The present study summarizes our ten-year (1985-1995) experience with endomyocardial biopsy (EMB) in patients with idiopathic congestive heart failure (CHF), with specific reference to frequency of myocarditis, treatment policy, relative benefits, and follow-up. Of the 601 patients who constituted our series, 38 were clinically suspected of having myocarditis on the bases of a very recent onset of congestive heart failure and/or of arrhythmias and/or of conduction disturbances, and of a close-to-recent history of flu-like febrile illness. Corresponding EMBs showed myocarditis in 16 of the 38 cases (42.1%). A further 10 EMBs, from patients with a recent onset of congestive heart failure without prior infection episodes, showed myocarditis. Therefore, biopsy-proven myocarditis occurred in 26 of the 601 patients (4.3%). Of the 26 cases, 21 were lymphocytic, 1 was necrotizing granulomatous, 1 was eosinophilic and occurred in a patient who later developed overt zoonosis, 1 had some giant cells within endocardial inflammatory infiltrates, and 2 were borderline forms. In active myocarditis, inflammatory cells mostly constituted of T-lymphocytes (CD45RO+) with sparse macrophages (CD68+) and a few B cells (CD20+). B-lymphocytes and macrophages, along with activated T-lymphocytes, all expressed MHC class II HLA DR molecules, which were also expressed "de novo" by activated endothelial calls of capillaries and of small intramural vessels. HLA DR revealed itself as a very useful marker for the detection of activated inflammatory and endothelial cells. We also noted an increase in the number of perivascular and interstitial mast cells. Ultrastructural study was helpful for the characterization of myocyte damage and of interactions between inflammatory cells and myocytes. In 4 cases (1 of whom was later revealed as HIV positive, and subsequently died of AIDS), we found microreticulotubular structures in endothelial cells of small vessel and capillaries; in 7 cases, there were myocyte changes similar to those described in polymyositis; in 1 case, we observed subplasmalemmal buddings, but no viral particles; in 6 cases, there was extensive myocyte damage with myofibrillar lysis and focal adipous metaplasia; the remaining 6 cases showed myocyte damage of differing extent and severity; in the borderline forms, such damage coexisted with interstitial fibrosis. One of the 21 lymphocytic myocardites was not treated because during hospital screening the patient proved to be HIV positive; of the remaining 20 active myocardites, 11 were treated with a 6-month tapered steroid and azathioprine protocol (one was treated for 24 months), while 9 were not treated. The corresponding follow-up was: 6 deaths (congestive heart failure), 2 cardiac transplants and 3 survivals (1 with pace-maker) in the treated group, and 3 deaths (2 of congestive heart failure and 1 of sudden death), 1 cardiac transplant and 5 survivals (1 on the waiting list for transplantation) in the non-treated group. One of the 2 patients with borderline myocarditis died of congestive heart failure, and 1 is alive. Of the 22 patients with clinical diagnosis of myocarditis and negative biopsy, 7 died of congestive heart failure (2 on the waiting list for transplantation), 4 underwent cardiac transplantation, and 11 are alive (1 is awaiting transplantation). Of the 20 patients currently alive, 1 was originally in NYHA class III, 15 were in class II and 4 were in class I. Of the 20 overall patients who died, 12 were originally in NYHA class IV, 6 in class III, 2 in class II; of the 8 patients who underwent transplantation, 6 were originally in NYHA class IV and 2 in class III. Our overall experience shows that the frequency of myocarditis diagnosed according to Dallas criteria is high in patients with clinical diagnosis of myocarditis, while it is extremely low in dilated cardiomyopathy patients. This finding suggests that, although non-specific, recent onset of symptoms and prior febrile infe