The topic of this lecture is complex acid-based disorders. I appreciate the invitation by Society of Critical Care Medicine for this lecture. I have no conflict of interest related to this activity. The objectives of this talk is related to metabolic acidosis, including interpretation of anion-gab and non-anion-gab acidosis, along with metabolic alkalosis and its pathogenesis. We start with metabolic acidosis. For evaluation of acid-based disorders, history and physical examination play a very important role. Indeed, with history and physical examination, we can identify different sources of acid-based disorders. For example, GI tract, renal respiratory neurologic, consuming drugs or taking toxins, volume status and respiratory rate are very helpful. The next step is to assess arterial pH. If pH is less than 7.37, patient is acidemic. If it is more than 7.43, patient is alkalemic. Obviously, Henderson-Hasselbalch formula can help us to understand acid-based disorders. Certainly, you do not need to memorize this formula. Just the ratio between bicarbonate to PCO2 as determining factors for pH is important to memorize. For acidemia, there are two different case scenarios. Bicarbonate decreases, which would be metabolic acidosis. PCO2 increases, which would be respiratory acidosis. In the setting of alkalemia or high pH, if bicarbonate increases, it is metabolic alkalosis. If PCO2 decreases, it would be respiratory acidosis. After identifying the source of acidosis or alkalosis, metabolic and respiratory, it is important to understand what is expected compensation by the other organ. Lungs are able to compensate for metabolic acidosis and metabolic alkalosis almost immediately. However, kidneys take about 24 to 72 hours in order to compensate for changes in respiratory function. So, for example, if patient is respiratory acidosis in acute fashion, for each mmHg increase in PCO2, bicarbonate only increases by 0.1 mEq per liter. For chronic setting, respiratory acidosis for each mmHg increase in PCO2, bicarbonate increases by 0.3 mEq per liter. The same for respiratory alkalosis. In acute respiratory alkalosis, for each mmHg decrease in PCO2, bicarbonate decreases by 0.2 mEq per liter. And for chronic setting, for each mmHg decrease in PCO2, bicarbonate decreases by 0.4 mEq per liter. For metabolic acidosis, each mmHg decrease in bicarbonate would be associated with 1.2 mEq decrease in PCO2. And for metabolic alkalosis, for each mmHg increase in bicarbonate, PCO2 increases by 0.7 mEq. In the next step, it is important to calculate an ion gap in order to identify the type of acid base or electrolyte imbalance. So, the formula for calculation of an ion gap is sodium minus sum of chloride with bicarbonate. An ion gap basically is the difference between unmeasured anions and unmeasured cations. Usually, the normal range of an ion gap is 8 to 12. Unmeasured anions include albumin, phosphate, sulfate, lactate, and pyruvate. And unmeasured cations include potassium, calcium, magnesium, and immunoglobulins. Albumin is one of the most important unmeasured anions that participates in an ion gap, normal level anion gap of 10 to 12. Therefore, with fall-in albumin, an ion gap range should be corrected. So, every one gram per deciliter fall-in serum albumin level, we should add 2.5 mEq per liter to an ion gap. Next step in acid base disorder assessment is to identify if an ion gap acidosis is associated with a secondary primary process or not. This is where delta-delta calculation comes to fruition. When an unmeasured anion like lactic acid is added to the body water. After ionization, each millimole of that acid would be able to be buffered by one millimole of bicarbonate. Therefore, although the direction of change in the acid and bicarbonate is in different directions, the extent of changes are different. Indeed, for every 1 mEq per liter of acid added to the circulation, serum bicarbonate decreases by 1 mEq per liter. Therefore, an ion gap increases by 1 mEq per liter. Therefore, delta-anion gap ratio to delta-bicarbonate should be around 1. To calculate delta-delta, the ratio of changes in anion gap to changes in bicarbonate should be calculated. So, as normal, anion gap is about 10. Anion gap minus 10 divided by 24 minus bicarbonate can help us to understand delta-delta ratio. When the ratio of delta-anion gap to delta-bicarbonate is 1, it is a simple anion gap acidosis. If the ratio is less than 1, we need to suspect a superimposed non-anion gap acidosis. When the ratio is more than 1, we need to suspect a superimposed metabolic alkalosis. For high anion gap metabolic acidosis, there are multiple reasons that the anion gap can increase. The mnemonic of Goldmark could be used to summarize these anion gap metabolic acidosis causes, including ethylene glycol, propylene glycol, 5-oxopyraline or pyroglutamic acid, lactic acidosis, which would be L and D lactic acidosis, we will focus on this later during this lecture, methanol, aspirin, chronic kidney disease, or renal, and then ketoacidosis. Let's start with ketoacidosis. Ketone bodies are water-soluble material that produce in perivenular hepatocytes. After adipose tissue release fatty acids and blood glucose enter liver, they can enter Krebs cycle. And along with amino acids coming from muscle mass generally, they can generate ketone bodies, including beta-hydroxybutyric acid, acetoacetate, and acetone, that would end up going back to bloodstream. Acetone has a fatter and is least abundant ketone available. Beta-hydroxybutyric acid to acetoacetate ratio could help us to identify the cause of ketoacidosis. For example, in normal situation, this ratio is about 1 to 1. In diabetic ketoacidosis, it is about 3 to 7 to 1. And for alcoholic ketoacidosis, the highest ratio of 10 to 1. For ketoacidosis, glycolysis needs to be shifted toward ketogenesis. There are several different factors that can make that shift. For example, decreased glucose and insulin and increasing glucagon and fatty acids can do that. There are factors that can impact this balance between insulin and glucagon, including age, fasting, using micronutrients, glycogen storage, and the muscle amino acid metabolization. When ketogenesis happens, then ketolysis is used in order to produce energy in different organs. So peripheral tissue mitochondrial generation of acetyl-CoA, ATP, bicarbonate, and water in the liver would result in generational ketone bodies that are used in kidney, muscle, and brain as a source of energy production. In ICU, ketoacidosis is often multifactorial. Majority of our patients are under stress with significant increase in stress hormones. In addition, majority of them are not fed, therefore starvation ketoacidosis adds to the picture. Some have been alcohol consumers, some are diabetic, therefore have higher risk of developing ketoacidosis accordingly. Some drugs like theophylline, salicylate, HGL-2 inhibitors, and acetone can increase ketogenesis. And finally, some of these patients who receive a very small amount of carbohydrate as a part of their diet, they can develop ketoacidosis with the same mechanism as ketogenic diet. We'll start with diabetic ketoacidosis. There are multiple processes that end up ketoacidosis in diabetic patients. So in peripheral tissues, by decreased glucose utilization, hyperglycemia would result in osmotic diuresis and hypovolemia. In addition, in liver, by increased glucose production, hyperglycemia would result in osmotic diuresis and hypovolemia. In addition, liver shifts glycolysis to ketogenesis, therefore ketones are generated, and that results in ketoacidosis, decreased base reserve, and metabolic acidosis. On the other hand, adipose tissue releases free fatty acids in order to provide energy to other organs, which results in ketoacidosis and metabolic acidosis. In management of diabetic ketoacidosis, expansion of intravascular volume is extremely important. As I mentioned earlier, due to hyperglycemia and osmotic diuresis, often these patients are very dry. Therefore, replacement of intravascular volume with isotonic normal saline is very important. In addition, when glucose level reaches to around 200, that's the time to switch to D5 normal saline in order to avoid hyperglycemia while we continue our insulin. Insulin is an important and mainstay of treatment, as insulin would change the balance between insulin and glucose in order to avoid further progression of diabetic ketoacidosis. Potassium often needs to be replaced to 4-5 mmol per liter. However, phosphorus, although decreases to a certain degree, often does not need replacement. Bicarbonate sometimes are given if blood pH, arterial pH is less than 6.95 in order to maintain pH above 7. However, in 12 randomized clinical trials, there was no advantage found in using bicarbonate in management of diabetic ketoacidosis patients. Indeed, bicarbonate may be associated with risk of hypokalemia and worsening cerebral edema. Starvation ketoacidosis is discussed in ICU setting very often, because a lot of our patients do not receive feeding particularly the first few days of their critical illness. However, it is important to understand the process of starvation ketoacidosis. It takes 12-14 hours for ketone body start appearing in the bloodstream. Of about 1 mmol per liter. Then it takes another 2-3 weeks for gradual increase in ketone bodies and drop in bicarbonate. After 3-4 weeks, peak ketone bodies happen with 8-10 mmol per liter with bicarbonate decrease of 7-8 mEq per liter. This is the maximum of starvation ketoacidosis. So it's usually mild stable elevation due to inhibitory effect of lipolysis and increased tissue utilization of ketone bodies. And that is why during the prolonged starvation, ketone bodies do not increase beyond 8-10 mmol per liter. It is important to know that during starvation ketoacidosis, plasma-free fatty acids increase at the beginning. Then plasma acetone followed by beta-hydroxybutyric acid and blood acetoacetate increase accordingly. The highest amount ketone body in bloodstream would be beta-hydroxybutyric acid instead of starvation ketoacid. The rate of increase in ketone bodies among patients with different age are different. Among younger individuals, particularly in pediatric and neonatal range, rate of increase in beta-hydroxybutyric acid after starvation are within 1 hour. However, for adults, rate of increase significantly slower, takes 3-4 weeks for reaching a plateau. There are some exceptions in increase in ketone bodies during starvation. For example, pregnant women and lactating women, because they have increased counter-regulatory hormones, they have decreased ketone-induced lipolysis inhibition. Therefore, their beta-hydroxybutyric acid increases faster in comparison with other adults under our care. If patients receive carbohydrates less than 40 grams per day or less than 10% of their caloric needs, starvation ketoacidosis can result rapid increase in ketone bodies. Also, among ICU patients, due to increased counter-regulatory hormones, this phenomenon is also more significant. Drug-induced ketoacidosis is a well-known phenomenon in outpatients in the hospital or ICU. Alcohol, sodium glucose co-transported to inhibitors or SGL2 inhibitors. Aspirin, theophylline, and acetone are all associated with ketoacidosis. Alcoholic ketoacidosis is usually in a setting of chronic alcoholism plus malnutrition. Between men and women, the rate of alcoholic ketoacidosis is the same. It takes about 1 or 2 days after binge drinking to develop ketoacidosis. Usually, these patients are hypoglycemic, hypokalemic, hypophosphatemic, and hypomagnesemic. There are several mechanistic reasons for alcoholic ketoacidosis. By increased alcohol ingestion, which results decreased gluconeogenesis, would resolve the need for use of other sources of energy, including fatty acids. In addition, alcohol ingestion directly can decrease insulin and increase glucagon. Patients who are alcoholic also have decreased food intake, which results decreased glucogen storage, and that would lead to increase in ketone body production and metabolic acidosis. In addition, these patients have metabolic stress and dehydration, which increases contra-regulatory hormones, including cortisol, norepinephrine, and growth hormone, which again leads to production of ketone body. Treatment of alcoholic ketoacidosis includes using vitamin replacement, including thiamine, giving them glucose to increase insulin, decrease lipolysis, treat volume depletion, replace magnesium and potassium, and there is absolutely no role for insulin or bicarbonate in alcoholic ketoacidosis. Other drug-induced ketoacidosis include the one that is reported in sodium glucose co-transporters to inhibitors. These drugs inhibit glucose reabsorption in proximal tubules, therefore result glucosuria, therefore can control blood glucose level among diabetics. A possible mechanism of ketoacidosis after taking this medication could be increased reabsorption of ketone bodies in proximal tubules or increased lipolysis. Treatment is to stop SGL-2 inhibitors, hydrodyspation, and insulin for blood glucose control should be used. In salicylate toxicity, 90% of patients have anion-gate metabolic acidosis. 40% of patients have ketoacidosis, particularly in children and younger adults. Lactic acidosis is also very common. Theophylline toxicity could result lactic acidosis or ketoacidosis due to increased lipolysis and free fatty acid chain availability. For treatment of theophylline-induced ketoacidosis, we have to stop theophylline, provide supportive care, including potassium supplements, and severe toxicity hemodialysis could occur. Other drug-induced ketoacidosis includes acetone ingestion, which is well-known in toxicology literature, usually found in nail polish. These patients present with ketosis, they have osmolar gap, however, it may not be present at the time of presentation, and acetone can have a prolonged half-life of 7 to 27 hours. Usually, treatment is supportive, and dialysis also could occur. Ketogenic diet is also very important. When we severely restrict carbohydrate diet and fat is used as a source of energy, we can develop ketoacidosis due to ketogenic diet. There are several diets available, including classic ketogenic diet, medium-chain triglyceride diet, modified Atkins diet, and low-glycemic index treatment, most of which have very low amount of carbohydrate percentage of the total caloric intake. A scenario that mimics ketogenic diet in ICU is when patients do not receive any carbohydrate, however, they receive clavidipine or propofol for sedation or blood pressure management. We know that clavidipine is within 20% fat emulsion and propofol is within 10% fat emulsion. They deliver significant amount of calorie to patients, but no carbohydrate. This can mimic what ketogenic diet can mimic in other patients. There are several drugs that we use routinely within ICU setting that contain significant amount of fat emulsion, which we need to be very aware of when these patients particularly do not receive any carbohydrates. There is also stress ketoacidosis, which is due to increased counter-regulatory hormones, which increases ketone bodies due to increased insulin resistance, increased lipolysis, and adrenaline per se would decrease acetyl-CoA carboxylase, and that would result in additional ketone body production. Now, switching gears to our lactic acidosis. This is most common cause of anion gap acidosis within ICU setting. Lactic acid increases anion gap when its serum concentration is usually more than 5 mEq per liter. Lactic acid exists in two different optical isomers. L-isomer is picked up by standard assays and is produced by L-lactate dehydrogenase only found in humans. However, D-lactate acid is produced by D-lactate dehydrogenase, which is only in bacteria. L-lactate acidosis are two different types. In type A, usually there is inadequate oxygen delivery. In type B, there is increased production of lactic acid with preserved oxygen delivery. D-lactate acidosis, however, is produced by bacteria and usually seen in a setting of short or blind bowel movement. Type A lactic acidosis is seen when there is a mismatch between oxygen delivery and consumption, which results in increased anaerobic glycolysis, therefore lactic acid production. It can happen in a shock state or non-shock state, examples of which are severe hypoxemic respiratory failure, receiving toxins including cyanide, severe anemia, or thromboembolism. Not all lactic acidosis are considered detrimental to the health of patients. For example, in exercise-induced lactic acidosis, there has been reports of a pH as low as 6.97 with severe L-lactate elevation. In patients with ground mouth seizure, also pH could be very low, with low bicarbonate, but significant increase in L-lactic acid level, which usually recovers within 60 minutes after ground mouth. Type B lactic acidosis have different reasons. Etiologies include diabetes mellitus, glycogen storage disease, ethanol, hepatic failure, malignancy, and drop. In setting of liver failure, lactic acidosis is rare if liver failure not associated with other organ failures. Decreased oxidization of lactate to glucose would result in gluconeogenesis. Decreased capacity of lactate utilization also decreases lactate metabolism, therefore lactic acidosis. Metformin-associated lactic acidosis is a well-described entity in the literature. It also can happen in the setting of use of antiretroviral medications or propofol. Propofol results in decreased oxidative phosphorylation of bimicrochondria, therefore decreased ATP availability. It is usually dose-dependent and increases with other comorbid conditions. In the setting of salicylate toxicity, uncoupling oxidative phosphorylation would result in lactic acid production. Increased anaerobic glycolysis and glucose utilization would add to the problem, which results in increased anion gap, which is usually due to unmeasured anions, including ketone bodies and lactate. Tumors also can result in lactic acidosis. Some of these tumors have internal ischemic events. Bleeding inside a tumor or tumor is large without appropriate vascular supplement would result in anaerobic metabolism, therefore lactic acidosis. Limitations in oxidative phosphorylation in some tumor cells and liver metastasis and failure are additional reasons for tumor-associated lactic acidosis. D-lactic acid is not produced by human cells. It's only produced by bacteria in the gut. As the human body is not able to metabolize it, if the source continues producing it, it accumulates in the body. It's been seen in short bowel or malabsorption. It commonly recurs and male individuals have higher tendency to develop D-lactic acidosis. Its symptoms is not only related to D-lactic acid but also due to simultaneous generated neurotoxins by bacteria in the gut, which includes causal breathing, nausea or vomiting, impaired alertness, confusion, gait disturbances, slurred speech, and involuntary eye movement. Antimicrobials against gram-positive competitors, which results increase in lactobacillus, would increase chances of D-lactic acidosis. Ingestion of propylene glycol, which is a vehicle of some drugs, also is a risk factor with D-lactic acidosis. Decreased intestinal motility and intestinal malabsorption would also increase bacterial overgrowth and therefore risk of D-lactic acidosis. Switching gears to metabolic acidosis with anion gap and osmolar gap. This table shows differential diagnosis of patients who have anion gap and osmolar gap. So if patient has anion gap metabolic acidosis and osmolar gap is normal, we need to think about anions like salicylate. When anion gap acidosis is accompanied by osmolar gap, then alcohol intoxication should be suspected, including ethanol, ethylene glycol, propylene glycol, and methanol. In intoxication with isopropanol, osmolar gap is present and elevated. However, there is no anion gap metabolic acidosis. Ethylene glycol is metabolized in the liver and as a result would lead to production of calcium oxalate, which is eliminated by the kidney. Ethylene glycol per se is not toxic. However, calcium oxalate is toxic after precipitation in interstitium of multiple tissues. In human urine following ethylene glycol ingestion, crystals could be observed, as it is obvious in the panel on the left, and also calcium oxalate can precipitate not only in the tubular section but also interstitium in the kidney. Ethylene glycol toxicity is one of the major causes of osmolar and anion gap metabolic acidosis. It follows several different phases. In phase 1, which is neurological phase, happens within half an hour to 12 hours after ingestion. These patients are inebriated, they have slurred speech, they are ataxic, they have nausea, vomiting. They usually have somnolence or they are in coma. Seizure and myoclonic jerks are very common. Anion gap metabolic acidosis could be found, including lactic acidosis and glycolic acid. In phase 2 is cardiopulmonary. It happens usually 12 to 36 hours after ingestion. These patients are tachycardic, they have hypertension followed by hypotension. They are tachypneic, they have cosmos breathing, they may suffer from congestive heart failure or pulmonary edema. And finally, in phase 3, which happens in 12 to 72 hours after ingestion, we have patients that have progressive acute kidney injury and ATN, hematuria, flank pain, ileoaneuria, and calcium oxalate crystal deposition. Symptoms of metanotoxicity is usually related to neurological symptoms. They have CNS depression, progressive metabolic acidosis, tachypnea as a result, coma, nausea, vomiting, severe abdominal pain, and confusion are described. Visual changes including experiencing scotomas, dilated non-reactive pupils, hyperemia of optic discs, retinal edema, and optic nerve cupping have been described. Prognosis is usually not very bad if pH remains above 7.2. Visual changes in that case would be transient. These are different presentations of optic examination, retinal examination. In comparison with normal examination, we can see optic disc hyperemia with retinal edema, papilledema, lots of optic nerve and optic cupping have been described, all in the setting of metanotoxicity. Ethylene glycol and methanol toxicity treatment would be related to inhibition of alcohol dehydrogenase which converts non-toxic material methanol and ethylene glycol to more toxic material including formic acid and calcium oxalate. Dialysis also could be used. Ethanol and formaldehyde are alcohol dehydrogenase inhibitors and dialysis obviously removes alcohols and their derivative products. Treatments of ethylene glycol and methanol toxicity need to be considered when there is non-ingestion without rapid access to serum level determination. Patients have an on-gap metabolic acidosis with elevated osmolar gap with strong clinical suspicion or they have methanol more than 8 and ethylene glycol more than 3.2 mmol with sign of end-organ damage including eyes and kidneys. Dialysis needs to be considered when metabolic acidosis and end-organ damage present. Serum level of more than 15 for methanol, more than 6 mmol for ethylene glycol or osmolar gap of more than 20 should indicate dialysis. Switching gears toward metabolic acidosis and non-anion gap. Non-anion gap or hyperchloramic metabolic acidosis happens due to three different reasons. One is gain of acid by exogenous sources or endogenous sources including ammonia, chloride intake or lactic acidosis. Acid retention or under excretion in a setting of distal RTA or chronic kidney disease. And finally, bicarbonate loss with diarrhea or proximal RTA. To determine renal tubular acidosis, it is important to be able to calculate urine anion gap. Urine anion gap is calculated as sodium plus potassium minus chloride which are the most important anions and cations of the urine. We know that in the urine, total anions should be equal to the total cations. And total anions include measured anions and unmeasured anions and total cations are measured cations and unmeasured cations. Therefore, measured cations minus measured anions should be equal to unmeasured anions minus unmeasured cations which is definition of urine anion gap. As the most abundant cations in the urine are sodium and potassium, measured cations could be sum of sodium and potassium. And also as the most abundant anion in the urine is chloride, sodium plus potassium minus chloride could be used to measure urine anion gap. Urine anion gap also means unmeasured anions which include sulfate, phosphate, bicarbonate, and organic anions minus unmeasured cations which is calcium, magnesium, or ammonia. The most important ones are bicarbonate and ammonia that can change during renal tubular acidosis. In a setting of acidemia, urine bicarbonate should be very negligible. Urine ammonia should be very high. Therefore, urine anion gap should be very low. Indeed, in a study of patients who had non-anion gap metabolic acidosis due to diarrhea, they found to have very negative urine anion gap in comparison with those who had distal RTA, they had positive urine anion gap. There are two major types of RTA, proximal RTA which is type 2, and distal RTA which includes classical distal type 1 RTA, hyperkalemic distal, and type 4 RTA. We know that bicarbonate is reabsorbed mostly in proximal tubular cells. And we also know that a proton and bicarbonate can be excreted in distal tubular collecting ducts. In addition, ammonia is excreted inside collecting ducts as well in order to excrete titratable acids. These are clinical features of non-anion gap metabolic acidosis. In a setting of diarrhea, potassium is low, urine anion gap is negative, urine pH is variable, and based on the extent of diarrhea, severity of acidosis is different. In proximal RTA, serum potassium is low, urine anion gap could be variable along with urinary pH, and it's usually seen in a setting of Fanconi syndrome. These patients have proximal tubular dysfunction in order to absorb bicarbonate. Also, they have glycosuria, they have aminoaciduria, have phosphaturia, and other functions of proximal tubular cells are disturbed. In distal RTA or type 1 distal RTA, serum potassium is low, urine anion gap is positive, urine pH is more than 5.5. Often, these patients present with nephrocalcinosis. Type 4 RTA, they have high serum potassium, increased positive urine anion gap, urine pH is less than 5.5. And hyperkalemic distal RTA is the same as type 1, however, in a setting of very advanced CKD or other potassium-retaining diseases, which would result increasing potassium. Treatment of renal tubular acidosis is also different based on the diagnosis of proximal versus distal. In proximal RTA, bicarbonate sodium or bicarbonate potassium need to be given a significant amount in order to avoid drop in blood bicarbonate level, usually above 4 mEq per kg per day. Type 1 RTA also requires potassium bicarbonate or citrate at a dose of 1 to 3 mEq per kg per day. In type 4 and hyperkalemic distal, controlling hyperkalemia is often sufficient, which can happen with diuretics or use of resins, bowel resins, in order to avoid potassium absorption. Now, a few words on metabolic alkalosis. For a patient to develop metabolic alkalosis, there are two different factors engaged. One is induction factor, and the second one is maintenance factor. So, for induction of metabolic alkalosis, ingestion of alkali, including antacids and citrate, could result, citrated blood could result in metabolic alkalosis. For loss of acid, GI and renal losses could be considered. GI loss happens in the setting of vomiting or nasogastric suctioning, which would result significant acid loss and therefore metabolic alkalosis. In the setting of use of diuretics, syndromes like Barter and Gitelman and hyperaldosteronism, by significant loss of acid through the kidney, patient develop metabolic alkalosis. And finally, cellular shift could be considered as another induction factor for metabolic alkalosis. Now, maintenance of alkalosis requires decreased kidney excretion of excess bicarbonate, which can happen in a setting of volume contraction, in a setting of vomiting and diuretics use, hyperaldosteronism, hypokalemia, or kidney failure. After loss of chloride and water in excess of bicarbonate, as described in induction reasons for alkalosis, in a setting of dehydration, kidney decreases bicarbonate excretion from proximal tubules by increasing sodium and bicarbonate reabsorption. Also, due to decreased intravascular volume, secondary hyperaldosteronism, increased distal tubular proton excretion that would lead to further development of alkalosis. With that, I thank you for your attention.

acid-based disorders

metabolic acidosis

metabolic alkalosis

anion gap

compensation mechanisms

renal tubular acidosis

kidney excretion