (1970) 11:131–43.[Abstract]
10. Del Boca J, Flatt J. Fatty acid synthesis from glucose and acetate and the control of lipogenesis in adipose tissue. Eur J Biochem (1969) 11:127–34.[Web of Science][Medline]
11. Clausen T, Everts ME. Regulation of the Na,K-pump in skeletal muscle. Kidney Int (1989) 35:1–13.[Web of Science][Medline]
12. Brand MD, Esteves TC. Physiological functions of the mitochondrial uncoupling proteins UCP2 and UCP3. Cell Metabolism (2005) 2:85–93.[CrossRef][Web of Science][Medline]
13. Davids MR, Segal AS, Brunengraber H, Halperin ML. An unusual cause for ketoacidosis. Q J Med (2004) 97:365–76.[Web of Science]
14. 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]
15. Sapir DG, Owen OE. Renal conservation of ketone bodies during starvation. Metabolism (1975) 24:23–33.[CrossRef][Web of Science][Medline]
16. Halperin ML. How much ‘new’ bicarbonate is formed in the distal nephron in the process of net acid excretion? Kidney Int (1989) 35:1277–81.[Web of Science][Medline]
17. Halperin ML, Kamel KS, Ethier JH, Magner PO. What is the underlying defect in patients with isolated, proximal renal tubular acidosis? Am J Nephrol (1989) 9:265–8.[Web of Science][Medline]
18. 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.
19. Halperin ML, Bear RA, Hannaford M. Selected aspects of the pathophysiology of metabolic acidosis in diabetes mellitus. Diabetes (1981) 30:781–9.[Abstract]
20. Corey HE. Stewart and beyond: New models of acid-base balance. Kidney Int (2003) 64:777–87.[CrossRef][Web of Science][Medline]
21. Stewart PA. How to understand acid-base. A quantitative acid-base primer for biology and medicine (1981) New York: Elsevier.
22. Dyck R, Asthana S, Kalra J, West M, Massey L. A modification of the urine osmolal gap: an improved method for estimating urine ammonium. Am J Nephrol (1990) 10:359–62.[Web of Science][Medline]
23. Fox IH, Halperin ML, Goldstein MB, Marliss EB. Renal excretion of uric acid during prolonged fasting. Metabolism (1976) 25:551–9.[CrossRef][Web of Science][Medline]
24. Oh M, Halperin ML. Toxin-induced metabolic acidosis. In: Acid-Base Disorders and their Treatment—Gennari FJ, Adrogue HJ, Galla JH, Madias NE, eds. (2005) Boston, MA, USA: Marcel Decker. 377–409.
25. Jungas RL, Halperin ML, Brosnan JT. Lessons learnt from a quantitative analysis of amino acid oxidation and related gluconeogenesis in man. Physiol Rev (1992) 72:419–48.
26. Halperin ML, Goguen JM, Scheich AM, Kamel KS. Clinical consequences of hyperglycemia and its correction. In: Clinical Disturbances of Water Metabolism—Seldin DW, Giebisch G, eds. (1993) New York: Raven Press. 249–72.
27. 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]
28. McGee S, Abernethy WB, Simel DL. Is this patient hypovolemic? J Am Med Assn (1999) 17:1022–9.
29. Gerrard S, Khayam-Bashi H. Negative interference with the Ektachem (Kodak) enzymic assay for creatinine by high serum glucose. Clin Chem (1984) 30:1884. (letter).
30. Nanji A, Campbell D. Falsely-elevated serum creatinine values in diabetic ketoacidosis—clinical implications. Biochem (1981) 14:91–3.
31. Gowrishankar M, Kamel KS, Halperin ML. Buffering of a H+ load; A ‘brain-protein-centered’ view. J Am Soc Nephrol (2007) 18:2278–80.
32. Allon M, Shanklin N. Adrenergic modulation of extrarenal potassium disposal in men with end-stage renal disease. Kidney Int (1991) 40:1103–9.[Web of Science][Medline]
33. Cheema-Dhadli S, Lin S-H, Halperin ML. Mechanisms used to dispose of a progressively increasing alkali load in the rat. Am J Physiol (2002) 282:F1049–55.[Web of Science]
34. Glaser N, Barnett P, McCaslin I, Nelson D, Trainor J, Louie J, et al. Risk factors for cerebral edema in children with diabetic ketoacidosis. N Engl J Med (2001) 344:264–9.
35. Edge JA, Hawkins MM, Winter DI, Dunger DB. The risk and outcome of cerebral oedema developing during diabetic ketoacidosis. Arch Dis Child (2001) 85:16–22.
36. Edge JA, Jakes RW, Roy Y, Hawkins MM, Winter D, Ford-Adams ME, Murphy MP, Bergomi A, Widmer B, Dunger DB. The UK case-control of cerebral oedema complicating diabetic ketoacidosis in children. Diabetologia (2006) 49:2002–9.[CrossRef][Web of Science][Medline]
37. Halperin ML, Kamel KS, Maccari C, Carlotti ACP, Bohn D. Strategies to reduce the danger of cerebral edema in a pediatric patient with diabetic ketoacidosis. Ped Diabetes (2006) 7:191–5.[CrossRef]
38. 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.[Web of Science][Medline]
39. Edge JA. Cerebral oedema during treatment of diabetic ketoacidosis: are we any nearer finding a cause? Diabetes Metab Res Rev (2000) 16:316–24.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Lee and L. V. Campbell Diabetic Ketoacidosis: the Usual Villain or a Scapegoat?: A novel cause of severe metabolic acidosis in type 1 diabetes Diabetes Care, March 1, 2008; 31(3): e13 - e13. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||