AMERICAN JOURNAL OF CLINICAL PATHOLOGY
Vol. 33, No. 1, January, 1960, pp. 6-13
Printed in U.S.A.
THROMBIN GENERATION AND CLOT RETRACTION STUDIES IN
PAROXYSMAL NOCTURNAL HEMOGLOBINURIA
J O H N M. F L E X N E R , M . D . , J O S E P H V. A U D I T O R E , P H . D . ,
AND R O B E R T C. H A R T M A N N , M . D .
Hematology Service, Department of Medicine, Vanderbilt University
of Medicine and Hospital, Nashville,
Tennessee
School
Thrombocytopenia of some degree is a generation studies were performed in plasma
common occurrence in paroxysmal nocturnal specimens from 4 PNH patients, 3 of whom
hemoglobinuria (PNH*). Crosby has sug- have thrombocytopenia. Furthermore, ingested that the PNH platelet has a stromal asmuch as the ability to produce clot redefect similar to that of the PNH erythro- traction is a critical test as to whether or not
cyte, and that for this reason PNH platelets the platelets are "intact," 8 such studies were
may be susceptible to injury by the plasma performed in 5 patients with PNH. A prehemolytic system.2 According to this belief, liminary report regarding these findings has
platelet lysis by the plasma lytic system is been presented elsewhere.6
responsible for the thrombocytopenia in
METHODS AND MATERIALS
PNH.
Mendel and associates13 reported obPatient material. The clinical protocols
servations regarding the possibility that free of 6 patients with PNH have been docplatelet products are present in the circu- umented. 6 The same case numbers have
lating plasma in PNH. Platelet-rich plasma been used in this report as in the previous
from a PNH patient with thrombocytopenia one. (Case 2 was not studied in the present
generated as much thrombin during clotting investigation.) The studies were performed
as did normal platelet-rich plasma. Further- in Cases 5 and 6 during a period of increased
more, PNH platelet-free plasma generated activity of the disorder. In Case 5 this was
much more thrombin than did normal plate- manifested by pronounced hemoglobinuria,
let-free plasma, and virtually the same and in Case 6 by a further fall in peripheral
amount as did PNH platelet-rich plasma. blood hematocrit and rise in plasma hemoThese workers concluded that PNH platelets globin.
undergo lysis in the circulation resulting in
Preparation of platelet-rich and plateletthe release of platelet substances into the free plasma. Specimens of plasma were preplasma.
pared by means of previously described
In the present investigation thrombin methods. 7,8 In brief, these methods were
based on the following principles: expeReceived, July 21, 1959; accepted for publicaditious and gentle handling of blood and
tion September 2.
plasma, and the use of disposable blood
Dr. Flexner is U. S. Public Health Service redonor sets,f silicone-treated equipment,
search fellow with the Hematology Service; Dr.
and
refrigerated containers and centrifuge.
Auditore is Instructor in Pharmacology in Medicine with the Hematology Service; and Dr. H a r t - Anticoagulants used were acid-citrate-dextrose (ACD) solution:): in a proportion of 1
mann is Associate Professor of Medicine and
part to 5 parts of blood, and M/10 sodium
Director of Hematology Laboratories, Vanderbilt
University School of Medicine.
oxalate in a proportion of 1 part to 9 parts
This study was supported by Research Grants
of blood.
H-4135 and H-3509 from the Division of Research Grants, National Institutes of Health, U. S.
Public Health Service, Department of Health, Education, and Welfare, and by the John A. H a r t ford Foundation, Inc., New York.
* The abbreviation P N H will be used to designate t h e hemolytic disorder paroxysmal nocturnal hemoglobinuria.
t Saftidonor sets were kindly provided by Dr.
E. B . McQuarrie, Biochemical Research Division,
C u t t e r Laboratories, Berkeley, California.
% Acid-citrate-dextrose (ACD) solution consisted of sodium citrate, 1.32 Gm.; citric acid,
0.48 Gm.; and dextrose, 1.47 Gm. in distilled
water to a total of 100 ml.
6
Jan. I960
T H R O M B I N G E N E R A T I O N AND CLOT R E T R A C T I O N I N
Platelet counts were performed according
to the phase microscopy technic of Brecher
and Cronkite.1 Platelet-free plasma prepared by our method in most instances contains less than 100 platelets per cu. mm. 8
Platelet concentrations were also adjusted
to the desired values by dilution of plateletrich plasma with an appropriate amount of
platelet-free plasma.
Thrombin generation test. The thrombin
generation test was performed according
to the method of Sj0lin.14 Both oxalated
and ACD plasma were used for this test.
In several instances native (without anticoagulant) plasma was prepared and tested.
The test was performed in glass tubes at
37 C.
Plasma clot retraction. Plasma clot retraction was tested by the method of Hartmann and Conley.8 ACD was used as the
anticoagulant. PNH and normal ACD platelet-rich and platelet-free plasmas were
mixed in varying proportions to obtain specimens with serial dilutions of platelet concentrations. Bovine thrombin § was then
added to clot the plasma, and the degree of
clot retraction in silicone-treated tubes at
the end of 60 min. incubation at 37 C. was
noted. The peripheral blood platelet concentration in Case 3 was so severely depressed that it was difficult to obtain a
plasma specimen with sufficient platelet
concentration for the clot retraction test.
In this instance platelets were concentrated
by using large volumes of blood, harvesting
the platelet-rich plasma, and centrifuging
the latter at high speed. The platelet "button" was then resuspended in a smaller
volume of platelet-free plasma to obtain a
final concentration of 96,000 per cu. mm.
7
PNH
within these areas represent the mean values
for platelet-rich and platelet-free plasma,
respectively. The platelet concentration in
platelet-rich plasma ranged from 200,000 to
350,000 per cu. mm. Thrombin generation
was slow during the first few minutes, and in
the instance of platelet-rich plasma, reached
a maximum in 4 to 9 min., following which
there was an abrupt decrease. Little thrombin was present after the 10 to 12 min.
readings. In contrast, thrombin generation
was greatly reduced or virtually absent in
recalcified platelet-free plasma. The peak of
64-i
56-
48-
40-
32600/t
24-
16-
«:':i!|i
RESULTS
Thrombin Generation Studies
Normal subjects. The thrombin generation
curves obtained from both ACD and oxalated platelet-rich and platelet-free plasmas
of 7 normal persons are graphically presented in Figure 1. All of the curves fell
within the shaded areas, and the heavy lines
§ Thrombin was generously supplied by Parke,
Davis & Company, Detroit, Michigan.
JLJUk
4
8
TIME
O
12
16
20
22
IN MINUTES
PLATELET
RICH PLASMA
E 3 PLATELET
FREE PLASMA
FIG. 1. Thrombin generation in normal
platelet-rich and platelet-free plasma.
P R P * ACD
40
40n
32
32
6 0 0 / 24
6 0) 0 / ;2 4
P R P * OXALATE
%
16
16-
8"
-^-—i
4
8
1
12
r
16
-i
20
4
T I M E IN MINUTES
14,000 per c.mm.
32-
PFP
32
ACD
600/ 24-
8
12
16
20
TIME IN MINUTES
*A. 2 8 , 0 0 0 \ „ , _ m m
B. 1 4 , 0 0 0 J r
'
PFP
OXALATE
600/24-
16
16-
4
8
12
4
T I M E IN MINUTES
8
12
T I M E IN MINUTES
FIG. 2. Thrombin generation in PNH Case 3. PRP = platelet-rich plasma,
PFP = platelet-free plasma, and ACD = aeid-citrate-dextrose solution.
PRP* ACD
PRP* OXALATE
40
32-
600
600>
X
24\6-\
8
4
40"
-i
8
r
12
16
PFP ACD
1
1
1
8
12
16
20
TIME IN MINUTES
* 230,000 per c.mm.
40-1
32
PFP OXALATE
600/ 2 4 -
6 0 0 / 22 4 -
/f
1
4
20
TIME IN MINUTES
230,000 per c.mm.
32
Q--Q
1
16
16
8"
4-
4-T— - l
4
8
12 16 20
TIME IN MINUTES
~-oooo
orn-n,-g°,
,_
4
8
12 16 20
TIME IN MINUTES
FIG. 3. Thrombin generation in PNH Case 4. PRP = platelet-rich plasma,
PFP = platelet-free plasma, and ACD = acid-citrate-dextrose solution.
Jan.
1960
THROMBIN GENERATION AND CLOT RETRACTION IN PNH
PRP* ACD
40
40-1
6oa
%
PRP* OXALATE
32
32"
60Q,
24-
%
9
o
24-
Q
'
168—I
1
1
1
—i
1
4
8
12
16 20
TIME IN MINUTES
230,000 per c.mm.
1
1
40
PFP ACD
32 I
PFP OXALATE
X
6 0 0 / 24
600/24
X
1
8
12
16 20
TIME IN MINUTES
230,000 per c.mm.
40
32
1
4
16-
16
0*t
8-
8-1
o-o
o-oj
8
12
16
TIME IN MINUTES
4
-r—
12
16
20
TIME IN MINUTES
FIG. 4. Thrombin generation studies in PNH Case 5. PUP = plateletrich plasma, PFP = platelet-free plasma, and ACD = acid-citrate-dextrosesolution.
thrombin generation was delayed, as compared with that obtained with platelet-rich
plasma. These findings are in agreement with
those of Macfarlane and Biggs.11 Whereas
the highest value for the platelet-free plasma
approached the lowest value obtained with
platelet-rich plasma, even in such an instance the peak obtained with platelet-free
plasma was considerably delayed, as compared with that obtained with plateletrich plasma. In none of the paired plateletrich and platelet-free specimens were the
respective curves in close approximation.
PNH patients. The thrombin generation
test was performed on both oxalated and
ACD platelet-rich and platelet-free plasmas
obtained from PNH Cases 3 to 6. The results
are graphically illustrated in Figures 2 to 5.
Platelet-rich plasma revealed normal thrombin generation in all respects. The results
were similar whether oxalate or ACD was
used as the anticoagulant, or whether native plasma (without anticoagulant) was
used. Generation of thrombin in the plateletrich plasma of P N H Case 3 was poor. This
was probably a reflection of the low platelet concentrations, which were only 28,000
and 14,000 per cu.mm.The general configuration of the curves, however, was not significantly different from normal. Furthermore, when the concentration of platelets
in normal plasma was artificially reduced to
those present in the platelet-rich plasma of
PNH Case 3, the curves were almost identical. Case 3 has had persistent and severe
thrombocytopenia; Case 1 moderate thrombocytopenia; Case 5 mild thrombocytopenia;
and Case 6 has always had an adequate peripheral blood platelet count.6
10
Vol. S3
FLEXNER ET AL.
64
56-
PRP
ACD
48
48
40
40600,
60O
% 32-
% 32
0~r
24-
24
16 •
16
8"
8
—\
4
8
12 16 20
TIME IN MINUTES
* 2 3 0 , 0 0 0 / C. mm.
48
PRP*
OXALATE
56
PFP
4
1
1
1
1
8
12 16 20
TIME IN MINUTES
3 5 0 , 0 0 0 / c.mm.
ACD
48-
PFP
OXALATE
40"
40
24
600/32't
24-
16
16-
T
600/32-1
8-
o-o-o-o-o-o-o-o
T
- J —
8
12 16
TIME IN MINUTES
i
4
i
i
i
i
12 16 20
8
TIME IN MINUTES
FIG. 5. Thrombin generation studies in PNH Case 6. PRP = platelet-rich
plasma, PFP = platelet-free plasma, and ACD = acid-citrate-dextrose solution.
Significantly, when PNH plasma was
rendered platelet-free, thrombin generation
was conspicuously reduced, and the latent
period was appreciably prolonged. The results were similar whether oxalate or ACD
was used as the anticoagulant, or whether
native plasma (without anticoagulant) was
used. Thus, there was no difference between
the amount of thrombin generated by PNH
or normal platelet-free plasma.
Clot Retraction Studies
The results of clot retraction studies were
repeatedly normal in PNH Cases 1 and 3 to
6. A typical result is illustrated in Figure 6
(PNH Case 1). The platelet-rich plasma of
PNH Case 3 revealed normal retraction
when the platelets were concentrated to
96,000 per cu. mm. by special technic (see
"Methods and Materials"). Serial dilutions
of this platelet-rich plasma with plateletfree plasma demonstrated a normally graded
diminution of clot retraction with decreasing
platelet concentration. Furthermore, the
survival of PNH platelets at 37 C. in vitro,
as judged by platelet counts and preservation of clot retraction function, was the
same as that noted previously for normal
platelets.8
Jan. I960
THROMBIN GENERATION AND CLOT RETRACTION IN PNH
11
NORMAL
PLATELETS (1000s percmm)
160
80
40
20
0
i
!
i
PAROXYSMAL NOCTURNAL
HEMOGLOBINURIA
(Case I )
PLATELETS (1000s per cmm)
160 80
'40
20
0
FIG. 6. Clot retraction in PNH Case 1
DISCUSSION
The etiology of the thrombocytopeni a
that frequently occurs in PNH remains obscure. Crosby believes that there is no clearcut evidence for decreased production of
platelets by the marrow megakaryocytes.
This investigator also noted that when in
vitro hemolysis of PNH blood was prevented
by the use of sodium citrate as the anticoagulant, platelets apparently did not disappear:2 On the other hand, when the in
vitro hemolysis of PNH blood was permitted
t o occur by the use of dilute heparin as the
anticoagulant, the platelets either were
clumped or disappeared. Furthermore, when
in vivo hemolysis in a PNH patient with
thrombocytopenia was suppressed by the
use of dicumarol, the peripheral blood platelet concentration rose to normal values. Such
observations were interpreted as suggesting
that PNH platelets may be lysed concomitant with the hemolysis of PNH erythrocytes by the plasma lytic system.
The studies of Mendel and associates13
were interpreted as providing further evidence for this belief. These workers demonstrated that PNH platelet-free plasma could
generate virtually as much thrombin as
PNH platelet-rich plasma. Their results
were interpreted as indicating that platelet
12
F L E X N E R ET
substances are present free in the circulating
plasma of PNH patients, and that lysis of
platelets in the plasma accounts for the
thrombocytopenia frequently seen in this
disorder. By contrast, in our studies PNH
platelet-free plasma generated very little
thrombin, and the amount was of the same
order of magnitude as that produced by
normal platelet-free plasma. Furthermore,
this was true whether or not the PNH patients were in hemolytic crisis, and whether
or not peripheral blood thrombocytopenia
was present.
During the course of our investigations,
McKellar and Dacie12 also reported normal thrombin generation in platelet-poor
citrated plasma of 5 PNH patients. The
platelet-rich plasma of their PNH patients, however, did not generate quite as
much thrombin as did normal platelet-rich
plasma. Such a finding could be interpreted
as evidence for some platelet defect, particularly as this difference in behavior between normal and PNH specimens was not
present in their platelet-free plasma. In
our studies, however, the platelet concentrations of platelet-rich plasma from PNH
and normal persons were adjusted to virtually identical values (Figs. 3, 4, and 5).
Under such circumstances, PNH plateletrich plasma yielded the same degree of
thrombin generation as normal platelet-rich
plasma. There is no evidence that platelet
counts were performed on platelet-rich
plasma specimens and platelet concentrations adjusted accordingly in the studies of
McKellar and Dacie.12 Furthermore, thrombin generation in the plasma of our PNH
Case 3, with only 14,000 to 28,000 platelets
per cu. mm. was equal to that of normal
plasma with a similar thrombocyte concentration (Fig. 2). This observation provides
even more critical evidence that the contribution of PNH platelets to thrombin generation is quantitatively similar to that of
normal platelets.
AL.
Vol. S3
tion period. Similar findings were noted in
other hemolytic disorders. These workers
suggested that the results obtained by
Mendel and associates might have been
caused by a delay in separation of the PNH
red cells from the plasma.
The studies of McKellar and Dacie, however, do not preclude some damage to PNH
platelets during hemolysis, inasmuch as
direct enumeration of platelets and assessment of their function were not performed.
In our studies, platelets removed from PNH
patients, regardless of the severity of the
disorder at the time, demonstrated normal
clot retraction function. Yet these thrombocytes might merely represent a proportion
of platelets that are less sensitive to the
plasmalytic system, and consequently survived the in vivo plasmalytic reaction. Such
variability in sensitivity to lysis is well
known in the instance of PNH erythrocytes.2
Furthermore, the presence of clot retraction
function does not necessarily preclude some
degree of damage to the platelet, inasmuch
as thrombocytes with normal clot retraction
function may still not circulate normally in
a recipient.8 Finally, damage to platelets in
vivo need not necessarily result in a so-called
"hypercoagulable" state in blood and plasma
removed from the patient. Under certain experimental conditions the intravenous injection of platelet extracts has had no influence on blood coagulation modalities.3
The amount and rate of release of thromboplastic material into the blood stream may
be critical factors as to whether or not blood
coagulation will be influenced.9 Even in the
case of intravenous infusion of tissue thromboplastin into animals, the amount and rate
of infusion determined whether intravascular thromboses, a hypocoagulable state, or
no effect was seen. For these reasons the
possibility that PNH platelets are damaged
intravascularly still can not be excluded.
For the moment, the most feasible approach
would seem to be to assess the effect of PNH
hemolysis in vitro on the clot retraction
McKellar and Dacie12 also noted that function of PNH platelets exposed to this
plasma obtained from PNH blood that had reaction. Such studies are in progress in our
been incubated at 37 C. for 1 hr. possessed laboratory.
increased thromboplastic activity. They
concluded that this was owing to the liberaEven if it can be demonstrated that platetion of "non-haemolytic thromboplastic lets are damaged during the PNH hemolytic
activity" from PNH cells during the incuba- reaction, this may not necessarily indicate a
Jan.
1960
THROMBIN
GENERATION A N D CLOT RETRACTION
direct and specific effect of the plasma lytic
system on the PNH platelet. The occurrence
of thrombocytopenia in hemolytic transfusion reactions has recently been reemphasized.4'10 The most likely explanation is
that this is one of the many consequences of
the liberation of thromboplastin-like activity from the hemolyzed red cells.10 Thus,
quite apart from any specific effect of the
plasmalytic system on presumably abnormal
platelets in PNH, the thrombocytopenia
could result from the liberation of the "nonhaemolytic thromboplastic activity" of
destroyed erythrocytes or be a nonspecific
consequence of any hemolytic reaction. 10 ' 12
Generation of thrombin in platelet-rich
and platelet-free plasma of 4 patients with
paroxysmal
nocturnal
hemoglobinuria
(PNH) was demonstrated to be the same as
in normal persons. There was the normally
occurring difference in the amount of thrombin generated in each of these 2 compartments, the results being related to the platelet concentration. The ability of P N H
platelets to promote clot retraction was
normal in all 4 patients studied. This investigation provides no evidence for the hypothesis that by virtue of a stromal defect similar
to that of the PNH erythrocyte, the PNH
platelet is lysed by the same plasma system
that hemolyzes the red cells.
SUMMARIO IN INTERLINGUA
Esseva demonstrate que le generation de
thrombina in plasma ric in plachettas e in
plasma libere de plachettas non differeva
inter 4 patientes con nocturne hemoglobinuria paroxystic (NHP) e apparentemente normal subjectos de controlo. In
ambe categorias, normal variationes esseva
constatate in le quantitate del thrombina
generate. Iste variationes es relationate al
concentration del plachettas. Le capacitate
del plachettas de promover le retraction del
coagulo esseva normal in le 4 patientes con
NHP in iste studio. Le resultatos del investigation non supporta le hypothese que
plachettas de patientes con N H P ha un
defecto stromal simile a illo del erythrocytes
13
de patientes con NHP e que in consequentia
de ille defecto stromal le plachettas de patientes con NHP suffre un lyse via le mesme
systema plasmatic que es responsabile pro le
lyse del erythrocytes.
REFERENCES
1. B R E C H E R , G., AND C R O N K I T E , E . P . : M o r p h o l -
ogy and enumeration of human blood platelets. J . Appl. Physiol., 3 : 365-377, 1950.
2. CROSBY, W. H . : Paroxysmal nocturnal hemoglobinuria. Relation of t h e clinical manifestations t o underlying pathogenic mechanisms. Blood, 8: 769-812, 1953.
3. E P S T E I N , E . , AND Q U I C K , A. J . : Effect of i n -
jecting platelet extract
intravenously.
Proc. Soc. Exper. Biol. & Med., 8 3 : 453454, 1953.
4. F B I E S E N ,
SUMMARY
IN PNH
S. R., H A R S H A , W. N . , AND M C -
CROSKEY, C. H . : Massive generalized
wound bleeding during operations with
clinical and experimental evidence of blood
transfusion reactions. Surgery, 32: 620629, 1952.
5. HARTMANN, R. C . : D a t a presented a t t h e
inaugural meeting of t h e American Society
of Hematology, April 7, 1957, Boston,
Massachusetts.
6. HARTMANN,
R.
C,
AND A U D I T O R E ,
J.
V. :
Paroxysmal nocturnal hemoglobinuria. I .
Clinical studies. Am. J . Med., 27: 389-400,
1959.
7. H A R T M A N N ,
R.
C,
AUDITORE,
J.
V., AND
JACKSON, D . P . : Studies on thrombocytosis. Hyperkalemia due to release of potassium from platelets during coagulation.
J. Clin. Invest., 37: 699-707, 1958.
8. H A R T M A N N , R . C , AND C O N L E Y , C. L . : Clot
retraction as a measure of platelet function.
I. Effects of certain experimental conditions on platelets in vitro. Bull. Johns
Hopkins Hosp., 9 3 : 355-369, 1953.
9. HARTMANN, R . C , C O N L E Y , C. L., AND K R E -
VANS, J . R . : T h e effect of intravenous infusion of thromboplastin on "heparin
tolerance." J . Clin. Invest., 30: 948-956,
1951.
10. K R E V A N S , J .
R., JACKSON, D . P . ,
CONLEY,
C. L., AND HARTMANN, R. C.: The n a t u r e of
the hemorrhagic disorder accompanying
hemolytic blood transfusion reactions in
man. Blood, 12: 834-843, 1957.
11. MACFARLANE, R . G., AND B I G G S , R . : A t h r o m -
bin generation test. T h e application in
haemophilia and thrombocytopenia. J .
Clin. P a t h . , 6: 3-8, 1953.
12. M C K E L L A R , M., AND D A C I E , J . V . : T h r o m b o -
plastic activity of the plasma in paroxysmal
nocturnal haemoglobinuria. Brit. J . H a e mat., 4 : 404-415, 1958.
13. M E N D E L , J . L . , W I L L I A M S , M . J . , AND C L A P P ,
M. P . : Thrombin-generation studies on a
case of paroxysmal nocturnal haemoglobinuria. Brit. J . Haemat., 2 : 194-196, 1956.
14. S J 0 L I N , K . E . : The thrombin generation test
in the diagnosis of classical hemophilia and
Christmas disease. Scandinav. J . Clin.
& L a b . Invest., 8: 138-144, 1956.