Red Blood Cell Distribution Width in Untreated
Pernicious Anemia
SUNITA SAXENA, M.D., JOHN M. WEINER, DR. P.H., AND RALPH CARMEL, M.D.
The red blood cell distribution width (RDW) was studied in 26
unselected patients with untreated pernicious anemia. RDW
changes were also sequentially followed after therapy in 12
patients. The mean (±1 SD) RDW values were significantly
higher in pernicious anemia patients than in controls (21.7
± 9.1% vs. 13.2 ± 1.1%, P < 0.0001). Nevertheless, 31% of the
patients had normal RDWs. There were no consistent findings
among those who had normal RDW. Most of them were in the
early stages of deficiency, but some had advanced deficiency.
Over half of those with normal RDW also had normal mean
corpuscular volume (MCV). Overall, 9 of the 26 patients
(35%) had normal MCV. Of eight patients whose RDW fell
with therapy, some showed a steady fall while others had a
transient rise followed by a progressive drop. Despite current
advocacy that a high RDW is a sensitive and consistent finding
in vitamin B12 deficiency, our findings show that a large proportion of untreated pernicious anemia patients have normal
RDWs and that in contrast to iron deficiency, elevation of
RDW is not necessarily the earliest indicator of vitamin B12
deficiency. (Key words: RDW; Pernicious anemia; Vitamin
B12; Hb; MCV) Am J Clin Pathol 1988;89:660-663
THE USEFULNESS of red blood cell distribution width
(RDW), an index of red blood cell size heterogeneity, in
the classification and workup of microcytic anemias has
been well documented. 1 , 2 , 4 ' 0 '" 1 6 Its usefulness in the
macrocytic anemias is less clear. According to some, all
patients with vitamin B12 (cobalamin) deficiency have
increased RDW 4,10 and can thus be differentiated from
macrocytic anemias with normal RDW, such as aplastic
anemia. However, others in brief or preliminary communications were unable to support this finding.8,14 It
has also been suggested that the RDW increases before
the mean corpuscular volume (MCV) in megaloblastic
anemia 2,4 and could therefore serve as a sensitive index
of early deficiency. However, data for such claims are
not available and preliminary observations led us to believe this was not always the case. Therefore, we studied
the RDW in unselected patients with untreated pernicious anemia. We also sequentially followed RDW
changes in some of the cases after therapy was initiated.
Received June 29, 1987; accepted for publication July 31, 1987.
Address reprint requests to Dr. Saxena: LAC/USC Medical Center,
Room 2900 General Hospital, 1200 North State Street, Los Angeles,
California 90033.
Interpretive Clinical Pathology Unit, Section of Laboratories
and Pathology, and the Department of Medicine, Los
Angeles, County University of Southern California
Medical Center, and Departments of Pathology
and Medicine, University of Southern California
School of Medicine, Los Angeles, California
Materials and Methods
Since the inclusion of RDW in the routine cell profile
at the Los Angeles County-University of Southern California Medical Center, 26 patients with pernicious anemia have been identified. The diagnosis of pernicious
anemia was established by a low serum vitamin B12
level, accompanied by one or more of the following: an
abnormal Schilling test result that was corrected with
exogenous intrinsic factor, absent gastric intrinsic factor
secretion despite betazole stimulation, or presence of
anti-intrinsic factor antibodies in the blood.
The cell profiles including hemoglobin (Hb), MCV,
and RDW were performed on the Coulter Counter®
Model S Plus IV (Coulter Diagnostics, Hialeah, FL) according to the manufacturer's instructions. Serum bilirubin and lactate dehydrogenase (LDH) determinations
were done on the Technicon's® Autoanalyzer SMAC-II
(Tarrytown, NY). Blood samples were analyzed within
30 to 60 minutes of phlebotomy.
The coefficients of variation on 110 repeated determinations of normal and abnormal pools respectively were
1.0% and 1.4% for Hb, 0.6% and 0.6% for MCV, and
1.7% and 2.8% for RDW.
The reference range for RDW was determined by
analysis of results from 313 employees. All employees
seen in the clinic for either pre-employment or annual
physical examination were considered for this analysis.
Regular users of any medication and iron-deficient subjects were excluded. Iron deficiency was defined as having <16% (. 16) transferrin saturation or <10 jtg/L (<10
ng/ml) serum ferritin. Their mean age was 30 years, with
99% of the employees between the ages of 16 and 51
years. There were 199 women and 114 men; 89 were
white, 88 black, 62 Latin American, 41 Oriental, and the
remaining 33 were from other or unknown ethnic backgrounds.
660
I
BRIEF SCIENTIFIC REPORTS
Vol. 89 • No. 5
Data from employees were first converted into logarithms. Mean and standard deviation (SD) were determined, which were found to be almost identical to published values by others.2 The mean RDW and the reference range (mean ± 2 SD) were 13.2% and 11.7-14.9%.
Student's Mest was used to compare log RDW results in
the controls and the pernicious anemia patients. To
study the relationship between RDW and severity of
anemia, log correlations between RDW and LDH, bilirubin, MCV, and hemoglobin were performed.
Results
Analysis of RDW in Untreated Pernicious Anemia
The pretreatment hematologic data on the 26 patients
with pernicious anemia are shown in Table 1. The mean
(± 1 SD) RDW in pernicious anemia patients was significantly higher than in our control population (21.7%
± 9.1 vs. 13.2% ± 1.1, P < 0.0001). Nevertheless, RDWs
were within our normal range in 8 (patients 9-16) of the
26 cases (31%). Even if the reference range recommended by the manufacturer (11.5-14.5%) is used instead of our own established range, a similarly high proportion of the patients (27%) had normal RDWs.
No consistent clinical or laboratory findings emerged
to set the patients who had normal RDWs apart from
the others. Most patients with normal RDW seemed to
be in the early stages of deficiency, as evident by the
mildness or absence of anemia (patients 9, 10, 12-15)
but others had advanced deficiency (patients 11, 16).
Patient 11, although not anemic, had severe neurologic
abnormalities, and patient 16 had severe anemia in addition to having clearly megaloblastic red cells. Over half
of the patients with normal RDW also had a normal
MCV (80-100 FL), but in others (patients 9, 12, 16) the
MCV was high.
This poor relationship between the MCV and the
RDW was demonstrated by statistical analysis which
showed no correlation (r = 0.2) between the two indices.
Although no relationship was noted between RDW and
MCV, the RDW correlated inversely with Hb (r
= -0.76) and directly with LDH (r = 0.72) and bilirubin
(r = 0.54).
Response to Therapy
Of the 26 patients, 12 received no transfusion and had
adequate follow-up periods that ranged from 6 weeks to
almost 20 months. In the remaining 14 cases, no followup was obtained because in most of these cases patients
were transfused after admission and the others were lost
to follow-up. The sequential effect of therapy on RDW
in 12 patients is shown in Figure 1 with changes shown
as percent of original value (which is taken as 100%).
661
Table 1. Red Blood Cell Data Before (Pre) and After
(Post) Therapy In Patients With Pernicious Anemia
Hb (g/L)
MCV (fL)
RDW (%)
Case
No.
Sex/Race
Pre
Post
Pre
Post
Pre
Post
1
2
3
4
5
6
7
8
9
10
11
12
F-LTN
F-LTN
F-BLK
F-WHT
M-LTN
F-BLK
F-LTN
F-LTN
M-LTN
F-BLK
M-WHT
F-WHT
55
97
123
94
92
79
96
109
133
127
152
120
141
147
155
163
149
134
120
113
159
126
148
149
126
140
102
137
123
121
107
115
121
86
90
101
86
84
87
91
88
92
77
89
92
86
88
101
33.9
25.6
22.0
20.8
19.0
18.0
17.8
16.7
14.9
14.2
13.5
11.8
14.9
13.7
13.6
15.5
13.6
13.2
15.0
14.5
13.7
14.1
13.9
13.5
13
14
15
16
17
18
19
20
21
22
23
24
25
26
M-LTN
F-WHT
F-LTN
M-WHT
F-LTN
M
F-LTN
F-LTN
M-WHT
F-BLK
F-LTN
F-BLK
M-BLK
F-LTN
137
128
139
54
92
121
98
60
141
59
43
24
43
58
—
—
—
—
—
—
—
—
—
—
—
—
—
96
98
96
122
80
99
120
124
113
127
98
97
113
116
—
—
—
—
—
—
—
—
—
—
—
—
—
12.4
13.2
13.4
14.1
15.8
17.9
18.2
18.2
24.9
31.1
33.2
38.4
39.4
43.4
—
—
—
—
—
—
—
—
—
—
—
—
M = male; F = female; BLK - black; WHT = white; LTN = Latin American.
Reference Range: MCV = 80-100 fL; RDW = 11.7-14.9%; Hb - 140-180 (male) and
120-160 g/L (female).
Conversion Factor for SI Unis: Hb (g/dL) = 0.1; MCV (um3) = 1.
Two patterns of RDW response emerged. In most
patients, the RDW dropped more than 10% below the
pretreatment level. The other pattern of response
showed minimal or no decrease in the RDW in the first
three months of therapy. Patients 10, 11, and 12 who
had presented with normal RDWs seemed to fall in the
latter category.
Within the group whose RDW fell with the therapy,
two types of responses were seen. In the first 9-10 weeks
after the onset of treatment, patients 3, 5, 6, and 7
showed a steady fall with the onset of therapy, while
others like patients 1, 2, 4, and 8 had a transient rise in
RDW followed by a progressive drop.
Discussion
Our study of 26 unselected patients with pernicious
anemia shows that a significant proportion of patients
with vitamin B12 deficiency present with normal RDW.
In this study, 31% (or 27%, depending on the normal
range used) of patients had normal RDW. This finding
contradicts previous surveys',4'8,15 and reviews2,10 that
suggested that vitamin B12 deficiency always causes an
elevation of RDW. It should be noted that one of those
SAXENA, WEINER, AND CARMEL
662
0
2
4
6
8
10
12
14
WEEKS
16
18
20
22
24
i"i7
2ff
8
9
A.J.C.P.-May 1988
10
11 12
MONTHS
13
14
15
1 6 " l 9 2 0
FIG. 1. Effect of therapy on RDW in patients with pernicious anemia (numbers in the graph refer to patient numbers in Table 1.)
Solid lines: patients with elevated RDW; broken lines: patients with normal RDW.
surveys actually showed one of 11 patients with pernicious anemia to have a normal red blood cell volume
range.15 Moreover, Perry and associates in, a preliminary report of 97 patients with macrocytosis of various
causes, claimed that only 5 had elevated RDW but gave
no details or diagnoses.I4
We agree with Bessman and colleagues4 that the average RDW is high in pernicious anemia and that vitamin
B12 deficiency normally produces anisocytosis. Moreover, our data suggest that even a normal RDW can
sometimes hide mild anisocytosis since a "normal"
RDW fell with therapy in some cases. However, it is
clear that many patients with pernicious anemia have
RDW values within the normal range and that RDW is
not a more sensitive index of vitamin B ^deficient hematopoiesis than is MCV.
It is evident from the correlation of RDW with the
degree of anemia and with LDH and bilirubin values
(but not with MCV) that the RDW rise is associated
with increasing ineffective erythropoiesis. This indicates
that a high RDW is not necessarily the earliest feature of
cobalamin deficiency, but that it progresses in severity
along with other reflections of disturbed erythropoiesis.
Moreover, cases 9, 16, and perhaps 12 showed that high
MCV can appear before a high RDW. High RDW,
never considered a specific finding for megaloblastic
anemia, is not the highly sensitive index it was once
thought to be either.2,4 In sharp contrast to iron deficiency, a rise in RDW is not an early change in vitamin
B12 deficiency.
Normal RDW can be seen at any stage of the development of pernicious anemia. One of our patients with
normal RDW had neither clinical nor hematologic evidence of the disease, indicative of a very early stage of
the deficiency. Five other patients were in a little more
advanced state since their MCVs were high even though
they had not yet developed anemia. Finally, two other
patients had progressed to severe anemia or neurologic
manifestations of pernicious anemia despite a normal
RDW. Thus, there were no consistent clinical or hematologic abnormalities associated with the normal RDW.
Although thalassemia often produces a normal RDW,4
no evidence was found for coexisting thalassemia in
these patients. Hemoglobin electrophoresis excluded
coexisting /3-thalassemia and the absence of post-treatment microcytosis militated against a-thalassemia.
When patients with normal RDW were treated no
significant change in RDW was usually observed,
BRIEF SCIENTIFIC REPORTS
Vol. 89 • No. 5
whereas those with high RDW displayed a drop with
specific therapy. Some of the latter patients had a transient rise in RDW in the first few weeks of treatment
before a progressive fall occurred as recovery continued.
Several authors studying the red blood cell volume distribution in treated and untreated pernicious anemia
patients described the appearance of a bimodal curve
during recovery from macrocytic anemia.3'5,7,13 This
curve represented the emergence of a new population of
normocytes in the few days after initial therapy. Such
coexistence of two populations results in increased anisocytosis and is probably responsible for the initial transient rise in RDW that we sometimes observed.
The incidental observation that normal MCV was
present in 10 of our 26 patients with untreated pernicious anemia is also worthy of note, although two of
these patients had coexisting iron deficiency. It adds to
the increasing awareness6'912'17'8 that the classical presentation of pernicious anemia with macrocytic anemia
does not hold true in a large proportion of cases and that
the diagnosis must be considered even in patients with
normal or low MCV.
References
1. Bessman JD: Heterogeneity of red cell volume: Quantitation,
clinical correlations, and possible mechanisms. The Johns
Hopkins Medical Journal 1980;146:226-230.
2. Bessman JD: Automated blood counts and differentials: A practical guide. Baltimore, The Johns Hopkins University Press,
1986,40-42.
663
3. Bessman JD: Erythropoiesis during recovery from macrocytic
anemia: Macrocytes, and microcytes. Blood 1977;50:9951000.
4. Bessman JD, Gilmer PR, Gardner FH: Improved classification of
anemias by MCV and RDW. Am J Clin Pathol 1983,80:322326.
5. Bessman JD, Johnson RK: Erythrocyte volume distribution in
normal and abnormal subjects. Blood 1975;46:369-379.
6. Carmel R, Karnaze DS: Physician response to low serum cobalamin levels. Arch Intern Med 1986;146:1161-1165.
7. England JM, Down MC: Red-cell-volume distribution curves and
the measurement of anisocytosis. Lancet 1974;1:701-703.
8. Farley PC: RDW misleading? Am J Clin Pathol 1984;81:415-416
(letter).
9. Green R, Kuhl W, Jacobsoh R, Johnson C, Carmel R, Beutler E:
Masking of macrocytosis by a a-thalassemia in blacks with
pernicious anemia. N EngJ Med 1982;307:1322-1325.
10. Karnad A, Poskitt TR: The automated complete blood cell count.
Arch Intern Med 1985;145:1270-1272.
11. McClure S, Custer E, Bessman JD: High RDW is the earliest
predictor of iron deficiency. Blood 1983;62(suppl l):51a.
12. Oscier DG, Hamblin TJ: Megaloblastic anaemia with normal
mean cell volume (letter). Lancet 1979;1:389-390.
13. Patel A, Chanarin I: Restoration of normal red cell size after
treatment in megaloblastic anemia. Br J Haematol
1975;30:57-63.
14. Perry DJ, Lowe DAS, Wilson CID, Harris Rl, Stevens MJ: The red
cell distribution width in macrocytosis. Br J Haematol
1985;61:559 (Abstract).
15. Proctor SJ, Cox JR, Sheridan TJ: Anisocytosis and the C-1000
channeiyzer in macrocytic anemia. J Clin Pathol
1976;29:719-723.
16. Roberts GT, El Badawi SB: Red blood cell distribution width
index in some hematologic diseases. Am J Clin Pathol
1985;83:222-226.
17. Solanki DL, Jacobson RJ, Green R, McKibbon J, Berdoff R:
Pernicious anemia in blacks. Am J Clin Pathol 1981,75:96-99:
18. Spivak JL: Masked megaloblastic anemia. Arch Intern Med
1982;142:2111-2114.
Monoclonal Antibody to Fibrin D-Dimer (DD-3B6) Recognizes
an Epitope on the y-Chain of Fragment D
DANA V. DEVINE, PH.D. AND CHARLES S. GREENBERG, M.D.
Laboratory determination of fibrinolysis has been facilitated
by diagnostic tests that use monoclonal antibody DD-3B6 to
measure fibrin D-dimer levels in plasma. When DD-3B6 is
reacted with soluble fibrin fragments, it is specific for fragment
D-dimer. The authors have used immunoblot analysis of
DD-3B6 binding to purified fragment D and fragment D-dimer
to localize the binding site of the antibody. Although DD-3B6
recognizes only fragment D-dimer in solution, it binds to both
immobilized fragment D-dimer and fragment D in immuno-
Received July 7, 1987; accepted for publication August 10, 1987.
Address reprint requests to Dr. Devine: Canadian Red Cross Blood
Transfusion Service, Vancouver Centre, 4750 Oak Street, Vancouver,
British Columbia V6H 2N9 Canada.
Departments of Medicine and Pathology, Duke University
Medical Center, Durham, North Carolina
blots. When immunoblots were performed using protein which
was electrophoresed under reducing conditions, DD-3B6
bound to the Y-chain of fragments Dl, D2, and D3. Therefore,
the epitope recognized by DD-3B6 resides between amino
acids 86 and 302 in the 7-chain of fragment D. This epitope is
masked in soluble non-cross-linked or nondegraded fibrin, but
becomes expressed after cross-linked fibrin has been cleaved
by plasmin. (Key words: Fibrinolysis; Antibody DD-3B6; Fibrin D-dimer; Plasmin; Coagulation) Am J Clin Pathol
1988;89:663-666