1. European HIT mastercases
Paris, May 18th 2017
Guidelines on heparin-induced thrombocytopenia: what is missing for their clinical implementation?
F.Mullier
Université catholique de Louvain, CHU UCL Namur
Namur Thrombosis and Hemostasis Center

2. Outline
Summary of Guidelines (from 2009 to nowadays): Many GRADE 2C, some disagreement between ACCP and BCSH
Absence of adherence to ACCP Guidelines
Underuse of platelet count monitoring
Underutilization of 4T score and overutilization of screening assays
Underutilization and limitations of functional assays
Treatment of HIT without a thrombotic complication: prophylactic or therapeutic?
Underdiagnosis
Overdiagnosis
Conclusions and perspectives

3. Summary of Guidelines
Summary of Guidelines (from 2009 to nowadays)
- ACCP (Linkins LA et al. Chest Feb;141(2 Suppl):e495S-530S, Junqueira DR.Chest Apr;143(4):1190-1).
BCSH (Watson H, et al Br J Haematol Dec;159(5):528-40)
Consensus European meeting (Alatri A, Thromb Res Apr;129(4):426-33)
Many weak recommendations with low level of evidence (GRADE 2C)

4. Absence of adherence to ACCP Guidelines
the CATCH Registry (Crespo EM et al. Am Heart J 2009;157:651-7)
Among a subset of patients whose clinical presentation suggested they were at high risk for HIT, suspicion for HIT was uncommon (prolonged (for ≥96 hours) heparin stratum 19.8%, cardiac care unit stratum 37.6%) and often did not arise until ≥1 day after patients developed thrombocytopenia.
Often patients were not evaluated for HIT until after they had had a thromboembolic complication (prolonged heparin stratum 43.8%, CCU stratum 61%).
Even after HIT was suspected, patients often continued to receive heparin.
Direct thrombin inhibitor use was infrequent (prolonged heparin stratum 29.4%, CCU stratum 35.6%). »
Crespo EM et al. Am Heart J 2009;157:651-7

5. Guidelines: Baseline platelet count
Chest: Obtaining a baseline platelet count prior to initiating anticoagulant therapy for a patient with VTE is considered standard medical practice (no GRADE)
BCSH: Patients who are to receive any heparin should have a baseline platelet count (2C).
Linkins LA et al. Chest 2012 Feb;141(2 Suppl):e495S-530S.
Watson H et al. Br J Haematol 2012 Dec;159(5):

6. Guidelines: platelet count monitoring
For patients receiving heparin in whom clinicians consider the risk of HIT to be > 1%, we suggest that platelet count monitoring be performed every 2 or 3 days from day 4 to day 14 (or until heparin is stopped, whichever occurs first) (Grade 2C) .
Linkins LA et al. Chest 2012 Feb;141(2 Suppl):e495S-530S.

7. Guidelines: platelet count monitoring
For patients receiving heparin in whom clinicians consider the risk of HIT to be <1%, we suggest that platelet counts not be monitored (Grade 2C).
Linkins LA et al. Chest 2012 Feb;141(2 Suppl):e495S-530S.

8. Guidelines: platelet count monitoring
Post-operative patients including obstetric cases receiving unfractionated heparin (UFH) should have platelet count monitoring performed every 2–3 d from days 4 to 14 or until heparin is stopped (2C).
Post-cardiopulmonary bypass patients receiving low molecular weight heparin (LMWH) should have platelet count monitoring performed every 2–3 d from days 4 to 14 or until heparin is stopped (2C).
Post-operative patients (other than cardiopulmonary bypass patients) receiving LMWH do not need routine platelet monitoring (2C).
Post-operative patients and cardiopulmonary bypass patients who have been exposed to heparin in the previous 100 d and are receiving any type of heparin should have a platelet count determined 24 h after starting heparin (2C).
Medical patients and obstetric patients receiving heparin do not need routine platelet monitoring (2C).
Watson H et al. Br J Haematol 2012 Dec;159(5):

9. Underuse of platelet count monitoring
Ten Berg MJ et al. Ann Pharmacother 2009;43:

10. Guidelines: Laboratory diagnosis
Chest: A negative ELISA or PaGIA in a patient with a low pretest probability of HIT excludes the diagnosis of HIT. A positive ELISA or PaGIA in a patient with a low pretest probability of HIT should not be interpreted as diagnostic for HIT and requires confirmation with a functional assay. (no GRADE)
BCSH:HIT can be excluded by a low pre-test probability score without the need for laboratory investigation (2B).  disagreement between Chest and BCSH
BCSH: If the pre-test probability of HIT is not low, heparin should be stopped and an alternative anticoagulant started in full dosage whilst laboratory tests are performed (1C).
Linkins LA et al. Chest 2012 Feb;141(2 Suppl):e495S-530S.
Watson H et al. Br J Haematol 2012 Dec;159(5):

11. Guidelines: Laboratory diagnosis
BCSH: HIT can be excluded in patients with an intermediate pre-test score who have a negative particle gel immunoassay (2B).
BCSH: HIT can be excluded in all patients by a negative antigen assay of high sensitivity (1A).
Linkins LA et al. Chest 2012 Feb;141(2 Suppl):e495S-530S.
Watson H et al. Br J Haematol 2012 Dec;159(5):

12. Guidelines: Laboratory diagnosis
BCSH: Non-expert laboratories should use an antigen assay of high sensitivity. Only the IgG class needs to be measured. Useful information is gained by reporting the actual optical density, degree of inhibition by high dose heparin, and the cut-off point for a positive test rather than simply reporting the test as positive or negative (1B).
BCSH: In making a diagnosis of HIT, the clinician’s estimate of the pre-test probability of HIT, together with the type of assay used and its quantitative result [enzyme-linked immunosorbent assay (ELISA) only] and information on reversal using higher doses of heparin should be used to determine the post-test probability of HIT (2B).
Watson H et al. Br J Haematol 2012 Dec;159(5):

13. Underutilization of 4 T score and overutilization of screening assays
Samhouri Y. J Community Hosp Intern Med Perspect  2016 Sep 7;6(4):32522.  .

14. Performances of the 4T score
NPV of a low 4T score remained high irrespective of the party responsible for scoring, the prevalence of HIT
Cuker A et al. Blood 2012 Nov 15;120(20):4160-7

15. Limitations of the 4T score
The PPV of an intermediate and high probability 4Ts score was limited
Patients with intermediate and high probability scores require further evaluation.
Cuker A et al. Blood 2012 Nov 15;120(20):4160-7

16. Limitations of the 4T score
Limited sensitivity:
Low 4Ts score  1.5% of HIT ( patients who present with, thrombosis in the absence of thrombocytopenia, missing platelet count data points)
Linkins et al. 81.3% (95% CI, %) and specificity was 63.8% (95% CI, %).
Limited interobserver agreement (Linkins et al. κ = 0.43; 95% CI, ).
The fourth domain of the 4Ts score yielded the greatest disagreement among raters.
M. Crowther et al. J Crit Care 2014 Jun;29(3):470.e7-15
Linkins LA et al. Blood 2015 Jul 30;126(5):

17. Limitations of the 4T score: « Q4 Other »
Definite drugs: Quinine, quinidine, trimethoprim/sulfametoxazole, vancomycine, penicilline, rifampicine, carbamazepine, ceftriaxone, ibuprofen, mirtazapine, oxaliplatin, suramin, abciximab, tirofiban, eptifibatide and heparin
Arnold DM J Thromb Haemost 2013;11:

18. Limitations of the 4T score
Sometimes difficult to calculate the 4T score:
missing platelet count data points
unclear history of recent heparin exposure
Nevertheless, the frequency of HIT in patients with a moderate and/or high PTP based on the 4Ts was sufficiently high to support discontinuation of heparin and consider instituting an alternate anticoagulant.
Favaloro EJ Blood Jul 30;126(5):563-4.

19. A.Greinacher N Engl J Med. 2015 Jul 16;373(3):252-61.

20. Guidelines: Laboratory diagnosis
BCSH: Platelet aggregation assays using platelet-rich plasma (PRP) lack sensitivity and are not recommended (2C).
BCSH: Platelet activation assays using washed platelets [heparin- induced platelet activation assay (HIPA) and serotonin release assay (SRA)] have a higher sensitivity than platelet aggregation assays using PRP and are regarded as the reference standard, but are technically demanding and their use should be restricted to experienced laboratories (2C). other limitations of SRA (delay, lack of standardization (cut-off, positive and negative controls,..) borderline, intermediate or non-specific results  risk of false negative and false positive)
Watson H et al. Br J Haematol 2012 Dec;159(5):

21. Guidelines: Treatment with danaparoid
BCSH: Danaparoid in a therapeutic dose regimen is a suitable alternative anticoagulant for use in patients with HIT (1B).
BCSH: Danaparoid at prophylactic doses is not recommended for the treatment of HIT (1B).
BCSH: Monitoring the anticoagulant effect of danaparoid using an anti-Xa assay with specific danaparoid calibrators should be considered in patients >90 kg and in patients with renal impairment (glomerular filtration rate <30 ml/min) (2C).target range : 0,5-0,8 anti Xa units/ml? .
Watson H et al. Br J Haematol 2012 Dec;159(5):

22. Guidelines: Treatment with argatroban
BCSH: An argatroban infusion adjusted to an activated partial thromboplastin time (APTT) ratio of 1.5–3.0 (but not exceeding 100 s) is a suitable alternative anticoagulant for the treatment of patients with HIT (1C). ACCP: no mention of a maximum of 100 s
“Hemoclot(®) thrombin inhibitor assay should be considered in patients on argatroban, particularly if there is concern the APTT may not be reflective of the degree of anticoagulation with argatroban due to other factors including coagulopathies in critically ill patients, the presence of a lupus anticoagulant or very high FVIII levels”  APTT: risk of underdosing of argatroban with risk of thrombosis
No consensus on the therapeutic range: 0,25–1,5 µg/ml? .
Watson H et al. Br J Haematol 2012 Dec;159(5):
Guy S et al. Int J Lab Hematol 2015 Dec;37(6):834-43
Guy S et al. Br J Haemato 2016 Oct 26. doi: /bjh
Tardy-Poncet B et al. Critical Care 2015, 19, 396–407.

23. Guidelines: treatment of HIT without thrombotic complication
Patients should be therapeutically anticoagulated for 3 months after HIT with a thrombotic complication (1A) and for 4 weeks following HIT without a thrombotic complication (2C).
Therapeutic-dose for treatment of HIT without complication is based one non-randomized retrospective study and retrospective studies:
Klarenbeek and Eikenboom: data from randomized trials that directly compare prophylactic regimens with therapeutic regimems are lacking
A.Greinacher:
risk of new symptomatic thrombosis among patients who did not receive anticoagulation was 5 to 10% per day over the 2 first days of treatment and 38% to 53% at 1-month follow-up
%thrombosis prophylactic > therapeutic
Prophylactic dose with regular assesment for new thrombosis is reasonable in patients with a high risk of bleeding or in patients with only a low or intermediate likelihood of HIT
Linkins LA et al. Chest 2012 Feb;141(2 Suppl):e495S-530S.
Greinacher A. N Engl J Med Jul 16;373(3): and N Engl J Med Nov 5;373(19):
Klarenbeek NB, Eikenboom JC, N Engl J Med 2015 Nov 5;373(19):   

24. New treatments: DOACs
Promising but lack of information about efficacy/safety
Linkins (multicenter, single-arm, prospective cohort study, rivaroxaban): One patient had recurrent thrombosis, one HIT-positive participant required limb amputation despite platelet recovery, and 9/10 patients with thrombocytopenia had platelet recovery)
Clinical settings: initial management of massive or submassive pulmonary embolism, limb-threatening arterial thrombosis, vena caval thrombosis, or cerebral thrombosis
Optimal duration of therapy : not known (at least 3 months)
Idarucizumab is available but andexanet alpha and ciraparantag are not yet approved
Tran PN, Tran MH, Clin Appl Thromb Haemost 2017
Ong SY et al. Ann Hematol Mar;96(3): ,
Linkins LA et al J Thromb Haemost 2016 Jun;14(6):

25. Underdiagnosis of HIT Lack of epidemiological data about HIT
Underuse of platelet count monitoring and determination of clinical scoring system
The characteristic timing of onset of thrombocytopenia is sometimes absent (delayed-onset HIT, persisting HIT, spontaneous HIT syndrome, fondaparinux-associated HIT (Warkentin Thromb Haemost  2016 Oct 28;116(5): ), Protamine (heparin)-induced thrombocytopenia:J. Thromb Haemost 2016 Sep;14(9):  

26. Overdiagnosis of HIT
Thrombocytopenia in the setting of a proximate heparin exposure is common
Low specificity of HIT antibodies
Long turn around time for ELISAs and reference functional assays (SRA, HIPA)
Reference: expert opinion or serotonin release assay?
Economic burden from unnecessary HIT testing and treatment
Cuker A et al. Thromb Haemost 2011; 106: 993–994,
Harada MY et al Crit Care Med 2017 Jan;45(1):28-34.
Cuker J Thromb Haemost 2017 Feb;15(2):
McMahon CM, Tanhehco YC, Cuker A, J Thromb Haemost 2017 Feb;15(2):   
Tardy B J Thromb Haemost 2011 Aug;9(8):1667-9
Smythe MA Thromb Haemost 2012 Nov;108(5):992-8 .

27. Solutions (1/3)
Avoid-heparin program (Mc Gowan KE, Blood 2016 Apr 21;127(16):1954-9)
Improvement of the identification of HIT
Educational campaigns (Warkentin TE. Think of HIT. Hematology Am Soc Hematol Educ Program 2006: 408– Horner BM, Myers SR. Don’t miss HIT (heparin induced thrombocytopenia). Burns 2004; 30: 88–90.)
Alerts (Riggio et al J Thromb Thrombolysis 2009, Hanatani T, J Clin Pharm Ther 2013 Oct;38(5): Smythe MA, Am J Health Syst Pharm 2012 Feb 1;69(3):241-8, Hasan M, J Thromb Thrombolysis 2016 Oct;42(3):441-6.)
Communication with the physician to calculate 4T score before ordering HIT screening assays (Burnett AE et al. J Thromb Thrombolysis 2016 Nov;42(4):471-8)
Communication between pharmacy ordering systems and informatic laboratory system

28. Alerts Increased laboratory testing (functional assays)
Changes in clinical reactions to decreasing platelet counts
Riggio et al J Thromb Thrombolysis 2009

29. Solutions (2/3)
Use of screening assays with high specificity (only IgG specific antibodies, high-dose heparin confirmation step, determination of the H/PF4 antibody titre) and shorter TATs
Development of apps which combine clinical scoring systems with immunoassays (Cuker A. Blood  2016 Feb 4;127(5):522-4)
Development of available, standardized functional assays (like HIMEA) with shorter TAT (Morel-Kopp MC et al. J Thromb Haemost 2016 Dec;14(12):  )

30. Solutions: Combination of scoring systems with immunoassays
Nagler M et al. Blood  2016 Feb 4;127(5):

31. Solutions: Combination of scoring systems with immunoassays
Nagler M et al. Blood  2016 Feb 4;127(5):

32. Solutions (3/3)
Cuker A. Blood  2016 Feb 4;127(5):522-4

33. Conclusions and perspectives
Guidelines: Many GRADE 2C, some disagreement between ACCP and BCSH  Clear guidances are needed to improve adherence
Platelet count monitoring: baseline, D5,D7,D9
Clinical scoring system: improvement of 4T score is required (especially in ICU, hemodialysis)
Integration of clinical scoring systems with immunoassays is an attractive option
Prospective evaluation of HIMEA consensus protocol is required
Additional data are required about treatment of HIT especially for HIT without thrombotic complication

34. Questions? Comments?