Thrombotic Microangiopathy After Allogeneic
Hematopoietic Stem Cell Transplantation:
An Autopsy Study
Koushan Siami,1 Kiarash Kojouri,2 Karen K. Swisher,3 George B. Selby,2 James N. George,2
and Zoltan G. Laszik1,4
Background.Posttransplantation thromboticmicroangiopathy (PTMA) is a complication of allogeneic hematopoietic
stem cell transplantation (HSCT). However, limited autopsy data are available, and it remains unclear whether PTMA
is a discrete clinical and pathologic entity. The aims of this autopsy study were to determine the correlation between
clinical and pathologic diagnosis of PTMA, to define the precise morphologic spectrum of PTMA, and to seek for
potential etiologic factors.
Methods. The study included 20 consecutive patients with HSCT autopsied at the University of Oklahoma, between
1994 and 2005. Applying strict clinical-laboratory criteria, 6 patients were diagnosed clinically with PTMA and treated
with plasma exchange. Clinical variables, including underlying disease, conditioning regimen, stem cell donor status,
duration and serum level of cyclosporine, infections, and acute graft versus host disease were compared statistically in
patients with histologic signs of PTMA (n
8) with those without PTMA (n
12).
Results.PTMAwas verified histologically in all 6 patients with a clinical diagnosis of PTMAbut only 2 of the 14 patients
who were not clinically diagnosed had histologic evidence of PTMA (P0.0001). Kidneys were affected in all 8 patients
with PTMA, and limited extrarenal involvement by PTMA was observed in 3 of these 8 patients. No statistically
significant differences in relevant clinical and morphologic variables were identified between the PTMA and non-
PTMA groups.
Conclusions.This study documents a strong correlation between the clinical andmorphologic diagnosis of PTMA.The
kidney is the primary target of PTMA, with dominant glomerular and arteriolar involvement. The etiology is likely to
be multifactorial.
Keywords: Posttransplantation thrombotic microangiopathy, Allogeneic hematopoietic stem cell transplantation.
(Transplantation 2008;85: 22–28)
Thrombotic microangiopathy (TMA) has been consideredto be a complication of allogeneic hematopoietic stemcell
transplantation (HSCT) (1–3), which is described as post-
transplantation TMA (4). However, there is little evidence to
establish posttransplantation TMA as a specific clinical or
pathologic entity (5). After allogeneic HSCT, patients may
havemany critical complications,making the clinical diagno-
sis of TMA uncertain (5). This uncertainty is emphasized by
the extreme variability of the reported frequency of the diag-
nosis of posttransplantation TMA after allogeneic HSCT,
ranging from 0.5% to 63.6%, and the use of many different
sets of diagnostic criteria (5). This striking variation may be
caused by the difficult clinical diagnosis of TMA in these pa-
tients. Microangiopathic hemolytic anemia and thrombocyto-
penia, the principal clinical and laboratory diagnostic features of
TMA, can be mimicked by common complications in patients
after HSCT, such as opportunistic infections, chemotherapy
regimen-related toxicity, radiation-related injury, and acute
graft-versus host-disease (GVHD).
In addition to the uncertain clinical data on posttrans-
plantation TMA, detailed morphologic descriptions of autopsy
findings in patients with the clinical diagnosis of posttransplan-
tation TMAare sparse, and the extent andmorphologic severity
of TMA have not been well-characterized. A recent systematic
review identifiedreportsof autopsies inonly35patientswhohad
been diagnosedwithTMAafter allogeneicHSCT; none had sys-
temic TMA; when TMA was described, it was localized to the
kidneys; infection was the most commonly reported cause of
death (5).
To determine whether the pathologic features of TMA
are related to the clinical diagnosis of posttransplantation
TMA, we analyzed the clinical and pathologic data of 20 con-
secutive patients who had autopsies after allogeneic HSCT.
Six of the 20 patients had been clinically diagnosed with post-
transplantation TMA and treated with plasma exchange. The
aims of the studywere (1) to correlate the clinical diagnosis of
posttransplantation TMA with the autopsy findings, (2) to
correlate clinical parameters with autopsy findings of TMA to
identify potential etiologic factors for the pathogenesis of
posttransplantation TMA, (3) to compare the autopsy find-
ings of posttransplantation TMA with the classic morpho-
logic findings of thrombotic thrombocytopenic purpura
(TTP) (i.e., multiple organ involvement by systemic TMA)
and hemolytic uremic syndrome (HUS; i.e., dominant renal
TMA with limited involvement of other organs), and (4) to
1 Department of Pathology, University of OklahomaHealth Sciences Center,
940 Stanton L. Young Blvd., Oklahoma City, OK 73190.
2 Section ofHematology/Oncology,University ofOklahomaHealth Sciences
Center, Oklahoma City, OK.
3 Department of Medicine, University of Oklahoma Health Sciences Center,
Oklahoma City, OK.
4 Address correspondence to Zoltan G. Laszik, Department of Pathology,
University of California San Francisco, 513 Parnassus Ave., Room S566,
San Francisco, CA 94143.
E-mail: zoltan.laszik@ucsfmedctr.org
Received 30 April 2007. Revision requested 28 June 2007.
Accepted 19 September 2007.
Copyright © 2008 by Lippincott Williams & Wilkins
ISSN 0041-1337/08/8501-22
DOI: 10.1097/01.tp.0000297998.33418.7e
22 Transplantation  Volume 85, Number 1, January 15, 2008

analyze the renal morphologic findings of TMA in patients
with posttransplantation TMA to determine which (if any)
are the dominant morphologic patterns (i.e., glomerular ver-
sus arterial).
MATERIALS AND METHODS
Patients
All patients who died and underwent autopsy after re-
ceiving allogeneic HSCT at the Department of Pathology,
University of Oklahoma Health Sciences Center between
April 25, 1994, the date of the first autopsy examination of a
patient after allogeneic HSCT, and December 31, 2005 were
identified for this analysis.
Clinical Data
Age, gender, underlying disease and its status at the
time of HSCT, conditioning regimen, use of total body irra-
diation as part of conditioning regimen, stem cell donor (re-
lated vs. unrelated), level of HLA antigenmismatch, duration
of cyclosporine exposure, peak serum cyclosporine level, sys-
temic cytomegalovirus (CMV) infection, systemic fungal in-
fection, and acute GVHD, its maximum overall grade, and
presence of active GVHD at the time of death were recorded
for each patient.
Systemic CMV infection/reactivation was diagnosed
clinically and required positive CMV antigenemia or viral
culture. Systemic fungal infection was diagnosed clinically
and required positive culture, staining, or histologic evidence
of invasive fungal infection.
Acute GVHD was diagnosed based on clinical criteria
(involvement of skin, gastrointestinal tract, and liver) and, if
available, supported by histology. ActiveGVHDat the time of
death was also separately recorded. GVHD was considered
active if symptoms/signs attributable to GVHD were present
and the patient received immunosuppressive therapy for
treatment of GVHD within 30 days before death. Patients
were classified as TMA and non-TMA groups using indepen-
dent clinical and histologic criteria.
Histologic Data
Hematoxylin and eosin-stained sections of paraffin-
embedded autopsy tissues were reviewed by two pathologists
(K.S. and Z.G.L.) independently for histologic evidence of
thrombotic microangiopathy (TMA) in a blinded fashion,
without clinical information regarding the presence or ab-
sence of a clinical diagnosis of posttransplantation TMA. Sec-
tions from heart, lungs, liver, pancreas, kidneys, adrenals,
thyroid gland, spleen, esophagus, stomach, intestines, blad-
der, bone marrow, and brain were available for histologic
evaluation from all 20 cases. Prostate and testes were also
examined in males and ovaries and uterine tissues in females.
Sampling of lungs included sections from each lobe, at least
one section from both kidneys, and multiple sections from
the brain; the rest of the organs were sampled representa-
tively. Fromeach kidney, periodic acid Schiff-stained sections
were also prepared and examined. Histologic diagnosis of
TMA in the kidney required the presence of thrombotic le-
sions and/or any other classic glomerular or vascular features
of TMA, as listed in Table 1. Detailed morphologic assess-
ment of the kidneys included examination of at least 200 glo-
meruli, 20 arterioles, and 20 interlobular arteries in each case.
The diagnosis of TMA in other organs was based on throm-
botic lesions in the capillaries and/or larger vessels.
Fungal infections suspected on hematoxylin and
eosin-stained slides were confirmed by Grocott methena-
mine silver special stain. Amongpatientswith systemic fungal
infection, the subset of patients who had evidence of angio-
invasive fungal infection (AFI) was further determined. AFI
was defined as presence of fungal hyphae in the lumen of
capillaries or small vessels, and/or traversing blood vessel
walls.
Clinical Classification
Patients who had received plasma exchange for treat-
ment of clinically diagnosed posttransplantation TMA (6)
were assigned to the Clinical TMA group; the remainder of
the patients comprised the Clinical non-TMA group. Neces-
sary criteria for the clinical diagnosis of posttransplantation
TMA were thrombocytopenia and microangiopathic hemo-
lytic anemia without an apparent alternative etiology (3, 7, 8).
The criterion of plasma exchange treatment as the basis for
assignment of patients to Clinical TMA and Clinical non-
TMA groups was established because it indicated a firm
commitment by the physician to the diagnosis of post-
transplantation TMA (7, 8).
Histologic Classification
On the basis of the presence or absence of histologic
evidence of TMA in the kidneys and other organs patients
were also assigned to histologic TMA and histologic non-
TMA groups. This assignment was independent of and with-
out knowledge of the patient’s clinical classification. For the
histologic TMA group, percentages of renal glomeruli, arte-
rioles and interlobular arteries showing features of TMAwere
calculated to assess the extent of involvement. The number of
organs showing histologic features of TMAalsowas recorded.
Statistical Analysis
Association between clinical and histologic classifica-
tions was analyzed with Fisher’s exact test. Histologic TMA
TABLE 1. Histologic features of TMA in the kidney
Glomerular changes
Capillaries and capillary walls
Thickening of the wall and double contours of basement
membranes
Narrowing and closure of capillary lumina
Fragmenting of red blood cells
Fibrin thrombi/fibrinoid necrosis
Mesangium
Edema and fibrillary appearance
Deposition of fibrin and fragmented red blood cells
Mesangiolysis
Vascular changes
Intimal swelling
Fibrinoid necrosis and/or thrombosis
Fragmented red blood cells in the wall
Myointimal cellular proliferation
© 2008 Lippincott Williams & Wilkins 23Siami et al.