Inapparent Polycythemia Vera: An Unrecognized Diagnosis

Inapparent Polycythemia Vera: An Unrecognized Diagnosis T. Lamy, A. Devillers, M. Bernard, A. Moisan, I. Grulois, B. Drenou, L. Amiot, R. Fauchet, P.Y. Le Prise, Rennes, France

PURPOSE: The Polycythemia Vera Study Group (PVSG) has established useful criteria for the diagnosis of polycythemia vera. In some circumstances, an increase of plasma volume (PV) masks that of red cell mass (RCM), with hemoglobin (Hb) and hematocrit (Ht) remaining normal. This defines the concept of inapparent polycythemia. PATIENTS AND METHODS: One hundred and three patients seen in the hematology unit with the diagnosis of polycythemia vera were studied. There were 55 males and 48 females with a median age of 59 years. Ninety-five patients fulfilled the PVSG criteria. Spontaneous erythroid colonies and low serum erythropoietin level confirmed the diagnosis in the 8 other cases. Patients were classified according to Hb and Ht level. RESULTS: Group A consisted of 85 patients with increased Hb and Ht defined, respectively, by Hb ú18 g/dL, Ht ú0.52 in males and Hb ú16 g/dL, Ht ú0.47 in females. Group B included 18 patients (17%) with inapparent polycythemia vera (IPV) defined by a normal Hb and Ht value at diagnosis. In this group, the reasons to perform RCM were as follows: splenomegaly associated with increased platelets and/or leucocytes counts (n Å 8), portal vein thrombosis (n Å 5), increased platelets or leucocytes counts without splenomegaly (n Å 3), and isolated splenomegaly (n Å 2). The two groups were balanced in terms of age, sex, leucocyte, serum iron, and platelet level. Hemoglobin and Ht levels were significantly different between the two groups. The difference between the PV was indeed highly significant. The mean PV increase was / 9.5% (nL õ/20%) in group A versus / 36.3% in group B (P õ0.00005). Red cell mass was not different between the two groups.

From the Department of Hematology (TL, MB, IG, PYLP), Hoˆpital Pontchaillou, Department of Nuclear Medicine, Centre Eugen`e Marquis (AD, AM) and Laboratory of Hematology (BD, LA, RF) Hopˆital Pontchaillou, Rennes, France. Requests for reprints should be addressed to T. Lamy, MD, Service d’He´matologie Clinique, Hopˆital Pontchaillou, 35033 Rennes, France. Manuscript submitted November 12, 1995 and accepted in revised form October 7, 1996.

14

CONCLUSIONS: Increased Hb or Ht should not constitute the sole criteria for RCM determination. In the context of portal vein thrombosis, isolated hyperleucocytosis, thrombocytosis, or splenomegaly, a RCM should be performed. The frequency of IPV remains to be specified but the diagnosis of polycythemia vera is probably underestimated. Q 1996 by Excerpta Medica, Inc. Am J. Med. 1997;102: 14–20

P

olycythemia vera is a clonal disease of the stem cell.1,2 It is a myeloproliferative disorder characterized by an increased total red cell mass (RCM) and often associated with splenomegaly as well as increased leucocytosis and platelets. The Polycythemia Vera Study Group (PVSG) established diagnostic criteria for polycythemia vera that are relatively simple to implement.3 However, in a certain number of cases, this diagnosis is confirmed when the required PVSG criteria are not present.4 Absence of splenomegaly is noted in about 30% of cases and leucocytosis and platelet counts can remain normal.3 Leucocyte alkaline phosphatase (LAP) scoring is becoming increasingly rare in common practice. In vitro cultures of erythroid progenitors and radioimmunoassay of erythropoetin (EPO) can contribute to making a diagnosis of polycythemia vera.4 – 7 These two assays may be useful when there is only isolated elevation of the red cell mass and all the usual causes of secondary polycythemia have been excluded.8,9 Reduced EPO rate is frequently observed in polycythemia vera; however, this level can also be low in secondary polycythemia. The evidence of endogenous erythroid colonies (EEC) represents a better index of polycythemia vera. The development of EEC without the addition of EPO is frequently observed in this disease, but this assay alone cannot be used to make the diagnosis. Firstly, EEC are also seen in significant portions of patients with essential thrombocytopenia and in some patients with chronic myeloid leukemia (CML).10,11 Secondly, EEC are not observed when serum free media is used.12 Nevertheless, this assay is still useful in cases of polycythemia not fulfilling PVSG criteria. Red cell mass measurement has been defined in accordance with international standards either in mL/kg or as a percentage of a theoretical standard.13,14 The diagnosis is confirmed, respectively,

Q1997 by Excerpta Medica, Inc. All rights reserved.

/ 220a 0258 Mp

14

Sunday Jan 05 10:03 PM

0002-9343/97/$17.00 PII S0002-9343(96)00351-8

EL–AJM (v. 102, no. 1)

0258

INAPPARENT POLYCYTHEMIA/LAMY ET AL

for values above 36 mL/kg in men and 32 mL/kg in women, or above 120% of the normal theoretical value. The polycythemia diagnosis is considered when a hematocrit level, respectively, greater than 0.50 in men and 0.45 in women is observed on blood count.15,16 The reference hematocrit is centrifuged hematocrit, but this is not commonly used. The hematocrit used is that calculated from the mean red cell volume and the blood count given by electronic counters.17 Above a hematocrit value of over 0.50 in men and over 0.45 in women, there is a possibility of increased red cell mass linked to the raised hematocrit.18 However, in certain situations, hemoglobin and hematocrit levels can remain normal even though there is a genuine case of polycythemia. The concept of inapparent polycythemia vera (IPV) is therefore defined in cases where increased red cell mass is associated with a marked rise in the plasma volume, which explains why hemoglobin values and hematocrit levels remain normal. This situation was reported in cases of portal hypertension secondary to portal or suprahepatic venous thrombosis.19 The real frequency of IPV is not yet established, and the number of cases of polycythemia vera could be underestimated. In this study we analyzed 103 cases of polycythemia vera recorded in our department between 1988 and 1994. We isolated 18 cases (17%) of IPV and specified the circumstances that led to the study of red cell mass and confirmation of diagnosis.

MATERIALS AND METHODS Patients From 1988 to 1994, 103 consecutive patients seen in our department and diagnosed for polycythemia vera were included in this study. There were 55 males and 48 females, with a median age of 59 years (range 27 to 83) at the time of diagnosis. None of them declared a family history of polycythemia. The initial signs that led to a suspected polycythemia vera diagnosis were as follows: neurological or cardiovascular symptoms (n Å 42), pruritus or headaches (n Å 20), systematic blood examination (n Å 16), portal vein thrombosis or Budd-Chiari syndrome (n Å 15), splenomegaly (n Å 10). In 95 patients, polycythemia vera was diagnosed according to the stringent PVSG criteria as previously described.3 These criteria are grouped into two categories: (1) three major criteria: A1 Å increased red cell mass (males ú 36 mL/kg; females ú32 mL/kg), A2 Å normal arterial oxygen saturation ú92%, and A3 Å splenomegaly (confirmed by ultrasound); and (2) four minor criteria: raised platelet and or leucocyte counts, increased leucocyte alkaline phosphatase or serum vitamin B12. Arterial oxygen saturation was determined in all patients and was superior to 92% in all

Figure. Distribution of the diagnostic criteria of the 103 patients.

cases. Fourteen of these 95 patients were not subjected to RCM measurement owing to a hematocrit ú0.62: the prevalence of an excessive RCM is indeed 100% when the hematocrit is ú0.60.18 These 14 patients showed the other PVSG criteria, either A2 / A3, A2 or A3, and at least two minor criteria. Eight patients showed the two A1 / A2 criteria but no splenomegaly and a single or no minor criteria. In these 8 patients, a diagnosis of polycythemia vera was considered as highly possible owing to the following reasons: (1) In all cases the existence of spontaneous erythroid colonies (CFU-E) with a ratio ranged from 5% to 30% (see materials and methods) was observed; (2) each of them had a decreased EPO level (õ10 IU/L); (3) the bone marrow biopsy performed in 4 of these 8 cases showed a megakaryocytic hyperplasia with dysmegakaryopoeisis, granulocytic and erythroid hyperplasia, and no storage iron; and (4) the karyotype showed absence of Philadelphia chromosome in the 2 cases associated with isolated hyperleucocytosis. The distribution of diagnostic criteria in the 103 patients is illustrated in the Figure. Portal vein thrombosis and Budd-Chiari syndrome in the absence of renal or hepatocellular carcinoma were diagnosed in 16 patients using angiography, ultrasound, and doppler. There was no protein S, protein C, or antithrombine III deficiency in these cases.

Methods On performing red cell mass measurement, Ht was measured according to the micro-hematocrit method. Venous and somatic Ht were calculated from the read hematocrit (read Ht): venous hematocrit (Hv) Å read Ht 1 0.98; somatic hematocrit (Hs) Å Hv 1 0.92. Normal theoretical Ht values vary from one study to another. French consensus conference recommendations and the Marsh study set the limit at 0.50 in males and 0.45 in females.15,16 These upper limits were considered in this study. Red cell mass was measured directly by labelling red blood cells with radioactive tracer chrome (Na2CR51O4) or indirectly by using plasma volume January 1997 The American Journal of MedicineT Volume 102

/ 220a 0258 Mp

15

Sunday Jan 05 10:03 PM

EL–AJM (v. 102, no. 1)

0258

15

INAPPARENT POLYCYTHEMIA/LAMY ET AL

after dilution of human serum albumin labelled with Iodine 125 (SAH.I 125). In the direct method (n Å 80 cases), the Cr-51 labelling method is the reference method recommended by the study group for standardization in hematology.13 A 10 mL blood sample is taken on heparin and red blood cells are isolated by centrifugation and labelled by 800 kBq of Na2CR51O4 for 30 minutes at 377C. Samples are collected 10, 20, and sometimes 40 minutes (in the case of splenomegaly) after injection. Radioactivity is assessed on samples of 1 mL total blood (Q1). As the Qo activity injected is known, RCM is calculated as follows: RCM Å Qo/Q1 1 Hv. In the indirect method (n Å 89 cases), plasma volume measurement is based on the use of serum proteins such as serum albumin labelled with Iodine 125 after injection of 185 kBq. Radioactivity is determined from blood samples taken 10, 20, and sometimes 40 minutes (in the case of splenomegaly) after injection of the tracer (Qo). The plasma is separated from the red blood cells by centrifugation. Plasma volume (PV) is then measured based on the radioactivity of the 1 mL plasma samples (Q1) as follows: PV Å Qo/Q1. After plasma volume has been determined, RCM is calculated in view of the somatic hematocrit (Hs): RCM Å PV /1-Hs. Simultaneous plasma volume and red cell volume measurement is possible after successive injection of the 2 tracers. The results are given in relation to the normal reference values according to the latest ICSH recommendations.20 For the erythroid colony assay, 5/mL bone marrow aspirate and 20 mL peripheral blood were collected in heparinized flasks. Mononuclear cells were separated by FICOLL gradient density centrifugation (d Å 1.077) (Pharmacia, Upsalla, Sweden). Cells were plated at 2.105 cells/mL in a plasmaclot culture system as previously described.21 Cells were cultured at 377C in 5% CO2, in complete ISCOVE Dulbecco’s medium (Gibco-BRL, Life Technologies. Pontoise, France) containing 10% bovine plasma, 10% fetal calf serum, and 10% human serum albumin (not considered as strictly EPO free media), in the presence of 3 IU EPO or without EPO. At, respectively, 7 and 14 days of culture, CFU-E and BFU-E were scored after glutaraldehyde fixation and benzidine staining; CFUE colony was defined as the presence of at least 8 erythroblasts and BFU-E as two or more clusters of at least 50 erythroblasts. The ratio of the number of erythropoietin-independent colonies observed in sera containing assays to the number of colonies in the presence of EPO was defined as the coefficient of spontaneous growth. A ratio above 5% was taken to indicate a primary myeloproliferative disorder. Forty-three patients underwent erythroid colony assay. 16

TABLE I Reasons To Perform Red Blood Cell Volume In Patients with Inapparent Polycythemia Vera Portal vein thrombosis* Isolated splenomegaly Splenomegaly / (platelets ú500 109/L)† / (leucocytes ú 12 109/L)‡ / Both increased platelets/leucocytes‡ Platelets ú500 109/L† without enlarged spleen† Both increased platelets ú500 109/L and leucocytes ú12 109/L without enlarged spleen‡

16

Sunday Jan 05 10:03 PM

5 2 1 2 1

* All of these 5 patients had splenomegaly. † Normal serum iron. ‡ Karyotype normal. Spleen size was determined in each case by clinical examination and ultrasound: In the 15 cases in which a splenomegaly was detected, the spleen size was ú15 cm.

Serum erythropoietin levels were estimated by a radioimmunoassay (EPO-TRAC, Incstar Corp., Sorin, Antony, France). This assay is based on the principle of the competition between unknown EPO concentration in the sample serum and radiolabeled 125 I-recombinant human EPO for a limited quantity of polyclonal goat antiserum that was raised against human urinary EPO.22 Normal EPO values are 10 to 19 UI/L in cases with hematocrit higher than 0.43. Forty-five patients had EPO dosage. Spleen size was determined clinically and by ultrasound in all patients. Spleen length ú12.5 cm was taken to indicate spleen enlargement.

Statistical Study A comparative study was conducted on the clinical and hematological parameters of the two groups with or without inapparent polycythemia vera (IPV) using the chi-square and Mann-Whitney tests.

RESULTS Patients The patients were classified into two groups. Group A consisted of 85 patients with evident polycythemia vera. All patients had increased hematocrit (ú0.52 in males, ú0.47 in females). All but 3 males had an increased Hb level (ú18 g/dL). All but 2 females had Hb ú16 g/dL. Forty-four (52%) had splenomegaly, 4 of whom with a spleen size ú19 cm. In 4 cases a Budd-Chiari syndrome and in 6 cases a portal vein thrombosis represented inaugural manifestations of polycythemia vera. Group B consisted of 18 patients with inapparent polycythemia vera (IPV) defined by a hemoglobin and hematocrit value, respectively, of õ16 g/dL and õ0.46 in females and õ17 g/dL and õ0.50 in males. In this group, the reasons for RCM determination are given in Table I: investigation of portal vein throm-

January 1997 The American Journal of MedicineT Volume 102

/ 220a 0258 Mp

5 2

EL–AJM (v. 102, no. 1)

0258

INAPPARENT POLYCYTHEMIA/LAMY ET AL

bosis, increased leucocyte or platelet counts with or without splenomegaly, and isolated splenomegaly. Clinical and biological characteristics of group B patients are given in Table II. As indicated, all of them fulfilled PVSG criteria. Bone marrow biopsy was performed in 10 of the 18 patients and showed in each case classical histological features of polycythemia vera. The P50 values (normal level 28 { 1 mm Hg) and the concentration of 2-3 DPG (normal level 2 { 0.5 mmole/L) determined in 14 cases (except case 2, 4, 6, 10, 18) were normal. The clinical and biological features of the 103 patients according to their inclusion into the groups A or B are summarized in Table III. The two groups were balanced in terms of age, sex, leucocyte, and platelet levels. Hemoglobin and Ht levels were significantly different in the two groups: the mean of Hb value was 19.1 g/dL versus 15.1 g/dL (P õ0.01) in males, 16.7 g/dL versus 13.7 g/dL (P õ0.01) in females; the mean of Ht was 0.58 versus 0.44 (P õ0.0001) in males, 0.52 versus 0.41 in females (P õ0.001). A statistical difference was also seen between the red cell count, which was lower in group B than in group A. In spite of the difference of Hb and Ht between the two groups, red cell mass measurements did not vary significantly from one group to the other, being, respectively, 48.2 mL/kg versus 43.3 mL/kg in males and 40.1 mL/kg versus 37.3 mL/ kg in females (P Å ns). This is explained by the considerable difference between the plasma volume values between the two groups for both sexes. Indeed, the mean PV increase compared with theoretical standards is / 9.5% in group A versus / 37.2% in group B (P õ0.00001). In group A, 44 of 85 (52%) of patients showed an enlarged spleen versus 15 of 18 (88%) in group B (P õ0.005). However, in each group, there was no correlation between spleen size and PV value (data not shown). In group A, the percentage of patients with an increased PV did not differ between those with or without splenomegaly, respectively, 6 of 44 (14%) versus 7 of 41 (17%). Three patients in group B did not have an enlarged spleen (cases 9, 11, and 14). They had thrombocytosis (546 1 109/L, 581 109/L, 1.400 1 109/L) with essential thrombocythemia (ET) as an initial diagnosis. It emphasizes the necessity of performing a RCM determination in cases of idiopathic thrombocytosis to eliminate a masked polycythemia vera. One patient (case 10) had a large isolated splenomegaly without portal vein hypertension, and the diagnosis of polycythemia vera was confirmed by an increased RCM, an EPO level of 2/U/L, colony erythroid assay, and histological findings. There were 5 cases of portal vein thrombosis that revealed the diagnosis of polycythemia vera, RCM being systematically requested in the face of this diagnosis. In 1 other case, the di-

agnosis of portal vein thrombosis was made 2 years later after the diagnosis of polycythemia vera despite a correction of RCM obtained with hydroxyurea. In the other 9 cases, patients had splenomegaly with increased platelet or WBC counts. A normal karyotype and the increased RCM eliminated the diagnosis of CML and ET, respectively. No difference in the level of serum protein was observed between group A and B, since it has been shown that hyperproteinemia can determine an increase of plasmatic volume. The serum iron did not differ between the two groups. Microcytosis and hypochromia were seen in only 1 case (no. 6). In group B, besides the 6 patients with portal vein thrombosis, 3 patients had vascular manifestations (angina pectoris, pulmonary embolism, dysesthesia).

Red Cell Mass Study Red cell mass was measured in 89 patients. Eighty of these (including 15 of 18 in group B) had RCM measurement concomitantly determined by the two techniques (CR51 and SAH.I125). The other 9 patients had RCM by indirect method. We observed a very good correlation between the two methods of analysis (r Å 0.93, P Å 0.00005).

DISCUSSION Polycythemia vera diagnosis entails recognition of an abnormal increased red cell volume. When hematocrit levels are between 0.50 and 0.60 in males and 0.45 and 0.53 in females, RCM implementation is preferable to confirm the diagnosis of polycythemia and eliminate relative or apparent polycythemia.23 A hematocrit value in excess of 0.60 in males and 0.53 in females almost certainly confirms polycythemia.18 When the hematocrit is under 0.50 in males and 0.45 in females, the probability of polycythemia is very low, and the measurement of RCM is generally considered to be of little use. Very few data have been reported on clinical situations where polycythemia vera was diagnosed when the hematocrit and hemoglobin were within normal ranges. Murphy et al24 stress that a high RCM is extremely rare when the Hb level is under 13 g/dL. In the PVSG series reported by Berlin et al3 on 325 patients, it is noted that 7.8% of these have a hematocrit level ranging from 0.40 to 0.52: however, some of these have previously undergone phlebotomies or had gastrointestinal bleeding. We present our series because of its unexpected frequency of IPV, 18 out of 103 patients, ie, 17%. Cases of IPV may be observed in two circumstances. The first is in cases where an increased plasma mass masks an increased red cell mass. This is observed in cases of portal hypertension secondary to suprahepatic or portal vein thrombosis.19,25 We observed January 1997 The American Journal of MedicineT Volume 102

/ 220a 0258 Mp

17

Sunday Jan 05 10:03 PM

EL–AJM (v. 102, no. 1)

0258

17

18

/ 220a 0258 Mp

18

January 1997 The American Journal of MedicineT Volume 102

Sunday Jan 05 10:03 PM

EL–AJM (v. 102, no. 1)

0258

52 53 42 52 38 28 82 33 45 75 76 42 42 75 67 58 36 56

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

F F F F F F F F F F F M M M M M M M

Sex Y Y Y Y Y Y Y Y N Y N Y Y N Y Y Y Y

Splenomegaly — Y — Y — Y — Y — — — Y Y — — — — —

PVT 4.39 4.94 4.70 3.90 5.40 5.50 4.10 4.90 5.20 4.70 4.63 5.50 5.37 5.20 4.80 5.20 4.70 5.90

RBC 110.9/L 12.5 14 15 12.1 15.9 13.7 11.2 13.3 15.2 14 13.7 14.9 15.5 15.1 13.8 15 14.2 15.7

Hb g/ dL 0.37 0.42 0.44 0.36 0.45 0.40 0.34 0.40 0.45 0.46 0.41 0.45 0.46 0.44 0.40 0.44 0.41 0.49

HT 34.2 33.8 35.6 38.4 34 40.5 43.2 37.3 40.2 46 34 50 43.3 43 42.1 45.9 41.1 60.8

RCM mL/kg 17.3 8.2 9.9 6.4 8.2 8.9 11.3 12.1 11.2 9.1 8.7 8.3 5 12.7 7.8 14.1 2.3 34

WBC 110.9/L 891 498 692 442 920 630 940 451 546 284 581 200 168 1,400 879 207 530 496

Platelets 110.9/L — — — 6% — 34% — 10% 21% 5% — 10% 8% — — — 9% —

ECA* 3 — — — — — — 7 5 4 5 5 5 5 — — 16 5

EPO IU/L

Clinical and Biological Characteristics of Group B Patients with Inapparent Polycythemia Vera

120 114 120 127 109 80 30 150 110 155 — 30 112 — 110 — — —

LAP Score

612 316 340 677 920 420 940 786 — 468 315 576 360 1,195 255 990 — 960

B12 pg/L

28 — 27 — 27 — — 27 27 — 28 27 27 27 27 27 27 —

P50 mm Hg

— 5 — 5 — 5 — 5 — 5 4 5 5 — — 5 — 5

Bone Marrow (grade)†

* Erythroid colony assay results are indicated as the ratio of the number of erythroid colonies (CFU-E) growing without addition of EPO to the number of those growing after addition of EPO. † For bone marrow grade, each of the following lesions was scored 1: erythroblastic hyperplasia, granulocytic hyperplasia, megakaryocytic hyperplasia with dysmegacaryocytopoiesis, no storage iron, marrow fibrosis. ‡ PVT Å portal vein thrombosis; LAP Å leukocyte alkaline phosphatase (score ú100 is elevated; P50 Å partial pressure of oxygen at 50% saturation; RBC Å red blood cells; Hb Å hemoglobin; HT Å hematocrit; RCM Å red cell mass; WBC Å white blood cells; ECA Å erythroid colony assay; EPO Å erythropoietin; Y Å yes; N Å no.

Age

Case

TABLE II

INAPPARENT POLYCYTHEMIA/LAMY ET AL

INAPPARENT POLYCYTHEMIA/LAMY ET AL

TABLE III Clinical and Biological Characteristics of the 103 Patients

Age (yr)* Sex M F Splenomegaly Portal vein thrombosis or Budd-Chiari syndrome Number of patients with Leucocytes ú12 1 109/L Platelets ú500 1 109/L Red blood cell count 109/L* M F Hemoglobin g/dL* M F Hematocrit* M F Red cell mass† mL/kg* M F Increase/theoretical norm (%)‡ Plasma volume* Increase/theoretical norm (%) Serum iron (mmol/L)*

Group A (n Å 85)

Group B (n Å 18)

P Value

61 (27–83)

52 (28–82)

ns

48 (56%) 37 (42%) 44 (52%) 10 (12%)

7 (39%) 11 (61%) 15 (83%) 6 (33%)

ns ns õ0.005 õ0.02

31 (36%) 40 (47%)

5 (28%) 10 (56%)

6.2 (4.9–7.4) 6 (4.2–7.35)

5.2 (4.7–5.9) 4.7 (3.7–5.5)

19.1 (17–23.9) 16.7 (15.5–23.9)

15.1 (13.8–15.7) 13.7 (11.2–15.9)

õ0.01 õ0.01

0.58 (0.52–0.74) 0.52 (0.47–0.67)

0.44 (0.40–0.49) 0.41 (0.31–0.46)

õ0.0001 õ0.001

48.2 (36.5–60.5) 40.1 (32–59) /53 ({26.6)

43.3 (41.1–60.8) 37.3 (34–46) /51.5 ({19)

ns ns ns

/9.5 (011, /61) 13.2 (5–25)

/36.3 (/20, /98) 12.3 (4–30)

õ0.00001 ns

ns ns õ0.0002 õ0.003

Group A includes patients with apparent polycythemia vera (n Å 85); group B includes patients with inapparent polycythemia vera (n Å 18). * Results are indicated as median (range). † Fourteen patients in group A with hematocrit ú62% did not have RCM determination. ‡ Mean //0 SD. ns Å not significant.

5 cases of inaugural portal vein thrombosis associated with splenomegaly where the blood count did not show increased Hb and Ht levels. The search for a myeloproliferative disorder should be systematic in the case of portal or suprahepatic venous thrombosis, for its frequency is estimated at up to 48%.19 In the series by Valla et al19 among 4 patients with portal vein thrombosis and increased PV, the maximum hemoglobin value observed is 14 g/dL. Gastrointestinal bleeding and/or an increase in plasma volume during portal vein hypertension often explains the apparent normal result of the blood count. The second circumstance is observed in the event of a standard iron deficiency during polycythemia vera and is partly explained by the high frequency of gastric ulcers and gastritis estimated at 20% to 30%. Resulting iron deficiency can be accompanied by hypochromia and microcytosis and occasionally a substantial decrease in hemoglobin levels. In this situation, polycythemia vera is not always easy to diagnose especially in the absence of splenomegaly and hyperleucocytosis. Thrombocytosis may be present but secondary to iron deficiency. Iron supplementation and treatment of the local cause of bleeding enable the diagnosis to be reestablished, for

the increase in hematocrit and hemoglobin levels is rapid and abnormal. When iron is given to such patients, a close follow-up is necessary so that the rapid increase in RCM does not produce thrombosis. However, we did not find any difference between group A and B regarding iron levels: five of the 18 patients in group B had low serum iron versus 16 of 85 in group A (ns). Shih et al26 reported masked polycythemia vera in patients with marked idiopathic thrombocytosis. Among 30 patients with a platelet count greater than 1.000 1 109/L and normal or reduced hemoglobin levels, 13 had EEC, 11 of whom developed polycythemia vera. The main differences observed between the two groups concern the PV measurement and the presence or absence of splenomegaly. It is clear that increased PV in group B was the cause of polycythemia masking on blood count. However, the cause of increased PV is not yet evident. The proportion of patients with splenomegaly is higher in group B than in group A but no correlation was found between spleen and plasma volume, as already suggested.27 For example, 3 patients in group B did not have splenic hypertrophy yet had a very high PV. January 1997 The American Journal of MedicineT Volume 102

/ 220a 0258 Mp

19

Sunday Jan 05 10:03 PM

EL–AJM (v. 102, no. 1)

0258

19

INAPPARENT POLYCYTHEMIA/LAMY ET AL

This study also confirms the need for RCM measurements in patients with increased leucocyte and/ or platelet levels irrespective of the existence of splenomegaly. The PVSG recommends an RCM before diagnosing ET in patients with over 13 g/dL.24 Indeed, in 9 patients with a clinicobiological profile suggestive of ET, the existence of an increased RCM made it possible to reestablish the diagnosis of polycythemia vera. In 2 patients with isolated hyperleucocytosis associated with splenomegaly and a normal karyotype, an increased RCM enabled the diagnosis of chronic granulocytic leukaemia (CGL) to be eliminated. For some authors, raised Ht is correlated with the risk of vascular thrombosis. However, an Ht level of under 0.50 does not exclude vascular risk, as we have seen in group B patients.28 This underlines the importance of confirming the diagnosis of polycythemia vera in these situations. In conclusion, this study shows that the diagnosis of polycythemia vera is probably underestimated. The problem lies in determining the circumstances leading to RCM measurement. Hematocrit and Hb levels above theoretically accepted values should not constitute the sole criterion. We think an RCM should be performed in the case of isolated and unexplained thrombocytosis, hyperleucocytosis, or splenomegaly. We also confirm the importance of RCM in the case of Budd-Chiari syndrome or portal vein thrombosis. The concept of IPV is a reality, and its frequency calls for greater recognition on large series.

ACKNOWLEDGMENTS We thank Professor Y. Najean for his critical reading of the manuscript.

REFERENCES 1. Adamson JW, Fialkow PJ, Murphy S, et al. Polycythemia vera: stem-cell and probable clonal origin of the disease. NEJM. 1976;295:913–920. 2. Prchal JT, Prchal JF. Evolving understanding of the cellular defect in polycythemia vera: implications for its clinical diagnosis and molecular pathophysiology. Blood. 1994;83:1–4. 3. Berlin NI. Diagnosis and classification of the polycythemias. Semin Hematol. 1975;12:339–351. 4. Lemoine F, Najman A, Baillou C, et al. A prospective study of the value of bone marrow erythroid progenitor cultures in polycythemia. Blood. 1986;68:996–1002. 5. Casadevall N, Lacombe C, Varet B. Erythroid cultures and erythropoietin assay: clinical and diagnostic value. Nouv Rev Fr Hematol. 1990;32:77–81. 6. Reid CDL, Fidler J, Kirk A. Endogenous erythroid clones (EEC) in polycythemia and their relationship to diagnosis and the response to treatment. Br J Haematol. 1988;68:395–400.

20

7. Najean Y, Schlageter MH, Toubert ME, Podgorniad MK. Radioimmunoassay of immunoreactive erythropoietin as a clinical tool for the classification of polycythemias. Nouv Rev Fr Hematol. 1990;32:237–240. 8. Messinezy M, Sawyer B, Westwood NB, Pearson TC. Idiopathic erythrocytosis: additional new study techniques suggest a heterogenous group. Eur J Haematol. 1994;53:163–167. 9. Partanen S, Juvonen E, Ikkala E, Ruutu T. Spontaneous erythroid colony formation in the differential diagnosis of erythrocytosis. Eur J Haematol. 1989;42:327–330. 10. Juvonen E, Partanen S, Ruutu T. Colony formation by megakaryocytic progenitors in essential thrombocythemia. Br J Haematol. 1987;66:161–164. 11. Eaves AJ, Eaves CJ. Abnormalities in the erythroid progenitor compartiments in patients with chronic myelogenous leukemia (CML). Exp Hematol. 1979;7(suppl 5):65–75. 12. Correa PN, Eskinazi D, Axelrad AA. Circulating erythroid progenitors in polycythemia vera are hypersensitive to insulin-like growth factor I in vitro: studies in an improve serum-free medium. Blood. 1994;83:99–105. 13. International Committee for Standardisation in Haematology. Recommended methods for measurement of red-cell and plasma volume. J Nucl Med. 1980;21:793–800. 14. Pearson TC, Glass UH, Wetherley-Mein G. Interpretation of measured red cell mass in the diagnosis of polycythemia. Scand J Haematol. 1978;21:153–162. 15. Marsh JCW, Liu Yin JA, Lewis SM. Blood volume measurement in pseudopolycythemia: when and why? Clin Lab Haemat. 1987;9:115–122. 16. Ferrant A. What clinical and laboratory data are indicative of polycythemia and when are blood volume studies needed? Nouv Rev Fr Hematol. 1994;36:151–154. 17. Guthrie DL, Pearson TC. PVC measurement in the management of polycythemic patients. Clin Lab Haematol. 1982;4:257–261. 18. Pearson TC, Botterill CA, Glass UH, Wetherley-Mein G. Interpretation of measured red cell mass and plasma volume in males with elevated venous PCV values. Scand J Haematol. 1984;33:68–74. 19. Valla D, Casadevall N, Huisse MG, et al. Etiology of portal vein thrombosis in adults. A prospective evaluation of primary myeloproliferative disorders. Gastroenterology. 1988;94:1063–1069. 20. Pearson TC, Guthrie DL, Simpson J, Chinn S. Interpretation of measured red cell mass and plasma volume in adults: expert panel radionuclides of the international council for standardization in haematology. Br J Haematol. 1995;89:748–756. 21. Shteeve MN, Axelrad AA. Improved plasma culture system for production of erythrocytic colonies in vitro: quantitative assay method for CFU-E. Blood. 1974;44:517–534. 22. Schlageter MH, Toubert ME, Podgorniak MP, Najean Y. Radio-immunoassay of erythropoietin: analytical performance and clinical use in hematology. Clin Chem. 1990;36:1731–1735. 23. Pearson TC. Apparent polycythemia. Blood Rev. 1991;5:205–213. 24. Murphy S, Iland H, Rosenthal H, Laszlo J. Essential thrombocythemia: an interim report from the polycythemia vera study group. Semin Hematol. 1986;23:177–182. 25. Drenou B, Le Prise PY, Poilane B, et al. Polycythemia and portal vein thrombosis. Nouv Rev Fr Hematol. 1992;34:399–403. 26. Shih LY, Lee CT. Identification of masked polycythemia vera from patients with idiopathic marked thrombocytosis by endogenous erythroid colony assay. Blood. 1994;83:744–748. 27. Zhang B, Lewis SM. The splenomegaly of myeloproliferative disorders: effects on blood volume and red blood count. Eur J Haematol. 1989;42:250–253. 28. Berk PD, Goldberg JD, Donovan PB, et al. Therapeutic recommendations in polycycythemia vera based on polycythemia vera study group protocols. Semin Hematol. 1986;23:132–143.

January 1997 The American Journal of MedicineT Volume 102

/ 220a 0258 Mp

20

Sunday Jan 05 10:03 PM

EL–AJM (v. 102, no. 1)

0258