Changes in Dlco following HSCT for individual patients are plotted in Figure 2
. Two patients were unable to complete Dlco measurements in spite of best effort at pre-HSCT evaluation. Of the 28 patients who have now been monitored > 12 months after HSCT, 8 patients had persistent lupus activity requiring therapy (BILAG grade A). Twenty patients did not. Among the eight patients with persistent disease activity requiring therapy, Dlco was improved over 12 months in two patients (25%) and was stable in six patients (75%). Among the 20 patients with no disease activity on extended follow-up, 5 patients had normal Dlco at study entry and these 5 patients retained normal Dlco at 12 months after HSCT. Of the remaining 15 individuals with freedom from BILAG grade A, the Dlco increased by ≥ 12% of the predicted value in 11 of 15 patients by 1 year after HSCT (73%). Of the remaining four individuals without active SLE and abnormal Dlco at entry, who did not show improvement in Dlco by 12 months, three patients had improved by ≥ 12% of the predicted value by 2 years; the other patient has remained stable. Therefore, of the 20 individuals evaluated for > 1 year, who have experienced sustained freedom from active SLE, 93% of those with a low Dlco have improved (14 of 15 patients), 5 of the 5 patients with a normal Dlco stabilized, and 1 of 15 patients with an abnormal Dlco neither improved nor declined by 2 years (7%). In contrast, only two of the eight patients with persistent active lupus requiring therapy experienced an improved Dlco at 1 year (25%). Among these 28 patients, by 12 months after HSCT, no statistically significant difference was observed on the proportion of clinically significant improvement in Dlco between patients with and without persistent SLE activity (2 of 8 patients vs 11 of 20 patients, p = 0.2213). Statistically significant differences in the proportion of clinically significant improvement in Dlco, however, were observed between patients with persistent SLE activity, who had an abnormal Dlco at study entry and those who had an abnormal Dlco at study entry, and did not experience persistent SLE activity (2 of 8 patients vs 11 of 15 patients, p = 0.0393), and between patients who remained free of SLE activity and had a normal FVC at study entry and patients who remained free of SLE activity and had an abnormal FVC at study entry (0 of 5 patients vs 11 of 15 patients, p = 0.0081). For the total 34 patients in this study, based on the analysis of the mixed-effect general linear model for repeated measurements, a statistically significant increasing trend in Dlco over time was observed (p < 0.0001). With the adjustment of age and the status of disease, post-HSCT Dlco statistically significantly increased at month 6 (n = 34; p = 0.0410), month 12 (n = 28; p < 0.0001), month 24 (n = 21; p < 0.0001), month 36 (n = 8; p < 0.0001), month 48 (n = 6; p = 0.0010), and month 60 (n = 5; p = 0.0006) compared with the pre-HSCT Dlco. With the adjustment for age and month, no statistically significant difference in Dlco between patients with disease remission after HSCT and those patients without disease remission after HSCT (p = 0.2755).