Medical Mystery II

An incarcerated man in his 30s comes to us from a prison hospital for work-up of intractable nausea, vomiting, and abdominal pain. He's had gastrointestinal symptoms from childhood, but they worsened in the last 4 years. Since then, he's had periodic episodes of severe nausea, vomiting, and abdominal pain requiring multiple hospitalizations. He's been diagnosed with gastroparesis, peptic and duodenal ulcers, and Mallory-Weiss tears. His symptoms have been so severe in the past that he's required two jejunostomy tubes for feeding. At an outside hospital, he had hematemesis and melena, and when he was hemodynamically stable, he was transferred to us because he was unable to eat. He presented to us with excruciating mid-epigastric and RUQ pain radiating to the back, worse with vomiting, minimally relieved by opiates, no identifiable triggers (no relation to meals). His past medical history is otherwise unremarkable. He takes some opiates for pain at baseline and a PPI but has no other medications. His family history is significant for gastroparesis in a grandparent secondary to diabetes, but otherwise negative. He has been incarcerated for several years now at a maximum security prison. He does not have a history of alcohol, drinking, or IV drug use. He is afebrile, BP in the 100s/60s but otherwise vital signs stable. To me, on presentation he looked almost like a surgical acute abdomen; he was rigid, legs drawn up, visibly distressed, exquisitely tender to palpation. He actually spent a night in the ICU for pain control.

Now, the obvious differential would be an acute abdomen (peritonitis, appy, perforated ulcer) vs. peptic ulcer, pancreatitis, cholecystitis, hepatitis. But his labs came back stone cold normal. No leukocytosis, normal chemistries, amylase, lipase, LFTs. They were normal at the outside hospital too. A CT of the abdomen and pelvis was completely unremarkable. A RUQ ultrasound showed no gallstones. A CXR was normal. A urinalysis was benign. EKG showed sinus bradycardia with a first degree heart block.

A medical mystery! I was excited (even though I'm not usually crazy about GI). I started with the obvious, calling in a GI consult. They did an EGD (upper endoscopy) which showed a normal esophagus, normal GE junction, mild gastritis, and mild duodenitis, no ulcers. This could not explain his symptoms. He was in unbelievable pain, requiring sky-high doses of narcotics each night.

My attending then did something quite smart and laudable. This patient was taking up a lot of resources. Over the last four years, he was hospitalized many times to no avail. Finally, he had made it to a tertiary care center. We had an obligation to rule out as many esoteric things as we could and hopefully make a diagnosis. It would not be enough simply to control him symptomatically and send him back to prison. So the attending asked me to compile the most thorough, comprehensive list of causes of abdominal pain, nausea, and vomiting that I could imagine. My assessment and plan for that progress note was ridiculous; it was 4 pages long. It pretty much included everything but ectopic pregnancy and salpingitis.

There's an interesting socio-economics question here. An inmate, this man's health care was paid for by taxpayer dollars. Is it fair to us to be paying for an extensive workup of this patient's symptoms? He's getting better health care than law-abiding citizens. Is that fair? To me, the answer is simple. As part of his medical team, he is my patient and I am his advocate. After many hospitalizations at other places without an answer, he deserved at least a decent attempt to decipher his problem.

The patient was HIV negative, RPR non-reactive, tissue transglutaminase and gliadin antibodies negative. His sedimentation rate was low, ANA negative, C4 normal. A work-up for acute intermittent porphyria was negative. A nuclear medicine gastric emptying study showed mildly delayed gastric emptying for liquids but not solids, but this was done on high dose opiates. A blood lead level was negative. A CT head showed no acute intracranial process.

Of course, the question of secondary gain arose. Pain and nausea are highly subjective symptoms (the patient did vomit a few times in hospital though). While he was in the hospital, he got his own room, had a flat screen TV, and enjoyed better food (he was soon able to take a soft gastroparietic diet) than at the prison. But the team and I spent a lot of time with him. I got to know him well, and my assessment was that secondary gain was highly unlikely.

The next day, I went to a medical student lecture on adrenal insufficiency, which can often present with nausea, vomiting, abdominal pain. The patient did not have electrolyte abnormalities but was receiving IV fluids. He was not overtly hypotensive, but he ran low pressures normally. A 4:45AM cortisol was 1 (not an ideal time for measurement, but low, especially if he was acutely in pain). An 8:50AM cortisol was 4. A high dose cosyntropin stim test went from 1 to 9 to 13. His ACTH was low normal at 9 and 15. Interestingly, a brain MR was done showing a possible old hemorrhage into the pituitary; no masses were identified. Dexamethasone made a stunning improvement in the patient's symptoms (moreso than expected in a normal person).

GI consult service could not identify a cause of this patient's symptoms and felt adrenal insufficiency was worth pursuing. Otherwise, they suggested a garbage diagnosis: nonulcer dyspepsia with visceral hypersensitivity. Unfortunately, when we called endocrine consult, they were not impressed for whatever reason.

At the time of this post, a final diagnosis has not been made. We're hard pressed to convince endocrine. However, in my mind, adrenal insufficiency could easily be the diagnosis. One thing to remember is that as common diseases are ruled out, the uncommon diseases increase in likelihood. At the beginning, pancreatitis, cholecystitis, hepatitis, ulcer disease, and gastritis probably added to 80% likelihood. But once we convinced ourselves those weren't right, oddballs like Familial Mediterranean Fever, abdominal migraine, and indeed, adrenal insufficiency have to increase in likelihood. I think the consult services lose sight of that since they worry only about their domain of diseases. But after labs and imaging suggest against GI causes, I think other organ systems causing GI symptoms should be taken more seriously.