Purpura, thrombotic thrombocytopenic

* Context.-Thrombotic thrombocytopenic purpura (TTP) and the hemolytic uremic syndrome (HUS) share many clinical features and have been difficult to separate into distinct entities. Histologic examination of organs from autopsied patients suggested that TTP and HUS have dissimilar lesions of different severity and distribution.

Objective.-To perform a retrospective observational review of autopsied patients with UP or HUS to compare the nature and severity of the lesions found.

Design.-To examine the pathologic features of these conditions, we reviewed all cases among 51 350 indexed autopsies at The Johns Hopkins Hospital (Baltimore, Md) diagnosed with UP or HUS, and included those showing multiple arteriolar thrombi or their sequela.

Results.-The 56 cases that met the inclusion criteria fell into 2 distinct groups, based on distribution and severity

of arteriolar lesions. In 25 patients classified as having TTP, platelet-rich thrombi were present-in decreasing severity-in heart, pancreas, kidney, adrenal gland, and brain. In 31 patients with HUS, fibrin/red cell-rich thrombi were present, largely confined to the kidney and often severe, and only 6 cases showed pancreas involvement, 4 adrenal gland involvement, 2 brain involvement, and 1 heart involvement.

Conclusion.-Despite similar clinical features and therapeutic approaches, TTP and HUS each have a characteristic constellation of histopathologic findings. This observation suggests that TTP and HUS are 2 distinct disease entities with different pathophysiologies, and that they do not represent a spectrum of the same disease process.

(Arch Pathol Lab Med. 2003;127:834-839)

Thrombotic thrombocytopenic purpura (TTP) was initially described in 1924 by Moschcowitz,1 who described a young woman with fever, hemolytic anemia, thrombocytopenia, renal failure, and neurological symptoms. Thrombotic thrombocytopenic purpura was given its present name by Singer et al in 1947.2 In 1955, Gasser and colleagues3 used the term hemolytic uremic syndrome (HUS) to describe a similar condition seen in children presenting primarily with acute renal failure. Since that time, there has been much debate over defining TTP and HUS as 2 distinct entities or the same entity along a clinical spectrum, prompting some clinicians to use the combined diagnosis TTP/HUS.4-6

Autopsy examinations of patients diagnosed with either TTP or HUS had suggested that, despite clinical similarities, there could be different patterns of pathologic changes in the 2 entities, and thus different pathophysiologies. This idea is supported by recent studies identifying different levels of von Willebrand factor-cleaving metalloproteinase activity in acute idiopathic TTP versus acquired HUS.7,8

MATERIALS AND METHODS

All cases among 51350 indexed autopsies at The Johns Hopkins Hospital (Baltimore, Md) diagnosed with TTP or HUS were reviewed. Fifty-six cases were identified with appropriate material for review, 25 with a diagnosis of TTP and 31 with HUS. All of these cases contained routine hematoxylin-eosin-stained sections of the kidney and other available organs. Phosphotungstic acid-- hematoxylin (PTAH), Mallory trichrome, and immunohistochemistry for factor VIII were present on kidney sections in a minority of cases and were used for illustrative purposes. Histology of the kidney, pancreas, brain, adrenal gland, and heart from each case was reviewed for the presence or absence of vascular lesions or their sequelae, including ischemic changes, necrosis, or resolving lesions. Slides were reviewed without knowledge of the individual patient's clinical course. Following the analysis, available clinical information was reviewed for the presence of thrombocytopenia, hemolytic anemia, purpura, fever, evidence of neurological dysfunction, acute renal failure, and survival time. Each feature was scored a "+" if it was present, "-" if it was absent, and "?" if the status was unknown. All clinical features were considered to be present if they had been found at the time of new onset in the setting of TTP or HUS and could not be definitively explained by other processes (infection, malignancy, etc). Comparisons of the TTP and HUS clinical and pathologic features were assessed for statistical significance using a standard Chi^sup 2^ test for binary data or the unpaired Student t test for numerical data.

RESULTS

All 56 autopsied patients had thrombocytopenia and hemolytic anemia. The 25 patients with TTP were autopsied between 1932 and 2001 (Table 1). These patients averaged 37 +/-15 years in age (range, 13-65 years), 15 were female, and they survived an average of 29 +/-33 days from diagnosis (range 5-126 days). All patients with a diagnosis of TTP showed evidence of at least 4 of the clinical pentad of microangiopathic hemolytic anemia (100%), thrombocytopenia (100%), fluctuating or predominant neurological symptoms (88%), fever (96%), and severe renal dysfunction (20%). Acute renal failure was by far less common than the other clinical features (5/25, 20%).

The 31 patients with HUS were autopsied between 1960 and 1999 (Table 2). They averaged 34+/-23 years in age (range, 0.33-75 years), 15 were female, and they survived an average of 41+/-38 days from diagnosis (range, 1-135 days). These patients commonly had severe renal dysfunction (28/31, 90%). Neurological dysfunction (17/31, 55%), purpura (16/31, 52%), and fever (7/31, 23%) were often present, but were less common than in TTP patients (summarized in Table 3).

Pathologic changes were frequent in all 5 organs examined in patients with Tm The arterial thrombi seen in TTP consisted of finely granular eosinophilic plugs (Figures 1-5). In the acute phase, these lesions appeared to contain primarily platelets, supported by factor VIII immunohistochemical staining (Figure 1, D). In some patients, active lesions were seen along with organizing thrombi and ischemia-associated changes (Figure 1, A and B), depicting the temporal changes seen in TTP

Patients with HUS all had lesions in the kidney (Figure 6), with many cases of extreme bilateral cortical necrosis (Figure 6, A and B). The PTAH stain (Figure 6, E and F) showed the common HUS lesions to contain a large component of fibrin, with few platelets demonstrated by factor VIII staining (not shown).

Statistical comparison of clinical and pathologic features of TTP and HUS is shown in Table 3. Among clinical features that were significantly more common, fever and neurological dysfunction were found more often in patients with TTP, and acute renal failure was noted in those with HUS. The presence of pathologic lesions in pancreas, brain, adrenal glands, and heart was significantly more common in TTP Thrombotic thrombocytopenic purpura and HUS had comparable frequencies of renal involvement.

COMMENT

The distinction between TTP and HUS has historically been difficult to make on clinical grounds. Thrombotic thrombocytopenic purpura is traditionally defined as a systemic adult disease that relatively spares the kidneys, whereas HUS occurs in children with severe renal involvement. However, close review of Moschcowitz's original TTP patient showed clinically evident renal disease, and Gasser's original HUS patients had neurological impairment.13 Many authors have since reported exceptions to these definitions, including cases of adults with HUS, children with TTP, HUS patients with multiorgan disease, and TTP patients with significant renal disease.4,9,10

Our review of 56 autopsied patients diagnosed with either TTP or HUS reveals patient populations with clinical features reflecting the traditional definitions of these diseases. One notable exception, however, is the average age of the HUS patients. Our HUS patients had an average age of 34 years, comparable to the average age of 37 years for our TTP patients. This observation likely reflects the selection bias of analyzing only autopsy patients. Hemolytic uremic syndrome is typically a disease of young children; however, these patients have a much better prognosis than adults and therefore would not proportionally enter into the autopsy files. Additionally, although the TTP patient group commonly had multisystem disease and the HUS patient group had a clinical course dominated by renal dysfunction, there was notable clinical overlap between the groups. For example, 55% of HUS patients had neurological dysfunction, 52% had purpura, and 23% had fever, obscuring the clinical boundaries of HUS and TTP.

The pathologic features, however, appear to differ in distribution, character, and consequences between the 2 entities. Thrombotic thrombocytopenic purpura shows striking involvement of myocardial arteries and variable degrees of vascular involvement in kidney, pancreas, brain, and adrenal glands. The thrombi of TTP are composed predominantly of platelets with a trivial component of fibrin, consistent with previous studies.11-13 Permanent injury in TTP seems limited to ischemic foci, generally mild, in the heart and pancreas. Hemolytic uremic syndrome shows marked involvement of the kidneys; infrequent occurrence of vascular lesions in pancreas, brain, and adrenal glands; and almost no myocardial disease.

The thrombi of HUS appear to be composed predominantly of fibrin with few platelets.14 Dramatic destruction of the renal cortex may occur in HUS. These histologic findings are consistent with the general clinical course seen with our patient populations.

Recent studies have identified differences in von Willebrand factor-cleaving metalloproteinase activity in patients with acute idiopathic TTP versus HUS. Patients with TTP had severely decreased metalloproteinase activity, whereas this activity was normal in HUS. The reduced metalloproteinase activity is thought to prevent the breakdown of von Willebrand factor multimers, explaining the presence of large multimers in the plasma of TIP patients.15 These findings may also explain why patients with acute idiopathic TTP respond more favorably than patients with HUS to current plasma-exchange therapy.16,17 Our files included considerably fewer autopsy subjects with TTP than with HUS during the past decade (see Table 1), likely reflecting the dramatic improvement of plasma-exchange therapy for TIP subjects during this period.

The differences in the character and distribution of arterial lesions found in this study suggest that TTP and HUS are distinct pathophysiologic entities. Lesions in multiple organs, with cardiac involvement in all cases, characterized TTP, whereas patients with HUS had severe renal disease and infrequent lesions in other organs. In our review, the most useful histologic discriminator between TTP and HUS on autopsy material was the plateletrich, fibrin-poor character of TTP lesions and the fibrin-- rich character of HUS lesions as shown by PTAH staining. In our experience, PTAH seemed more reliable than immunohistochemistry when applied to autopsy tissues.

This study was limited by being a retrospective review of existing information and histologic materials in autopsied patients. More detailed prospective clinical and pathologic studies may assist in clarifying the etiology and pathogenesis of these disorders.

References

1. Moschcowitz E. Hyaline thrombosis of the terminal arterioles and capillaries: a hitherto undescribed disease. Proc NY Pathol Soc. 1924;24:21-24.

2. Singer K, Bornstein FP, Wiles A. Thrombotic thrombocytopenia purpura. Blood. 1947;2:542-554.

3. Gasser VC, Gautier E, Steck A, Siebenmann RE, Oechslin R. Hamolytischuramische Syndrome: Bilaterale Nierenrindennekrosen beiakuten erworbenen hamolytischen Anamien. Schweiz Med Wochenschr. 1955;85:905-909.

4. Remuzzi G. HUS and TTP: variable expression of a single entity. Kidney Int. 1987;32:292-308.

5. Kaplan BS, Drummond KN. The hemolytic-uremic syndrome is a syndrome [editorial]. N Engl J Med. 1978;298:964-966.

6. Wardle EN. "Thrombotic micro-angiopathy" [letter]. Clin Nephrol. 1983; 20:323.

7. Furlan M, Robles R, Galbusera M, et al. von Willebrand factor-cleaving

protease in thrombotic thrombocytopenic purpura and the hemolytic-uremic syndrome. N Engl Med. 1998;339:1578-1584.

8. Tsai HM, Lian EC. Antibodies to von Willebrand factor-leaving protease in acute thrombotic thrombocytopenic purpura. N Engl J Med. 1998;339:1585-1594.

9. Gianantonio CA, Vitacco M, Mendilaharzu F, Gallo GE, Sojo ET. The hemolytic-uremic syndrome. Nephron. 1973;11:174-192.

10. Sheth Ki, Swick HM, Haworth N. Neurological involvement in hemolyticuremic syndrome. Ann Neurol 1986;19:90-93.

11. Baehr G, Klemperer P, Schifrin A. An acute febrile anemia and thrombocytopenic purpura with diffuse platelet thromboses of capillaries and arterioles. Trans Assoc Am Physicians. 1936;65:43-58.

12. Asada Y, Sumiyoshi A, Hayashi IF; Suzumiya J, Kaketani K. Immunohistochemistry of vascular lesion in thrombotic thrombocytopenic purpura, with special reference to factor VIII related antigen. Thromb Res. 1985;38:469-479.

13. Ridolfi RL, Hutchins GM, Bell WR. The heart and cardiac conduction system in thrombotic thrombocytopenic purpura: a clinicopathologic study of 17 autopsied patients. Ann Intern Med. 1979;91:357-363.

14. Morel-Maroger L, Kanfer A, Solez K, Sraer JD, Richet G. Prognostic importance of vascular lesions in acute renal failure with microangiopathic hemolytic anemia (hemolytic-uremic syndrome): clinicopathologic study in 20 adults. Kidney Int. 1979;15:548-558.

15. Moake JL. Moschcowitz, multimers and metalloprotease [editorial]. N Engl J Med. 1998;339:1629-1631.

16. Bell WR, Braine HG, Ness PM, Kickler TS. Improved survival in thrombotic thrombocytopenic purpura-hemolytic uremic syndrome: clinical experience in 108 patients. N Engl Med. 1991;325:398-403.

17. Baker KR, Moake JL. Thrombotic thrombocytopenic purpura and the hemolytic uremic syndrome. Curr Opin Pediatr 2000;12:23-28.

Gregory A. Hosler, MD, PhD; Ana M. Cusumano, MD; Grover M. Hutchins, MD

Accepted for publication January 30, 2003.

From the Department of Pathology, The Johns Hopkins Medical Institutions, Baltimore, Md.

Reprints: Grover M. Hutchins, MD, Department of Pathology, The Johns Hopkins Hospital, 600 N Wolfe St, Baltimore, MD 21287 (e-mail: ghutchi@jhmi.edu).

Copyright College of American Pathologists Jul 2003
Provided by ProQuest Information and Learning Company. All rights Reserved