A 26-year-old woman with end-stage lung disease secondary to cystic fibrosis, cirrhosis secondary to hepatitis C, and insulin-dependent diabetes mellitus presented with acute pneumonia. She was treated with intravenous piperacillin/tazobactam and tobramycin. Other medications included dexamethasone, bumetanide, pantoprazole, dornase, and insulin. At admission, the patient's hematocrit was 30.8% (Table 1). On day 8 of hospitalization, she developed severe anemia with a hematocrit of 11.5%. Her blood type was group A, Rh D-positive, and the results of her antibody screen (performed in solid-phase, low-ionic-strength solution), which had previously been positive for only anti-E antibody, were now positive with a panreactive pattern. The result of the direct antiglobulin test (DAT)3 was 2+ (moderately positive) for IgG, and the eluate was nonreactive. Six units of red blood cells (RBCs) were requested for transfusion. Her plasma was weakly to moderately cross-match incompatible with all group A–positive, E antigen–negative units tested. Given the severity of her anemia, 6 of these units were emergently released for transfusion to stabilize the patient while the investigation of her anemia and the new serologic findings continued.
Only 2 mechanisms were likely responsible for such an abrupt and severe decrease in the hematocrit: bleeding and hemolysis. Bleeding was initially suspected; however, no source of hemorrhage was found. On the other hand, laboratory findings revealed high lactate dehydrogenase (LDH) activity, hemoglobinuria, and unconjugated hyperbilirubinemia, all of which were consistent with hemolysis. The positive DAT result indicated an immune-mediated etiology. The panreactive pattern of the antibody panel results suggested warm autoantibodies, but the nonreactive eluate eliminated this possibility because panreactivity should also be observed in the eluate. The anti-E antibody first detected at admission indicated that the patient had received a transfusion or had become pregnant at some point in her life. There was no history of recent transfusion, however, and the eluate was nonreactive—results that rendered a delayed hemolytic transfusion reaction highly unlikely.
QUESTIONS TO CONSIDER
What tests are used to diagnose hemolytic anemia?
What laboratory test suggests the hemolytic anemia in this patient is immune mediated?
What is unusual about a panreactive pattern on the antibody screen with a negative eluate result?
How does cirrhosis of the liver affect the laboratory evaluation of hemolysis?
DIAGNOSIS OF HEMOLYSIS
Fig. 1 is a simplified diagnostic algorithm for hemolytic anemia. Hemolysis is classified as either intravascular, occurring within the blood vessels, or extravascular, mediated by phagocytic destruction in the reticuloendothelial system. Recognizing hemolysis requires a careful review of clinical history, signs, symptoms, and laboratory test results. Loss of RBCs, whether due to hemorrhage or hemolysis, causes a compensatory increase in reticulocytes, usually within 3–5 days (1). During intravascular hemolysis, free hemoglobin is released into the plasma and binds to haptoglobin, leading to decreased haptoglobin concentrations. Excess free hemoglobin is eliminated in the urine, where it is detected by urinalysis and identified clinically as a dark, tea-colored urine. LDH is a ubiquitous enzyme found in tissues and the blood that increases during cellular injury and hemolysis, particularly intravascular hemolysis. In extravascular hemolysis, splenic macrophages phagocytose RBCs, releasing hemoglobin, which is then broken down into heme. Heme is converted to unconjugated bilirubin, which is secreted into the plasma, bound to albumin, and transported to the liver, where it is conjugated to glucuronic acid and excreted in bile. The excess unconjugated bilirubin overwhelms the liver, leading to unconjugated hyperbilirubinemia and the clinical sign of jaundice.
DIAGNOSIS OF IMMUNE-MEDIATED HEMOLYSIS
A positive DAT result is the hallmark of immune-mediated hemolysis. The DAT detects IgG antibodies and complement protein C3d bound on the surface of RBCs. In sufficient quantity, bound complement mediates intravascular hemolysis. The main diagnostic consideration is cold agglutinin disease. In contrast, IgG-coated RBCs are targeted primarily for extravascular destruction. A positive DAT result for IgG prompts an elution test, whereby acid or heat is used to elute the antibody off the RBCs and a panel of reagent RBCs of known phenotype is used to determine its specificity. The most common causes of IgG-mediated hemolysis are warm autoimmune hemolytic anemia (WAIHA) and delayed hemolytic transfusion reaction (DHTR). In WAIHA, the antibody (autoantibody) is directed against RBC epitopes found on most RBCs, including the patient's own. Therefore, the autoantibody is panreactive to a panel of reagent RBCs and the patient's own RBCs. The presence of warm autoantibodies, however, is not always clinically important, and correlation with clinical and laboratory data is required to determine if they are associated with WAIHA. In DHTRs, an antibody (alloantibody) is directed against a specific RBC antigen. It reacts only with RBCs bearing that antigen and is not reactive with the patient's own RBCs.
DRUG-INDUCED IMMUNE HEMOLYTIC ANEMIA
Drug-induced immune hemolytic anemia (DIIHA) is rare, with an estimated incidence of 1 in 1 × 106 people (2). It typically presents as hemolytic anemia occurring 6 or more days after initial drug exposure, with a severity that ranges from mild anemia to fatal hemolysis. The diagnosis requires a high index of suspicion and should be considered in any patient with unexplained hemolysis after drug therapy. Historically, DIIHA was described to occur with methyldopa and high-dose penicillin administration. Although >125 drugs have since been implicated (3), the majority of contemporary cases are caused by second- and third-generation cephalosporins, with cefotetan and ceftriaxone being responsible for approximately 70% and 10% of all cases, respectively (4, 5). Piperacillin and β-lactamase inhibitors are the next most frequently reported drugs and are responsible for up to 9% and 7% of cases, respectively (5). Drugs like penicillin and cefotetan bind covalently to RBC membrane proteins. If a patient makes an IgG antibody to the drug, that antibody can bind to the drug-coated RBC and facilitate extravascular hemolysis (2). The DAT result is usually strongly positive for IgG and negative for C3d. In contrast to WAIHA and DHTRs, the eluate is nonreactive, because the reagent RBCs lack the drug required for the antibody to bind. Drugs such as ceftriaxone bind noncovalently to RBCs, leading to an immune response and the production of a drug-dependent antibody, typically IgM (2). These antibodies activate complement, leading to intravascular hemolysis. The DAT result is usually weakly positive for only C3d. Many drugs are capable of inducing both IgG and IgM antibodies, with concomitant extravascular and intravascular hemolysis, as well as a DAT result positive for both IgG and complement. Interestingly, piperacillin appears to bind covalently to RBCs (6); however, antipiperacillin antibodies are best demonstrated through the complement-associated immune complex method.
Drug-dependent antibody testing can confirm DIIHA but is usually available only at specialized reference laboratories at blood centers. Two primary test methods have been described and are detailed in the American Association of Blood Banks Technical Manual (7). The basic methodology of the “immune complex” method requires mixing the patient's serum (which contains drug antibodies) with the drug in question and pooled group O reagent RBCs. The second method uses RBCs coated with the drug in question and the patient's serum. In either case, agglutination is interpreted as a positive result. DIIHA is managed primarily by discontinuation of the offending drug and transfusion of RBCs as needed. The hemolysis usually resolves within a few days (2).
The patient's laboratory findings including unconjugated hyperbilirubinemia, increased LDH, and hemoglobinuria were consistent with hemolysis, and it was readily recognized despite her liver disease. It is not uncommon for patients with end-stage liver disease to have low haptoglobin, owing to dysfunction in its synthesis. In such cases, total protein and albumin would also be decreased. Cirrhosis can cause hyperbilirubinemia of predominantly the unconjugated fraction (whereas most other liver diseases cause increases primarily in conjugated bilirubin). The patient's baseline low albumin concentration was compatible with end-stage liver disease, and the increases in aspartate and alanine aminotransferases on day 8 indicated acute liver injury. In the setting of most liver diseases, hemoglobinuria, increases in primarily unconjugated bilirubin, and otherwise unexplained high LDH measurements generally are the most useful in diagnosing hemolysis. The patient's positive DAT result for IgG supported an immune-mediated etiology. The negative eluate result was not compatible with WAIHA or a DHTR. The authors suspected DIIHA, particularly considering the piperacillin therapy initiated 8 days prior. Serum samples were tested at the American Red Cross Reference Laboratory (Pomona, CA) for drug-dependent antibodies by the immune complex method. Two volumes of patient serum were combined with 2 volumes of piperacillin solution (1 g/L suspended in PBS). To this mixture was added 1 volume of group O, E antigen–negative, and enzyme (ficin)-treated reagent RBCs suspended in a 50-g/L NaCl solution. The mixture was incubated at 37 °C for 1–2 h and centrifuged. Strong (3½+) agglutination was observed. No agglutination was seen in the absence of piperacillin. This finding confirmed DIIHA due to piperacillin. Piperacillin was discontinued on day 12 of hospitalization, and the patient's hemolysis resolved quickly. Unfortunately, the patient died on day 19 from complications of her liver and lung disease.
Piperacillin DIIHA is well reported in adults with cystic fibrosis (8, 9) and typically presents as complement-mediated intravascular hemolysis with a DAT result positive for C3d. In this case, the DAT detected only IgG, a result that occurs in a minority of cases of piperacillin DIIHA. This case may represent one with a “falsely” negative C3d result, which can occur in intravascular hemolysis if all of the complement-coated RBCs have been lysed in vivo before sample collection. Another unusual but previously described finding is the positive antibody screen with a panreactive pattern (9). Piperacillin is metabolized by the liver, and in view of the patient's cirrhosis, the authors hypothesize that her plasma contained high concentrations of piperacillin in addition to the antipiperacillin antibodies. In essence, the antibody screen was functionally equivalent to the immune complex drug-dependent antibody assay, and this explanation accounts for the initially positive antibody screen with a panreactive pattern that disappeared after the drug was discontinued. In the appropriate setting, such an unusual finding may suggest piperacillin DIIHA, which should be confirmed by drug-dependent antibody testing.
POINTS TO REMEMBER
Hemolysis is characterized by reticulocytosis, unconjugated hyperbilirubinemia, high LDH, low or undetectable haptoglobin, and hemoglobinuria. The findings tend to be more dramatic in intravascular hemolysis and more subtle in extravascular hemolysis.
In the setting of most liver diseases, hemoglobinuria, increases of primarily unconjugated bilirubin, and otherwise unexplained high LDH measurements are the most useful in diagnosing hemolysis.
In the setting of hemolysis, a positive DAT result suggests an immune-mediated etiology. The main differential diagnosis for immune-mediated hemolysis includes WAIHA, DHTRs, cold agglutinin disease, and DIIHA.
DIIHA should be considered in any case of unexplained immune-mediated hemolysis, particularly when the DAT result is positive and the eluate is nonreactive. The drugs most often implicated in DIIHA are cefotetan, ceftriaxone, and piperacillin.
↵3 Nonstandard abbreviations:
- direct antiglobulin test;
- red blood cell;
- lactate dehydrogenase;
- warm autoimmune hemolytic anemia;
- delayed hemolytic transfusion reaction;
- drug-induced immune hemolytic anemia.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
- Received for publication April 25, 2011.
- Accepted for publication August 3, 2011.
- © 2012 The American Association for Clinical Chemistry