High-dose immunosuppressive therapy and autologous hematopoietic cell transplantation for severe systemic sclerosis: long-term follow-up of the US multicenter pilot study

Abstract

More effective therapeutic strategies are required for patients with poor-prognosis systemic sclerosis (SSc). A phase 2 single-arm study of high-dose immunosuppressive therapy (HDIT) and autologous CD34-selected hematopoietic cell transplantation (HCT) was conducted in 34 patients with diffuse cutaneous SSc. HDIT included total body irradiation (800 cGy) with lung shielding, cyclophosphamide (120 mg/kg), and equine antithymocyte globulin (90 mg/kg). Neutrophil and platelet counts were recovered by 9 (range, 7 to 13) and 11 (range, 7 to 25) days after HCT, respectively. Seventeen of 27 (63%) evaluable patients who survived at least 1 year after HDIT had sustained responses at a median follow-up of 4 (range, 1 to 8) years. There was a major improvement in skin (modified Rodnan skin score, -22.08; P < .001) and overall function (modified Health Assessment Questionnaire Disability Index, -1.03; P < .001) at final evaluation. Importantly, for the first time, biopsies confirmed a statistically significant decrease of dermal fibrosis compared with baseline (P < .001). Lung, heart, and kidney function, in general, remained clinically stable. There were 12 deaths during the study (transplantation-related, 8; SSc-related, 4). The estimated progression-free survival was 64% at 5 years. Sustained responses including a decrease in dermal fibrosis were observed exceeding those previously reported with other therapies. HDIT and autologous HCT for SSc should be evaluated in a randomized clinical trial.

Figures

Figure 1

Dermal fibrosis grades 0 to 5. Each pair represents a different dermal fibrosis grade. The left-hand column of each pair containing low-power views (original magnification, × 2.5), illustrates the spectrum of change in the overall dermal breadth and distribution of fibrosis. The right-hand column views for each pair (original magnification, × 10), illustrate the qualitative size, shape, distribution, and density in the dermal collagen bundles and surrounding extracellular matrix. All sections are stained with H&E. Grade 0 is thin curlicue collagen bundles with only minimal focal papillary dermal sclerosis. Grade 1 is less than 25% sclerosis with focal dense deep dermal fibrosis. In grade 2, the dense zone of fibrosis within the hypodermis comprises less than 50% of the overall dermal breadth. The remaining reticular dermis is composed of fine collagen bundles with actinic change (blue) in the upper dermis. Higher power shows the distinction between the sclerotic and normal collagen bundles. Grade 3 is extensive sclerosis—more than 50%—throughout the dermis with thickened collagen bundles admixed with thinner bundles and intervening extracellular matrix. Grade 4 is pandermal fibrosis with loss of most extracellular matrix without extension into hypodermis (or entrapment of eccrine coils) or surrounding fat. Grade 5 is markedly fibrotic and thickened—more than 4 mm—dermis replaced with dense waxy collagen with deep extension into hypodermis with fibrous incorporation of a muscular artery. See “Patients and methods; Assessment of skin biopsies” for more information on images.

Figure 2

Change in modified Rodnan skin score (mRSS), modified Health Assessment Questionnaire (mHAQ) score, and organ function after high-dose immunosuppressive therapy (HDIT) and autologous hematopoietic cell transplantation. A determination was made as to whether a parameter value was statistically significantly increasing or decreasing over time using a generalized estimating equation (GEE) model. The bold black solid line represents the mean value over time for the parameter of interest. The bold black dotted line represents an estimate of the modeled linear relationship between the parameter value and time and summarizes the results of the GEE models. The gray solid lines are parameter values for individual patients. The mean mRSS and mHAQ values statistically significantly decreased with time after HDIT (both P < .001; panels A and B, respectively). The mean values for diffusion capacity of the lung for carbon monoxide (DLCO) adjusted for hemoglobin levels did not statistically significantly change (P = .50, panel C), and forced vital capacity statistically significantly increased with time (P = .01, panel D). The mean values for serum creatinine and ejection fraction statistically significantly increased and decreased with time (P = .01 and P = .04; panels E and F, respectively).

Figure 3

Resolution of dermal fibrosis after HDIT and autologous HCT. Full-thickness skin biopsies from patient no. 11, collected at baseline and then at 1 and 5 years after HDIT, were histologically evaluated (H&E). The skin biopsies performed after HDIT were performed at a site adjacent to the baseline skin biopsy. The biopsies in the left-hand column were taken at original magnification × 5, and those in the right-hand column were taken at original magnification × 20. All sections were stained with H&E. (A) Skin biopsy was obtained before HDIT and autologous HCT. Pandermal sclerosis from the dermal-epidermal border to the hypodermis (subcutaneous fat) was observed. The epidermis is mildly acanthotic (thickened) with loss of rete ridges. The reticular dermis is replaced by dense compact collagen without normal fascicular bundles or dermal appendages. This pretransplantation skin biopsy was determined as grade 5 dermal fibrosis. The thickness of the dermis was measured at more than 2 mm. (B) In the higher-power magnification, the straightened dermal-subcutaneous border demonstrates the abnormal, densely packed, homogenized collagen. (C) The skin biopsy at 1 year after HDIT was determined to be a grade 2 dermal fibrosis and has less fibrosis than at baseline. The low-power magnification view shows crowded collagen fascicles with focal areas of residual thickened bundles. (D) A higher-power view of the 1-year skin biopsy from panel C shows thin and collagen bundles admixed with residual thick straightened hypereosinophilic collagen bundles without dense homogenization at baseline. The residual eccrine unit lacks any surrounding adipose tissue. (E) The skin biopsy at 5 years shows complete resolution of the dermal fibrosis (grade 0) with a reduction in the thickness of the dermis from baseline to 1 mm. The collagen bands in the dermis are thin with a relative increase in the intervening extracellular matrix (space between the collagen bands). The dermal-epidermal border remains straightened with loss of rete ridges. (F) A higher-power view of collagen in the lower reticular dermis demonstrates the change to thin wavy bundles separated by increased ground substance. See “Patients and methods; Assessment of skin biopsies” for more information on images.

Figure 4

Change in dermal fibrosis grade and mRSS after HDIT and autologous HCT. The change in dermal fibrosis grade and mRSS was evaluated in patients from whom we had skin biopsies from both before and after HDIT and autologous HCT (n = 10). The mean decrease in skin score from baseline after HDIT was 19.7 (P < .001; left horizontal line) and was very comparable to the result for the overall group. The mean decrease in the dermal fibrosis grade from baseline was 3.1 (P = .001; right horizontal line) from a mean baseline grade of 4.3. Seven of the 10 patients had a dermal fibrosis grade of 0 to 1 at the time of last skin biopsy. The median follow-up was 4 (range, 1 to 6) years after HDIT and autologous HCT.

Figure 5

Survival and progression-free survival after high-dose immunosuppressive therapy and autologous hematopoietic cell transplantation for systemic sclerosis. The estimated probability of survival and progression-free survival at 5 years was 64%. Tick marks represent censored observations.