Resurgence of Shoshin beriberi during the COVID-19 pandemic

Summary

This report describes two patients who presented with severe type B lactic acidosis and shock, initially thought to be due to bowel ischaemia/myocardial infarction and pulmonary sepsis, respectively. This led to a delay in the diagnosis of thiamine deficiency. In both cases, there was a dramatic response to intravenous thiamine, confirming the diagnosis of Shoshin beriberi. Both patients admitted to drinking home-brewed alcohol during the time of COVID-19 restrictions on alcohol consumption. These cases highlight the need for early diagnosis and immediate empirical treatment with intravenous thiamine in patients presenting with unexplained severe metabolic acidosis and circulatory shock.

Acute pernicious (Shoshin) beriberi is a rare, fulminant type of cardiovascular beriberi characterised by hypotension and/ or circulatory shock with severe lactic acidosis. Knowledge around the clinical characteristics, management and outcomes of patients with this condition is limited to isolated case reports and case series.^1-3 It was last described in South Africa in the 1980s in patients with a history of excessive alcohol consumption. All patients responded to administration of intravenous thiamine.¹ Here, we describe two cases of Shoshin beriberi that will alert the clinician to the pitfalls in the early diagnosis and management of this condition.

Case report 1

A 54-year-old unemployed black female patient presented to the emergency department (ED) of a regional-level hospital in KwaZulu-Natal with a five-day history of right lower-quadrant abdominal pain associated with nausea, vomiting and loss of appetite. There was no accompanying fever, diarrhoea or constipation. She denied alcohol and herbal medication intake or substance abuse.

The patient was hypotensive on admission (blood pressure 86/54 mmHg, pulse rate 88 beats/min). She had Kussmaul’s breathing with a respiratory rate of 25 breaths/min. Her oxygen saturation was 97% on room air.

Finger-prick glucose was 8.9 mmol/l. Physical examination revealed mild bilateral parotidomegaly. The abdomen was soft and not distended. Tenderness was elicited in the right lower quadrant, but there were no features of peritonitism. No organomegaly was noted and normal bowel sounds were present. Other systems were normal.

Blood gas analysis revealed a severe metabolic acidosis: pH 7.12, serum lactate 21 mmol/l, bicarbonate 3 mmol/l and PO2 157 mmHg. The full blood count was as follows: haemoglobin level 11.4 g/dl, white cell count 8.47 × 109 cells/l, platelets 126 × 109 cells/l and mean corpuscular volume 99.1 fl.

Blood chemistry was as follows: sodium 124 mmol/l, potassium 3.1 mmol/l, chloride 89 mmol/l, bicarbonate 3 mmol/l, urea 2.3 mmol/l and creatinine 100 μmol/l. Serum transaminases were mildly elevated: alanine aminotransferase 72 U/l (normal 7–35 U/l) and aspartate aminotransferase 105 U/l (normal 13–35 U/l).

Initial imaging demonstrated mild cardiomegaly on chest X-ray and a distended loop of bowel with a ‘coffee-bean’ appearance on abdominal X-ray (Fig. 1). Because of the abdominal symptoms and the abdominal X-ray findings, a provisional diagnosis of lactic acidosis secondary to sigmoid volvulus or mesenteric ischaemia was made.

Fig. 1.

Abdominal radiograph of index patient showing distended loops of bowel on the left, interpreted as a sigmoid volvulus.

The patient was reviewed by the general surgeon who felt that the abdominal X-ray showed excess gas in the descending colon with no features of sigmoid volvulus or bowel obstruction. The surgeon was of the opinion that the abdomen was ‘non-surgical’ and advised that other causes be considered.

Non-ST-segment elevation myocardial infarction (NSTEMI) was also considered in the differential diagnosis when a troponin I estimation, taken two hours after arrival in the ED, returned a result of 1 358 ng/l. However, the patient experienced no chest pain symptoms and the admission electrocardiogram (ECG) showed no evidence of myocardial infarction. The troponin I elevation or ‘leak’ was attributed to the presenting problem of profound hypotension with severe acidosis.

A computed tomography (CT) scan of the abdomen showed mildly thickened caecal and ascending colon walls. The radiologist suggested that ischaemic colitis be considered since mural calcifications were noted in multiple abdominal blood vessels, suggesting the presence of extensive atherosclerotic disease. A review was performed by the surgeon, who subsequently planned for an exploratory laparotomy, but the patient’s family refused consent for surgery.

A point-of-care ultrasound (POCUS) assessment showed that the inferior vena cava was collapsed and there was no free fluid in the abdomen. The cardiac chambers appeared normal with good contractility and there was no pericardial effusion. Initial measures included volume replacement. Despite adequate fluid resuscitation with plasmalyte B (a repeat POCUS examination showed that the inferior vena cava was no longer collapsible), the patient remained hypotensive with persistent acidosis six hours after the initial assessment and had begun to show signs of mental confusion (Table 1). She was placed on inotropic support with an adrenaline infusion.

Table 1. The metabolic response to thiamine.

	Hours after commencing thiamine
Variables	7 h	10 h	17 h	27 h
Case 1
pH	6.94	7.06	7.39	7.45
Lactate (mmol/l)	23	19	10.2	1.7
Base excess	-28.2	-24.9	-9.9	1.3
HCO (mmol/l)	6.0	7.8	15.1	25.3
	1 h	1.5 h	3 h	5 h
Case 2
pH	7.015	7.205	7.24	7.23
Lactate (mmol/l)	14	7.1	5.3	3.5
Base excess (mmol/l)	-23	-13.1	-8.5	-7.3
HCO (mmol/l)	8.9	14.7	17.3	18.3

Case 1: There was early haemodynamic improvement but the improvement in acidosis and fall in lactate levels occurred 17 hours after the commencement of thiamine.

Case 2: A dramatic improvement in acidosis and fall in lactate levels occurred within 1.5 hours of commencing thiamine.

A repeat ECG recording performed one hour after commencement of the adrenaline infusion (seven hours after initial assessment) showed ST-segment elevation in leads V3–V6 with reciprocal ST-segment depression in aVR (Fig. 2). The patient was treated as having an acute ST-elevation myocardial infarction (STEMI) and was given aspirin 300 mg and clopidogrel 300 mg orally as well as enoxaparin 30 mg intravenously, and the internal medicine team was consulted.

Fig. 2.

Case 1: A. Baseline ECG recorded at the same time as the troponin I estimation (1 358 ng/l). showed no evidence of myocardial infarction. B. ECG done one hour after commencement of adrenaline infusion shows ST-segment elevation in V3–6. C. Recording two days after intravenous thiamine shows ST-segment resolution with T-wave inversion in precordial leads V1–4.

A type II myocardial infarction was suspected in view of the temporal relationship between the initiation of adrenaline and the development of the ECG changes. In the absence of markers of sepsis (normal white cell count and C-reactive protein = 6 mg/l) the internal medicine team considered a metabolic cause for the acidosis and made a diagnosis of type B lactic acidosis secondary to thiamine deficiency, and thiamine 300 mg was administered intravenously. The decision was made to continue observing the patient while on inotropic support after thiamine had been administered.

Four hours after intravenous thiamine administration, the ST-segment elevation on the ECG subsided and T-wave inversion appeared in the precordial leads V1–4 (Fig. 2). Repeat troponin I estimations remained elevated (1 275 ng/l). The patient’s inotrope requirement gradually decreased and she was fully weaned off adrenaline infusion after 24 hours. Her level of consciousness also returned to normal over 24 hours.

The lactic acidosis steadily improved after intravenous thiamine and had completely resolved 27 hours after commencing thiamine (Table 1). At this stage, it was also noticed that the patient developed jaundice with worsening derangements in the liver enzymes. Repeat estimation 72 hours later showed that these also had largely resolved while receiving daily thiamine.

Upon discharge, the patient disclosed that she did take alcohol, but was guarded about the type and amount of alcohol she had been consuming (probably due to the South African lockdown regulations during the COVID pandemic, which prohibited the purchase/brewing of alcohol at that time).

Case report 2

A 42-year-old black male presented in extremis at 21:40 to the ED of a regional-level hospital. There was a collateral history of four days of progressive shortness of breath associated with an intermittent non-productive cough. He had been receiving antiretroviral therapy in the fixed-dose combination of tenofovir/ emtricitabine/efavirenz for an unspecified duration.

He was restless and confused with cold, clammy peripheries. He had Kussmaul’s breathing with a respiratory rate of 40 breaths/ min, pulse 104 beats/min and blood pressure 86/40 mmHg. The arterial blood gas revealed a severe lactic acidosis: pH 7.037, PaCO2 12.7 mmHg, PaO2 211 mmHg, serum bicarbonate 7.2 mmol/l, base deficit 27.2 mmol/l and lactate 20 mmol/l.

Physical examination revealed bilateral basal crackles with no other evidence of cardiac failure. Apart from the confused state, the neurological examination was unremarkable.

The patient required urgent resuscitation with intravenous fluids and a dose of ceftriaxone was administered based on a presumed diagnosis of severe sepsis/hypotension from an underlying HIV-related pneumonia that resulted in lactic acidosis. The blood pressure did not respond to volume expansion and shortly after arrival he was placed on inotropic support with an adrenaline infusion. The patient also required intubation for increased work of breathing and potential respiratory fatigue. Despite stabilisation of his blood pressure with inotropes and maintained oxygen saturations, the acidosis continued to worsen in the ensuing eight to 10 hours.

A 12-lead ECG showed sinus rhythm and ST-segment depression in leads V4–V6. A POCUS examination revealed normal cardiac chamber sizes with good contractility, and a non-collapsible inferior vena cava with no pericardial or pleural effusions was noted. Blood results revealed the following: white cell count 13.58 × 109 cells/l, C-reactive protein 6 mg/l, alanine transaminase 163 U/l (normal 10–40 U/l), aspartate transaminase 258 U/l (normal 15–40 U/l), alkaline phosphatase 237 U/l (normal 53–128 U/l) and gamma-glutamyl transferase 642 U/l (normal < 68 U/l).

The chest radiograph showed no evidence of pneumonia and in the absence of markers of sepsis, Shoshin beriberi was considered as the cause of the lactic acidosis/shock and he was given 300 mg of intravenous thiamine at 9:30. This was followed by a dramatic improvement in lactate levels and rapid resolution of acidosis (Table 1).

Repeat blood gas three hours after thiamine admission showed the following: pH 7.24, bicarbonate 17.3 mmol/l and lactate 5.3 mmol/l. Adrenalin infusion was weaned off and he was extubated at 13:27. His blood gases continued to improve as well as the derangements in liver enzymes.

Upon recovery, he admitted to drinking four jugs of homebrewed Zulu beer daily for six days a week. Repeat ECGs on discharge showed widespread T-wave inversion.

Discussion

These two cases typify the presentation of acute pernicious beriberi (Shoshin), which is characterised by severe lactic acidosis with circulatory shock secondary to severe thiamine deficiency.² Thiamine deficiency leads to decreased activity of pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase, both important enzymes in aerobic metabolism. There is subsequent accumulation of pyruvate and lactate in the tissues, resulting in lactic acidosis with peripheral vasodilatation and cardiac dysfunction.^2,3 The Shoshin type is an uncommon form of beriberi due to advanced thiamine deficiency, which contrasts markedly with the classic presentation of high-output cardiac failure with features of a hyperdynamic circulation.¹

Shoshin beriberi was first reported in early studies in the Far East secondary to chronic malnutrition and excessive consumption of polished, refined rice and cereals.³ In the Western world, Shoshin beriberi has been described mainly in alcoholics and in the critical-care setting in patients receiving total parenteral nutrition without adequate vitamin supplementation.^2-4

Upon recovery both our patients admitted to excessive consumption of alcohol in the form of Zulu beer, which was probably home brewed in the light of the current restrictions on alcohol sales during the COVID-19 pandemic. Home brewed Zulu beer is usually made with fermented maize, which may contain thiaminases that remove thiamine in the brew.⁵ Clinical markers of excessive alcohol intake were present in both patients and included parotidomegaly, raised mean corpuscular volume and elevated gamma-glutamyl transferases.

These patients usually have non-specific symptoms such as fatigue, loss of appetite, vomiting, abdominal pain and dyspnoea. Although features of cardiac failure typically form part of the clinical presentation, there have been cases reported with the absence of cardiac failure.² Neither of our patients displayed overt signs of cardiac failure. In the first patient this was likely due to volume depletion from vomiting and the associated ileus. Given the non-specific characteristics of clinical presentation of this condition, it is often confused with other acute medical problems such as massive pulmonary embolism, myocardial infarction or sepsis with multi-organ failure.^1-3

Of importance, several clinical and radiological factors led the clinician away from suspecting thiamine deficiency in both cases. In the first patient, the gastrointestinal symptoms and abdominal distension together with the X-ray appearance of the bowel suggested a sigmoid volvulus, all of which resolved completely after thiamine admission.

In retrospect, these manifestations were most likely precipitated by vomiting, which led to hypokalaemia and paralytic ileus, resulting in gaseous abdominal distension. The CT abdomen showing thickened bowel loops was likely due to bowel wall oedema. Furthermore, accompanying co-morbidities may direct attention away from a metabolic aetiology for the tachypnoea.

In the second patient, the HIV status and the presence of lung crackles suggested infection and directed the clinician’s attention away from a diagnosis of type B lactic acidosis due to thiamine deficiency. Furthermore, the raised troponin levels and ST-segment elevation on the electrocardiogram in the first patient led the clinician to also suspect a diagnosis of myocardial infarction.

The electrocardiographic changes that have been described in Shoshin beriberi are non-specific and have varied from ST-segment elevation or depression to T-wave inversion.^1-4,6,7 A notable feature that has been described is T-wave inversion in the precordial leads that occurs after thiamine replacement, as was the case in both of our patients. This transient repolarisation abnormality following reversal of acidosis is thought to be due to marked shifts in myocardial cell polarity.^1-4,6

The actual mechanism of myocardial dysfunction in Shoshin beriberi has not been fully studied. In the first patient, the admission echocardiogram showed preserved left ventricular function and there was no obvious evidence of a wall-motion abnormality, suggesting myocardial infarction due to underlying atherosclerotic disease. It is well established that elevated troponins may be caused by several factors other than myocardial infarction.⁸

The transient ST-segment elevation that resolved with thiamine replacement therapy was most likely due to acute myocardial ischaemia secondary to the effect of severe acidosis and accompanying hypotension on the myocardial cell. Previous studies also suggest that the true myocardial contractility in beriberi is concealed by the systemic vasodilatation and may become apparent on recovery when the systemic resistance elevates back towards normal levels.⁹ Patients who abuse alcohol may have an underlying alcoholic cardiomyopathy that may become unmasked during recovery when the systemic resistance returns to normal.¹⁰

Laboratory tests that are considered diagnostic (red cell transketolase activity and blood thiamine concentration) are not readily available or have prolonged turn-around times and are not practical for use in the acute-care setting of medical emergencies that require immediate empiric management before laboratory results become available.^2,3

Key points

• Shoshin beriberi presents as severe metabolic acidosis with/without shock.

• Metabolic acidosis should be suspected in any patient with severe tachypnoea and relatively clear lung fields.

• The lactic acidosis in Shoshin beriberi is refractory to resuscitative measures and responds only to thiamine.

• Shoshin beriberi can mimic severe medical problems such as massive pulmonary embolism, extensive myocardial infarction or sepsis.

• Intravenous thiamine should be administered immediately to patients presenting with severe metabolic acidosis, while they are being resuscitated.

Administration of intravenous thiamine is the only definitive treatment for Shoshin beriberi and the response to thiamine is generally dramatic, with improvement in haemodynamic parameters and resolution of lactic acidosis within minutes to hours if the diagnosis is not delayed. Rapid response to thiamine confirms the diagnosis in the acute setting,² as was the case in the second patient who received empirical thiamine at admission and recovered almost immediately (Table 1).

Conclusion

Shoshin beriberi is an easily treatable condition that may masquerade as other life-threatening medical conditions. Treatment with intravenous thiamine must be empirical to avoid delay in diagnosis and potential fatalities. It should be administered to any patient who presents with circulatory shock and unexplained severe metabolic acidosis.