Ann Hematol (1991) 63:101-106

Annals of

Hematology 9 Springer-Verlag 1991

Original article Incidence and clinical risk factors for bleeding and thrombotic complications in myeloproliferative disorders A retrospective analysis of 260 patients A. Wehmeier, I. Daum, H. Jamin, and W. Schneider Medizinische Klinik und Poliklinik der Heinrich-Heine-Universit/~t,Abteilung for H~matologie, Onkologie und klinische Immunologie, Moorenstrasse 5, W-4000 D0sseldorf, FRG Received March 1, 1991/Accepted May 31, 1991

Summary. Bleeding and thrombosis are frequent complications in myeloproliferative disorders (MPD) and are associated with severe organ damage and a high mortality. Elevated platelet count, elevated hematocrit, and patient age are regarded as risk factors for bleeding and thromboembolic events in MPD, although the significance of these parameters was not confirmed by clinical studies. We retrospectively analyzed vascular complications in 260 patients with MPD and tried to identify parameters predictive for bleeding and thromboembolic events. Our cohort consisted of 115 patients with chronic myeloid leukemia (CML), 84 patients with polycythemia vera (PV), 26 with essential thrombocythemia (ET), 25 with osteomyelofibrosis (OMF), and 10 patients with unclassifiable MPD. During a median follow-up period of 31 months, 126 patients with chronic MPD suffered bleeding or thrombotic events. Bleeding was observed in 57% of patients with OMF, 23 o70 with PV, 20~ with chronic phase CML, and 16~ with ET. Thrombotic events were most common in patients with PV (36~ of patients), followed by ET, OMF, and chronic phase CML (20O7o, 17O7o,and 6~ of patients, respectively). Recurrent thrombotic episodes frequently occurred in patients with PV and ET, whereas patients with OMF often had more than one bleeding event. Thirty patients died of thrombohemorrhagic complications during follow-up. Multivariate analysis, including all patients with chronic MPD, revealed that elevated red blood cell count, higher hemoglobin level, and increased percentage of segmented neutrophils at the time of diagnosis were associated with thrombosis, whereas patients with bleeding complications were characterized by low red cell count, lower hemoglobin, and a lower percentage of segmented neutrophils. However, when analyzed by MPD subgroup, none of these parameters retained a predictive value for bleeding or thrombotic events. Moreover, elevated platelet count and patient age Address for correspondence: A. Wehmeier, Medizinische Klinik

und Poliklinik der Universit~t, Abteilung for H~matologie, Onkologie und klinische Immunologie,Moorenstrasse5, W-4000D0sseldorf 1, FRG

were not risk factors for bleeding complications. Thrombotic events were less frequent in patients below the age of 40, and were increased in patients aged 70 and above. However, this was primarily due to the high percentage of elderly patients in subgroups mainly affected by thrombosis (PV and ET). In most MPD subgroups, the rate of bleeding and thrombosis was highest just before and during the first months after diagnosis, and declined thereafter. Thrombohemorrhagic complications were less frequent after phlebotomy in PV and after therapy with alkylating agents in CML. The institution of cytoreductire therapy soon after the diagnosis was made may explain the reduced incidence of complications later in the disease. We conclude that morbidity and mortality from thrombohemorrhagic complications are high in myeloproliferative disorders. Subgroup and stage of the disease and previous complications are helpful in estimating the risk of a bleeding or thrombotic event. Conventional cytoreductive therapy may result in a reduction of these complications.

Key words: Myeloproliferative disorders - Bleeding Thrombosis - Clinical risk factors

Introduction Morbidity and mortality in myeloproliferative disorders (MPD) are related mainly to bleeding and thrombotic events. However, many clinical studies on MPD have not evaluated thrombohemorrhagic complications or have analyzed only selected subgroups, mainly polycythemia rubra vera (for review see [24]). Platelet function tests may contribute to the identification of subgroups at increased risk of bleeding or thrombotic complications [7, 22, 28] but are unable to predict an event in an individual patient [29]. Several risk factors for thrombohemorrhagic complications in MPD have been proposed, including patient age [15], increased platelet count [18], and elevated hematocrit. These parameters are used clinically as indi-

102 cators for t h e r a p e u t i c interventions, a l t h o u g h their validity is uncertain. In order to identify risk factors for bleeding a n d t h r o m b o t i c events, we retrospectively investigated the incidence, l o c a l i z a t i o n a n d o u t c o m e o f t h r o m b o h e m o r rhagic c o m p l i c a t i o n s in 260 patients with m y e l o p r o l i f e r a tive disorders. We a n a l y z e d the r e l a t i o n o f these c o m p l i cations to clinical a n d l a b o r a t o r y p a r a m e t e r s at the time o f diagnosis a n d evaluated the influence o f cytoreductive therapy.

Patients and m e t h o d s Patients with myeloproliferative disorders were identified by reviewing the bone marrow registry of the Department of Hematology, Oncology and Clinical Immunology at the University of Diisseldorf in the years 1976-1986. Of 331 patients with suspected MPD, we were able to confirm the diagnosis and obtain follow-up data on 260. One hundred and fifteen patients were classified as chronic myeloid leukemia (CML) based on a complete blood count, leukocyte alkaline phosphatase score, the results of a bone marrow biopsy, and the clinical course of the disease. The Philadelphia chromosome was determined in 52 patients and was positive in 46. Patients in blast crisis were analyzed separately unless otherwise stated (e.g., survival curve). The diagnosis of polycythemia rubra vera (PV, 84 patients), essential thrombocythemia (ET, 26 patients), and osteomyelofibrosis (OMF, 25 patients) was made according to established clinical criteria [6, 16,20] and the results of a bone marrow biopsy. Ten patients were regarded as unclassified MPD because they did not fulfill the diagnostic criteria of the defined subgroups. Median follow-up time was 31 months (range: 1 - 227 months, CML: 28 months, PV: 46 months, ET: 41 months, OMF: 31 months), and short follow-up indicates early death. Patients were not treated according to a prospectively defined protocol. However, most patients with CML received busulfan or hydroxyurea. Patients with PV were usually treated by phlebotomy, and alkylating agents were administered to some later in their disease. More than half of the patients with ET were not treated with cytoreductive therapy, and patients with OMF often received supportive care or corticosteroids (Table 1).

Table 1. Treatment in 260 patients with chronic myeloproliferative disordersa CML

n

BS Mono 59 HU Mono 7 BS/HU 17 BS/SR 9 BS/other 10 Other 9 ASA

5

PV PT Mono BS Mono MP Mono CA Mono PT/BS PT/MP PT/CA pT/32p SE or SR ASA

n 35 2 1 1 13 11 7 4 6 19

ET

n

OMF

BS Mono MP Mono CA Mono

5 5 1

ASA

8

C Mono 6 BS Mono 2 MP Mono 2 SE 4 Other 3 ASA

n

Data were analyzed on a microcomputer (Apple Macintosh, Cupertino, CA, USA) using StatView (Abacus Concepts, Berkeley, CA, USA) and Systat 5.0 (Evanston, IL, USA) statistical software.

Results

Incidence and localization of bleeding and thrombotic complications T h r o m b o h e m o r r h a g i c c o m p l i c a t i o n s were observed in 126 patients with chronic M P D d u r i n g follow-up (Table 2). Bleeding was m o s t frequent in p a t i e n t s with O M F (57% o f patients, recurrent b l e e d i n g episodes: 22%), whereas t h r o m b o t i c c o m p l i c a t i o n s prevailed in patients with PV (36% o f patients, recurrent t h r o m b o s i s : 18%). I n ET, t h r o m b o t i c c o m p l i c a t i o n s (20% o f patients, recurrent events: 12%) were m o r e frequent t h a n bleeding episodes (16% o f patients). I n c h r o n i c - p h a s e C M L the incidence o f b l e e d i n g was low (20% o f patients), a n d t h r o m b o s i s was a rare event. O n l y 6.5% o f M P D patients, m a i n l y those with PV a n d O M F , h a d b o t h bleeding a n d t h r o m b o t i c episodes (Table 2). F o r t y - n i n e C M L patients went into blast crisis, a n d one p a t i e n t with PV a n d one with O M F developed acute l e u k e m i a d u r i n g follow-up. Two patients in blast crisis suffered f r o m t h r o m b o s i s a n d 21 p a t i e n t s h a d (often multiple) b l e e d i n g episodes. O n l y two C M L p a t i e n t s bleeding in blast crisis h a d also experienced bleeding in chronicphase CML. There were 88 d o c u m e n t e d events o f bleeding a n d 79 t h r o m b o t i c events in patients with chronic M P D . G a s t r o intestinal b l e e d i n g was the m o s t c o m m o n c o m p l i c a t i o n (22 events), followed by skin a n d m u c o s a l bleeding (18 events) a n d epistaxis (11 events), Severe bleeding c o m p l i cations were observed in 37 cases (11 episodes o f intracerebral bleeding, ten cases o f p o s t o p e r a t i v e bleeding c o m p l i c a t i o n s , nine cases o f o r g a n bleeding u n r e l a t e d to surgery, a n d seven cases o f retinal bleeding). D e e p vein t h r o m b o s i s (23 events), cerebral i n f a r c t i o n or i s c h e m i a (18 events), a n d m y o c a r d i a l i n f a r c t i o n (13 cases) were c o m m o n t h r o m b o t i c c o m p l i c a t i o n s . In a d d i t i o n , there were eight cases o f p u l m o n a r y e m b o l i s m , five splenic infarc-

Table 2. Incidence of bleeding and thrombosis in subgroups of chronic myeloproliferative disorders Bleeding

2

a Therapy of patients in blast crisis is not listed. Treatment modalities separated by "/" were administered successively during the course of the disease. CML, Chronic myeloid leukemia; PV, polycythemia vera; ET, essential thrombocythemia; OMF, osteomyelofibrosis; Ara-C, cytosinearabinoside; ASA, acetylsalicylic acid; BS, busulfan; C, corticosteroids; CA, cblorambucil; HU, hydroxyurea; Mono, monotherapy; MP, melphalan; PT, phlebotomy; SE, splenectomy; SR, splenic irradiation; 32p, radiophosphorus

CML n = 115 PV n = 84 ET n = 26 OMF n = 25

Thrombosis

Both

20 (3)a

6 (2)

1

23 (5)

36 (18)

6

16 (0)

20 (12)

4

57 (22)

17 (9)

9

a Numbers for bleeding and thrombotic events represent 070 of patients with complications in each subgroup. Numbers in brackets indicate recurrent complications

103 tions, six cases of peripheral gangrene, and two cases of major arterial embolism. Portal vein thrombosis, a complication regarded as a characteristic feature of MPD, was observed in only four cases. Patients with ET mostly suffered from microcirculatory disorders (burning hands and feet), but no ET patient died of thrombosis. In addition to these complications, 37 bleeding events and two splenic infarctions were observed in CML blast crisis.

Incidence of bleeding 0,7 0,6 0 5

OMF

0,40,3

0,1 0,0

In 33 cases, the diagnosis of MPD was established because of a bleeding or thrombotic event. The risk of such a complication was highest at the time of diagnosis (_+ 1 month). The incidence of bleeding at that time was 5.8% per patient month in ET, 5.4% in CML, 3.2~ in PV, and 2.2% in OMF. In addition, the initial risk of thrombosis was 8.2% per patient month in PV, 5.8% in ET, 2.2% in OMF, and 0.4~ in CML. The rate of bleeding complications in CML and PV and the rate of thrombotic events in PV declined a few months after diagnosis (Fig. 1). In contrast, the rate of bleeding in OMF was almost constant during the follow-up period. In CML, the incidence of bleeding increased sharply when patients entered the accelerated phase or blast crisis (Fig. 1 a). Patient age, sex, liver and spleen size, and standard laboratory parameters at the time of initial diagnosis were included in an analysis of risk factors for bleeding and thrombotic complications. When all patients except those in blast crisis were pooled, patients with thrombotic events showed significantly elevated red blood cell count, hemoglobin, and percentage of segmented neutrophils at the time of diagnosis, whereas patients with bleeding complications had significantly decreased values for these three parameters, compared with patients without complications (p < 0.05, Mann-Whitney U-test). Univariate discriminant analysis revealed that red blood cell count, hemoglobin, and percentage of segmented neutrophils (at the time of diagnosis) separated a group of patients with thrombotic complications from those without complications and patients with bleeding from those without bleeding episodes. Using multivariate discriminant analysis, the sensitivity of this set of parameters in correctly identifying patients with thrombosis was 67% and in identifying patients with bleeding complications 62%. However, when analyzed by MPD subgroup, none of the parameters had a predictive value in identifying patients with bleeding or thrombosis. The evaluation of clinical and laboratory data obtained shortly after a bleeding or thrombotic event did not reveal any additional risk factors. Patient age was not a significant risk factor for bleeding or thrombosis in this series. The age distribution of patients at the time of a thrombohemorrhagic event almost paralleled the age distribution of patients at the time of diagnosis (Fig. 2). However, thrombotic events were less frequent in patients below the age of 40 (15% of patients were below 40 years of age at diagnosis but only 4% of thrombotic events occurred in that age

I

PV

0,2

Risk factors for thrombohemorrhagic events: subgroup and course of the disease

V

i 1000

i

30'00

2000

40'00

Days after diagnosis

0,8" Incidence of thrombosis 0,7" 0,60,5PV 0,4. 0,3-

I ET

O ' 2 ~ ~ S L 0,1 0,0 I' 0

i

1000

2000

i

3000 40'00 Days after diagnosis

Fig. 1 a, b. Cumulative incidence of bleeding (a) and thrombotic (b) complications in subgroups of myeloproliferative disorders (including CML blast crisis)

group), whereas there was some excess thrombosis in patients over 69 years of age.

Therapy Twenty CML patients treated with busulfan experienced a bleeding episode in the chronic disease phase. Bleeding occurred before therapy was initiated in 13 patients and in seven patients thereafter. Of eight CML patients with a thrombotic event, three had thrombosis before and five after initiation of busulfan therapy. Laboratory data obtained a median of 87 days after the introduction of busulfan showed a reduction of mean WBC by 74%, with a relative reduction of immature precursors and an increase in neutrophils and lymphocytes in the blood count (data not shown). Of 29 polycythemic patients with thrombotic complications, ten suffered from thrombosis before and 12 after venesection therapy was begun. Phlebotomy resulted in a significant reduction of mean hematocrit by 24%, whereas mean platelet count increased by 22%.

104 100

70 Patients surviving

60

(%) 50 Number of patients

I=T 80

4O 60

I

3O 2O

L

l_

40

10 20

0 69 i

i

1000

2000

30~00

4000

Days after diagnosis

60T b

Fig. 3. Comparative survival of patients with myeloproliferative disorders (according to the method of Kaplan and Meier [14])

50 84

Number of events 30

[ ] Thrombosis [ ] Bleeding

2O 10

Table 3. Frequency and localization of lethal thrombohemorrhagic

complications in 30 patients with myeloproliferative disorders who died of bleeding or thrombosis during follow-up Bleeding

69 Patient age (years)

Fig. 2 a, b. Comparison of the age distribution of 260 patients with myeloproliferative disorders at diagnosis (a) to the age distribution of patients at the time of a bleeding (n = 88) or thrombotic (n = 79) event (b) in chronic MPD (excluding CML blast crisis)

Acetylsalicylic acid (ASA) was administered to 34 patients (Table 1). In six patients, bleeding episodes occurred during ASA therapy.

Survival and cause o f death During the follow-up period, mortality among patients with C M L was 55% (63 patients, 18 in chronic phase and 45 in blast crisis), 23% in PV (19 patients), 53% in O M F (13 patients), and only 8% in ET (2 patients). Comparative Kaplan-Meier plots of patient survival are depicted in Fig. 3. In 30 cases, death was caused by bleeding or thrombotic complications. Intracerebral hemorrhage was the most frequent cause of death in these patients, followed by gastrointestinal bleeding and pulmonary embolism (Table 3).

Discussion

Bleeding or thrombotic events were observed in 48% of patients with chronic MPD. This is in agreement with other larger series ( 1 - 4 , 8 - 1 0 , 12, 19, 2 5 - 2 7 ) , where the incidence of thrombohemorrhagic complications varied between 30% and 70%. The frequency of bleeding and

n

CML, chronic phase: cerebral bleeding 3 retroperitoneal bleeding 1 CML, blast crisis: cerebral bleeding 10 subdural hematoma 1 bleeding after splenectomy 1 diffuse bleeding (DIC) 1 gastrointestinal bleeding 1 PV: cerebral bleeding 2 abdominal bleeding 1 OMF: gastrointestinal bleeding cerebral bleeding diffuse bleeding Total:

Thrombosis

n

pulmonary embolism

1

pulmonary embolism cerebral infarction myocardial infarction

2 1 1

2 1 1 25

n, Number of patients; DIC, disseminated intravascular coagulation

thrombosis probably depends mainly on patient selection, length of follow-up, and also on the intensity of therapeutic intervention. Characteristic of our patient population was the evaluation of unselected M P D patients who presented to one center over a 10-year period. The frequency of bleeding and thrombosis in our cohort was determined primarily by the distribution of M P D subgroups in patients referred to our department, which may roughly reflect the distribution in the general population. Patients in the largest subgroup (CML) were rarely affected by thrombosis. The incidence of bleeding was also moderate in chronic-phase C M L (20% of patients), but rose to 53% of patients in accelerated phase and blast crisis. Patients in blast crisis were evaluated separately be-

105 cause factors unrelated to thrombohemorrhagic complications in chronic phase, namely thrombocytopenia (platelet count < 20x 109/1 in 30 bleeding episodes), and procoagulant and fibrinolytic activities of blast ceils (review in [17]) may be the predominant causes of hemorrhage in advanced disease. The group of patients with OMF was small but showed the highest incidence of bleeding (57% of patients), and almost half of these patients bled more than once. Thrombosis was the major complication in PV and ET, and here also half of the affected patients suffered more than one episode. The concurrence of bleeding and thrombotic events in one patient is considered a characteristic feature of MPD but was found in only 17 patients (6.5%), mainly with OMF and PV. In an effort to identify independent risk factors for bleeding and thrombosis in chronic MPD, we evaluated clinical and laboratory parameters at the time of initial diagnosis. Using multivariate analysis, elevated red blood cell count and hemoglobin and increased percentage of segmented neutrophils were associated with thrombosis, whereas patients with bleeding complications had decreased values of these parameters. These three variables combined were able to identify patients with thrombosis or bleeding complications, although the sensitivity was rather low (thrombosis: 67%, bleeding: 62%). However, when analyzed by MPD subgroup, none of the parameters (or any combination) was able to discriminate patients with bleeding or thrombosis from patients without complications. In retrospective studies such as ours, the frequency of complications may be underestimated through incomplete observation, and increased variability of data due to several sources and retrospective acquisition may influence the significance of statistical tests. In addition, the number of patients in some subgroups (ET, OMF) may be too small for statistical analysis. However, the distribution of bleeding and thrombosis among MPD subgroups and the analysis of potential risk factors indicate that the information required for the determination of MPD subgroup contains most clinical data predictive for thrombohemorrhagic complications, so MPD subgroup itself may be one of the most important prognostic factors for bleeding and thrombotic events. In agreement with most other reported work [2, 8, 10, 13], we did not find a relationship between elevated platelet count at the time of diagnosis and subsequent bleeding or thrombosis, and platelet count was not increased at the time of thrombotic or hemorrhagic events compared with the time of diagnosis. Patient age has also been regarded as a risk factor for bleeding and thrombosis in MPD. We were not able to confirm the findings of Kessler et al. [15], who reported bleeding episodes only in patients over 50 years of age. In contrast, the age distribution of patients at the time of a bleeding event paralleled the age distribution of the total cohort at the time of diagnosis. The frequency of thrombotic events was decreased below the age of 40 and increased in patients over 69 years of age. A low rate of thrombohemorrhagic complications in young MPD patients was also observed by others [11, 15,21,23]. However, the age difference between patients with and those without thrombosis was

not statistically significant in our series, and age was not a risk factor for thrombosis in discriminant analysis. The age distribution of patients with thrombosis in our series probably just reflects the age distribution of the MPD subgroups mainly affected by thrombotic complications (PV and ET, Fig. 2). The Polycythemia Vera Study Group reported that among patients with PV treated with phlebotomy alone, thrombotic events were especially common in elderly patients and those with a history of previous thrombotic events, or in patients who required multiple phlebotomies [5]. Age, a previous thrombotic event, and inadequate treatment of thrombocytosis were also reported as risk factors for thrombosis in a large cohort of patients with ET [10]. Our data confirm a high percentage of recurrent thrombotic episodes in PV and ET (Table 3), but we were unable to evaluate the frequency of phlebotomy retrospectively. In our polycythemic patients we saw a reduction of thrombohemorrhagic episodes after cytoreductive therapy, especially after phlebotomy. In PV, there was one complication in 7.4 patient months before venesection and one in 78 patient months after the initiation of venesection. In CML patients treated with alkylating agents the reduction was smaller: one complication in 24.1 patient months before, versus one in 44.3 patient months after chemotherapy. Further analysis of the influence of various therapeutic strategies on thrombohemorrhagic complications was not carried out because therapy was not prospectively defined. The present study emphasizes the importance of thrombohemorrhagic complications as the main cause of morbidity and mortality in myeloproliferative disorders. On the basis of subgroup and stage of the disease and a history of previous events, the risk of bleeding or thrombosis can be estimated. Conventional cytoreductive therapy (alkylating agents and venesection) led to a reduction in the rate of thrombohemorrhagic complications, irrespective of a reduction in platelet count.

Acknowledgements. We wish to express our gratitude to numerous colleagues who referred patients or contributed follow-up data to this study. References

1. Anger B, Haug U, Seidler R, Heimpel H (1989) Polycythemia vera. A clinical study of 141 patients. Blut 59:493-500 2. Barbui T, Cortelazzo S, Viero P, Bassan R, Dini E, Semeraro N (1983) Thrombohemorrhagic complications in 101 cases of myeloproliferative disorders: relationship to platelet number and function. Eur J Cancer Clin Oncol 19:1593-1599 3. Bellucci S, Janvier M, Tobelem G, Flandrin G, Charpak Y, Berger R, Boiron M (1986)Essential thrombocythemias. Clinical evolutionary and biological data. Cancer 58:2440-2447 4. Berk PD, Goldberg JD, Silverstein MN, Weinfeld A, Donovan PB, Ellis JT, Landaw SA, Laszlo J, Najean Y, Pisciotta AV, Wasserman LR (1981) Increased incidence of acute leukemia in polycythemia vera associated with chlorambucil therapy. N Engl J Med 304:441-447 5. Berk PD, Goldberg JD, Donovan PB, Fruchtman SM, Berlin NI, Wasserman LR (1986) Therapeutic recommendations in polycythemia vera based on polycythemia vera study group protocols. Semin Hematol 23:132-143 6. Berlin NI (1975) Diagnosis and classification of the polycythemias. Semin Hematol 12:339-351

106 7. Boneu B, Nouvel C, Sie P, Caranobe C, Combes D, Laurent G, Pris J, Bierme R (1980) Platelets in myeloproliferative disorders. I. A comparative evaluation with certain platelet function tests. Scand J Haematol 25:214-220 8. Buss DH, Stuart J J, Lipscomb GE (1985) The incidence of thrombotic and hemorrhagic disorders in association with extreme thrombocytosis: An analysis of 129 cases. Am J Hematol 20:365-372 9. Chievitz E, Thiede T (1962) Complications and causes of death in polycythemia vera. Acta Med Scand 172:513-523 10. Cortelazzo S, Viero P, Finazzi G, D'Emilio A, Rodeghiero F, Barbui T (1990) Incidence and risk factors for thrombotic complications in a historical cohort of 100 patients with essential thrombocythemia. J Clin Oncol 8:556-562 11. Davis RB (1985) Acute thrombotic complications of myeloproliferative disorder in young adults. Am J Clin Pathol 84: 180-185 12. Fenaux P, Simon M, Caulier MT, Lai JL, Goudemand J, Bauters F (1990) Clinical course of essential thrombocythemia in 147 cases. Cancer 66:549-556 13. Hehlmann R, Jahn M, Baumann B, KOpcke W (1988) Essential thrombocythemia. Clinical characteristics and course of 61 cases. Cancer 61:2487-2496 14. Kaplan EL, Meier P (1958) Nonparametrie estimation from incomplete observations. J Am Stat Assoc 53:457-481 15. Kessler CM, Klein HG, Havlik RJ (1982) Uncontrolled thrombocytosis in chronic myeloproliferative disorders. Br J Haematol 50:157-167 16. Laszlo J (1975) Myeloproliferative disorders (MPD): myelofibrosis, myelosclerosis, extramedullary hematopoiesis, undifferentiated MPD, and hemorrhagic thrombocythemia. Semin Hematol 12:409-432 17. Lisiewicz J (1988) Disseminated intravascular coagulation in acute leukemia. Semin Thromb Hemost 14:339-350 18. Mason JE, DeVita VT, Canellos GP (1974) Thrombocytosis in

19.

20.

21.

22.

23.

24. 25. 26. 27.

28.

29.

chronic granulocytic leukemia: incidence and clinical significance. Blood 44:483-487 Monfardini S, Gee T, Fried J, Clarkson B (1973) Survival in chronic myelogenous leukemia: influence of treatment and extent of disease at diagnosis. Cancer 31:492-501 Murphy S, Iland H, Rosenthal D, Laszlo J (1986) Essential thrombocythemia: an interim report from the polycythemia vera study group. Semin Hematol 23:177-182 Najean Y, Mugnier P, Dresch C, Rain J (1987) Polycythemia vera in young people: an analysis of 58 cases diagnosed before 40 years. Br J Haematol 67:285-291 Pareti FI, Gugliotta L, Mannucci L, Guarini A, Mannucci PM (1982) Biochemical and metabolic aspects of platelet dysfunction in chronic myeloproliferative disorders. Thromb Haemost 47:84-89 Randi ML, Fabris F, Girolami A (1990) Thrombocytosis in young people: evaluation of 57 cases diagnosed before the age of 40. Blur 60:233-237 Schafer AI (1984) Bleeding and thrombosis in the myeloproliferative disorders. Blood 6 4 : 1 - 1 2 Stroebel CF, Hall BE (1951) Evaluation of radiophosphorus therapy in primary polycythemia. JAMA 146:1301-1307 Videbaek A (1950) Polycythemia vera. Course and prognosis. Acta Med Scand 138:179-187 Wasserman LR (1954) Polycythemia vera - its course and treatment: relation to myeloid metaplasia and leukemia. Bull N Y Acad Med 30:343-375 Wehmeier A, Scharf RE, Fricke S, Schneider W (1989) Bleeding and thrombosis in chronic myeloproliferative disorders: relation of platelet disorders to clinical aspects of the disease. Haemostasis 19:251-259 Wehmeier A, Fricke S, Scharf RE, Schneider W (1990) A prospective study of haemostatic parameters in relation to the clinical course of myeloproliferative disorders. Eur J Haematol 45:191-197