Among the many advances in our understanding of the various physiologic phenomena,
few can match that of fibrinolysis. This is the process in which a fibrin clot is
removed by the body when its hemostatic function is completed. The fibrinolytic system
plays an important role in many of the physiologic functions of the body as well as
in the pathogenesis of many diseases. Our knowledge of this system, also known as
the plasminogen–plasmin system (P–P system), has made enormous advances from regulation
of hemostasis to the pathogenesis of a wide range of diseases including atherosclerosis,
obesity, cancer, and even autoimmune disorders, as well as neuronal degeneration.
New findings recently obtained have enormous therapeutic implications. Components
of the fibrinolytic system can be selectively targeted by new inhibitors. On the 40th
anniversary of Seminars in Thrombosis & Hemostasis , it seems appropriate to provide our readers with a historical overview of this remarkable
development, along with some interesting personal observations and an outlook into
the enormous translational potential of this system.
Early Observations
Postmortem Fibrinolysis
This phenomenon was first observed in postmortem blood by both Morgagni[1 ] in 1769 and later John Hunter[2 ] in 1794. The latter wrote:
In many modes of destroying life the blood is deprived of its power of coagulation,
as happens in sudden death produced by many kinds of fits, by anger, electricity or
lightning; or by a blow on the stomach, etc. In these cases we find the blood, after
death, not only as fluid a state as in the living vessels, but it does not even coagulate
when taken out of them.
Virchow noted that “capillary blood in the cadaver was always fluid and incoagulable
and that the blood in the veins was more often than not incoagulable.” From this astute
observation, he hypothesized that liquefaction of blood originated from the endothelium.[3 ] Morawitz observed that in sudden death, there was no fibrinogen in the blood, that
such blood contained a lysin that could destroy the fibrinogen and fibrin in normal
human blood.[4 ] Yudin made use of these discoveries and published the results of transfusion of
cadaver blood in Russia in 49 clinical cases by selecting subjects that died from
sudden death.[5 ] Blood was collected without preservatives from the jugular vein to avoid the infected
mesenteric blood and could be stored in the refrigerator for up to 4 weeks. He and
his assistants Skundina and Rusakov were able to observe the process of fibrinolysis
in postmortem blood under a microscope.
In Vitro Observations
In 1838, Denis observed that blood collected by wet-cupping first clotted and then
spontaneously dissolved in 12 to 24 hours.[6 ] In 1887, Green noted that fibrin disintegrated in saline without obvious bacterial
action.[7 ] In 1893, the term “fibrinolysis” was first given by Dastre,[8 ] who noted that the process was a source of error in the measurement of fibrin in
plasma. With dog's blood, he found an average loss of 8% in weight of fibrin on incubation
for 18 hours. While many investigators were pursuing the isolation of the enzyme “fibrinolysin”
from the plasma,[9 ]
[10 ]
[11 ] Tillet and Gardner made a notable contribution by observing that a filtrate from
streptococcus could activate the process of fibrinolysis and went on to isolate the
enzyme streptokinase (SK).[12 ] Milstone showed that the globulin fraction of plasma could produce fibrinolysis.[13 ] Kaplan and later Christensen and McLeod found that this protein is inactive itself,
but it is the precursor of the protease. The precursor of this was then termed plasminogen
and the active enzyme plasmin.[14 ]
[15 ]
[16 ] Biochemical studies led to discovery of plasminogen activators (PAs). They are naturally
occurring within our bodies as tissue plasminogen activator (tPA) and urokinase, also
known as urokinase-type plasminogen activator (uPA). They are also found in bacteria,
such as SK in β-hemolytic streptococcus and staphylokinase in staphylococcus, and
in other animals, such as desmoteplase in the vampire bat saliva. They were put to
therapeutic use for thrombolysis, with SK being first used to break down fibrinous
pleural adhesions in 1949[17 ] and in acute myocardial infarction.[18 ]
In Vivo Observations
Many studies were performed during this period in experimental animals, showing that
fibrinolytic activity can be induced by anaphylactic shock,[19 ]
[20 ] severe hemorrhage,[21 ] and electric convulsion.[22 ] In man, increased fibrinolytic activity was found during surgical operations,[23 ]
[24 ] severe hemorrhage,[21 ]
[23 ] strenuous exercise, and the injection of adrenaline.[25 ] It was believed that adrenaline might have been responsible, as conditions such
as alarming suggestions under hypnosis, anxiety in students about to take part in
examinations, and patients awaiting gastroscopy, could all induce fibrinolysis. The
finding that excessive fibrinolysis could cause major clinical bleeding was observed
in many disorders, and was referred to as “fibrinolytic purpura.”[26 ] It was seen in transfusion reactions,[27 ] severe burns,[28 ] metastatic carcinoma of the prostate,[29 ] obstetrical complications such as abruption placentae and amniotic fluid embolism,[30 ]
[31 ]
[32 ]
[33 ] and in many types of surgical operations,[33 ]
[34 ]
[35 ] some complicated with fatal intraoperative hemorrhage. In liver diseases, spontaneous
plasma fibrinolysis was first noted by Goodpasture.[36 ] He devised a simple test for fibrinolysis by observing the dissolution of a clot
formed from recalcified blood over 24 hours. This “Goodpasture test” was used for
many years in North America. Soon after, Ratnoff confirmed presence of fibrinolytic
activity in cirrhosis but not in acute hepatitis.[37 ]
Personal Observations
My introduction to this topic was serendipitous. In 1955, while performing the one-stage
thrombin time, using Quick original method, as part of the work up of a cirrhotic
patient with massive intraoperative hemorrhage, I encountered difficulty in obtaining
the end point. A thin wisp of fibrin was formed but quickly disappeared under my eyes.
Such was the dramatic effect of the excessive fibrinolysis activated by surgery in
cirrhosis,[38 ]
[39 ] thus confirming observations by others.[36 ]
[37 ] As the origin of the fibrinolytic activity was unknown at the time, we used direct
approaches studying venous blood in vivo in man and found that fibrinolytic activity
could be released from veins after various stimuli.[40 ]
[41 ] Likewise, similar results were found in experimentally induced venous thrombi in
rabbits.[42 ] As discussed earlier, Virchow, by noting that blood in small blood vessels was more
likely fluid and incoagulable than blood collected from larger vessels,[3 ] had suspected that blood vessels were the origin of fibrinolysis. But our findings
were the first direct observation that fibrinolytic activity was derived from veins.
Several interesting aspects of our studies are noteworthy. First, we observed that
stimulation of one venous segment could release fibrinolytic activity from another
vein located far from the site of stimulation, indicating that the stimulus could
be transmitted via perivascular sympathetic nerves.[41 ] At the time, we were mystified and had no other explanation for this phenomenon.
Some 50 years later, Jim O'Rourke came to me at a meeting and excitedly told me that
he had demonstrated that the perivascular sympathetic pathway was indeed responsible
for this signal transmission.[43 ]
[44 ]
In our studies on animals, experimentally induced venous thrombi were produced in
the marginal veins of rabbit's ears.[42 ] Parallel studies using these stimuli were done using the lysis of the thrombi as
the end point. The findings verified that we had the ones observed in the human veins.[42 ]
[45 ]
Next Five Decades (1960–2010)
While Astrup and his coauthors were able to show the amounts of PA in different organs
by extraction,[46 ]
[47 ]
[48 ]
[49 ]
[50 ]
[51 ]
[52 ]
[53 ]
[54 ] the pursuit of knowledge soon turned to finding components of the P–P system at
the cellular level. A fibrin slide method for histologic localization was first designed
by Todd,[55 ] who used a modification of Astrup fibrin plate method for pinpointing areas of lysis
in histologic sections. With this method, fibrinolytic activity was localized to the
endothelium in both normal[56 ]
[57 ]
[58 ]
[59 ] and pathologic tissues.[60 ]
[61 ]
[62 ]
[63 ] It appeared that the activity was most intense in young regenerating endothelial
cells, as shown in newly formed capillaries growing with granulation tissues,[59 ] in revascularized myocardium following infarction[60 ] and in coronary atherosclerotic lesions.[62 ] These findings were subsequently confirmed by more sophisticated methods such as
in situ hybridization.
Using cell cultures in vitro, studies of the PAs from various cells were made to determine
their function in physiology and in pathology.[64 ] These studies revealed that fibrin is not the sole substrate for plasmin. As a protease,
plasmin can break down extracellular matrix, thereby enabling cell movements. Plasmin
can also activate latent metalloproteinases. Thus, plasmin participates in a wide
range of processes that involve cell migration. As the role of plasmin is not limited
to the lysis of fibrin, the fibrinolytic system is more appropriately referred to
as the P–P system.[65 ] This period of development is also noted for the discovery of other members of the
P–P system,[65 ]
[66 ] including cell surface receptor for uPA (uPAR),[67 ] for tPA, known as annexin A2[68 ]
[69 ]
[70 ] as well as a surface protein S-100A10 that is colocalized on the cell surface with
receptors for plasminogen.[71 ]
[72 ] In addition, the inhibitors of fibrinolysis were also identified, including several
PA inhibitors (PAI),[73 ]
[74 ]
[75 ]
[76 ] of which PAI-1 has an especially important biological role. The present day concept
of the P–P system is shown in [Fig. 1 ].
Fig. 1 The present day concept of the plasminogen–plasmin system. The factors listed on
the left are activators and those on the right (in italics) are inhibitors, kept in
balance under physiologic conditions. APC, activated protein C; AT, antithrombin;
PAI, plasminogen activator inhibitor; TAFI, thrombin-activable fibrinolysis inhibitor;
tPA, plasminogen activator; uPA, urokinase-type plasminogen activator.
Through the proteolytic action of plasmin, the P–P system was found to be regulating
many physiologic processes, including embrogenesis,[77 ] ovulation,[78 ]
[79 ] neuron growth,[80 ]
[81 ] brain function,[82 ]
[83 ]
[84 ] catecholamine secretion,[85 ] activation of inflammatory cells,[86 ] wound healing,[87 ]
[88 ] and skeletal muscle regeneration.[89 ] Notable advances were made in linking the function of this system with the pathogenesis
of a wide range of diseases, among which are cancer[90 ]
[91 ]
[92 ] and vascular diseases.[93 ] Many new thrombolytic agents were also developed.[94 ]
[95 ] Much information was made available in several reviews[65 ]
[96 ] including an issue of the Seminars of Thrombosis & Hemostasis in 1991[97 ] and more recently in 2013.[98 ] In the following sections, progress in several of these developments has been selected
to be reviewed in greater depth.
Cancer
The earliest observation of association of fibrinolysis and cancer was made by Carrel
and Burrows, who observed liquefaction of growth media by malignant tumors,[99 ] whereas clinical record of fibrinolytic bleeding was made by Tagnon et al in patients
with metastatic carcinoma of prostate.[29 ] Evidence of a possible causative role of PA in malignancy was shown by the sharp
increase in fibrinolytic activity in viral transformed fibroblasts.[100 ] This PA was identified by Astedt and Holmberg as uPA.[101 ] Subsequently, uPA, uPAR, and PAI-1 were also found by tissue extraction, immunohistochemical
staining, and in situ hybridization to be greatly increased in many forms of cancer.
These include cancer of the breast, stomach, colon and rectum, esophagus, pancreas,
glioma, lung, kidney, prostate, uterine cervix, ovary, liver, and bone.[92 ]
[100 ]
[102 ] The P–P system participates in multiple steps in cancer from carcinogenesis to growth
and metastasis.[92 ]
[103 ]
[104 ] The complex interactions involved in these processes are beyond the scope of this
article.
The levels of uPA, uPAR, and PAI-1 have been found to be correlated to the aggressiveness
and the metastatic potential of many tumors both in tumor cell cultures and in tumor
tissues.[104 ]
[105 ] These are used as biomarkers in the risk stratification of several cancer, especially
in carcinoma of breast. Higher levels of uPA and PAI-1 are associated with worse prognosis
in carcinoma of breast.[106 ]
[107 ]
[108 ]
[109 ] The incorporation of the uPA/PAI-1 status into the treatment algorithms has been
shown to be helpful in deciding which patients can be spared from the more aggressive
chemotherapy.[110 ] Clinical validation of the usefulness of these biomarkers are currently being performed
in other types of cancer as well.[104 ]
[105 ]
[111 ]
[112 ]
[113 ] In carcinoma of pancreas, the postoperative survival of those with high expression
of uPA and uPAR was 9 months compared with 18 months in those without expression of
both markers or of only one marker.[114 ] In small cell carcinoma of lung, those with high levels of uPAR predicted poor response
to chemotherapy.[115 ] More studies are needed to verify the clinical utility of these biomarkers.
The use of inhibitors in retarding tumor growth has some successes in experimental
animals. For example, transfection of PAI-1 to prostate cancer cells impaired growth
and metastasis in mice.[116 ] Anti-uPAR antibody blocks prostate cancer invasion, migration, growth, and experimental
skeletal metastasis in vitro and in vivo.[117 ] The development of drugs targeting uPA, uPAR, and PAI-1 has become an exciting area
of investigation. A wide spectrum of monoclonal antibodies, targeted toxins, synthetic
small molecules and peptides, and antisense molecules are now known to have antitumor
effects in human cancer. Several promising drugs include a small molecule targeting
the active site in the S1 pocket of uPA. This agent, known as WX 671, has antitumor
activity in carcinoma of head and neck,[118 ] pancreatic carcinoma,[119 ] and carcinoma of breast.[120 ] Hopefully in the near future, more agents will become available, and prove to be
effective as anticancer agents.
Thrombolysis
The earliest thrombolytic agent was SK, an activator of plasminogen.[12 ]
[17 ]
[121 ]
[122 ] It was first used to breakdown fibrinous pleural exudates.[17 ] This same agent was then employed in thrombolysis of experimental clots in rabbits,[123 ] and SK-activated plasmin (also known as “fibrinolysin”) was subsequently used to
lyse experimental venous and arterial thrombi in animals.[124 ]
[125 ] In man, SK-activated plasmin was found to be effective in lysis of thrombi in several
clinical studies.[18 ]
[126 ]
[127 ]
[128 ] One notable study was performed on human volunteers, in whom experimental thrombi
were produced by a dental broach in an arm vein,[129 ] a feat unlikely to be repeated today.
The delivery of the thrombolytic agents was for many years performed by the intravenous
route. Before long, this method was found to be problematic, as the plasmin was rapidly
inhibited by the circulating antiplasmin. The next phase was the development of several
PAs, including tPA and uPA. Again, their therapeutic life span was short lived as
they are also promptly inhibited by PAI-1 and antiplasmin. The inhibition, however,
can be reduced if the agent is rapidly bound to the fibrin thrombus thereby improving
its thrombolytic efficacy.[94 ] Many mutants of recombinant tPA were developed to improve these pharmacologic features.
Successful thrombolysis depends on age and content of the thrombus, and accessibility
of the thrombolytic agent. One notable example for the importance of early thrombosis
is that of ischemic stroke, with convincing evidence that brain function restoration
occurs only in early lysis within 3 hours.[130 ]
[131 ]
[132 ]
To circumvent the inhibitors of fibrinolysis, the thrombolytic agents can be delivered
by a catheter directly to the thrombus. Direct delivery of thrombolytic agents to
the occluded coronary artery was first performed by Boucek and Murphy in 1960[132 ] and in peripheral arterial thrombosis by Dotter in 1974.[133 ] The practice today uses improved techniques such as percutaneous endovascular insertion
of a catheter advanced to the site of the thrombus, under direct radiologic imaging.
The catheter can be left in place for a slow delivery of the agent.[134 ]
[135 ] Catheter-directed thrombolysis can also be performed in conjunction with thrombectomy
or with high-frequency low-intensity ultrasound waves to accelerate clot dissolution
by dissociating the fibrin strands.[135 ] Today, the indications for thrombolysis cover practically all forms of thrombi.
Further discussion of this topic is beyond the scope of this article.
Inhibitors of PAI-1 as New Drug Target
Among the various components of the P–P system, PAI-1 has been found to be involved
in the pathogenesis of a variety of disorders.[136 ]
[137 ]
[138 ]
[139 ] These include thrombotic disorders, cancer, metabolic syndrome and obesity, polycystic
ovarian disease, alopecia, pulmonary fibrosis, nephrosclerosis, and myelofibrosis.
It is also implicated in aging and Alzheimer disease. Thus, intensive efforts are
being performed in the search for an inhibitor of PAI-1. Both neutralizing monoclonal
antibodies against PAI-1[140 ] and small peptide molecules[141 ]
[142 ] have been developed. Their clinical activities are yet to be determined.
Conclusion
Since the observations of postmortem fibrinolysis over two centuries ago, our understanding
of the fibrinolytic system has evolved immensely. From having a limited function of
dissolution of a fibrin clot, the components of the P–P system are now known to be
involved in many physiologic and pathologic processes. Such knowledge has enabled
the development of effective treatment of many thrombotic disorders. It will be of
great interest to watch further development of new drugs based on our increasing knowledge
of the action of the components of this system, in particular of PAI-1.