# Thrombocytopenia: Inpatient Differential and Management

## Overview

Thrombocytopenia is defined as a platelet count below 150,000/mcL and is extremely common in hospitalized patients, affecting 25 to 40 percent of ICU patients. Severity is categorized as mild (100,000-150,000), moderate (50,000-100,000), severe (below 50,000), and critical (below 10,000, which carries risk of spontaneous bleeding). A structured approach based on mechanism — decreased production, increased destruction, sequestration, or dilution — guides efficient diagnosis.

## Diagnostic Framework

### Step 1: Rule Out Pseudothrombocytopenia

EDTA-dependent platelet clumping is the most common cause of pseudothrombocytopenia. The peripheral blood smear should always be reviewed when thrombocytopenia is unexpected. If clumping is seen, the blood should be redrawn in a citrate or heparin tube.

### Step 2: Mechanism-Based Classification

Decreased production results from bone marrow failure (aplastic anemia, myelodysplastic syndromes), marrow infiltration (leukemia, lymphoma, metastatic cancer, myelofibrosis), nutritional deficiency (B12, folate, copper), infections (HIV, hepatitis C, parvovirus B19), medications (chemotherapy, linezolid, valproic acid, alcohol causing direct marrow suppression), and radiation.

Increased destruction can be immune-mediated or non-immune. Immune-mediated causes include ITP, drug-induced thrombocytopenia (heparin/HIT, quinine, vancomycin, linezolid), SLE, antiphospholipid syndrome, and post-transfusion purpura. Non-immune consumption occurs in DIC, TTP, HUS, HELLP syndrome, and with mechanical heart valves. Infection-associated destruction is seen with sepsis (through both consumption and marrow suppression), malaria, dengue, and rickettsial diseases.

Sequestration occurs with splenomegaly from portal hypertension, cirrhosis, lymphoma, or myeloproliferative disorders. Normally about one-third of platelets are sequestered in the spleen, but this proportion increases substantially with splenomegaly.

Dilutional thrombocytopenia results from massive transfusion or large-volume fluid resuscitation.

<image>Mechanism-based diagnostic algorithm for thrombocytopenia showing decreased production, increased destruction (immune vs. non-immune), sequestration, and dilutional causes with key laboratory distinguishing features</image>

## Heparin-Induced Thrombocytopenia (HIT)

### Pathophysiology

HIT involves IgG antibodies against PF4-heparin complexes that activate platelets, paradoxically causing thrombosis rather than bleeding. HIT Type 1 is benign, non-immune, produces a mild platelet drop of less than 30% within 1 to 2 days, and has no clinical significance. HIT Type 2 is the immune-mediated, clinically significant form — the true "HIT."

### Clinical Features

The platelet count typically drops more than 50% from baseline, occurring 5 to 10 days after heparin exposure (or within 1 day if prior heparin exposure occurred within the last 100 days). The nadir is usually 50,000-80,000 and rarely falls below 20,000. Thrombosis (arterial or venous) occurs in 30 to 50 percent of cases, manifesting as PE, DVT, limb ischemia, stroke, or MI. Skin necrosis at injection sites is another characteristic finding.

### 4T Score (Pre-test Probability)

| 4T Score | Probability | Action |
|----------|------------|--------|
| 0-3 | Low (>99% NPV) | HIT excluded; PF4 testing unnecessary |
| 4-5 | Intermediate | Send PF4 antibody (ELISA); consider empiric treatment |
| 6-8 | High | Begin empiric non-heparin anticoagulation while awaiting results |

The 4T score assesses thrombocytopenia (timing and magnitude of platelet drop), timing of onset, thrombosis or other clinical sequelae, and other causes of thrombocytopenia. A score of 0-3 indicates low probability with greater than 99% negative predictive value, making PF4 testing unnecessary. A score of 4-5 indicates intermediate probability and warrants sending PF4 antibody testing. A score of 6-8 indicates high probability and empiric treatment should begin while awaiting results.

### Diagnosis

The PF4/heparin ELISA has high sensitivity (approximately 99%) but lower specificity, with many false positives especially at low-positive optical density values. The serotonin release assay (SRA) is the gold standard with over 95% specificity but is not widely available and takes days to result. Functional assay confirmation is recommended for intermediate ELISA results.

### Management

All heparin must be immediately stopped, including flushes and heparin-coated catheters. Alternative anticoagulation must be started immediately since leaving the patient without anticoagulation carries high thrombosis risk. Options include argatroban (IV, hepatically cleared, preferred in renal failure, requires aPTT monitoring), bivalirudin (IV, short half-life, preferred for patients needing procedures), and fondaparinux (subcutaneous, increasingly used off-label, does not cross-react with HIT antibodies in a clinically significant way). DOACs have emerging evidence for use after initial parenteral treatment in stable patients but are not yet guideline-endorsed for the acute HIT phase.

Warfarin must not be given until platelets recover above 150,000, because protein C depletion in acute HIT creates risk of venous limb gangrene and skin necrosis. Once platelets exceed 150,000, transition to warfarin with overlap of the alternative anticoagulant, or transition to a DOAC. Duration is a minimum of 3 months if thrombosis occurred, or 4 weeks for isolated HIT without thrombosis.

<image>4T scoring system for HIT showing point assignments for thrombocytopenia severity, timing of platelet drop, thrombosis presence, and other causes, with management pathway based on score</image>

## Thrombotic Thrombocytopenic Purpura (TTP)

### Pathophysiology

TTP results from severe deficiency of ADAMTS13 (below 10%), causing ultra-large von Willebrand factor multimers to persist in the circulation and trigger platelet aggregation in the microvasculature, producing MAHA and thrombocytopenia. The acquired form (autoimmune, with anti-ADAMTS13 antibodies) is far more common than hereditary congenital ADAMTS13 deficiency.

### Clinical Features — Classic Pentad (full pentad rare)

The classic pentad includes microangiopathic hemolytic anemia (with schistocytes on smear, low haptoglobin, and elevated LDH), thrombocytopenia (often severe, below 30,000), neurologic symptoms (confusion, headache, focal deficits, seizures), renal dysfunction (usually mild), and fever. However, the full pentad is present in only a minority of cases.

### PLASMIC Score

The PLASMIC score predicts the likelihood of ADAMTS13 below 10%, with a higher score indicating higher likelihood. Its components include platelet count, hemolysis markers, absence of active cancer, absence of organ transplant, MCV, INR, and creatinine.

### Management

Plasma exchange (TPE) is the first-line treatment, removing anti-ADAMTS13 antibodies and ultra-large vWF multimers while replacing ADAMTS13. It must be started immediately if clinical suspicion is high — clinicians should not wait for ADAMTS13 results since mortality without treatment is 90% compared to less than 20% with TPE. Corticosteroids (methylprednisolone 1 mg/kg/day or pulse dose) are given concurrently. Caplacizumab, an anti-vWF nanobody that blocks platelet-vWF interaction, reduces time to platelet recovery and recurrence as shown in the HERCULES trial, though it is expensive. Rituximab targets anti-ADAMTS13 antibody production and is used for refractory or relapsing TTP. Platelet transfusion should not be given unless there is life-threatening hemorrhage, as it may worsen thrombosis by adding fuel to the fire.

## Immune Thrombocytopenia (ITP)

### Diagnosis

ITP is a diagnosis of exclusion characterized by isolated thrombocytopenia with no other clear cause. There is no specific confirmatory test. The peripheral smear shows large platelets without schistocytes or blasts. HIV, HCV, H. pylori, SLE, CVID, and lymphoproliferative disorders must be ruled out.

### Management

Patients with platelets above 30,000 and no bleeding may be observed without treatment. First-line therapy is corticosteroids, either dexamethasone 40 mg for 4 days or prednisone 1 mg/kg for 2 to 4 weeks with taper. IVIG at 1 g/kg for 1 to 2 days provides rapid platelet increase and serves as a bridge to other therapies or before procedures. Second-line options include rituximab, TPO receptor agonists (eltrombopag, romiplostim, avatrombopag), and splenectomy. For life-threatening bleeding, emergency treatment combines platelets, IVIG, corticosteroids, and tranexamic acid.

<image>Management algorithm for immune thrombocytopenia showing observation thresholds, first-line corticosteroids, IVIG for urgent situations, and second-line options including TPO agonists and rituximab</image>

## Clinical Pearls

When a hospitalized patient develops thrombocytopenia, the first three considerations should be sepsis/DIC, medications (especially heparin), and dilutional/consumptive causes, as these account for the vast majority of cases. The 4T score is essential for HIT workup — a low score of 0-3 has greater than 99% negative predictive value and PF4 testing is not needed. In TTP, plasma exchange should never be delayed while waiting for ADAMTS13 results since mortality is 90% without treatment and the assay takes days. Platelet transfusion in TTP is contraindicated unless there is life-threatening hemorrhage because it worsens microvascular thrombosis. Drug-induced thrombocytopenia is more common than HIT, with frequent culprits including vancomycin, piperacillin-tazobactam, linezolid, GPIIb/IIIa inhibitors, and quinine. ITP is a diagnosis of exclusion, and HIV, HCV, and peripheral smear should always be checked before making the diagnosis.

## References
- Neunert C, et al. ASH 2019 Guidelines for ITP. *Blood Adv*. 2019;3(23):3829-3866.
- Cuker A, et al. ASH 2018 Guidelines for Management of VTE: HIT. *Blood Adv*. 2018;2(22):3360-3392.
- Scully M, et al. Caplacizumab Treatment for Acquired TTP (HERCULES). *N Engl J Med*. 2019;380:335-346.
- Warkentin TE. HIT: Pathogenesis and Management. *Br J Haematol*. 2003;121:535-555.
- Bendapudi PK, et al. Derivation and External Validation of the PLASMIC Score. *Lancet Haematol*. 2017;4:e157-e164.