Fishing for complements

In this issue of Blood, Jodele et al¹  build upon earlier work on the role of complement in the pathophysiology²  and therapy³  of transplant-associated thrombotic microangiopathy (TA-TMA) by exploring genetic predisposition for developing this complication. Earlier work in atypical hemolytic uremic syndrome showed a number of alterations in genes that had been previously associated with the disease, along with gene variants in coagulation and other pathways that were likely associated with this condition.⁴  In a similar fashion, Jodele et al hone in on genetic risk for TA-TMA by looking for variants in 17 genes specifically known to play a role in complement activation in a cohort of 77 subjects undergoing hematopoietic cell transplantation, 34 of whom had developed TMA.¹ 

The tricky part with this approach was defining what a “variant” means—in other words, picking through minor or major differences in and around the 17 genes chosen and deciding what variations noted could possibly predispose a patient to developing the disease. The authors worked hard to retain for analysis genetic changes that had a high chance of being clinically meaningful using a variety of in silico methods. Deep intronic variations along with commonly known single-nucleotide polymorphisms were removed, leaving 42 variants, 27 of which had not been previously reported. Of these, 8 were predicted or reported to be pathogenic, 15 were likely benign, and 19 had unknown significance. The authors consciously chose to retain the “likely benign” variations to avoid possibly excluding something from screening that could make a difference in risk, even if unlikely.

The method successfully identified broad associations of this group of variants with the target outcome: 65% of patients with TMA had 1 or more of the defined variants, as opposed to 9% of those who did not experience TMA (P < .0001). Eighty percent of white patients with at least 1 gene variant developed TMA, compared with 25% of those without variants (P < .0001). Patients with ≥3 variants had more severe disease, as noted by an increase in transplant-related mortality (57% vs 21%; P = .02).

Of note, the authors describe intriguing differences in both the incidence and number of gene variants by race (white vs nonwhite). Recipients who were white never had more than 1-2 variants present, with a median number of 0 vs 2.5 in nonwhite, who often had more than 2 and as many as 7, leading to worse survival outcomes (see figure). This prompted the investigators to study the incidence of TMA and survival in a larger cohort of 333 consecutive patients transplanted at their center. Consistent with their observation in the smaller genetics cohort, nonwhite patients had higher incidence of TMA (52% vs 29%; P = .008) and worse survival at 1 year (56% vs 81%; .0009).

Although this is a good start toward understanding susceptibility, screening, and potentially intervening in some way to prevent TMA in high-risk patients, his study raises several questions. First, the incidence of TA-TMA was very high at this center, likely because of the combination of a very broad definition of this clinical condition (developed largely through individuals associated with this work⁵ ) and a transplant population very different from many centers (largely nonmalignant disorders). Because the definition of TMA is based upon clinical observations in a population that is often very sick during the transplant procedure, it is likely that some of the patients on the mild end of the spectrum may have different clinical explanations for their observed symptomatology. On the other side of the equation of assigning risk, many of the variants occurred only once in this study, and because many are predicted to be benign, further testing may show them to not be of pathologic significance.

These issues could be addressed by a more stringent definition of the disease, but this may or may not reflect the true pathophysiology of the disorder—there simply may be mild variants of true TMA that have a more benign course and better prognosis. The second issue, assigning true pathological predictive power and significance to given gene variants, requires study of a much larger number of patients; only by assessment of large numbers of a given variant can one truly discern whether it contributes to predicting outcomes.

Atypical hemolytic uremic syndrome genetic susceptibility panels are already available at certified laboratories, and variations of these panels more closely tied to TA-TMA are likely to blossom soon. Along with refining genetic risk profiling, however, we need to develop reasons to screen—interventions that can meaningfully decrease risk of this potentially devastating complication. Eculizumab is the newest agent with a suspected a role in treatment,³  but the outcome in many patients treated with this very expensive drug is less than ideal, and it may be cost-prohibitive and potentially not effective to give as prophylaxis. Which of the many risk factors for TA-TMA⁶  (certain conditioning regimens, calcineurin inhibitors [especially when given with sirolimus⁷ ], HLA mismatch, graft-versus-host disease, and certain infections) can be modified in a patient identified at risk? How many variants are needed before intervention should occur, and which variants are most important? Many questions need to be addressed in this field, but hopefully in the near future we will be able to effect a decreased incidence and offer better therapy for this disease through screening, prevention, and more effective treatment approaches.