Plasma Hemoglobin and Hemoglobin
Fractions in Sickle Cell Crisis
HANS
N.
NAUMANN,
M.D.,
LEMUEL
W. Dices, M.D., Luis
AND BEVERLY J. WILLIAMS,
BARRERAS,
M.D.,
M.D.
Sickle Cell Center and Clinical Research Center, Department of Medicine,
University of Tennessee College of Medicine, Memphis, Tennessee 3S103
ABSTRACT
Naumann, Hans N., Diggs, Lemuel W., Barreras, Luis, and Williams, Beverly
J.: Plasma hemoglobin and hemoglobin fractions in sickle cell crisis. Amer.
f. Clin Path. 56: 137-147, 1971. The geometric mean of plasma hemoglobin
concentrations assayed by a modified benzidine procedure in 14 patients
with sickle cell anemia (S-S hemoglobin) during 18 painful crises was 8.5 mg.
per 100 ml. as compared with 5.9 mg. per 100 ml. during quiescent periods
and 0.3 mg. per 100 ml. in normal controls. On the first and second days of
crisis, plasma hemoglobin values were highest (15.3 and 10.1 mg. per 100 ml.,
respectively), dropping precipitously to steady state plateaus, sometimes with
secondary rises and usually paralleling clinical symptomatology. Methemalbumin, free hemoglobin, and haptoglobin-bound hemoglobin were estimated
by plasma electrophoresis and scan of benzidine-stained paper strips. Fraction
values during crisis were variable, with high initial methemalbumin or free
hemoglobin, or both, and often elevations of haptoglobin-bound hemoglobin.
The highest plasma hemoglobin observed, 40.8 mg. per 100 ml., could be
accounted for by intravascular hemolysis of only 20 ml. of blood. This indicates that "hyperhemolysis" in sickle cell crisis is minor, presumably secondary,
and different in principle from the massive intravascular hemolysis observed in
immunohemolytic anemias, hereditary enzyme deficiencies, or hemolytic
septicemias.
have reported increased serum or plasma hemoglobin as eviclence of intravascular hemolysis during the
steady state * in sickle cell anemia (Table
1). However, whether or not there is accel"
"
NUMEROUS INVESTICATORS
Received July 13, 1970; received revised manu-
script September 21, 1970; accepted for publication
October 2, 1970.
..„„,, c „
, ^
Supported in part by USPHS Research Grant
HE-07275 and HE-09495 and by USPHS Grant RR00211, University of Tennessee Clinical Research
Center.
* "Steady state" is defined as that status of sickle
cell anemia patients during intervals—days, weeks,
or months—between febrile and painful crises, when
acute symptoms and fever are absent, with persistence of chronic hemolytic anemia.
erated hemolysis ("hyperhemolysis" 13 ) during the recurrent painful and febrile crises
as compared with the steady state periods
is still debatable.
Early clinicians,' 1 ' 10 ' 43 '' 17 impressed by
similarities of acute abdominal episodes in
.
.
hereditary icterus and sickle cell anemia,
assumed that increased hemolysis played a
m a j o r r o l e i n t h e crises, w h i c h t h e y desig, ,,.
, .
.
,, , ,
•
,•
r
nated
h e m o l y t i c crises. C o n f i r m a t i o n ol:
s u c n a n a s s u m p t i o n t h r o u g h l a b o r a t o r y evidence by later o b s e r v e r s 1 7 ' 2 8 ' 3 7 ' 4 8 ' 0 0 was
. ,
, , .
.
, .
,, ~
,
e i t h e r
lacking or inconclusive.13 T r u e he137
138
NAUMANN ET AL.
molytic crises in sickle cell anemia, described by several authors, could be accounted for by superimposed conditions
such as severe infections 4 - "• 3 5 or hereditary enzyme deficiencies,36'38'44 and therefore were not due to sickle cell anemia per
se.
Previous studies of erythrocyte values
and concentrations of urinary and fecal
pigments 10 and of the concentrations of
serum hemoglobin, 22 performed in the laboratory of Diggs, revealed no evidence
of a significant increase in the rate of hemolysis during sickle cell crises. However,
the samples for serum hemoglobin determination, which were considered representative of a given crisis, were taken on random days, and usually on the later days
of crisis. Stutman and Shinowara 46 have
pointed out that plasma hemoglobin values
are changing rapidly and that values obtained on a given day or even at a certain
hour are not representative of other days.
Because the hemolytic process is of importance in sickle cell crisis, and because
the evidence for or against an increase of
hemolysis during crisis is conflicting, and
our previous estimations were derived from
random samples and analyzed by technics
that were less sensitive than those now in
use, additional, more detailed and better
controlled investigations were undertaken.
In addition to concentrations of total
plasma hemoglobin (benzidine-positive pigments), plasma hemoglobin fractions,
namely methemalbumin, free hemoglobin,
and haptoglobin-bound hemoglobin, were
estimated.
Material
The clinical material consisted of seven
male and seven female Negro patients, 13
to 44 years of age, with classical sickle cell
anemia, electrophoretic pattern of S-S hemoglobin, and less than 5% of fetal hemoglobin. All were admitted to the Clinical
Research Center of the University of Ten-
A.J.C.P.—Vol.
56
nessee because of severe bone or abdominal
pain, or both. Over a 2-year period, 18 entire crises, lasting as long as 8 days and
free of complications or superimposed lesions, were observed. In addition, 11 crises
were seen during one or two of the painful
days. Treatment consisted of codeine for
pain and intravenous fluids. Transfusions
were not given. The total numbers of
plasma hemoglobin determinations were
143 during crises and 54 during the steady
state. Fractions of plasma hemoglobin
amounted to 162 determinations during
crises and 60 during the steady state.
Normal controls for total plasma hemoglobin were provided by 33 healthy hospital employees or by medical students,
free of anemia or jaundice, and in the case
of Negro employees, free of sickle cells as
examined by the metabisulfite procedure.
For 10 of these normal individuals estimation of plasma hemoglobin fractions was
attempted, but traces present were at the
limit of sensitivity and, therefore, should
be considered not exact data but rather
approximate trace amounts.
Methods
Blood Collection and Plasma Separation.
Blood samples for estimation of plasma hemoglobin were taken within 15 hours after
onset of pain and prior to treatment. Additional samples were collected during the
first 4 days of crisis and at intervals thereafter.
Ten ml. of blood were drawn from a
cubital vein with 20-gauge needles and 20
ml. syringes wetted with heparin solution
(1,000 USP units per ml.). Blood was allowed to flow into the heparinized syringe
with minimum suction. After collection
the needle was removed, a small amount of
air admitted by slight withdrawal of the
plunger, and after closing the nozzle with
a Parafilm-covered finger, the blood and
heparin were intimately mixed by inverting the syringe repeatedly. The blood was
August
1971
139
PLASMA HEMOGLOBIN IN CELL CRISIS
Table 1. Plasma Hemoglot)in* in Sickle Cell Anemia
Authors
Number of
Cases
Plasma
Serum
Normal
1.0-4.0
6.0
0.5-3.7
1.0-5.0
5.0
0.2-3.4
1.6-4.6
0.3-2.5
1.0-5.0
<0.1-1.2
Steady state
Crosby and Dameshek"
Keitel and Blakely26
Herman21
Lathem and Jensen20
Whitten"
Upshavv and associates49
Iuchi and associates22
Naumann 82
Stutman and Shinowara40
Present authors
2
22t
17
10
43
29
37+
3
3t
14f
—
16.0-21.0
3.0-150.0
—
—
0.5-42.4
3.0-42.0
0.5-33.6
—
—
—
—
3.4-39.2
1.4-41.6
0.2-5.1
3.0-160.0
0.6-17.6
—
—
—
Painful crisis
Cockrell and Naumann 3
Iuchi and associates22
Stutman and Shinowara48
Present authors
1
17
3t
14f
7.0
—
33.0-72.0
1.6-40.8
—
4.8-45.9
—
—
0.3-2.5
1.6-4.6
1.0-5.0
<0.1-1.2
* All values in mg. of hemoglobin per 100 ml. of plasma or serum,
f More than one examination per patient.
then delivered gently along the wall into
a plastic tube containing 2 drops of heparin solution, covered with Parafilm, and
mixed again by inversion of the tube. After
centrifugation at 1,000 r.p.m. for 10 min.
the plasma was slowly aspirated by capillary pipette, the tip well above the packed
cell layer so as to guard against disturbing
it. The plasma then was recentrifuged at
2,500 r.p.m. for 10 min. and aspirated as
before.
In hemoglobin assays plasma is superior
to serum because increased erythrocytic
breakdown and spurious high results of
serum hemoglobin values are unavoidable
after moving or manipulating clotted
blood.*1 The most indispensable prerequisites are acid-cleaned, thoroughly rinsed
and dried glassware or disposable plastic
ware, a clean venipuncture, and clean aspiration of supernatant plasma. 3 ' 32 The
common pitfall of plasma hemoglobin determinations is failure to observe care in
each step of blood collection and plasma
separation, and this may explain some of
the high hemoglobin concentrations in
plasma or in serum previously reported. 6 ' *•
14, 22, 25, 26, 46, 40, 51
Assay of Total Plasma
Hemoglobin.
Total plasma hemoglobin (benzidine-positive pigments) in terms of mg. per 100 ml.
of plasma was determined by Naumann's
modification of the benzidine method, 3 - 32
designed to eliminate plasma interference that may lower plasma hemoglobin
values more than 50% as compared with
aqueous hemoglobin solutions. 5 ' ' • " The
principle of the method is based on the
observation that benzidine reacts with
heme only, if benzidine solution is added
first and H 2 0 2 thereafter. Contrariwise, if
H 2 O a is added first, followed by benzidine
solution, the heme reaction is inhibited.
This fact is utilized by placing equal
amounts of plasma in three tubes labelled
"unknown," "standard," and "blank," so
that plasma hemoglobin reacts in the "unknown" tube alone but is inhibited in
"standard" and "blank" solutions by prior
addition of H 2 Cv Ultimately, all three solutions are identical in volumes of reagents—only the sequence of addition dif-
140
NAUMANN ET
fers—and of plasma in the tubes with equal
interference which thus is cancelled out.
The sensitivity of the procedure is < 0.1
mg. of hemoglobin per 100 ml. of plasma
with a normal range of < 0 . 1 to 1.2 mg.
per 100 ml. Results were adequately reproducible in a series of 10 replicate samples
each of high and normal values, respectively. The coefficients of variation were
5% in ranges of 20 mg. of hemoglobin per
100 ml. of plasma and 20% in the range of
1 mg. of hemoglobin per 100 ml. of plasma
or less.
Assay of Plasma Hemoglobin Fractions.
The hemoglobin fractions were estimated
in terms of mg. per 100 ml. of plasma
after paper electrophoresis of plasma proteins with standard Spinco-Beckman equipment and a technic similar to the procedures of Lathem and Jensen 26>27 and Sears
and Huser. 39 ' 40 The dried paper strips after
electrophoresis were cut in half lengthwise.
One of the half-strips was stained with
routine bromphenol blue and scanned after
taping it on a black cardboard mask fitting
the strip holder of the "Analytrol" densitometer. The second half-strip was stained
by spraying with fresh benzidine solution
(0.7 Gm. of benzidine base (Fisher) dissolved in 10 ml. of glacial acetic acid and
mixed shortly before use with 10 ml. of
3 % H 2 0 2 , USP). A glass atomizer t connected to a compressed air outlet or to a
1-liter bottle with a one-hole rubber stopper and rubber bulb provided an air stream
of even pressure. Strips were scanned while
still wet, supported by the above cardboard
mask, and within 5 min. before the initial
blue color faded and gradually changed to
a final brown. In the presence of low total
plasma hemoglobin values, greater accuracy was achieved by repeat scannings or
repeat electrophoresis with two or three
plasma applications on the strip. Superimposing benzidine on bromphenol blue
tracings (Fig. 1) helped to identify fractions
f Fisher Scientifiic Co., Cat. No. 5-719-5.
AL.
A.J.CJP.—Vol. 56
as methemalbumin' 42 when coinciding with
the albumin peak, as free hemoglobin when
located within the nbrinogen-gamma complex, and as haptoglobin-bound hemoglobin when migrating with alpha 2 or between
the alpha 2 and beta fractions.34
The location of the three blue bands
varied somewhat with the degree of benzidine reaction and other experimental conditions. This did not prevent identification
of the terminal methemalbumin and free
hemoglobin bands but sometimes interfered with aligning haptoglobin-bound hemoglobin. In the interest of uniform interpretation, therefore, haptoglobin-bound
hemoglobin was defined as intermediate
between methemalbumin and free hemoglobin, i.e., the region from beta to alpha t .
Small peaks in the alpha^ near alpha 2 , and
beta regions were occasionally present (Fig.
1) but were considered insignificant.31-Sd
Areas below the benzidine tracings were
determined by actual count of sq. mm.
Corresponding to the method used for protein fractions, the hemoglobin fraction
areas were expressed as per cent of total
area, and values in mg. per 100 ml. calculated from total plasma hemoglobin values.
Statistical Analysis. Values of total plasma
hemoglobin plotted against frequencies revealed distribution curves that were positively skewed, i.e., presenting peaks shifted
to the left of the system of coordinates.
Asymmetric distributions of this type preclude application of the customary arithmetic means and standard deviations, and
require special statistical procedures. 2 ' s - 18
in, 29,43 Frequency distributions plotted on
semilogarithmic graph paper approached
normal Gaussian configurations, and it was
thus concluded that lognormal transformation and geometric means t would best reflect the central tendency of plasma hemoJ Geometric mean =
., l o g X 1 + l o g X a + l o g X , + . . . ,
analog
^
where X denotes the values found and N the number of observations. 8
August 1971
141
PLASMA HEMOGLOBIN IN CELL CRISIS
| I1
ALBUMIN
I
I
t
Hp-Hb
METHEMALBUMIN
FREE Hb
Hp-Hb
METHEMALBUMIN
FREE Hb
Fie. I. Elcctrophcrogram of plasma proteins (dashed line) with superimposed scan of ben/.idinc-slained paper strip (solid line), showing increases of methemalbumin and free hemoglobin
(left) and of haptoglobin-bound hemoglobin (right).
globin values, as has been demonstrated
for other blood constituents. 18 ' 19 ' 52 We
have, therefore, presented our results as
geometric means throughout this paper.
The fact that geometric means are smaller
than arithmetic means 8 is one more reason—besides those given above under technical pitfalls—why our values are lower
than most of those hitherto reported as
arithmetic means.
Results
Total Plasma Hemoglobin. Total plasma
hemoglobin values of 14 patients on 83
days of 18 painful crises ranged from 1.6
to 40.8 mg. per 100 ml., with a geometric
mean of 8.5 mg. per 100 ml. (Table 2). On
the other hand, values of 54 days of steady
state were moderately lower, ranging from
0.6 to 17.6 mg. per 100 ml. with geometric
mean of 5.9 mg. per 100 ml. The normal
range extended from trace amounts of < 0.1
to 1.2 mg. per 100 ml., with a geometric
mean of 0.3 mg. per 100 ml. Thus, the crisis
value, 8.5 mg. per 100 ml., was 1.4 times
greater than that of the steady state, 5.9
mg. per 100 ml. (p < 0.02), which in turn
was almost 20 times as great as the normal
value, 0.3 mg. per 100 ml. T o correlate the
customary arithmetic with the less common
geometric mean, it may suffice to note that
generally values of the former are slightly
to moderately greater. Our series of plasma
hemoglobin values in crisis, steady state,
and controls showed arithmetic means of
9.4, 6.5, and 0.4 mg. per 100 ml., respectively, compared with the above geometric
means of 8.5, 5.9, and 0.3 mg. per 100 ml.,
respectively.
142
A.J.C.P.—Vol.
NAUMANN ET AL.
56
Table 2. Plasma Hemoglobin in Patients with Sickle Cell Anemia*
Steady State
Crisis
Number
of
Patients
Range
Geometric
Mean
Number
of
Patients
Range
Normal Controls
Geometric
Mean
Number
of
Persons
Range
Geometric
Mean
Total plasma
hemoglobin
14
1.6-40.8
8.5
14
0.6-17.6
5.9
33
<0.1-1.2
0.3
Haptoglobinbound
hemoglobin
9
<0.1-16.1
4.6
8
0.2-8.0
1.9
10
<0.1-0.2
<0.1
Methemalbumin
9
<0.1-21.4
1.8
8
<0.1-6.8
0.5
10
<0.1-0.4
<0.1
Free
hemoglobin
9
<0.1-32.8
1.9
8
<0.1-4.5
0.5
10
<0.1-0.7
<0.1
' All values in mg. of hemoglobin per 100 ml. of plasma.
A more striking contrast between the
plasma hemoglobin concentrations during
crisis and those during steady state becomes
apparent from values on the first four days
of crisis compared with values of the steady
state using several statistical approaches
(Table 3).*. 8,18,10,4s The median of the
first day of crisis was 14.8 mg. per 100 ml.
(p < 0.01 vs. steady state value), decreasing
to 10.0 mg. per 100 ml. (p < 0.01 vs. steady
state value) on the second day and more
gradually to 8.1 and 6.7 mg. per 100 ml. on
the third and fourth day, respectively. The
last two values do not differ significantly
from 5.8 mg. per 100 ml., the median of the
steady state, (p > 0.05). The interquartile
ranges § characterize the degree of skewness
of distribution by inequality of their component segments, median — quartile 1 and
median — quartile 3. For instance, using
the first day values of Table 3, the segment
median (Q2) — quartile 1 (Q]), i.e., 14.8 —
11.7 = 3.1, is smaller than the segment
quartile 3 (Q3) — median (Q2), i.e., 2 1 . 7 § Considering a hypothetical series of 11 figures
arrayed from lowest to highest value:
Qi
02
Qa
1.5 4.3 5.1 8.4 11.5 17.8 21.2 25.6 30.5 33.5 39.3,
then three values:quartile 1 (5.1), quartile 2 or
median (17.8), and quartile 3 (30.5) divide the series into four segments consisting equally of two
figures each, as indicated. T h e interval from 5.1 to
30.5 = Qi — Qa is the interquartile range.
14.8 = 6.9. This is explained by shifting the
distribution peak to the left and by "compressing" the smaller values left of the median compared with "spreading" the larger
values right of the median conforming to
the tail of the skewed curve, thus indicating
the degree of skewness. At the same time,
the interquartile range Qx — Q 3 conveys an
impression of variation by the extension of
spreading to both sides of the point of central tendency, the median.
Statistical significance of geometric means
of the first four days of crisis was analyzed
by the Wilcoxon test.H4S Comparing geometric means of plasma hemoglobin values
(Table 3) on each of the first four clays of
crisis with those of the remaining three, it
was found that first-day values exceeded
any of the other three (p < 0.01). The values on the second day were greater than
those on the third and fourth days at the
same probability level. However, values on
the last two days were not significantly
different.
Further statistical evaluation of geometric means of the first four days of crisis
fl Both the Wilcoxon matched-pairs signed-rank
test and the Mann-Whitney U test are classified as
"nonparametric" statistics which have been used
increasingly when conditions for parametric tests
cannot be met, such as normal frequency distributions.2' " • "
August 1971
143
PLASMA HEMOGLOBIN IN CELL CRISIS
versus the value of the steady state and of
the latter versus that of the control group
was performed by the Mann-Whitney U
test.11dS The computations were based on
data, results of which are shown in Table
3. Conclusions are that plasma hemoglobin
values on the first and second day of crisis
are significantly greater than those of the
steady state (p < 0.01), whereas the values
of the third and fourth day are not. As is
expected, the geometric mean of the control group is significantly smaller than that
of the steady state (p < 0.01).
The general pattern of the first four days
of crisis is thus characterized by initial high
plasma hemoglobin values, followed by
rapid drop and approach to levels of the
steady state. This sequence is demonstrated
graphically by the 18 crises of our 14 patients. Figure 2, with minor variations, illustrates the high plasma hemoglobin values on the first or second day, or both,
followed by rapid falling within hours.
Sometimes there was a secondary rise. Generally, there was a decline to or below the
steady state plateau, which differed from
patient to patient but proved fairly constant in individual patients. 49 It is of interest that in two cases plasma hemoglobin
was determined fortuitously on the day
prior to the painful episode and that both
values previous to onset of pain were below
those during the first day of crisis.
Plasma Hemoglobin Fractions.W Haptoglobin-bound hemoglobin, methemalbumin,
and free hemoglobin were present with few
exceptions in each of 9 of our 14 patients
tested during sickle cell crisis (Table 2).
There was great variability of haptoglobinbound hemoglobin, methemalbumin, and
free hemoglobin in different patients and
in the same patients on different days (Fig.
3). During crises the geometric mean of
haptoglobin-bound hemoglobin was 4.6 mg.
|| T h e sums of geometric means of hemoglobin
fractions do not equal the geometric mean of total
plasma hemoglobin because the calculation of geometric means is based on summation of logarithmic values of each fraction divided by the number
of observations. Since log ab = log a + log b and
fractions are sums and not products, therefore, log
(plasma hemoglobin) 9^ log (hapotoglobin-bound
hemoglobin) + log (methemalbumin) -)- log (free
hemoglobin).
Table 3. Total Plasma Hemoglobin during First four Days of Sickle Cell Crisis,
in Steady States, and in Normal Controls*
Crises
First
Day
Number of
patient daysf
Second
Day
Third
Day
26
25
29
Fourth
Day
24
Steady States
Normals
54
33
Medians and interquartile rangesj
Medians
14.8
Quartilest
Qi-Q.
11.7-21.7
10.2
5.6-20.4
4.4-11.5
6.7
5.8
0.4
4.2-12.5
4.0-7.6
0.3 0.6
6.4
5.9
0.3
4.6-8.9
4.6-7.4
0.1-0.8
Lognormal transformation
Geometric
means
15.3
95% confidence
limits
11.8-20.3
10.1
8.8-11.6
7.6
5.9-9.8
* All values in mg. of hemoglobin per 100 ml. of plasma.
t Number of days on which all or part of 14 patients were observed during one or several of 29 crises (see
text under Material, paragraph 1).
t See text under Results, paragraph 2.
144
NAUMANN ET
per 100 nil., compared with 1.8 and 1.9 mg.
per 100 ml. for methemalbumin and free
hemoglobin, respectively. During the steady
state the geometric means were decreased in
roughly similar proportions. Plasma hemoglobin fractions in normal control samples
were often noticeable as faintly bluish, but
barely measurable, bands. During major
crises with estimations of hemoglobin fractions on successive days there was a general
tendency for methemalbumin and free hemoglobin to decline initially, and later for
haptoglobin-bound hemoglobin to increase
slightly and to parallel total plasma hemoglobin during some phases of the crisis.
Discussion
According to the evidence presented
above, there can be little doubt that superadded hemolysis or "hyperhemolysis" is
present in the painful crisis of sickle cell
anemia, particularly on the first day of
AL.
A.J.CP.—Vol.
56
crisis, less on the second, and not significantly on the third and fourth days. However, variability in different patients or
even in the same patient at different times
is such that there are exceptions to the
general rule. A similar variability was recently encountered by Neely and colleagues 33 in their important observations
on fluctuations of high levels of lactic acid
dehydrogenase found during sickle cell
crises but poorly correlated with variable
plasma hemoglobin concentrations. In some
of our patients, the peak of plasma hemoglobin values was on the second day of
crisis, followed by a drop and later a secondary rise, particularly in crises lasting
five days or longer. The gradient of decline
was usually precipitous, but may be protracted.
The amounts of plasma hemoglobin
could be accounted for by destruction of
small volumes of erythrocytes. For instance,
® PAINFUL SICKLE CELL
ANEMIA CRISES
*
FIG. 2. Total plasma hemoglobin
curves of 18 sickle cell crises related to
hours after onset of pain (solid line:
pain; dashed line: no pain) showing
highest values on first day of crisis,
precipitous fall on first to second day,
some followed by secondary rise, most
tapering to or below the level of the
steady state.
03
o
o
u
X
<
<
-J
0
24 48
HOURS
7Z 96 120 144 168
AFTER ONSET
August J971
145
PLASMA HEMOGLOBIN IN CELL CRISIS
PLASMA
HEMOGLOBIN
MG/IOO ML
Fie. 3. Total plasma hemoglobin
and hemoglobin fraction curves of two
sickle cell crises, illustrating variability
of hemoglobin fractions and incidental
features such as occasional high haptoglobins initially, more often rising
from low initial values and paralleling
total plasma hemoglobin in some
phases, initial high mcthemalbumin
dropping irregularly, and erratic free
hemoglobin curves.
—
TOTAL HB
-
METHEMALBUMIN
•-
HAPTOGLOBIN
<• FREE HB
1
2
3
4
5
DAY OF CRISIS
the highest plasma hemoglobin value observed by us, 40.8 mg. per 100 ml., based
on reports of infusions of hemoglobin solutions15- 20' 28 ' 27 could have been produced
by erythrocytic lysis of only 20 ml. of anemic blood. We realize, however, that such
an estimate must of necessity be vague, due
to the multitude of variables, such as rate
of hemoglobin injection or release from
cells, rate of hemoglobin production and
excretion or absorption, rate of hemoglobin
catabolism, time of sample withdrawal, patient's size, and decrease of plasma volume. 1
At best, an erythrocyte volume contained
in 20 ml. of blood measured as deficit of
total erythrocyte mass is obviously beyond
the sensitivity of hematocrit determinations
or erythrocyte counts, even by electronic
means. On the other hand, the increase of
plasma hemoglobin above the level of the
steady state, about 10 mg. per 100 ml.
(Table 3), is accurately measurable, provided the technic is sensitive and safeguards
against spurious hemolysis are observed
with care (see above). However, the most
flawless of technics is not the sole requirement, for proper timing of specimen collections is indispensable during the early
phases (within 15 hours of onset of pain)
and successive clays of crisis and the succeeding steady state. Values of total plasma
hemoglobin determined accurately and
timed correctly will yield characteristic
curves (Fig. 2).
Generally, there was parallelism between
the increase of total plasma hemoglobin
and the severity of symptoms, but here
again there were exceptions. In certain patients plasma hemoglobin values during
crisis were barely elevated above the levels
of the steady states and, vice versa, high
values have been found on occasion in
some patients free of symptoms and followed over many months (to be reported
later).
Combined clinical and anatomic evidence indicates that the painful and febrile
crises are associated with vascular occlusion
in capillaries, which may be sufficient to
cause local tissue infarcts.9- "• 1= The stasis
of blood and destruction of erythrocytes in
such infarcted areas is followed by release
of hemoglobin from injured erythrocytes
into collaterals and thence the systemic circulation. Various stages of such injuries
have recently been demonstrated by Jensen 23 by means of stereoscan electron microscopy.
This sequence of events explains the peak
of plasma hemoglobin on the first day of
crisis, depending on severity and extension
and the decline thereafter. The assumption
that occlusive lesions are primary and that
hemolysis is secondary was supported by