Response: Further examination of intravascular hemolysis (IVH) following intravenous anti-D for immune thrombocytopenic purpura (ITP)

Tarantino and colleagues, remind us of the need to further examine the frequency, pathophysiology, and prevention of serious complications of acute hemolysis associated with Rh_o(D) immune globulin intravenous (WinRho) treatment of immune thrombocytopenic purpura (ITP). Complications reported to date include compromising anemia, red blood cell (RBC) transfusion, renal failure, dialysis, disseminated intravascular coagulation (DIC), and death, which may occur concurrently.^1,2 

Understanding the etiology of WinRho-associated acute hemolysis is a critical need. This knowledge could offer insight about risk factors and preventive measures. More than one mechanism may account for platelet responses in WinRho-treated ITP patients.²  Similarly, multiple mechanisms may explain the acute hemolysis and could account for differences among patients. All theories warrant further consideration and evaluation.

We agree with the authors that “high titers of a single additional alloantibody, found in only specific lots of WinRho SDF,” is an improbable explanation for all occurrences of acute hemolysis. All WinRho lots contain anti-D, while anti-A, anti-B, anti-C, anti-E, and possibly other “unexpected” RBC antibodies (eg, anti-Fy^a) may vary in identities and titers.^1–4  This lot-to-lot variation, in conjunction with differing patient RBC phenotypes, could explain why only some patients experience lot-specific “acute hemolytic transfusion reactions.” The authors noted that “[o]nly 2 patients shared one lot of [WinRho] in the 2 reported series.” However, information about lot(s) administered was unavailable for 11 (52%) of 21 case series patients.^1,2 

The authors propose a “2-hit” model to explain the DIC and renal failure complications. This phenomenon may indeed apply. However, some case series patients experienced serious complications but were not reported to have other predisposing medical conditions. The authors further propose that patient RBCs may express a single high-density polymorphic determinant or sufficient copies of multiple relevant RBC antigens to trigger lot-dependent acute hemolysis. We are aware of testing for such determinants or antigens for only one patient who experienced acute hemolysis following WinRho administration. Although those findings were unremarkable, evaluation in other patients is nonetheless warranted.

Case series data^1,2  and the authors' collective experience support the recommendation that patients be monitored for 48 hours after WinRho administration for signs and symptoms of acute hemolysis and other potential complications. The characteristic hemoglobinuria of acute hemolysis tended to occur within 4 hours in the case series patients^1,2  and can generally be detected with routine urinalysis tests (eg, reagent strip for blood = large, together with microscopic = 0-5 RBC/high power field). When hemoglobinuria cannot be established (eg, in the presence of confounding hematuria), alternative laboratory tests (eg, serum hemoglobin) are available. Although renal failure tended to occur within 48 hours,²  data were unavailable to indicate whether 48 hours was a sufficient time frame for detection of DIC.¹ 

Only with further data can we examine the incidence, etiology, risk factors, complications, and prevention of acute hemolysis following WinRho administration for ITP.^1,2  Detailed adverse event reports (including medical history, pretreatment and posttreatment laboratory data, clinical findings, and WinRho lot number[s]) are needed^1,2,5  and can be submitted to the manufacturer/distributor⁴  and/or FDA.⁶