Background

Acute pancreatitis is a condition in which digestive proenzymes synthesised in the pancreas are activated before they are secreted, resulting in autodigestion of the organ itself. It often presents with severe epigastric pain, nausea and vomiting. Acute pancreatitis is responsible for 300 000 emergency department visits a year and is one of the leading gastrointestinal causes of hospital admissions in the USA.1 It is most often caused by gallstones (45%), alcohol use (20%) or hypertriglyceridaemia (10%).2 3 However, an estimated 0.1%4 to 2%5 of acute pancreatitis cases are drug-induced. While there is substantial evidence that the tetracycline class of antibiotics may be associated with drug-induced pancreatitis, there are only a handful of case reports specifically implicating doxycycline.6–16 These case reports implicating doxycycline as a culprit for drug-induced pancreatitis involved younger adults, with only two representing adults older than 65 and no reports involving an octagenarian.7 16 In this case report, we describe a case of a woman in her 80s who presented with severe abdominal pain, nausea and vomiting, found to have acute pancreatitis following doxycycline treatment.

Case presentation

A woman in her 80s presented to the emergency department for acute onset midepigastric and left upper quadrant pain with associated severe nausea and vomiting. Seven days prior to presentation, the patient began taking doxycycline for acute cystitis. On her sixth day of taking doxycycline, she woke up with sharp midepigastric and left upper quadrant pain that she described as 8/10 in severity. She was unable to tolerate oral intake and had intractable non-bilious emesis which led her to present to the emergency department. Her medical history was significant for cholelithiasis status post cholecystectomy, chronic kidney disease stage III, persistent atrial fibrillation status post ablation, breast cancer status post left mastectomy and radiation, hypertension, rheumatoid arthritis, major depressive disorder and fibromyalgia. She had no history of diabetes or pancreatitis. She denied alcohol use and quit smoking 38 years prior. Doxycycline was the only new medication she had started, and she did not have any recent changes in medication doses (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Timeline of when current medications were started compared with the timing of the acute pancreatitis episode described. Figure was created by MRG.

Her outpatient medications included:

• Amlodipine.

• Aripiprazole.

• B complex vitamins.

• Cholecalciferol.

• Dextroamphetamine.

• Docusate sodium.

• Apixaban.

• Gabapentin.

• Ipratropium bromide.

• Lactobacillus acidophilus.

• Losartan.

• Metoprolol succinate.

• Meclizine.

• Omeprazole.

• Oxycodone-acetaminophen.

• Trazodone.

Investigations

In the emergency department, her vital signs were: temperature of 37°C, blood pressure of 144/59, pulse of 75 beats per minute and respiratory rate of 16. White blood cell count was elevated at 21.4 ×10⁹/L . Her lipase was elevated at 4165 units/L. Her triglyceride level was normal at 45 mg/dL. A CT of the abdomen and pelvis without contrast was performed to assess for signs of acute pancreatitis, assess disease severity and assess for any alternative causes of severe abdominal pain, including pyelonephritis in the setting of her recent acute cystitis. The CT showed an oedematous pancreas with surrounding pancreatic stranding (figure 2). Of note, a CT of the abdomen and pelvis collected 2 months prior showed a normal-appearing pancreas. Troponin was 0.03 and ECG showed normal sinus rhythm with no concerning changes. Chest X-ray was unremarkable. Physical examination was significant for epigastric tenderness and a negative Cullen’s sign and Grey-Turner sign. Given her clinical presentation, leucocytosis, elevated lipase and characteristic findings in CT of the abdomen and pelvis, she was diagnosed with acute pancreatitis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

CT of the abdomen and pelvis without contrast obtained on day 1 of hospitalisation showing an oedematous pancreas with surrounding pancreatic stranding concerning for acute pancreatitis.

Differential diagnosis

Determining the aetiology of acute pancreatitis is a clinical diagnosis based on history. The most common causes of acute pancreatitis are alcohol use, gallstones and hypertriglyceridaemia.2 3 Given her lack of alcohol use, prior cholecystectomy and no evidence of gallstones or common bile duct dilation on imaging, and her normal triglyceride level, the most common aetiologies of acute pancreatitis were ruled out. Other less common causes were considered, including hypercalcaemia, recent endoscopic retrograde cholangiopancreatography, malignancy, congenital anomalies of the pancreas, infection and trauma. Her calcium was normal at 8.9 mg/dL. She had no history of ERCP. There was no evidence of malignancy on cross-sectional imaging or recent routine cancer screening at the time of her pancreatitis, nor was there evidence of congenital anomalies of the pancreas on imaging. There was no evidence of infection on urinalysis or respiratory pathogen panel, and no recent trauma. Idiopathic acute pancreatitis was also considered. However, given her polypharmacy and recent initiation of a new drug, and the importance of discontinuing the offending drug to prevent progression of disease in drug-induced acute pancreatitis, it was critical to identify any potential drug-related aetiologies. From her medication list, opioids, acetaminophen, omeprazole and doxycycline have been reported as causes of drug-induced acute pancreatitis.17 The patient began taking oxycodone-acetaminophen and omeprazole years prior to establishing care at our institution, over 3 years before her first acute pancreatitis episode (figure 1). She had no recent changes in the dosages of these medications or any other medications. In contrast, her symptoms began shortly after taking doxycycline. Therefore, the aetiology of her acute pancreatitis was ultimately determined to be doxycycline.

Treatment

Doxycycline was discontinued. She received intravenous fluid resuscitation, pain management and nausea management. Enteral nutrition was introduced as early as she could tolerate it. Her lipase decreased from 4165 units/L to 1144 units/L 36 hours after admission. Her leucocytosis improved from 21.4 ×10⁹/L on admission to within normal limits at 8.5 ×10⁹/L by discharge. Intravenous fluids were stopped on day 3 of hospitalisation as her oral intake improved. Pain and nausea management requirements decreased throughout her hospital stay. She was discharged on day 9 of hospitalisation with home health support.

Outcome and follow-up

12 days after discharge, she returned to the emergency room with persistent epigastric abdominal pain, nausea and vomiting that had not resolved since she was discharged. A differential diagnosis of chronic pancreatitis or complications including pancreatic pseudocyst, pancreatic abscess or pancreatic necrosis were considered. Her vital signs were stable and she did not have fever, chills or any other systemic symptoms. She had a lipase of 175 units/L (reference range 8–78 units/L) without lactic acidosis, leucocytosis or hypocalcaemia. This was a decrease from her most recent lipase of 1144 units/L a week before. Lipase is known to peak at 24 hours after onset and remain elevated for 2 weeks after the episode; therefore, this persistent elevation was not unexpected.18 In the setting of a downtrending lipase and the absence of clinical or laboratory findings concerning for complications, the team believed this to be a continuation of her initial episode of pancreatitis and decided to consider repeat imaging if she did not improve to her baseline within a few days. She was given intravenous fluids, ondansetron for nausea and intravenous hydromorphone for pain. She stayed in the hospital for one day and then was transferred to the Hospital at Home programme where she continued to receive fluids, receive multimodal pain control and advance her diet. On discharge from the programme, she was tolerating a regular diet without abdominal pain and thus, did not require follow-up imaging. She followed-up with her primary care physician with regular interval check-ins and resolution of symptoms. She has not taken doxycycline since, had no recurrent episodes of pancreatitis. Her clinical course is summarised in figure 3.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Timeline of the clinical course of illness. Figure was created by MRG. ED, emergency department; IV, intravenous; PCP, primary care physician; PO, oral.