11.3 Immune and Nonimmune Cytopenias

11.3 Immune and Nonimmune Cytopenias

Quick Answer: WAIHA is usually IgG, often autoanti-e / Rh, DAT IgG ± C3 — adsorb the auto to find an underlying alloantibody. CAD is IgM autoanti-I with a C3-only DAT. PCH is the Donath–Landsteiner biphasic IgG autoanti-P. Drug-induced hemolysis uses four named mechanisms: hapten (penicillin), immune complex, autoantibody, and nonimmunologic protein adsorption. Nonimmune hemolysis is mechanical, PNH, or an intrinsic red-cell defect. ITP ≠ PTP ≠ NAIT. Transfuse least-incompatible red cells only when oxygen delivery requires it.

Outline III.B.4. HDFN is Chapter 12 — do not turn every DAT-positive neonate into this section. Chapter 9 taught anti-I and anti-P as antigens; this section is the cytopenia and the product decision.

Warm autoimmune hemolytic anemia

WAIHA is the everyday immune hemolysis of adults. The autoantibody is usually IgG, reacts at 37 °C, and often has Rh specificity — autoanti-e is the classic named specificity, but autoanti-Rh, autoanti-c, or a panreactive warm auto are all in bounds. The DAT is IgG positive, sometimes IgG + C3. The eluate typically reacts with all panel cells (or with all e+ cells if the specificity is clean).

Two jobs sit on the blood bank, in this order:

1. Is there an underlying alloantibody? Recently transfused patients can hide an alloantibody under the auto. If the patient has not been transfused in the last three months, autoadsorption (often after enzyme or ZZAP pretreatment of autologous cells) strips the auto so the adsorbed serum can be tested for alloantibody. If the patient was recently transfused, do not autoadsorb donor cells and call it the patient’s phenotype — use allogeneic adsorption.
2. Does this patient actually need red cells? Steroids, treating CLL/lymphoma/SLE, and stopping a culprit drug fix more WAIHA than a cooler full of “least incompatible” units. Transfuse for symptomatic anemia or unstable oxygen delivery, not for a positive DAT.

If transfusion is required, provide Rh/Kell (and Kidd/Duffy when feasible) phenotype- or genotype-matched units to avoid making a new alloantibody while every crossmatch looks incompatible. When every unit is incompatible after alloantibody is excluded, issue least-incompatible (or “least serologically incompatible”) red cells and tell the team the incompatibility is the auto, not an uninvestigated allo. Do not delay a crashing patient for a perfect AHG-crossmatch.

CAD, PCH, mixed, and drug-induced

Cold agglutinin disease (CAD) is usually IgM autoanti-I with a wide thermal amplitude. The DAT is typically C3 only — IgM washes off during DAT washing. Pathogenicity is titer plus thermal amplitude, not the letter I (Chapter 9.1). Mycoplasma is the classic infectious pairing for anti-I; EBV pairs with autoanti-i. Keep the patient and the unit warm; issue through a blood warmer. Prewarm can hide a benign cold auto; it is the wrong sole method for pathogenic CAD and can miss a significant alloantibody.

Paroxysmal cold hemoglobinuria (PCH) is a different disease. The antibody is the Donath–Landsteiner antibody: a biphasic IgG with anti-P specificity that binds in the cold and lyses with complement when the system is warmed to 37 °C. The DAT is often C3 only. The named test incubates patient serum + P+ cells + complement at 4 °C, then at 37 °C, and looks for hemolysis only in the tube that saw both temperatures. PCH is typically a post-viral illness in children. Do not work it as a p-phenotype rare-donor problem, and do not call every cold hemolysin PCH.

Mixed autoimmune hemolysis has both warm IgG and a pathogenic cold antibody. The DAT is often IgG + C3, the eluate is warm-reactive, and a cold agglutinin has a dangerous thermal amplitude. Treat both mechanisms; do not pick one DAT story and ignore the other.

Drug-induced immune hemolytic anemia has four mechanisms the exam still names:

• Hapten / drug-adsorption (penicillin, some cephalosporins at high dose) — drug coats red cells; IgG against the drug-coated membrane. DAT IgG. The eluate (or serum) reacts with drug-treated cells, not untreated cells.
• Immune-complex (quinidine-type, some cefotetan/ceftriaxone stories) — drug–antibody complexes adsorb and bind complement. Often acute intravascular hemolysis. DAT usually C3.
• True autoantibody (methyldopa, some fludarabine/procainamide) — looks serologically like WAIHA; the antibody reacts without adding drug.
• Nonimmunologic protein adsorption (some cephalosporins) — proteins stick to the membrane; DAT is positive but the eluate is nonreactive and hemolysis may be absent.

Stop the drug. Transfuse only if the patient needs oxygen-carrying capacity.

Nonimmune hemolysis

A negative DAT with hemolysis is not a failed workup — it is a different outline line.

• Mechanical — prosthetic valves, microangiopathy (TTP/HUS/DIC), march hemoglobinuria. Schistocytes, not spherocytes. The blood bank may still be asked for plasma exchange in TTP; that is ADAMTS13 replacement, not an antibody panel.
• PNH — acquired PIGA mutation, loss of GPI-anchored complement regulators CD55 and CD59. Complement lyses the patient’s own cells. Dark morning urine, cytopenias, and thrombosis are the clinical triad. Historical Ham/sucrose tests are exam history; diagnosis is flow cytometry. Transfused cells have normal GPI anchors and survive; the disease is the patient’s clone.
• Intrinsic red-cell defects — membrane (hereditary spherocytosis), enzyme (G6PD, pyruvate kinase), hemoglobin (sickle, unstable hemoglobins). Oxidant stress (primaquine, dapsone, infection, fava beans) precipitates G6PD hemolysis. These units crossmatch compatible because there is no allo- or autoantibody against a blood-group antigen.

Also keep thermal, hypotonic, and bacterial hemolysis on the nonimmune list when the unit or the IV fluid is the culprit (Chapter 18), not the patient’s immune system.

Immune thrombocytopenia and neutropenia

ITP is autoantibody against platelet glycoproteins (often GPIIb/IIIa or GPIb). Isolated thrombocytopenia, often after a viral illness in children or chronic in adults. First-line therapy is steroids or IVIG, not a standing platelet order. Issue platelets for bleeding or an imminent procedure, knowing increments may be short.

Post-transfusion purpura (PTP) is not ITP. 5–10 days after transfusion, usually a multiparous HPA-1a–negative woman, an anamnestic anti–HPA-1a (or other HPA) destroys both donor and autologous platelets. Counts crash. Treat with IVIG. Random platelets are consumed; if platelets must be given, prefer HPA-compatible units. This is a transfusion-medicine emergency, not “ITP that happened to follow a unit.”

Neonatal alloimmune thrombocytopenia (NAIT) is the platelet analog of HDFN: maternal IgG against a paternal HPA antigen (classically HPA-1a) the fetus inherited. The neonate bleeds; the mother is fine. Treatment is HPA-1a–negative platelets or washed maternal platelets, plus IVIG in selected protocols — not random adult platelets as the planned product.

Immune neutropenia follows the same logic on granulocytes: autoimmune neutropenia, neonatal alloimmune neutropenia (maternal HNA antibody), and drug-induced immune neutropenia. Granulocyte transfusion is rare and always irradiated (Chapter 4.3). Most management is growth factor, antibiotics, and waiting — not a standing granulocyte order.

Worked scenario. An untransfused adult has hemoglobin 6.4 g/dL, DAT IgG 3+, panreactive eluate, and autoanti-e recovered from the eluate. Autoadsorb. If adsorbed serum is nonreactive, the panagglutinin was the auto. Transfuse because the hemoglobin is symptomatic, using e− or phenotype-matched units if the inventory allows, and document least-incompatible only after alloantibody is excluded. Do not wait 24 hours for a rare e− frozen unit while the patient is ischemic if e− blood is not immediately available and no alloantibody was found.

Exam traps. CAD is C3 / IgM anti-I; PCH is DL IgG anti-P — do not swap them. Autoadsorb only if not recently transfused. PTP is not ITP and is not treated with random platelets as the plan. A negative DAT does not mean “no hemolysis.” Least-incompatible is a last step after alloantibody exclusion, not a substitute for an antibody workup.