Persistent pleural effusions in primary systemic amyloidosis: etiology and prognosis.

Background: Restrictive cardiomyopathy frequently complicates primary systemic amyloidosis (AL), yet only a small number of these patients develop large pleural effusions refractory to diuretic therapy and thoracentesis. We hypothesized that disruption of pleural function by amyloid deposits underlies persistent pleural effusions (PPEs) in patients with AL disease.
Methods: We performed a retrospective study of AL patients with and without PPEs who had been referred to Boston University between 1994 and 2001. The presence of PPEs was defined by a failure to resolve the condition with thoracentesis and aggressive diuresis. AL cardiomyopathy patients without pleural effusions constituted the control (cardiac) group. Indexes of plasma cell dyscrasia, nephrotic syndrome, thyroid function, and echocardiographic measures of left and right ventricle performance were compared between groups. When available, closed needle biopsies and autopsy specimens of parietal pleura were examined for amyloid deposits.
Results: Among 636 patients with AL, 35 PPE patients underwent a median of three thoracenteses each. No statistical differences were found between the PPE and cardiac groups in echocardiographic measures of septal thickness, left ventricular systolic function, or diastolic compliance. Right ventricular (RV) hypokinesis occurred more often in PPE patients; however, nearly half of this group had normal RV systolic function. Renal function, plasma protein levels, and thyroid function were the same between groups. Nephrotic range proteinuria (ie, > 3 g/d) was more prevalent in the cardiac group than in the PPE group (44% vs 26%, respectively; p = 0.057). All pleural biopsies in the PPE group (six biopsies) revealed amyloid deposits. Autopsy samples of parietal pleura were negative for disease in two cardiac patients. Eighteen patients had chest tubes placed, and 11 underwent pleurodesis. PPE signaled limited survival among patients who were ineligible for treatment. Untreated PPE patients lived a median 1.8 months vs 6 months for untreated cardiac patients (p = 0.031). Survival after intensive chemotherapy and autologous stem cell transplantation was comparable in the PPE and cardiac groups (21.8 vs 15.6 months, respectively; p = 0.405).
Conclusion: In AL patients with cardiac amyloid, neither echocardiographic measures of ventricular function nor the degree of nephrosis distinguished those patients with PPEs. We conclude that pleural amyloid infiltration plays a central role in the creation and persistence of pleural effusions among patients with AL.