Heparin-Induced Thrombocytopenia

Heparin-induced thrombocytopenia has appeared several times in the CICM Fellowship Exam, both in the written paper (Question 9.1 from the second paper of 2017, Question 23 from the second paper of 2012) and as a viva station. In view of this, it has been apportioned its own revision chapter. As far as reading about this subject goes, the most effective use of one's time would be to read the 2015 review by Greinacher (NEJM). This article has added credibility because is Andreas Greinacher has been publishing HITS-related material since the nineties, and is something of a guru in the field. If the exam candidate has no constrains on their time or attention span, they can delve deeper with the somehow freely available third edition of Heparin-Induced Thrombocytopenia (2004) by Warkentin and Greinacher. Its not clear whether it was posted in this way intentionally.  

Definition and classification of HITS and HITTs

Correct nomenclature of HIT or HITT is hard to pin down. Apparently in 1998 a consensus statement of representatives from three research groups (Greinacher was one of them) had finally nailed down some sort of laboratory definition, but this statement is not available anywhere. 

For example, some authors separate HIT into types (Type 1 is not antibody mediated, whereas Type 2 is). UpToDate certainly makes that distinction. HITT is then a flavour of HIT Type II, which has both antibody-mediated thrombocytopenia and thrombosis. On the other hand, the NEJM reviews from 2015 or 2006 do not mention types or even HITT for that matter. The HIT to which Greinacher refers to in his 2015 article seems to be HITT by others' definitions (to quote him, "immune-mediated heparin-induced thrombocytopenia (HIT) ...results in a paradoxical prothrombotic state"). Others call HIT Type 1 just "HIT" and Type 2 automatically gets an extra "T" to indicate the prothrombotic tendency, as in Patel et al (2007) - "HIT is a benign clinical condition characterized by a mild drop in platelet count with no clinical significance. HITT is an immune-mediated reaction associated with a widespread “hypercoagulable” state resulting in arterial and venous thrombosis."

In short, it is probably best to know one system and to stick to it.

HIT Type I

• Mild transient thrombocytopenia, platelet count above 100
• Totally reversed by heparin cessation
• Occurs in up to 10% of patents
• NOT associated with an increased risk of thrombosis
• Probably not even immune in origin (likely it is some sort of pre-aggregatory effect on the platelets)

HIT Type II

• Nasty severe thrombocytopenia, platelet count might drop to nil
• Occurs in something like 1% of patents
• Associated with thrombosis in 30% of cases
• Requires anticoagulation

HIT Type II can be further split into several recognisable varieties:

• latent: antibodies without thrombocytopenia;
• HIT: antibodies with thrombocytopenia; and
• HITT: antibodies with thrombocytopenia and thrombosis.

Pathophysiology of heparin-induced thrombocytopenia

In the 2004 book by Warkentin and Greinacher, a chapter (Ch.7, Visentin et al) offers a detailed look at the molecular immunology of HIT. In summary:

• It is IgG mediated
• The antibodies form rapidly (within days)
• These antibodies are directed against PF4 complexed with heparin (PF4 is "Platelet Factor 4", a little cytokine which is normally released from platalet α-granules and which is meant to control the influence of various natural heparin-like molecules on platelet aggregation)
• When anti-PF4 antibodies bind to PF4 it  activating more platelets and triggers the release of more PF4, providing more substrate for antibody binding in a positive feedback loop
• The IgG-coated platelets are then removed by macrophages in the reticuloendothelial system. Thrombocytopenia ensues.
• The fact that anti-PF4 antibodies can activate platelets also leads to thrombosis (both arterial and venous); a second mechanism is the consumption of platelets in thrombi. The actual mechanism of this prothrombotic tendency is apparently unknown. 

Classical features and course of heparin-induced thrombocytopenia

Onset usually is within 5-10 days of heparin commencement. However:

• Major surgery can "reset the clock" - even patients chronically exposed to heparin can develop HIT 5-10 days after major surgery
• Rapid-onset HIT can occur within 90 days of heparin exposure, upon re-exposure
• HIT can continue getting worse after heparin has been discontinued

The 4T score, or "Four Score"

Basically it is a system of  four criteria, of which only three are really Ts: Thrombocytopenia, Timing, Thrombosis and "other causes of thrombocytopenia" A score of 0, 1 or 2 is attributed to each criterion. The table below has been stolen shamelessly from the 2015 Greinacher article.

 Lo et al (2006) evaluated this system and found that the maximum score of 8 was associated with "real" laboratory-confirmed HIT 100% of the time, whereas a low score essentially excluded it as a possibility. In fact a score of less than 4 has a negative predictive value of 97-99%.  The UpToDate authors have recommended that a low score should discourage one from pursuing any sort of formal laboratory testing.

Laboratory diagnosis of heparin-induced thrombocytopenia

There are two major tests for HIT. You can either try to detect the antibodies to PF4, or perform a functional platelet activation assay to measures the ability of patient serum to activate test platelets in the presence of heparin. A negative functional assay excludes HIT and is the gold standard, but is difficult to run, takes several days and requires a specialised laboratory. The anti-PF4 ELISA immunoassay is faster, but ore useless. The detection of a low level of anti-PF4 antibodies will be interpreted as a positive test, but the patient will not have clinically significant disease. This was demonstrated very well by Warkentin et al (2000), who found that among screened patients only 2-15% of patients who tested "positive" actually went on to develop any sort of clinically significant HITs.

In summary:

• Test the patients with the immunoassay only if there is a high pre-test probability, i.e. if the 4T score is high. 
• Perform a functional assay only if the 4T score is high and the immunoassay is equivocal

In Question 9.1 from the second paper of 2017, the ICU registrar has sent a premature HIT screen in a patient with low pre-test probability, which had come back positive. This result is to be ignored.