Rituximab Provides Steroid-Sparing Therapy in New-Onset Chronic Graft-Versus-Host Disease

Chronic graft-versus-host disease (cGVHD) remains a complication of allogeneic HCT and treatment remains prednisone dependent with starting dose of 1mg/kg/day. Rituximab shows efficacy in steroid-refractory cGVHD supporting B-cell pathogenesis. We hypothesized that anti-B cell therapy delivered at cGVHD development is effective and steroid-sparing. 35 patients (pts) with new-onset cGVHD, requiring systemic steroid treatment at 1mg/kg/day, received Rituximab 375-mg/m²weekly x4 within first month of diagnosis. A second course was allowed if pts had less than a partial response (PR) after 2 months.

Primary endpoints were six-month and 1y clinical responses including steroid requirements. All 35 pts are evaluable with 1.5y follow-up. Clinically meaningful responses (defined as pts with a PR or CR and prednisone <0.25 mg/kg/day) were seen in 15/35 pts (43%, 13 CR; 2 PR) at 6 months and remained 15/35 (43%, 11 CR, 4 PR) through 1y.  There were 6 deaths from cGVHD within 1y of Rituximab.

Overall, cGVHD clinical response was predicted by naïve CD19⁺CD38^-IgD⁺CD27^-B-cells pre- rituximab (p=0.03), and low BAFF (p=0.007). B-cell-alloreactivity was assessed by HY IgM, IgG, and HY specific B-cell quantification. 15 cGVHD patients were male with female donors (F→M) and 6 (38%) tested HY IgM positive and 9 (56%) were HY IgG positive at cGVHD diagnosis. Both HY IgM and IgG became undetectable in all responders.  In contrast, 3 of the 4 nonresponders remained IgG HY positive and the other 1 redeveloped HY IgG with cGVHD progression. Similarly, HY-antigen specific B-cells were detected in 8/15 (53%) before rituximab and recurred in 4/6 survivors with progression of cGVHD thus far.

This first reported clinical trial of Rituximab and corticosteroids for new-onset cGVHD showed 46% clinically meaningful responses through 1y with a steroid sparing effect. The redevelopment of alloreactive B cells and HY IgG in association with cGVHD progression suggests longer Rituximab treatment-schedules or more potent B-cell inhibitors may further improve cGVHD therapy.  

Clinically meaningful responses (CR or PR and prednisone<0.25 mg/kg)

	Responders   n=15	Nonresponders n =17	P-value
Median days to cGVHD	233 (136-988)	245 (138-357)	ns
Pre-ritux CD19+ B cells/microliter	110 (range 0-1723)	299 (14-1140)	0.09
Pre-ritux naïve CD19+ CD38-, IgD+ CD27- B cells/microliter	102 (0-1158)	58 (0-700)	0.03
Recovery of naïve B cells 1.5 y after Rituximab /microliter	63 (0-430)	8 (0-33)	0.05
Pre-ritux IgG (mg/dl)	645 (224-1800)	547 (242-1410)	ns
Pre-ritux BAFF (mg/dl)	2.8 (0.8-8)	7.3 (1-15.5)	p=0.007
F→M pairs n=15	11 (73%)	4 (27%)	p=0.01
9/15 FàM tested HY IgG positive preRitux	All 11 responders became seronegative	3 remained HY positive; one  redeveloped HY IgG