Prevalence and Reversibility of Cardiac Dysfunction in Systemic Lupus Erythematosus: Improved Ventricular and Valvular Function after Disease Remission Following Hematopoietic Stem Cell Transplantation.

Patients with cardiac abnormalities undergoing hematopoietic stem cell transplantation (HSCT) are considered at increased risk of cardiac toxicity which has traditionally precluded them from undergoing HSCT. However, cardiac involvement in patients with systemic lupus erythematosus (SLE) is not uncommon with a reported prevalence of up to 75%, often disease-related and may even be the indication for HSCT. Therefore, we undertook a retrospective analysis of all SLE patients who had undergone autologous HSCT (auto-HSCT) to determine the prevalence and nature of cardiac involvement, safety and effect of HSCT on cardiac function. Between April 1997 and January 2006, a total of 55 patients with SLE underwent auto-HSCT. Stem cell mobilization was with cyclophosphamide 2g/m^2 with subcutaneous filgrastim and peripheral blood stem cells (PBSC) were harvested via apheresis upon neutrophil recovery. The PBSC product was lymphocyte-depleted via CD34-positive selection with an Isolex cell separator in 51 patients and unmanipulated in the other 4 patients. HSCT conditioning regimens were cyclophosphamide 200mg/kg over 4 days and equine antithymocyte globulin (ATG) 90mg/kg over 3 days. The PBSC product was re-infused 36h after the completion of cyclophosphamide on day 0 and the patient received subcutaneous filgrastim from either day 0 or day+6 until neutrophil recovery. Fourteen patients were found to have baseline cardiac abnormalities, either clinically or on baseline radionuclide technetium multi-gated acquisition scans or two-dimensional echocardiogram. Of these, 13 were female and the median age at the time of HSCT was 39.8 years (range 15.7 to 55.5 years). The median baseline SLE Disease Activity Index of this cohort was 16 (range 10 to 35). The cardiac abnormalities were impaired left ventricular function (n=6), pulmonary hypertension (n=5), mitral valve dysfunction (n=3), significant pericardial effusion (n=1) and accelerated coronary artery disease (n=1). At a median follow up of 23.5 months (8 to 103 months), 2 of the patients are dead, one from disease progression/relapse at 11.5 months after HSCT and the other died from an accident at 8 months post-transplant. There were no cardiac or transplant-related deaths or significant cardiac toxicities. All patients with impaired ejection fraction (range 25% to 47%) stabilized or improved post-HSCT induced remission. The median pulmonary artery systolic pressure was 39mmHg (range 27 to 50 mmHg). Among the 5 patients with pulmonary hypertension, 3 improved paralleling remission from SLE. Again paralleling disease remission, 2 patients with severe disability from symptomatic mitral valve disease improved, obviating the need for valve replacement which had been planned pre-HSCT for one. In conclusion, we found that our patients with ventricular systolic and valvular dysfunction, and possibly pulmonary hypertension, improved following HSCT if they achieved remission from SLE. Our experience suggests auto-HSCT may be safely performed in selected patients with SLE with cardiac dysfunction and furthermore, improvement of cardiac function may occur following HSCT-induced disease remission. This warrants further study in a larger cohort of patients.