To the Editor:
von Willebrand factor (vWF)-cleaving protease has been shown to be deficient in patients with thrombotic thrombocytopenic purpura (TTP).1,2 The recent article by van der Plas et al3 confirms the observation that vWF-cleaving protease is severely decreased or missing in plasma of patients with classic TTP, whereas normal protease activities were found in patients with bone marrow transplantation (BMT)-associated TTP. The investigators found no significant difference between the 2 groups regarding the levels of cellular fibronectin. However, as shown in Table 1 of their report, there was a very significant difference in vWF:Ag and vWF:Rcof between classic TTP and BMT-associated TTP, even though the number of patients in each group was quite limited (5 and 8 patients, respectively). Thus, the median values of vWF:Ag were 88% in classic TTP and 292% in BMT-associated TTP. The respective median values of vWF:Rcof were 67% and 161%. The Mann-Whitney rank sum test resulted in P values of .010 (vWF:Ag) and .011 (vWF:Rcof) for the difference between classic and BMT group. These results suggest that, in contrast to the classic group, there was massive release of vWF from the endothelial cells in the BMT group, possibly due to graft-versus-host disease and/or cyclosporin treatment.4,5 It is conceivable that the excessive mobilization of abnormally adhesive very large multimeric forms in the BMT patients leads to aggregation of the circulating platelets and to thrombotic microangiopathy, even in the presence of vWF-cleaving protease.6
Thrombotic microangiopathy (TMA), which is associated with BMT, is sometimes labeled as TTP and sometimes as hemolytic-uremic syndrome (HUS). Hematologists seem to favor the diagnosis of TTP, whereas nephrologists tend to call it HUS. There is no distinct boundary between TTP and HUS: diagnosis of TTP is preferred for cases with neurologic disturbances, whereas cases involving predominantly renal impairment are classified as HUS. No information is given in the study of van der Plas3 regarding the clinical symptoms and criteria, according to which the TMA in their patients had been classified as TTP. It is obvious (Table 1) that the creatinine level was normal in patients with classic TTP (median value, 76 μmol/L), but significantly increased (P = .019) in patients with BMT-associated TTP (median value, 178 μmol/L). Therefore, the TMA in the BMT group should be classified as HUS rather than as TTP.
In our study on vWF-cleaving protease in patients with TTP and HUS,1 we observed deficient or strongly decreased protease activity in patients with familial and nonfamilial TTP, but normal protease activity in patients classified as having HUS. It should be added that 2 patients were included into the latter study who presented with symptoms of TMA after BMT: the patient no. 24A, classified as having TTP, had 25% protease activity, whereas the symptoms in the patient no. 35A were compatible with the diagnosis of HUS and the patient had normal activity of vWF-cleaving protease. Although there were no neurological symptoms observed in patient no. 24A, his diagnosis was apparently misclassified as TTP by the participating investigator, because the creatinine levels had been only moderately increased.
It appears that the discrepancy between conclusions of our report1 and the report of van der Plas et al3only arises from the differential diagnosis of the TMA in BMT patients. If the BMT cases had been classified by van der Plas et al3as HUS, which may be appropriate regarding the laboratory data, their report would perfectly confirm our observation that the vWF-cleaving protease is deficient in patients with TTP but present in patients with HUS.
The level of vWF in the BMT-associated TTP was indeed higher compared with the level in patients with classic TTP. Furlan states that the massive release of vWF from the endothelium in the BMT group could play a role in the development of the TTP. Endothelial activation with an increase in vWF is known to occur in patients using cyclosporin as well as in patients receiving a BMT.1-1,1-2 Not all of these patients develop TTP. Moreover, a lot of other circumstances, eg, sepsis and exercise, are also known to induce a massive increase in vWF without development of TTP.1-3 So, an increase of vWF is not enough to induce TTP, and other mechanisms are probably involved.
TMA is a clinical diagnosis that includes HUS and TTP. Both are characterized by microangiopathic hemolytic anemia, thrombocytopenia, microvascular thrombosis, and multiorgan failure.1-4 Typical verotoxin-associated HUS, mostly due toEscherichia coli 0157:H7 infection, is associated with prodromal diarrhea followed by acute renal failure and is considered a disease with a good outcome. Atypic HUS covers 2 clinical conditions. The first form is characterized by severe gastrointestinal prodromes, acute onset anuria, and neurological involvement. The second form is without diarrhea prodromes but with progressive renal function deterioration and neurological involvement that resembles TTP. Hence, TTP and HUS appear to be closely related and part of a disease continuum. Renal function was only slightly impaired in our patients.1-5 A rapid renal insufficiency with diarrhea was not observed. No spontaneous improvement occurred, and patients with an autologous BMT-associated TMA were successfully treated with cyclosporin. Our patients did not respond to plasmapheresis. This makes the BMT-associated TMA a specific subgroup. The normal vWF-cleavage protease activity in what is defined HUS and deficiency of the protease in TTP provides insight in a different pathophysiologic mechanism.1-6 Patients with BMT-associated TTP had normal vWF proteolysis, and this confirms the observation that TMA occurs with a normal vWF cleaving activity. It is of interest to study whether the ineffectiveness of plasmapheresis in our BMT-TTP patients can also be observed in the other TMAs with normal vWF protease activity. A clear clinical and pathophysiologic distinction between vWF-cleaving protease present and deficient TMA’s could define a better classification system.
Our studies were supported by grants from the Swiss National Science Foundation (Grant No. 32-47033.96); from the Central Laboratory, Blood Transfusion Service, Swiss Red Cross; from Immuno (Vienna, Austria); and from the Malcolm Hewitt Wiener Foundation (New York, NY).