Journal Logo

Diagnosis Deconstructed

Diagnosis Deconstructed: Years in the Making

Morchi, Ravi MD

Author Information
doi: 10.1097/01.EEM.0000418677.72342.9a

    Signout rounds. The team describes a 58-year-old man with back pain radiating down his thighs. No neurologic symptoms. He has a history of untreated prostatic malignancy. And he is thrombocytopenic to 30 for the past month.

    A CT abdomen-pelvis shows a pancreatic pseudocyst. But no blastic or lytic lesions to his spine. No pathologic fracture. Just degenerative joint disease in the lumbar area. He is admitted to the medical hospitalist for back pain, and remains in our ED awaiting an inpatient bed.

    Back pain is a symptom, not a diagnosis. What is the primary problem? Is the pseudocyst the reason for his back pain? Why is he thrombocytopenic? Right now, he is a collection of data points without synthesis. A disposition, yes. But no synthesis.

    Imaging. I scroll through his CT scan as we round. He has calcifications strung along the body of his pancreas. Evidence of chronic pancreatitis. He is probably a drinker. The pseudocyst is very small, and actually has air-fluid levels in it.

    A collection of pancreatic juice contained by retroperitoneal tissue, a pseudocyst represents extravasation from pancreatic cells or the duct itself due to insufficiency of the drainage system after prior episodes of inflammation and fibrosis. Because it cannot drain into the duodenal lumen, the pancreatic fluid tends to collect in a potential space within the retroperitoneum and sit there, sometimes for months. Lacking a true cellular lining, it cannot claim to be a true cyst. But an air-fluid level within it, that is unusual. We need to look further into the past. And he needs to stop drinking.

    Anaerobic infection of the contents? Or did someone try to drain the pseudocyst into his stomach or a loop of small bowel, with air representing free communication to an enteric lumen?

    Looking back at old CTs, the pseudocyst used to be much larger and without an air-fluid level. More recent scans show it has shrunk in size and begun to accumulate air. It has established communication with an enteric lumen.

    Labs. He is thrombocytopenic to 30, and is anemic to 30%. Without another reason, the combination of these two findings means reflexive consideration of one diagnosis. Regardless of symptoms or signs. Regardless of logical construct and our desire for everything to make sense. This diagnosis will otherwise creep up on us, wrapped in the packaging of what appears to be another clinical condition. Distracted by a vague headache or some vaginal spotting, we will miss it.

    Thrombotic thrombocytopenic purpura means that platelets are strung together by something other than -fibrin. An abnormally large Von Willebrand's factor. Clumped in this manner and lining the arterioles, masses of platelets tend to jut into the vascular lumen and slice red cells shooting by at systolic pressures.

    Does he have schistocytes on his smear? Even without fever, confusion, or new renal insufficiency, the presence of unexplained anemia and thrombocytopenia must have us first consider microangiopathic hemolysis. And without DIC or another process to cause this type of hemolysis, thrombotic thrombocytopenic purpura should come to mind. If only for a moment. Not because it logically fits into the history and physical exam, but rather because the treatment for this condition remains very different from any other emergent diagnosis we see.

    Scrolling down, he has no schistocytes. Without them, it is unlikely that TTP is behind his blood count derangements. With an INR of 1.3 and macrocytosis, it is more likely he has bone marrow suppression and possibly portal hypertension with an enlarged spleen sequestering platelets. He really needs to stop drinking.

    As rounds close, I suppose it is time to see where these pixels and numbers actually originate. What has generated all these data points?

    The patient. He is tachycardic but otherwise in no distress. His abdomen is mildly distended but nontender. He prefers to lie still because any movement he attempts, turning to his side or flexing at the hip, reproduces his back pain.

    So this is not solely his pseudocyst. With limitations to even a few degrees of motion, it seems more mechanical. Originating from the vertebral column.

    The pain is in the lumbar region. No radiation down either leg, front or back. No sensory disturbances. No lumbar or sacral distribution radiculopathy. His lower extremity strength is preserved, and his rectal tone is normal. He is not retaining urine. Spinal stenosis or degenerating discs?

    Normally I would be fine with this differential. But not for this man. This man has another issue, one particular to clotting. Hemostasis requires platelets and fibrin. And we know he has a problem with both.

    Looking over his skin, he is not littered with petechiae or purpura. No ecchymoses. His IV site is not oozing. It looks like he tolerates a platelet count of 30 and an INR of 1.3 quite well. But I have to say, even without evidence of red cells leaking into soft tissue and even with degenerative joint disease as a fallback diagnosis for his low back pain, I cannot be certain that red cells and plasma are not accumulating in the spinal epidural space.

    Sagittal reconstructions of the spine taken from his abdominal CT show degenerative joint disease in the lumbar region. With loss of disc space, osteophyte formation, and sclerosis of bone edges, we also note spondilolisthesis of L5 on S1. These are all chronic in nature, and implicate disc herniation or spinal stenosis as the cause of his musculoskeletal low back pain.

    On closer look, I do not see any bulge into the canal. Instead, what he seems to have are small flecks of lucency within his lumbar spine. This is not mentioned on the formal radiologist reading. Is this air? Air in the intervertebral space can be seen in the vacuum disc phenomenon of severe degenerative joint disease. But decompressed into the spinal canal? This would be spontaneous pneumorrhachis from a degenerated disc. A confluence of coincidental findings all consistent with degenerative joint disease? Or something more serious?

    I have to consider an infectious cause producing epidural air: a secondarily infected hematoma or a primary abscess. I look back at his blood work.

    His total white cells are normal quantitatively; qualitatively they are not. They are unhappy. They are working hard. Eating and digesting something bad. Is it in response to his prostatic malignancy or the infection that may be brewing in necrotic pancreatic tissue or his spinal canal? I do not know for sure. But his neutrophils, they know. They contain toxic granulations, vacuolation, and Döhle bodies.

    Synthesis. I notify the hospitalist that epidural hematoma and abscess are on the differential. High on the differential. He could have seeded a degenerated lumbar spine from hematogenous flow of bacteria whose origins lay in necrosed, infected pancreatic contents in communication with an enteric lumen. He receives parenteral antibiotics. The hospitalist agrees to arrange for an MRI.

    As an inpatient, his temperature spikes to 104°F rectal. His blood cultures grow 4/4 bottles of Klebsiella. And his MRI later reveals an epidural abscess extending from L3 to S1. With alcohol by his side, the origins of his bacteremia were years in the making. After laminectomy and decompression, he may need sampling and drainage of the pancreatic contents or even a necrosectomy.

    Click and Connect!Access the links in EMN by reading this issue on our website or in our iPad app, both available

    © 2012 Lippincott Williams & Wilkins, Inc.