CONTENTS

• Rapid Reference 🚀
• Overview & definition
• Differentials
  • Overall classification of thrombocytopenia in the ICU
  • Clinical clues & specific situations
• Investigation of thrombocytopenia
• Platelet transfusion
• Additional comments on specific causes of thrombocytopenia:
  • Drug-induced immune thrombocytopenia (D-ITP)
  • Transfusion related thrombocytopenia
  • Heparin-Induced Thrombocytopenia
• Podcast
• Questions & discussion
• Pitfalls

---

rapid reference

(back to contents)

---

approach to thrombocytopenia in critical illness ✅

investigations (more)

• Exam for signs of bleeding (e.g., purpura).
• Basic Thrombocytopenia Panel:
  • CBC in citrated blood tube (blue top).
  • INR, PTT, fibrinogen, D-dimer.
  • Blood smear.
• May consider also:
  • Heparin-PF4 antibody (only if 4T-score is 4-8).
  • Ultrasonography to evaluate for DVT.
  • Ferritin if concern for HLH.📖
  • Liver function tests if ? cirrhosis.
  • TEG may be helpful in more complex cases.

immediate concerns:  

• (1) Consider immediate risk of bleeding:
  • Review all anticoagulating medications and consider discontinuation (especially antiplatelet agents).
  • Consider platelet transfusion.📖
• (2) Consider HIT.📖
• (3) If schistocytes are present, consider TTP.📖

---

overview & definition

(back to contents)

---

overview

• Thrombocytopenia is extremely common in the ICU.  It is often a poor prognostic sign that is associated with systemic inflammation.
• Most cases of thrombocytopenia will resolve in parallel with the patient's overall recovery.

when to initiate an evaluation?

• Most ICU patients with mild thrombocytopenia don't require an exhaustive evaluation.  Potential indications to evaluate further might include:
  • Severe thrombocytopenia (e.g., below ~50,000).
  • Features of HIT (e.g., abrupt drop in platelet count by >50%, skin necrosis at site of heparin injection).
  • Clinical thrombosis.
  • The underlying disease process is unclear, raising a possibility of an underlying hematologic disorder (e.g., thrombotic thrombocytopenic purpura or hemophagocytic lymphohistiocytosis).

---

common causes of thrombocytopenia in the ICU

(back to contents)

---

Thrombocytopenia in critical illness may be divided into roughly four classes:

subacute/chronic thrombocytopenia:  Often present before critical illness.

• Cirrhosis.
• Alcoholism (direct marrow toxicity; can be suggested by macrocytosis; platelets should rise within 2-5 days of cessation).(31269407)  
• Bone marrow dysfunction (e.g. myelodysplastic syndrome, malignant infiltration).
• Bone marrow suppression due to medication (chemotherapy, some antiviral agents, linezolid, some penicillins/cephalosporins, trimethoprim/sulfamethoxazole, NSAIDs, thiazides).
• HCV, HIV.

non-immune consumption:  Often gradual decrease over several days, adequate response to platelet transfusion.

• Thrombotic microangiopathies (e.g., TTP, HUS).📖
• Disseminated intravascular coagulation (DIC), large PE/DVT.
• Catastrophic antiphospholipid antibody syndrome (CAPS).
• Sepsis:
  • Can occur with any pathogen.
  • Especially occurs with specific pathogens (ehrlichiosis, babesiosis, anaplasmosis, rocky mountain spotted fever, hantavirus, dengue).
• Surgery, trauma.
• Hemophagocytic lymphohistiocytosis (HLH).
• Devices:  intra-aortic balloon pump, hemodialysis, ECMO.

immune consumption:  Often occurs several days after ICU admission, decrease is rapid (>50% fall in 24-48 hours), often with severe thrombocytopenia which is poorly responsive to platelet transfusion.

• HIT.
• Drug-induced immune thrombocytopenia (D-ITP).
• Post-transfusion purpura.
• Passive alloimmune thrombocytopenia.
• Idiopathic thrombocytopenic purpura (ITP)(diagnosis of exclusion).

other

• Obvious causes:
  • Massive transfusion.
  • Hypothermia.
• Pseudothrombocytopenia:
  • This refers to artificially low platelet count due to in vitro aggregation induced by EDTA in blood tubes.
  • Pseudothrombocytopenia may be suggested by finding clumped platelets on blood smear.  Platelets may also coat the surface of neutrophils (platelet satellitism).(30579402)
  • Definitive diagnosis is based upon measuring a higher platelet count when blood is drawn in a citrated blood tube (“blue top platelet tube”).
  • Pseudothrombocytopenia is usually meaningless, but can be associated with lupus, vasculitis, or lymphoma.(28808508)

---

clinical clues & specific situations

(back to contents)

---

The differential diagnosis of thrombocytopenia is broad.  Below are some clues which may help point in the right direction.  These shouldn't be used to narrow the differential diagnosis, but rather merely to highlight possibilities that deserve particular attention.

thrombocytopenia with (paradoxical) clinical thrombosis

• Heparin-Induced Thrombocytopenia (HIT).
• Disseminated intravascular coagulation (DIC) of a pro-thrombotic type (e.g. associated with malignancy or sepsis)
  • Acute DIC/liver necrosis/limb necrosis syndrome:  Shock liver causes depletion of protein C and antithrombin, with subsequent microvascular thrombosis of extremities.(26539932)
• Catastrophic antiphospholipid antibody syndrome (CAPS)
• “Massive clot thrombocytopenia” – Venous thromboembolic disease itself may cause mild thrombocytopenia.(28808508, 27982413)
• Thrombotic thrombocytopenic purpura (TTP)
• Antiphospholipid antibody syndrome (APLS)
• Cirrhosis, in some patients
• Paroxysmal nocturnal hemoglobinuria (rare)

severe thrombocytopenia (e.g., <20,000)(29222318)

• Immune-mediated thrombocytopenia of any cause:
  • Drug-induced immune thrombocytopenia (D-ITP)
  • Transfusion-related (post-transfusional purpura, passive alloimmune thrombocytopenia)
  • HIT causes severe thrombocytopenia in ~10% of cases (generally when HIT causes simultaneous DIC).(25590528)
  • Idiopathic thrombocytopenic purpura (ITP)
• Thrombotic thrombocytopenic purpura (TTP).
• Hemophagocytic lymphohistiocytosis (HLH).
• Severe marrow failure (e.g. chemotherapy or leukemia).  Will usually see neutropenia here as well.
• (Usually not due solely to sepsis or DIC)

thrombocytopenia in septic-appearing patient 

• Disseminated intravascular coagulation (DIC) due to the infection.
• Specific infections (ehrlichiosis, babeosis, anaplasmosis, rocky mountain spotted fever, hantavirus, dengue).
• Hemophagocytic lymphohistiocytosis (HLH) – either mimicking sepsis, or HLH secondary to infection.📖

thrombocytopenia in the cardiac patient

• Heparin-Induced Thrombocytopenia (HIT).
• Cardiopulmonary bypass, ECMO, intra-aortic balloon pump (IABP).
• GPIIb/IIIa inhibitors.
• Thrombotic thrombocytopenic purpura (TTP) related to clopidogrel.

thrombocytopenia in pregnant patient

• Pre-eclampsia and/or HELLP syndrome (Hemolysis, Elevated LFTs and Low Platelets).
• Acute fatty liver of pregnancy.
• Thrombotic microangiopathy (including TTP, pregnancy-induced atypical HUS).📖
• Disseminated intravascular coagulation (may result from various obstetric catastrophes).
• Idiopathic thrombocytopenia (ITP).
• Gestational thrombocytopenia.

---

investigation of thrombocytopenia

(back to contents)

---

chart review

• Chronicity of thrombocytopenia?   (Obtaining baseline CBC from prior admissions or from other hospitals may be extremely helpful.)
• Medication review?  (Many drugs can cause thrombocytopenia, not all of which are listed above.)
• Calculate the 4T score for HIT.
• Recent events known to cause thrombocytopenia (e.g. massive transfusion, surgery, hemodialysis, intra-aortic balloon pump, ECMO, transfusion)?

physical examination

• Evaluate for petechiae or purpura (especially on the legs, around the blood pressure cuff, and in the mouth).(31269407)  Substantial petechiae/purpura has several implications:
  • (1) This supports that the thrombocytopenia is real (not spurious).
  • (2) This argues against processes which promote thrombosis (e.g., HIT).
  • (3) This may lend some support to benefit from platelet transfusion.📖

core investigations for thrombocytopenia

• Platelet count in citrated tube to exclude pseudothrombocytopenia (blue top tube).
• Blood smear:
  • Platelet clumps or satellitism (platelets clumped around neutrophils) suggests pseudothrombocytopenia.
  • Schistocytes may suggest thrombotic microangiopathy.📖
  • Blasts may reveal acute leukemia.
  • Atypical lymphocytes may suggest unusual infections.
  • Large platelets suggest a consumptive process leading to an increased fraction of immature platelets.(30579402)  
  • Spherocytes may suggest autoimmune hemolytic anemia with Evan's syndrome.
• Coagulation studies (INR, PTT).
• DIC labs (fibrinogen, D-dimer).

additional studies to consider

• Heparin-PF4 antibody to evaluate for HIT if the 4T score is 4-8
• Ferritin, if concern for hemophagocytic lymphohistiocytosis (HLH).📖
• Ultrasonography to evaluate for DVT (if HIT or other thrombogenic forms of thrombocytopenia are possible).
• Liver function tests, if concern for cirrhosis.
• Thromboelastography (TEG) may be useful, especially patients with DIC or cirrhosis.(29595638)

---

platelet transfusion

(back to contents)

---

relative contraindications to platelet transfusion 

• HIT.
• Thrombotic microangiopathy (e.g., TTP, HUS).
• DIC with clinical thrombosis.

platelet transfusion targets?

• Limitations in the concept of platelet transfusion target:
  • (1) Clinical bleeding doesn't track well with platelet count.  Numerous other factors affect the risk of bleeding (e.g., age, uremia, use of anticoagulant/antiplatelet agents).
  • (2) Platelet targets for various procedures are based mostly on convention, than actual evidence.
• Conventional transfusion targets:
  • Target >100 if active intracranial hemorrhage or pre-operative before neurosurgery/ophthalmic surgery.
  • Target >50 if active bleeding or pre-operative (non-neurologic surgery).
  • Target >20 possibly in patients with other coagulopathies (e.g., DIC, severe renal dysfunction), patients undergoing bronchoscopy with bronchoalveolar lavage.(29222318, 23860341)
  • Target >10 in most patients.
• An alternative approach is to use a TEG-based transfusion strategy.  Many cases patients with thrombocytopenia will still have a preserved maximal amplitude (MA), suggesting that platelet transfusion is unnecessary.📖  Ultimately, the risks and benefits of transfusion should be considered for each individual patient.

response to platelet transfusion

• Platelets may be given in one of the forms listed below, which are functionally equivalent.  Either one should increase platelet count by roughly 30,000-60,000.
  • 4-6 pack of pooled platelets from multiple donors.
  • One unit of platelets obtained using apheresis from a single person.
  • 💡 A more precise approach to determine whether the platelet count has increased appropriately is to use the Corrected Count Increment (CCI).🧮
• If the patient doesn't respond appropriately to platelet transfusion, this may be investigated by checking a platelet level one hour after the next platelet transfusion.
  • If the 1-hour platelet level fails to increase appropriately, this implies immune platelet consumption.   In addition to the causes of immune platelet consumption listed above, this can also be due to alloimmunization (antibodies against donor platelet antigens).  Alloimmunization may be managed by obtaining HLA-matched platelets (with the help of the blood bank).
  • If the 1-hour platelet level increases appropriately and then platelet count falls over the following day, this indicates non-immune platelet consumption (causes listed above).

strategies to improve platelet function (other than transfusion)

• Restoring platelet counts using transfusion may be the ideal solution to thrombocytopenia, but this is not always feasible for several reasons:
  • Platelets may be rapidly consumed.
  • Hospital may lack compatible platelets.
  • Some smaller hospitals may lack immediate access to platelet transfusion.
• Alternative approaches to improve hemostasis may include the following:
  • Desmopressin (DDAVP) may cause a temporary improvement in platelet function and hemostasis.📖 (30579402)
  • Targeting unusually high fibrinogen levels (e.g., >200 mg/dL) may be considered.  Platelets and fibrinogen function together to physically form a clot.  To a certain extent, a deficiency of one may be compensated for by an excess in the other.

---

drug-induced immune thrombocytopenia (D-ITP)

(back to contents)

---

clinical features

• Typically begins 1-3 weeks after starting new medication (but can occur within a day, if previously sensitized to the medication).
• Immune consumption often causes severe thrombocytopenia (e.g., <10,000-20,000) with bleeding.
• Often causes systemic symptoms, including fever and chills.(27982413)

commonly implicated medications

• Cardiac:  Abciximab, Amiodarone, Amlodipine, Captopril, Digoxin, Diltiazem, Eptifibatide, Hydralazine, Procainamide, Quinidine, Simvastatin, Tirofiban,
• Heme/onc:  Bleomycin, Fludarabine, Gemcitabine, Checkpoint inhibitors? (Ipilimumab), Oxaliplatin, Rituximab, Trastuzumab
• Infectious disease:  Amphotericin B, Beta-lactams (esp. ampicillin, ceftriaxone, penicillin, piperacillin), Clarithromycin, Fluconazole, Fluoroquinolones (Ciprofloxacin, Levofloxacin, Moxifloxacin), Indinavir, Interferon-alpha, Linezolid, Rifampin, Trimethoprim/Sulfamethoxazole, Vancomycin
• Nephrology:  Acetazolamide, Chlorothiazide, Desmopressin, Hydrochlorothiazide,
• Neuro/Psych:  Carbamazepine, Clozapine, Diazepam, Fluoxetine, Haloperidol, Lamotrigine, Levetiracetam, Olanzapine, Ondansetron, Phenytoin, Quetiapine, Valproic acid
• Rheum:  Allopurinol, Cyclosporin, Methotrexate, Tacrolimus, TNF-inhibitors (Adalimumab, Etanercept, Infliximab)
• General/Misc:  Acetaimonphen, Aspirin, Danazol, Famotidine, NSAIDs (esp. Diclofenac, Ibuprofen, Indomethacin, Naproxen), Pantoprazole, Quinine (found in tonic water, mixed drinks)
• More information including more detailed up-to-date lists here.  Bolded drugs seem to be more common causes of D-ITP.

treatment

• Discontinuation of the drug will usually lead to resolution within 4-8 days.(31498668)
• May require platelet transfusion (however, ongoing consumption will typically cause patients to be refractory to platelet transfusion).
• Intravenous immunoglobulin (IVIG) may be used in severe cases (e.g. 1 gram/kg on two consecutive days).
• Steroid doesn't seem to help.(30579402)

---

transfusion related thrombocytopenias

(back to contents)

---

post-transfusion purpura (PTP)

• Mechanism:
  • Patient was previously sensitized to platelet antigens (especially HPA-1a).
  • Transfusion leads to antibody against HPA-1a, which ends up destroying patient's own platelets as well.
• Epidemiology:
  • At-risk host:  Usually mothers (previously sensitized at pregnancy) or patients with history of transfusion or transplantation.
• Clinical presentation:
  • Occurs ~5-10 days after transfusion of platelets, packed RBCs, or whole blood.
  • Abrupt fall in platelets to <20,000 which can cause life-threatening bleeding.
  • May be refractory to platelet transfusion.
• Investigation:
  • Diagnosis supported by anti-HLA alloantibodies in patient's serum.
• Treatment:
  • IV immunoglobulin may be the front line therapy, possibly combined with steroid.(33707107)
  • Avoid unmatched platelets.  Transfusion of platelets from compatible donor may be more effective (HPA-1a antigen negative platelets).

passive alloimmune thrombocytopenia

• Rare, results from passive transfer of anti-HPA-1a antibodies.
• Immediate, severe thrombocytopenia occurs within hours of transfusion.
• Treatment is unclear, but IV immunoglobulin has been used.

---

heparin induced thrombocytopenia

(back to contents)

---

introduction

• HIT results from autoantibodies which react to an antigen created by the combination of heparin plus platelet factor 4 (PF4).  These antibodies cause activation of platelets, leading to arterial and venous thrombosis.  Consumption of activated platelets leads to thrombocytopenia.
• HIT is rare, yet important to recognize early.  HIT is challenging to diagnose, because the vast majority of patients with DVT or thrombocytopenia won't have HIT.
• In ~10% of cases, HIT leads to overt DIC.  These cases are distinguished by unusually low platelet counts (e.g. <20,000), lab derangements of DIC (e.g., elevated INR and low fibrinogen), and often microvascular thrombosis.  This combination of HIT and DIC can be difficult to diagnose, because the unusually low platelet count will confound diagnostic algorithms for HIT.

epidemiology of HIT

• Rare, with incidence varying based on the type of unit:
  • Surgical ICU (especially cardiothoracic and orthopedic surgery) has highest rate, up to ~2%.
  • Cardiac ICU:  Intermediate rate, may approach ~1%.
  • Medical ICU:  Lowest rate, <<1%.
  • Obstetric patients:  Rare.(34288442)
• Unfractionated heparin is almost ten times higher risk than low molecular-weight heparin.(32904403)
• Additional risk factors:  Female sex, older age, obesity, hemodialysis.
• Overall, HIT causes at most ~1% of thrombocytopenia which is seen in an ICU.(25590528)

when to evaluate for HIT

• Consider HIT in a patient recently treated with heparin who has:
  • New-onset thrombocytopenia or a platelet count that suddenly drops by 50%.
  • Skin necrosis at sites of heparin injection.
  • Anaphylactoid response to systemic heparin infusion.
  • Venous and/or arterial thrombosis.
• Evaluation begins with risk-stratification using the 4T score.(28808508)

interpretation of the 4T score:

• 0-3 points:  Low probability (<<1%):
  • No further workup of HIT is indicated (avoid checking heparin-PF4 antibodies).
• 4-5 points:  Intermediate probability (~10%):
  • Stop heparin & check anti-PF4 antibody titer.
  • Consider therapeutic anticoagulation with non-heparin agent (more on this below).
• 6-8 points:  High probability (50%):
  • Stop heparin & check anti-PF4 antibody titer.
  • Initiate therapeutic anticoagulation with a non-heparin agent.

PF4 antibody titer (ELISA for Heparin-PF4 antibody complex)

• This test detects any antibodies that bind the heparin-PF4 complex.  However, most antibodies that bind this complex don't cause clinical HIT.  Only IgG antibodies are pathologic, so assays specific for IgG anti-PF4 antibodies may have slightly higher specificity.(31069988)
• Overall, heparin-PF4 has high sensitivity but poor specificity (it's positive in ~15% of ICU patients).(19911252)  Therefore, a negative result largely excludes HIT, but a positive result doesn't prove the diagnosis of HIT.  Care should be taken to avoid ordering this test indiscriminately, because that will generate false-positive results.
• PF4 is often treated as a binary test (positive/negative), but this is a gross oversimplification.  In reality, the risk of HIT increases with increasing antibody titer as shown above.
• The best way to interpret the PF4 level is in combination with the 4T score, using Bayesian analysis:

Serotonin Release Assay (SRA) 

• This is the gold-standard diagnostic test with excellent sensitivity and specificity.
• Unfortunately, this test is expensive and usually takes several days to return.  In some cases, empiric therapy for HIT is indicated while waiting for the serotonin release assay to result.
• The sensitivity of the serotonin release assay isn't perfect.  Thus, if a patient has a very high anti-PF4 antibody titer and a negative serotonin release assay result, HIT may remain possible.(34288442)  

when to initiate empiric anticoagulation therapy for HIT

• (1) For patients with clinical thrombosis (e.g. DVT/PT) on therapeutic heparin anticoagulation, heparin should be switched to a non-heparin agent such as argatroban (more on this below).
• (2) For patients without clinical thrombosis who have a high likelihood of HIT (e.g. ~50%), empiric anticoagulation with a non-heparin agent should generally be started empirically before the diagnosis is confirmed.  Delaying therapy for HIT carries a significant risk of thrombosis, so anticoagulation is generally rational in high-likelihood patients.(26847066)
• (3) For patients without clinical thrombosis who have an intermediate likelihood of HIT (e.g. ~10-15%), management is controversial.(30502310)  If the patient is thrombocytopenic due to another cause, then anticoagulation could compound thrombocytopenia to cause major bleeding.(29253554)  Thromboelastography might theoretically be helpful here, because HIT seems to associate with a hypercoagulable pattern whereas a hypocoagulable pattern would argue against HIT and potentially signal harm from anticoagulation.(18056152, 17632008)  A sensible compromise might be to use a DVT-prophylactic dose of fondaparinux in this situation (2.5 mg sq. daily).(26539932, 32904403)  Ultrasonography of upper and lower extremities should be performed to exclude DVT if this hasn't already been done.  Ongoing evaluation for HIT should be expedited, with further management depending on the results of additional testing.

non-heparin anticoagulants for HIT 

• Argatroban 💉 is perhaps most commonly used (with asimilar option being Bivalirudin 💉).
• Fondaparinux 💉 could be considered in intermediate-likelihood situations, at a reduced dose (2.5 mg sq daily).

---

podcast

(back to contents)

---

Follow us on iTunes

---

questions & discussion

(back to contents)

---

To keep this page small and fast, questions & discussion about this post can be found on another page here.

• Avoid checking HIT labs if the 4T score is 0-3 points.  These patients are at very low risk for HIT, so if the heparin-PF4 test is positive it is probably a false-positive.
• Avoid platelet transfusions unless truly indicated.  Platelet transfusion has a significant side-effect profile, and transfused platelets often don't last long anyway.
• Warfarin is contraindicated in the acute (thrombocytopenic) phase of HIT.  Initiation of warfarin in this context may cause depletion of endogenous anticoagulants (proteins C and S), leading to ischemic limb necrosis.
• Avoid the use of unfractionated heparin for DVT prophylaxis if possible, given that unfractionated heparin causes a substantially higher risk of HIT compared to low molecular weight heparin (e.g. enoxaparin) or fondaparinux.(25590528)
• Don't assume that a “positive” anti-PF4 antibody proves a diagnosis of HIT.  The significance of a positive result depends on the antibody titer.  A validated Bayesian algorithm incorporating the 4T score and anti-PF4 antibody titer is probably the best way to approach this, see section above.(28622439)

Guide to emoji hyperlinks 

•  = Link to online calculator.
•  = Link to Medscape monograph about a drug.
•  = Link to IBCC section about a drug.
•  = Link to IBCC section covering that topic.
•  = Link to FOAMed site with related information.
•  = Link to supplemental media.

Going further: 

• Thrombocytopenia
  • Thrombocytopenia (Chris Nickson, LITFL)
  • Thrombocytopenia (WikEM)
• HIT
  • Heparin Induced Thrombotic Thrombocytopenia (Chris Nickson, LITFL)
  • HIT overview (PulmCCM)