When seizure control comes at a cost: valproic acid-induced necrotising pancreatitis in a patient receiving antiepileptic polypharmacy

Summary

We present a case of a man in his late 20s with psychiatric and metabolic comorbidities who developed acute necrotising pancreatitis in the setting of antiepileptic polypharmacy, including valproic acid. The patient presented with nausea, vomiting and progressive lethargy without the classic abdominal pain of pancreatitis, which contributed to diagnostic delay. He rapidly deteriorated into multiorgan failure and septic shock. Laboratory evaluation demonstrated markedly elevated lipase at 16 310 U/L, leucocytosis, metabolic acidosis, acute kidney injury and lactic acidosis. Serum triglycerides, calcium and glucose were within normal limits, excluding metabolic causes. Imaging confirmed necrotising pancreatitis with a large peripancreatic fluid collection. Following exclusion of alternative aetiologies, valproic acid was identified as the most likely cause. The patient improved with supportive management. This case underscores that valproic acid can cause life-threatening pancreatitis even at therapeutic levels and highlights the importance of reviewing medications, particularly in cases presenting atypically without abdominal pain.

Background

Necrotising pancreatitis is a severe form of acute pancreatitis associated with significant morbidity and mortality. Overall mortality is approximately 15%, rising to 30–39% in cases of infected necrosis.¹ Gallstones, alcohol use and hypertriglyceridaemia remain the most common causes; however, drug-induced pancreatitis is increasingly recognised due to expanding pharmacotherapy, though establishing causality remains challenging because of its idiosyncratic nature and reliance on exclusion of other aetiologies.²

Valproic acid is widely used in the treatment of epilepsy, bipolar disorder and other neuropsychiatric conditions due to its broad-spectrum efficacy and dual role as both an anticonvulsant and a mood stabiliser.³ This pharmacological versatility often makes it preferable to other antiepileptic agents in patients with comorbid psychiatric illness. Despite its clinical utility, valproic acid has been associated with rare but serious adverse effects, including hepatotoxicity, hyperammonaemia and acute pancreatitis.⁴ Notably, valproic acid-induced pancreatitis is classified among the strongest evidence class of drug–pancreatitis associations, yet it occurs in only approximately 1 in 40 000 patients on this medication.^{4 5}

We report a case of valproic acid-induced necrotising pancreatitis in a patient receiving antiepileptic polytherapy, highlighting diagnostic challenges and potential risks associated with combined therapy.

Case presentation

A man in his late 20s with a history of schizoaffective disorder, attention-deficit/hyperactivity disorder, autism spectrum disorder, obstructive sleep apnoea, pre-diabetes and hyperlipidaemia presented to the emergency department with a 1-day history of nausea, vomiting and progressive lethargy. Notably, he did not report abdominal pain, the classic presenting symptom of pancreatitis. Due to an altered mental status, history was obtained from his mother. There was no prior history of alcohol use, gallstones, known liver disease, intravenous drug use or recent infections. His home medications included bupropion, propranolol, olanzapine, naltrexone, lamotrigine, topiramate and valproic acid. It was noted that his valproic acid dose had been increased from 2 500 to 3 500 mg daily approximately 6 months before presentation.

On physical examination, the patient was obtunded and tachycardic. There were no cutaneous stigmata of chronic liver disease (no jaundice, spider angiomas, palmar erythema or leuconychia), no peripheral oedema and no skin lesions or track marks to suggest intravenous drug use or infectious source. Cullen’s sign and Grey Turner’s sign were both absent, consistent with the absence of haemorrhagic pancreatitis at the time of initial assessment.

Shortly after presentation, the patient developed seizure-like activity followed by rapid clinical deterioration with worsening respiratory failure requiring emergent intubation. He subsequently experienced a pulseless electrical activity arrest, with return of spontaneous circulation after one cycle of cardiopulmonary resuscitation. He was admitted to the intensive care unit in shock with multiorgan failure requiring vasopressor support.

Investigations

Initial laboratory evaluation revealed significant abnormalities, including leucocytosis with a white blood cell count of 17.5×10⁹/L, metabolic acidosis with a bicarbonate level of 13 mmol/L and an anion gap of 19 and acute kidney injury with a creatinine level of 2.2 mg/dL. Serum lactate was markedly elevated at 9.2 mmol/L, and lipase was significantly elevated at 16 310 U/L. The serum valproic acid level was 119 µg/mL, within the therapeutic range of 50–125 µg/mL. Importantly, serum triglycerides were within normal limits (162 mg/dL), serum calcium was normal (8.9 mg/dL) and fasting glucose was 104 mg/dL, excluding hypertriglyceridaemia, hypercalcaemia and hyperglycaemia as contributing aetiologies. Electrolytes, including sodium (138 mEq/L), potassium (4.1 mEq/L) and magnesium (1.9 mg/dL), were within acceptable ranges, making metabolic encephalopathy from electrolyte derangement unlikely.

The patient’s altered mental status was attributed to the combination of severe metabolic acidosis, markedly elevated lactate reflecting global tissue hypoperfusion, acute kidney injury and the effects of septic shock on cerebral perfusion, rather than primary neurological or hepatic encephalopathy. Liver enzymes were mildly elevated (aspartate aminotransferase (AST) 78 U/L, alanine aminotransferase (ALT) 52 U/L, alkaline phosphatase 94 U/L). Total bilirubin was 1.1 mg/dL, albumin 3.4 g/dL and prothrombin time-international normalized ratio (PT-INR) was 1.2, collectively indicating no evidence of acute liver failure despite the patient’s use of valproic acid and lamotrigine, both of which are known to have hepatotoxic potential. Immunoglobulin G4 (IgG4) serum levels were obtained and were within normal ranges.

Toxicology screening, including salicylate, acetaminophen, ethanol levels and urine drug screen, was unremarkable. Infectious workup, including blood cultures, respiratory cultures and viral serologies, was negative, with no identifiable infectious focus despite the clinical picture of septic shock.

The Bedside Index of Severity in Acute Pancreatitis (BISAP) score is a validated five-criterion bedside tool developed for early, objective mortality risk stratification in acute pancreatitis and requires no imaging or specialised calculation.⁶ It was applied at the time of initial evaluation in this patient. Of the five criteria, blood urea nitrogen (BUN) >25 mg/dL, impaired mental status, systemic inflammatory response syndrome (SIRS), age >60 years and evidence of pleural effusion, three were positive: BUN >25 mg/dL, impaired mental status and SIRS criteria were all met. Age (patient in his late twenties, ie, under 60) and pleural effusion criteria were not met and therefore did not add to the score. This yielded a BISAP score of 3 out of 5, which in the original derivation and validation cohort corresponds to a marked stepwise increase in mortality (approximately 5–20%) compared with scores of 0–2^{6 7} and supported the decision for early intensive care unit (ICU)-level monitoring. The Balthazar CT Severity Index (CTSI) was 8 out of 10, indicating severe pancreatitis with a correspondingly high risk of morbidity and mortality.^{6 7}

While an initial ultrasound showed no pertinent findings, given clinical deterioration, a non-contrast CT was subsequently obtained, which showed findings consistent with acute pancreatitis with no evidence of cholecystitis or cholelithiasis, excluding biliary aetiology. After a period of haemodynamic stabilisation (approximately 48–72 hours after admission), a contrast-enhanced CT was performed to assess for necrosis. While early contrast-enhanced CT has been debated given theoretical concerns about vasoconstriction potentially worsening ischaemia,⁸ this was performed after initial resuscitation when haemodynamic stability was achieved, and the clinical benefit of characterising the extent of necrosis was judged to outweigh the risk. This revealed progression to necrotising pancreatitis with areas of pancreatic hypoenhancement in the head (~3 cm) and distal body/tail (~6 cm), and a large developing peripancreatic fluid collection measuring approximately 10.9×7.2×3.8 cm along the greater curvature of the stomach (figure 1).

Figure 1. Contrast-enhanced CT imaging showing necrotising pancreatitis and associated complications. (A–C) Axial views show areas of hypoenhancement within the pancreatic head (A, ~3 cm) and distal body/tail (B, ~6 cm), consistent with pancreatic necrosis. (C and D) A large developing peripancreatic fluid collection is seen along the greater curvature of the stomach, measuring around 7.2×3.8×10.9 cm, seen on axial (C) and coronal (D) views.

Differential diagnosis

The differential diagnosis for acute pancreatitis in this patient was approached systematically, beginning with the most common aetiologies. Gallstone pancreatitis was excluded as there was no evidence of gallstones or biliary dilation. Alcohol-induced pancreatitis was excluded by the absence of history and a negative ethanol level. Hypertriglyceridaemia-induced pancreatitis was excluded by a normal fasting triglyceride level. Hypercalcaemia was excluded by a normal serum calcium. Infectious pancreatitis was considered given the septic presentation; however, all cultures and infectious serologies were negative, and no infectious focus was identified.

Autoimmune pancreatitis was considered given the patient’s young age. However, type 1 (IgG4-related) autoimmune pancreatitis classically follows a subacute or chronic course, often presenting with painless obstructive jaundice, a diffusely enlarged “sausage-shaped” pancreas, or a focal mass mimicking malignancy rather than the fulminant, necrotising course with septic shock seen in this patient. This made autoimmune pancreatitis inherently unlikely on clinical and temporal grounds alone. Nonetheless, autoimmune pancreatitis was included as part of our broader aetiologic workup, and serum IgG4 was obtained as it is an established component of the diagnostic criteria for autoimmune pancreatitis.⁹ This impression was further supported, rather than solely established, by a normal serum IgG4 level and imaging findings that were not consistent with IgG4-related disease; a normal IgG4 does not, by itself, fully exclude the diagnosis, but taken together with the atypical acute presentation, it made autoimmune pancreatitis a much less likely explanation than drug-induced injury.

Given the patient’s complex medication regimen, drug-induced pancreatitis was strongly considered, with valproic acid identified as the most probable causative agent based on clinical context and literature support.

Treatment

The patient was managed in the ICU in accordance with evidence-based guidelines for severe acute pancreatitis. Early aggressive intravenous fluid resuscitation was initiated, which represents the cornerstone of acute pancreatitis management. Mechanical ventilation was required due to respiratory failure. Vasopressor support was commenced for septic shock. Broad-spectrum antibiotics (piperacillin-tazobactam) were initiated, given the clinical picture of septic shock and the difficulty in clinically excluding infected necrosis at the time of presentation. It is recognised that antibiotics are not routinely indicated in uncomplicated acute pancreatitis; however, in this case, the degree of haemodynamic compromise and multiorgan failure justified their empirical use until infectious causes were excluded and the patient stabilised.

All suspected offending medications, including valproic acid, lamotrigine and topiramate, were discontinued. Given the presence of necrotising pancreatitis with a large fluid collection, interventional radiology and surgical teams were consulted. The patient was managed conservatively as he was not a candidate for immediate intervention in the context of ongoing haemodynamic instability. Oral feeding was withheld initially; enteral nutrition via nasogastric tube was initiated once haemodynamic stability was achieved.

Outcome and follow-up

During the hospital course, the patient demonstrated gradual improvement in renal, hepatic and respiratory function. Vasopressor support was successfully weaned, and metabolic derangements resolved. Serial imaging demonstrated stabilisation of the pancreatic necrosis and associated fluid collection. After 3 weeks of prolonged hospital stay, the patient was discharged in stable condition with plans for outpatient follow-up with gastroenterology and psychiatry to review alternative medication management.

Discussion

This case illustrates several important clinical lessons relevant to the diagnosis and management of drug-induced acute pancreatitis, particularly in atypically presenting patients.

Atypical presentation: acute pancreatitis without abdominal pain

A fundamental learning point from this case is that acute pancreatitis can present without the classic epigastric pain that clinicians rely on for early diagnosis. Our patient presented with nausea, vomiting and progressive lethargy, with the diagnosis only established biochemically and radiologically. Clinicians should maintain a high index of suspicion for acute pancreatitis in patients presenting with septic shock of unclear aetiology, particularly when no infectious focus is identified, and more obvious diagnoses have been excluded. This awareness is especially pertinent in patients with an altered conscious level in whom the history is limited.

Systematic exclusion of common aetiologies

In acute pancreatitis, it is standard and essential practice to exclude common aetiologies before attributing causality to rare causes such as drug toxicity. In this case, gallstone pancreatitis was excluded, alcohol use was absent, serum triglycerides were normal (162 mg/dL), serum calcium was normal (8.9 mg/dL), and infectious causes were excluded by comprehensive culture and serology data. Metabolic causes of encephalopathy were also systematically excluded: sodium, potassium, glucose and magnesium were all within acceptable ranges. Only after this systematic exclusion was drug-induced pancreatitis elevated to the primary diagnosis.

Importance of medication review in acute pancreatitis

A critical lesson from this case is that thorough medication review is essential in all patients presenting with acute pancreatitis, particularly when common aetiologies have been excluded. Drug-induced pancreatitis accounts for 0.1–2% of cases overall, but this figure may be underestimated due to underreporting and diagnostic anchoring on more common causes.⁵ Valproic acid is a commonly prescribed medication used for the treatment of generalised and focal epilepsy, trigeminal neuralgia, migraine, bipolar disorder and sometimes schizoaffective disorder, as in our patient.^{3 10} While most side effects are mild, valproic acid is a very rare cause of necrotising pancreatitis, categorised in the strongest class of medications associated with pancreatitis, underscoring its rarity and the diagnostic challenge it poses.^{3 5} Critically, in most reported cases, consistent with our patient (level 119 µg/mL, therapeutic range 50–125 µg/mL), valproic acid-induced pancreatitis occurs at therapeutic serum levels, meaning routine drug level monitoring does not predict or prevent this complication.¹¹

Antiepileptic polypharmacy and pancreatic toxicity risk

A review of 16 cases of valproic acid-induced pancreatitis found that 45% of adults were receiving concurrent antiepileptic therapy.³ While topiramate and lamotrigine have each been implicated in isolated case reports,^12–14 the concurrent use of all three agents in our patient raises the possibility that polytherapy may potentiate pancreatic toxicity, potentially through additive or synergistic mechanisms. This underscores the importance of careful medication reconciliation, particularly when adding or dose-escalating agents with known toxicity profiles. In our patient, the dose increase of valproic acid from 2,500 mg to 3,500 mg daily approximately 6 months prior to presentation may have been a contributing precipitant, even though the resulting level remained within the therapeutic range.

Mechanism of valproic acid-induced pancreatitis

The mechanism is incompletely understood, but leading theories implicate direct toxicity mediated by free radicals, resulting from depletion of antioxidant enzymes including superoxide dismutase, catalase and glutathione peroxidase, leading to pancreatic cellular injury and inflammation.³ Mitochondrial dysfunction has also been proposed as a contributing mechanism. Notably, the idiosyncratic and non-dose-dependent nature of this adverse effect means that monitoring serum valproate levels does not provide reliable protection. The time to onset is highly variable, ranging from days to years after initiation, requiring continued vigilance even in long-term therapy.

Factors associated with progression to necrotising pancreatitis

Necrotising pancreatitis results from pancreatic ischaemia and sustained inflammatory injury following enzymatic autodigestion. In our patient, several factors likely contributed to necrosis development: early circulatory shock with severe lactic acidosis (9.2 mmol/L), acute kidney injury, multiorgan dysfunction and elevated body mass index (BMI). The BISAP score of 3 correctly predicted a high-mortality risk trajectory, and the CTSI of 8 confirmed severe disease. These validated scoring tools should be applied early in the clinical course to stratify severity and guide decisions regarding ICU-level care and resource allocation.^{6 7}

Hepatotoxicity considerations with valproic acid and lamotrigine

Valproic acid and lamotrigine are both known to carry hepatotoxic potential. In this patient, liver enzymes were mildly elevated (AST 78 U/L, ALT 52 U/L), but bilirubin, albumin and PT-INR were within acceptable ranges, and there were no clinical features of acute liver failure. This is an important finding to document, as it supports the hepatic safety profile in this case and argues against hepatic encephalopathy as the cause of altered mental status.

Management principles of severe drug-induced necrotising pancreatitis

Evidence-based management of severe acute pancreatitis centres on early aggressive intravenous fluid resuscitation, analgesia and nutritional support, preferably enteral. Antibiotics are not routinely indicated in sterile pancreatitis; in this case, empirical antibiotics were initiated in the context of septic shock with no identifiable focus and were appropriately reviewed and de-escalated once cultures were negative.⁸ Regarding CT imaging, current guidelines recommend an initial abdominal ultrasound as the first-line imaging modality, followed by contrast-enhanced CT after 48–72 hours of symptom onset when there is diagnostic uncertainty or clinical deterioration. In our patient, after an abdominal ultrasound was inconclusive, the non-contrast CT was performed initially, and the contrast-enhanced CT was deferred until after haemodynamic stabilisation, consistent with this approach. The theoretical concern that contrast enhancement may worsen early pancreatic ischaemia is acknowledged; however, once haemodynamic stability was achieved, the clinical information gained from characterising the necrosis was essential for multidisciplinary surgical and interventional radiology decision-making.⁸

This case report describes a patient with multiple psychiatric and medical comorbidities on a complex regimen, including valproic acid, topiramate and lamotrigine, who developed necrotising pancreatitis. It highlights the importance of recognising valproic acid as a rare but serious cause of necrotising pancreatitis and demonstrates the risks of antiepileptic polypharmacy. This case demonstrates how crucial it is to carefully review medication lists when adding or increasing dosages of medications associated with toxicity risk.

Learning points.

• Acute pancreatitis should be considered in cases of septic shock where no infectious focus is found, particularly when the classic presentation of abdominal pain is absent. Biochemical and radiological evaluation is essential in such cases.

• Valproic acid-associated pancreatitis is idiosyncratic, has the potential to occur at therapeutic serum levels and can develop from days to years after treatment initiation.

• Thorough medication review is essential in all acute pancreatitis cases, especially when common aetiologies (gallstones, alcohol, hypertriglyceridaemia, hypercalcaemia) have been excluded.

• Proposed mechanisms include mitochondrial dysfunction, oxidative stress and free radical-mediated pancreatic injury, though the exact pathophysiology remains unclear.

• Concomitant use of other medications (eg, enzyme-inducing antiepileptics) may alter metabolism and contribute to toxicity risk.