Paroxysmal nocturnal hemoglobinuria and the risk of venous thrombosis: Review and recommendations for management of the pregnant and nonpregnant patient
Jg. Ray et al., Paroxysmal nocturnal hemoglobinuria and the risk of venous thrombosis: Review and recommendations for management of the pregnant and nonpregnant patient, HAEMOSTASIS, 30(3), 2000, pp. 103-117
Background Paroxysmal nocturnal hemoglobinuria is a rare, clonal primitive
hematopoietic cell disorder, often affecting middle-aged adults, including
women of reproductive age. Major morbidity and mortality with this disease
are often ascribed to the development of venous thromboembolism. We reviewe
d the current literature on the risk of venous thrombosis among nonpregnant
and pregnant patients, and generated recommendations for the prevention of
venous thromboembolism, as well as duration of treatment for affected pati
ents who develop thrombotic disease. Methods: We searched Medline for paper
s published between January 1966 and April 1999. We also requested relevant
unpublished data from speakers who attended a recent international worksho
p of paroxysmal nocturnal hemoglobinuria. References from all primary data
and review publications were also examined. Only English language publicati
ons were included. Event rates for venous thromboembolism and death were po
oled using a random effect technique. Reports of paroxysmal nocturnal hemog
lobinuria during pregnancy were summarized using descriptive statistics onl
y. Results: Thirteen retrospective studies of paroxysmal nocturnal hemoglob
inuria in nonpregnant individuals were found. The rates of venous thrombosi
s varied considerably, but were reported to affect 14.4% of all individuals
[95% confidence interval (CI) 7.6-25.5]. Among patients from western natio
ns, venous thromboembolism seemed to develop at a higher rate (30.3%, 95% C
I 26.1-34.9). The majority of venous thromboembolic events were intra-abdom
inal, principally within the hepatic and mesenteric veins. The likely cause
of death among patients with paroxysmal nocturnal hemoglobinuria was descr
ibed in nine studies: 22.2% of fatalities were due to venous thrombosis (95
% CI 11.8-38.0), more commonly in western countries (event rate 37.2%, 95%
CI 21.6-56.0). Another 20 published reports described the outcome of 33 pre
gnant women with paroxysmal nocturnal hemoglobinuria. Two women developed v
enous thromboembolism during pregnancy and another 2 during the postpartum
state for a combined event rate of 12.1% (95% CI 3.4-25.2), three of which
resulted in death. The all-cause mortality rate was 20.8% (95% CI 7.3-39.0)
. Both anemia (event rate 72.7%, 95% CI 56.5-86.3), and thrombocytopenia (e
vent rate 27.3%, 95% CI 13.7-43.5) were common, often necessitating red cel
l or platelet transfusions. Almost half of all infants (54.8%, 95% CI 36.1-
72.7) were delivered preterm, and had a mean live birth weight of 2,800 g.
Th ree of 34 reported births ended in death (perinatal mortality rate 8.8%,
95% C 1.9-23.7). Conclusion: In accordance with the apparently high rate o
f venous thrombosis among pregnant and nonpregnant individuals with paroxys
mal nocturnal hemoglobinuria, especially for fatal thrombosis, we developed
practical recommendations for the prevention and treatment of venous throm
boembolic disease in these groups. Copyright (C) 2000 S. Karger AG, Basel.