Clin. lab. Haemat. 1991, 13, 229-237

ADONIS 0 14198549l00031Z

The red cell distribution width in sickle cell disease-is it of clinical value? M I N E R V A T H A M E * , M B BS, D C H , Y V O N N E GRANDISON, DMT, KARLENE MASON, RT, MARVIN THOMPSON, D M T , DOUGLAS HIGGST, M R C P , J O A N N E M O R R I S , BSc, MSc, B E R Y L SERJEANT, FIMLS & GRAHAM SERJEANT, MD, FRCP *Department of Child Health and the M R C Laboratories (Jamaica), University of the West Indies, Mona, Kingston Jamaica, and TMRC Molecular Haematology Unit, John Radclife Hospital, Oxford, UK Accepted for publication 2 May 1991 Summary The red cell distribution width (RDW) has been studied during the clinical steady state in I121 patients with homozygous sickle cell (SS) disease, 344 with sickle cell-haemoglobin C (SC) disease, 68 with sickle cell-beta+ thalassaemia, 49 with sickle cell beta' thalassaemia and in 130 control subjects with a normal (AA) genotype. The mean RDW was moderately increased in Sbeta+ thalassaemia and SC disease and markedly increased in Sbeta' thalassaemia and SS disease. In SS, SC and Sbeta' thalassaemia genotypes, lower RDW values occurred in females and with alpha thalassaemia. The RDW correlated negatively with total haemoglobin, mean cell haemoglobin concentration, mean cell volume, and fetal haemoglobin (HbF) and positively with reticulocyte count in SS disease. A low RDW was associated with higher weight and less frequent dactylitis, painful crisis, acute chest syndrome, acute splenic sequestration, and hospital admissions. A low RDW in SS disease is consistent with a high total haemoglobin, high HbF, low reticulocyte count, alpha thalassaemia, and a more mild clinical course.

Keywords: RDW, sickle, SS disease

The red cell distribution width (RDW), as calculated by Coulter Counter Model S plus I1 and later models, is derived from the coefficient of variation around the mean red cell volume. It reflects the degree of anisocytosis and higher values have been reported with the abnormal erythrocyte morphology of sickle cell disease (Hammersley et al. 1981; Schweiger 1981; Bessman et al. 1983; Roberts & El Badawi 1985; Webster & Castro 1986; Monzon et al. 1987). The RDW correlates negatively with total haemoglobin (Webster & Castro 1986) and positively with reticulocyte count (Roberts & El Badawi 1985; Webster & Castro 1986), and a fall Correspondence: Professor G.R. Serjeant, The Medical Research Council Laboratories, (Jamaica), University of the West Indies, Mona, Kingston 7, Jamaica, West Indies.

229

230

M . Thame et al.

has been reported on recovery from the painful crisis (Lawrence et a/. 1985). Since the RDW is now widely available, the present study addresses the possible clinical significance and value of this index in the management of patients with sickle cell disease.

Materials and methods The patients attended the Sickle Cell Clinic of the University Hospital of the West Indies, Kingston, Jamaica, and had venepunctures for routine clinical care between 28/1/88 and 31/12/89. The study group was restricted to subjects with an AA genotype, homozygous sickle cell (SS) disease, sickle cell-haemoglobin C (SC) disease, sickle cell-beta+ thalassaemia (Sbeta + thalassaemia) and sickle cell-beta' thalassaemia (Sbeta' thalassaemia) diagnosed on standard criteria (Serjeant 1985). The term Sbeta+ thalassaemia applied to patients with high amounts of HbA (approximately 20-25%). The AA subjects were normal controls in a cohort study of sickle cell disease and were aged 12-16t years at the study date. The patients were drawn from both cohort and non-cohort clinics and were aged between 8 months and 7 3 years. The data were further restricted to the clinical steady state and to the first observation in individuals where multiple observations were available. Values within 3 months of transfusion were excluded. The final data set included 130 AA controls, 1121 patients with SS disease, 344 patients with SC disease, 68 with Sbeta+ thalassaemia, and 49 with Sbeta' thalassaemia. Venous blood was taken into evacuated tubes containing K EDTA (BectonDickinson) and analysed in a Coulter Model S plus IV usually within 2 h. Mean cell haemoglobin concentration (MCHC) was calculated from the spun microhaematocrit. Reticulocytes were counted after staining with 5% brilliant cresyl blue but were not measured in AA controls. Fetal haemoglobin (HbF) was assessed by alkali denaturation (Betke et al. 1959) but only values measured at the time of the RDW are presented. Irreversibly sickled cells (ISCs) were assessed on stored films in a subsample of patients with SS disease, chosen to give approximately 20 counts within each of the following RDW value bands: 20,20-24, 25-29; and 2 30 in order to study the relationship between ISC count and RDW. The number of alpha globin genes was determined by restriction endonuclease analysis of DNA obtained from the buffy layer (Higgs et al. 1981), and was performed routinely in cohort study children and for selected cases in the non-cohort clinic. The distributions of HbF and of reticulocyte count were skewed and were transformed (log,,[HbF 11, log,,[retics 11) prior to analysis. Differences in distributions were compared by Student's t test and the relationship between RDW and selected indices by the product moment correlation coefficient. Since haematological indices are frequently interrelated, partial correlation coefficient analysis was performed to investigate the relationship between RDW and haematological variables. Clinical correlates with RDW in SS disease were examined in the cohort study by defining subgroups of 30 subjects each with the highest and lowest RDW values. Dactylitis was defined as an acute, non-pitting swelling of the

+

+

RD W in sickle cell disease

231

digits of the hands and feet, and painful crisis as acute bone or abdominal pain of sufficient severity to limit activity and require hospital or clinic attendance. Acute chest syndrome was defined as shortness of breath, cough and fever in the presence of clinical or radiological evidence of pulmonary infiltrates. Acute splenic sequestration was defined by a fall in haemoglobin of 2 g/dl or more associated with acute splenic enlargement and an increase in reticulocyte or nucleated red cell count. Results EFFECT O F GENOTYPE

The RDW was moderately but significantly elevated in SC disease and Sbeta+ thalassaemia and markedly elevated in SS disease and Sbeta' thalassaemia compared with the AA controls (Table 1). All genotypes of sickle cell disease showed a wide range of values, overlapping the range in normal AA children. EFFECT O F AGE

There was no change of RDW with age in the AA controls but observations were available only over an age band of 44 years (12-164 years). In SS disease there was a significant age related decline in RDW ( F statistic for trend = 4.23, P < 0.OOl) but this was predominantly due to high RDW values in children under 10 years of age (Figure 1). Exclusion of this group from the analysis removed any 40

-

.

m m

m

m

m

35-9 mm

0..

m

m

-

. . I

-

30

2a

25

E

-I-

m o m

0

. I . m I 0-

.

I. I.m

I I .

m - m m 1 1 m m m m

. I

9

I.

m-mm

.

.

mm

- - . - 1 1 1 1 1 1 m I m Im I-I-m -I

I

mm

. . -

. I .

.I. 1)11.I

o m -

--I-

m m I . m m I

.m

Im

I

m m I

m

m

I .

.

m

.

mm I... I.I -1- m I 9 1

.I

mmm m -

mm

I

I

L

. ... . m

. I

I

1-

mm m m

..

. I I.

m m.

m m m

..-m--mm

m

I

mmm . m m m m I I I . I -..)-I mmm m m

I

15-.

m m

. . . I

- m I D - - m - m I m I

m

m

..I.

1 ) 1 1 ) 1 . ( 1 1 -

20

.

m mm

I . .

m

.

m

m

m

I

09

I

I

10-

'

I

.

I

1

.

1

I

.

l

.

1

.

232

M . Tharne et al. Table 1. Red cell distribution width (RDW) and genotype-ranked in order of RDW

Genotype

n

AA

130 68 344 1121 49

Sbeta+ thalassaemia SC disease SS disease Sbeta" thalassaemia

RDW mean SD

range

14.0 16.6 16.8 23.0 23.3

12-21 14-24 12-29 13-39 17-31

1.6 2.0 2.3 3.9 3.3

age related trend ( F = 1.58, P = 0.11) No age-related trend was apparent in SC disease and observations in Sbeta+ and Sbeta' thalassaemia were too few to test. E F F E C T OF SEX

RDW values were higher in males with SS and SC disease (Table 2). Whether this sex difference could be attributed to higher HbF levels in females was assessed by analysis of covariance which showed that the sex difference persisted after allowing for H bF in both SS disease ( F statistic = 6.4, P = 0.012) and SC disease ( F = 16.1, P < 0.001). EFFECT OF A L P H A THALASSAEMIA

The RDW did not vary with alpha globin gene number in AA controls although a highly significant trend between RDW and increasing alpha globin gene number occurred in SS and SC disease (Table 3). CORRELATIONS W I T H HAEMATOLOGICAL INDICES

The RDW correlated negatively with total haemoglobin (Hb) and with mean cell volume in AA, SS, and SC genotypes and with MCHC in AA and SS genotypes (Table 4). The RDW correlated negatively with HbF and positively with reticuloTable 2. Red cell distribution width (RDW) and sex analysed within genotypes ~

Genotype

AA SS disease SC disease Sbeta+ thal Sbeta" thal

n

Males mean

SD

n

60 527 168 31 26

14.0 23.7 17.4 16.7 24.2

1.6 3.7 2.3 2.3 3.5

70 594 176 37 23

Females mean SD 14.0 22.4 16.2 16.6 22.3

1.6 4.1 2.0 1.7 2.9

Significance

7' 0.15 5.75 5.11 0.36 2.02

P 0.9

< 0.001 < 0.001 0.7 0.049

RD W in sickle cell disease

233

Table 3. Red cell distribution width (RDW) and alpha thalassaemia

AA controls (total n = 129) n mean SD

Alpha globin genotype

4 39 84 2

a-/aCY - /at?

rwu/au ctaalac!

15.0 14.3 13.8 12.5

SS disease (total n = 487) n mean SD

0.0 1.6 1.7 0.7

32 168 283 4

19.5 22.5 23.3 26.0

3.5 3.6 3.8 5.2

SC disease (total n = 138) n mean SD 3 42 92 1

14.0 16.4 17.2 17.0

1.0 1.8 2.4 0.0

F statistic for trend

1.77

11.49

3.12

P

= 0.155

< 0.001

= 0.028

So thalassaemia (total n = 35) n mean SD

2 10 22

20.5 21.0 24.6

0.7 2.5 2.5

not tested

Table 4. Correlations between RDW and haematological indices in AA, SS, and SC genotypes

Variable

n

AA r

Hb MCHC MCV Retics HbF

130 130 130

-0.49 -0.43 -0.66

-

117

-

-0.06

ss P

n

r

P

n

SC r

< 0.001 < 0.001

1121

< 0.001 < 0.001