Transplant Associated Thrombotic Microangiopathy (TA-TMA) Commonly Occurs after Steroid-Refractory or Dependent Gastrointestinal (GI) Graft Versus Host Disease (GVHD) and Contributes to Post-Transplant Morbidity

Background: Hematopoietic cell transplantation associated thrombotic microangiopathy (TA-TMA) shares common features with atypical hemolytic uremic syndrome and thrombotic thrombocytopenic purpura. TA-TMA affects multiple organ systems resulting in multi-organ failure and non-relapse mortality. Recently, complement activation has been shown to be associated with TA-TMA leading to utilization of complement blockade as a therapeutic intervention in TA-TMA. We sought to evaluate the incidence of TA-TMA in patients with GVHD. We hypothesized that patients with gastrointestinal (GI) GVHD with transfusion dependence and renal dysfunction have a concomitant diagnosis of TA-TMA and that this contributes to worse outcomes.

Methods: We performed a retrospective review of 124 patients with grade 3/4 GI GVHD admitted October 2008-October 2016. Patients who did not achieve a response by 7 days of 2mg/kg steroids were defined as steroid-refractory and patients who had flare of GVHD symptoms requiring a second course of 2mg/kg steroids were defined as steroid-dependent GVHD. We collected transplant-related data, number of hospitalizations, infections, length of stay and number of platelet/ pRBC transfusions. We assigned patients to groups by likelihood of TA-TMA based on presence of two out of three criteria: (1) > 2 transfusions after first GI biopsy, (2) evidence of hemolysis (schistocytes or nucleated RBCs) with negative Coombs' test and (3) evidence of renal dysfunction. 29/124 patients had paired samples for measurement of complement activation at time of GVHD diagnosis and 2-6 weeks later. Markers of complement activation included C5a and C5b-9 (terminal complement pathway), C4d (Classical pathway), and BBPlus (Alternative pathway).

Results: The median age at transplant was 54 years (19-72) and 40.3% were female. Reduced intensity or non-myeloablative conditioning was used in 66.1% of patients and all but 10 patients received tacrolimus. The median time to GVHD onset was 38 days (range: 12-273).

TA-TMA and GVHD. A total of 84 (67.7%) subjects met at least 2 criteria for TA-TMA. TA-TMA typically occurred after GVHD at a median of 142 days post-transplant and 78 days post-GVHD. The median time to GVHD was 38 days (12-273). Patients with steroid-refractory GVHD received more platelet and pRBC transfusions than those with steroid-responsive GVHD (p = 0.0018 and 0.0068, respectively) at 2 and 4 weeks post GVHD diagnosis (Figure 1). Patients with steroid-refractory or dependent GVHD were more likely to develop TA-TMA than those with steroid-responsive GVHD. Regardless of responsiveness of GVHD to steroids and accounting for differences in follow-up, patients with TA-TMA had longer length of hospital stay (median: 53 v. 25.5 days, p < 0.001) than patients without TA-TMA.

Post-transplant infections in all patients included CMV reactivation (46.8%), other viral infection (60.5%), fungal infection (12.1%), and bacterial infection (65.3%). Renovascular complications included new or worsening hypertension (18.6%), acute kidney injury (45.2%), need for hemodialysis (16.9%), and proteinuria >30mg/dL (75.8%). In multivariable analysis, the risk factors for developing TA-TMA from time of transplant included acute GVHD, viral infection other than CMV, and fungal infection (Table 1).

At 2-4 weeks after GVHD diagnosis, differential upregulation of alternative pathway (BBPlus) was seen in patients who went on to develop TA-TMA. It is notable that no differences in complement activation were seen at initial GVHD diagnosis, suggesting that TA-TMA appears to be a secondary event that follows GVHD.

Discussion: We show that GVHD is a risk factor for development of TA-TMA and that patients with steroid-refractory/dependent GVHD are more likely to have TA-TMA than steroid-responsive patients. Patients with both GVHD and TA-TMA receive more transfusions and have longer hospitalizations than those with GVHD alone. Our work suggests that the alternative complement pathway (BBPlus) is upregulated in patients with GVHD and TA-TMA, suggesting that complement blockade may be evaluated as a therapeutic strategy in these patients. Prospective study is warranted to understand the triggers of TA-TMA in patients with GVHD, particularly the role of viral and fungal infections and whether complement biomarkers could differentiate between steroid-responsive and steroid refractory/dependent GVHD.

Figure

View largeDownload slide

Figure

View largeDownload slide

Close modal