Medical Mystery

I admitted my first patient on cardiology nearly four weeks ago, a transfer from an outside hospital. It took us over three weeks to make a diagnosis, a fascinating story which unfolded as each path we went down lead to a dead end. Hindsight is 20/20, but I think I should have gotten the diagnosis earlier. This post is technical and I'm not sure I have everything correct, but here are my thoughts.

A 63 year old woman with a history significant for a psoas abscess two years ago and a questionable exposure to TB presented to an outside hospital with two years of increasing anasarca - swelling of her entire body. This started since her hospitalization for the psoas abscess and since then, she's gradually retained fluid until her legs are so swollen she has severe pain and cannot walk anymore. The exam showed pitting asymmetric 4+ edema in both legs, a holosystolic blowing murmur best heard at the left lower sternal border, and a blood pressure of 100/40.

At the outside hospital, she was worked up for renal, hepatic, gastrointestinal, thyroid, and autoimmune causes for swelling, which were all negative. An echocardiogram showed preserved systolic function but right heart enlargement and severe tricuspid regurgitation. She was then transferred to our hospital for work-up for pulmonary hypertension.

Our repeat echo confirmed elevated pulmonary pressures and we decided to focus on that differential. I read up on the WHO classification and began to explore each possibility. The echo showed no left heart failure. CT and pulmonary function tests suggested it was not a primary lung problem. To rule out chronic thromboembolism, we did lower extremity dopplers (negative except for a small AV fistula) and a V/Q scan (negative). There were no systemic symptoms suggesting collagen vascular disease, sarcoidosis, or rheumatologic illness. She denied drugs, was HIV negative, and had no other medical conditions. At this point, I started feeling like this could be idiopathic pulmonary hypertension.

But one attending suggested looking at the heart. He felt this could be constrictive pericarditis, especially given a questionable history of TB. A cardiac MR was negative for pericardial disease so we decided to do an invasive right and left heart catheterization. This showed increased pulmonary pressures, physiology of a restrictive cardiomyopathy, and high output heart failure.

For the restrictive cardiomyopathy, we took several heart biopsies. I was convinced this was a disease like sarcoid or hemochromatosis or amyloid, and that this woman would need a heart transplant. But then the biopsy came back completely negative: no Congo Red staining, no iron deposition, just nonspecific fibrosis, "cardiomyopathy NOS" (not otherwise specified).

We were stuck, still no good answer. But word of this patient spread through the department and some of the most expert clinicians at UCSF came by to see her. Eventually, they felt that this was not a cardiomyopathy but simply "severe tricuspid regurgitation causing a reverse Bernheim effect." Bernheim was a pathologist who described in 1910 a theory that left ventricular hypertrophy could impinge on right ventricular filling. This turned out not to be the case. But the reverse Bernheim effect describes a scenario where right ventricular enlargement impinges on left ventricular filling. Indeed, this patient had a dilated RV and left diastolic dysfunction.

The proposed treatment was tricuspid valve annuloplasty, a highly risky operation. We consulted the cardiothoracic surgeons and they consented the patient for the procedure. Around this time, a student nurse who had just come onto the service told us he heard an abdominal bruit. We brushed it off at first, attributing it to hearing the tricuspid regurgitation in the abdomen of this very small woman. But this student nurse actually made the most critical finding.

An angiogram revealed this was a massive arterio-venous fistula probably originating from the incision and drainage of the psoas abscess a few years ago. It was so large that it shunted a lot of blood past capillary beds. As a result, there was low systemic resistance and a high output heart failure. The increased venous return led to right ventricular dilatation and tricuspid regurgitation. As blood was shunted away from the kidneys through the A-V fistula, the body began retaining fluid and she became overloaded and edematous.

Over the last three weeks, the proposed therapy was lung transplant (idiopathic pulmonary hypertension), then heart transplant (cardiomyopathy), then tricuspid valve annuloplasty, and now finally vascular surgery which may resolve all the abnormal physiology. The clues were in the history and physical: the edema began after this psoas abscess (initially dismissed as irrelevant), she had an unusually large pulse pressure, and the student nurse heard a bruit over her femoral artery. It was a fascinating case, but one that could have been solved earlier if we kept re-evaluating what was going on as new information came.