Brief History/Physical

It's about 11 P.M. on a busy night, a third of your shift is over, as you pick up the next chart. The triage note says this is a 44-year-old female with a two-day history of weakness, nausea, vomiting and abdominal pain. Vital signs: pulse 114, respiration rate 24, BP 140/96, temperature 98.2°F, O₂ sat 99% on room air.

When you enter the room, you see a thin white female who looks her stated age. She is pale with deep respirations and has recently vomited into a nearby emesis basin. She tells you that she began having upper abdominal pain about three days ago and started vomiting yesterday. She states that she has not been able to hold anything down for over 24 hours. She denies fever, chills, diarrhea or hematemesis. Her last menses was three weeks prior.

She denies any significant past medical history and lists tonsilectomy as her only prior surgery. She has smoked a pack a day for 24 years and admits to "a couple of drinks" a day with dinner. Family history is positive for breast cancer and heart disease after age 65.

On physical exam, she appears ill with dry mucus membranes and tenting of the skin. HEENT exam is otherwise within normal limits. The lungs are clear with equal breath sounds. The heart rate is tachycardic without murmur or rub. On abdominal exam, there is mild epigastric tenderness without rebound or guarding. The liver is somewhat enlarged, being palpable 3 cm below the costal margin. The stool is heme negative on rectal and the pelvic exam is unremarkable. Extremities are negative for edema, clubbing or cyanosis and the neurological exam is non-focal. The patient does complain of becoming lightheaded when you stand her to assess her gait.

Q. How would you proceed with the initial work-up of this patient?

A. Orthostatic vital signs will further assess the level of dehydration evidenced by dry mucus membranes and tenting skin. Initial laboratory orders should include urinalysis, a pregnancy test, electrolytes, BUN and creatinine. Because she is tachypneic, you suspect an acid-base abnormality and add electrolytes and a blood gas for pH. A white count, liver function studies and amylase and lipase are also indicated, based on clinical findings of epigastric tenderness and hepatomegaly. An acute abdomen series (AAS -- usually a chest X-ray, and a supine and erect abdominal film) is added to rule out ileus or bowel obstruction.

Q. What is your initial differential diagnosis? How do your laboratory findings narrow this list?

A. The initial differential diagnosis in this case is quite broad. Upper abdominal pain and vomiting can be caused by intrabdominal pathology, such as gastritis, peptic ulcer disease, cholecystitis, gallstones, hepatitis, pancreatitis, gastroenteritis, bowel obstruction or ileus. Extra-abdominal pathology, such as coronary artery disease, renal stones or infection, pregnancy and other gynecological conditions may also be responsible. Toxic ingestions, diabetic ketoacidosis and alcoholic ketoacidosis are possible systemic etiologies.

Laboratory Results

Orthostatics

• supine: BP 144/96, P 102;
• sitting: BP 140/98, P 114
• standing: BP 138/98, P 134.

Lab values

• Urinalysis: 1.015, 5.0, trace ketones, trace protein, micro negative for cells and crystals
• Urine Pregnancy Test: negative
• Electrolytes: Na+ 140, K+ 3.6, Cl^-104, HCO3^-10, BUN 45, Cr 1.5,
• Glucose 145
• CBC: WBC 12.7, normal differential, HGB 11.2, HCT 35, PLT 215
• LFT: AST 52, ALT 48, AP 50, bili 1.1
• Amylase: 32
• Lipase: 115
• ABG: 7.14/CO₂ 24/O₂ 84/96% on room air
• AAS: normal

These results are significant for dehydration and a metabolic acidosis. Dehydration is manifest by a 32-point increase in pulse rate from supine to standing and a BUN to creatinine ratio of 30. Findings indicative of metabolic acidosis are a pH of 7.14 and an HCO3^- of 10. The CO₂ of 24 and respiratory rate of 24 represent partial respiratory compensation. The anion gap is 26 (140-[104+10]) with the normal value being 12±2. Therefore, this patient has an elevated anion gap metabolic acidosis.

Discussion

Metabolic acidosis is generally divided into two types: those with a normal anion gap and those with an increased anion gap. The anion gap is used to estimate the concentration of anions other than Cl^- and HCO3^-. The anion gap (AG) is calculated by subtracting the sum of the measured Cl^- andHCO3^- from the Na+ value as below:

AG = [Na+] - ([Cl^-] + [HCO3^-])

The normal value for the AG is 12±2. The value for this patient is 26. An increased anion gap metabolic acidosis can be caused by generation of an increased amount of an endogenous anion, such an uremia or lactic acid, or ingestion of a substance known to produce an exogenous anion, such as methanol. The mnemonic MUDPILES may be helpful:

Methanol
Uremia
Diabetic ketoacidosis
Paraldehyde
Isoniazid
Lactic acidosis, a Lcoholic ketoacidosis
Ethylene glycol
Salicylates

Q. What, if any, further testing would you do based on the above results?

A. Further laboratory testing should be directed toward narrowing the differential diagnosis of the cause of the acidosis. Serum ketones, a salicylate level, measured osmolarity, an alcohol level and a lactate level would be helpful.

Results

• Serum ketones small
• ASA negative
• Alcohol (ETOH) level zero
• Measured osmolarity 308
• Lactate level 3.0 mmol/L (nl 1-3 mmol/L)

To narrow the differential for an elevated anion gap metabolic acidosis, we must rely on a combination of history, clinical findings and a few laboratory values. The patient has no history of taking isoniazid or paraldehyde and denies any recent aspirin ingestion. The BUN is only slightly elevated, thus eliminating uremia as a likely etiology. The urine was positive for ketones. Serum ketones would lend further credence to either DKA or AKA as an etiology. While the glucose is slightly high, the patient has no history of diabetes, making DKA a less likely explanation.

Lactic acidosis can be secondary to hypoperfusion or hypoxia from any condition. Some common etiologies of lactic acidosis include hypovolemic shock, sepsis, seizures, diabetes, liver disease, toxins, such as phenformin and ethanol, and certain congenital errors of metabolism. Most of these are evident by the history and physical. A lactate level can be measured if uncertainty exists.

Either ethylene glycol or methanol may cause a severe metabolic acidosis if ingested in small amounts. Most hospital laboratories don't have the capability to test for either of these toxins; however, both will increase the osmolal gap, if present.

The osmolal gap is calculated by subtracting the calculated osmolarity from the measured osmolarity and is normally <10. The osmolarity is calculated by the following formula:

Osmolarity = 2[Na+] + [glucose]/18 + [BUN]/2.8

If we use the above formula, the calculated osmolarity for this patient is 304. The measured osmolarity is 308, leaving a gap of 4, which rules out an exogenous anion as the source of the acidosis.

Either ethyl alcohol or isopropyl alcohol can also elevate the osmolal gap but neither causes a significant metabolic acidosis. Since ethanol is often co-ingested with other alcohols and can be directly measured, its effect on the osmolal gap can be accounted for by the following addition to the above formula:

Osmolarity = 2[Na+] + [glucose]/18 + [BUN]/2.8 + ETOH/4.6

The presence of calcium oxalate crystals in the urine, in the setting of an increased anion gap metabolic acidosis, should prompt one to suspect ethylene glycol toxicity. Ethylene glycol, commonly found in antifreeze, has also been reported to cause fluorescence of urine under a Woods lamp.

Q. What is your final diagnosis?

A. The correct answer is alcoholic ketoacidosis (AKA), a syndrome of an elevated gap metabolic acidosis and dehydration that usually occurs in chronic alcoholics after a period of binge drinking terminated by vomiting and decreased oral intake. The most common presenting complaints are nausea, vomiting and abdominal pain. Some authorities have suggested that AKA is more common in women alcoholics but other authorities dispute this(5)

Pathophysiology

The pathophysiology of AKA has never been proven but is thought to involve acute starvation, on top of chronic malnutrition, that results in glycogen depletion and production of ketones. There is felt to be a relative deficiency of insulin and increase in the counter-regulatory hormones glucagon, cortisol, growth hormone and catecholamines present during starvation that also plays a role in the development of ketosis.(5) Unlike diabetic ketoacidosis, there is often no inciting illness in AKA.

The primary ketones produced in AKA are b-hydroxybutyrate (bHB) and acetoacetate (AcAc). The ratio of bHB to AcAc depends on the ratio of NADH to NAD. Because of dehydration, the ratio of NADH to NAD in alcoholic ketoacidosis is often elevated, resulting in a higher percent of ketones as bHB.

History

The history is commonly one of binge drinking that is often superimposed on chronic alcohol abuse. The patient begins to complain of upper abdominal pain and nausea that leads to decreasing food intake, usually with continued alcohol intake. After two to three days, the patient begins profuse vomiting, resulting in dehydration and acidosis.

Physical

Patients often appear quite ill. Common vital sign abnormalities include tachycardia, tachypnea or Kussmaul respirations, orthostasis or frank hypotension, if the dehydration is severe. Fever is not a common finding and should promote the search for a complicating disorder, such as infection or delirium tremens. A fruity odor from ketosis may be noted on the breath. Alcohol odor may also be noted if recent intake has occurred. The abdomen is often tender but is usually not surgical. Common coexistent abdominal conditions in patients with AKA include hepatitis, cirrhosis, gastritis, peptic ulcer disease and pancreatitis.

Laboratory/Radiology/Special Tests

Alcoholic ketoacidosis is marked by several laboratory abnormalities. Chief among these is the presence of an elevated anion gap metabolic acidosis. The ketones produced in AKA are primarily b-hydroxybutyric acid, rather than acetone or acetoacetoic acid (AcAc). Because the nitroprusside test used for ketones detects only AcAc, patients with AKA may have falsely low to negative tests for ketones. The ratio of bHB to AcAc reverses with reversal of dehydration. Although the patient may be clinically improving, the resulting increase in AcAc can produce an apparently "worse" laboratory ketone value.

Occasionally, a patient with AKA may present with alkalemia or a mixed acid-base disorder, secondary to contraction alkalosis form profuse vomiting. A mild lactic acidosis, secondary to ethanol use, can also compound the patient's acid-base status.

Both hypo- and hyperphosphatemia have also been noted in patients with AKA.(2) As in all forms of acidosis, the serum potassium level may be falsely elevated. Mild elevations in liver function tests, as well as pancreatic enzymes, may be seen secondary to ethanol use.

Q. How would you treat this condition?

A. IV hydration is the mainstay of treatment for alcoholic ketoacidosis. As most alcoholics are glycogen depleted and the serum glucose level in most patients with AKA is normal to slightly elevated, a crystalloid with supplemental glucose, such as D5NS, is the fluid of choice. Thiamine 100 mg IV or IM should be given with glucose therapy in any alcoholic to prevent the development of Werkicke-Korsakoff syndrome. Multivitamins, magnesium and folate supplementation are indicated for chronic alcoholics who are usually malnourished. Potassium replacement should be started as soon as adequate urine output is established in patients with low to normal K+ values, as the K+ value will fall with correction of the acidosis. Bicarbonate is rarely, if ever, indicated in AKA. Hyperphosphatemia will usually resolve with treatment of the dehydration. Phosphorus supplementation is indicated in patients whose level is less than 1 mEq/L.

Insulin is not indicated in the treatment of AKA, unless there is pre-existing diabetes.

Benzodiazepines may also be required for the prevention of DTs.

Treatment should be continued until dehydration, acidosis and electrolyte abnormalities have been reversed and good oral intake is established. Most patients will correct their metabolic and fluid abnormalities with 12-24 hours IVF and glucose administration.