H&O What is the first step in evaluating a patient with hemolysis?
TK The first step is to take a history and conduct a physical examination. The patient’s age, the presentation and evidence of an acute onset or lifetime chronicity of the disease, and the family history are important factors that guide the differential diagnosis.
The next step is a laboratory evaluation. This starts with a complete blood cell (CBC) count and a differential with reticulocyte count (when the goal is to evaluate the erythroid lineage, the reticulocyte count is a must), smear evaluation, and direct antiglobulin test (DAT, also known as the direct Coombs test). In acute presentations that may require transfusion, a type and screen test is performed, with the “screen” also having the diagnostic value of being the indirect antiglobulin test (IAT, also known as the indirect Coombs test).
Hemolysis is defined as an increased turnover of red blood cells (RBCs), and the most reliable indicator of hemolysis is an increase in the reticulocyte count. Decreased haptoglobin is also a typical finding in patients with hemolysis and a useful test in patients older than 6 months; young infants tend to have a low haptoglobin level even without hemolysis because their relatively immature liver does not yet produce normal levels of this protein. Elevated unconjugated bilirubin and lactate dehydrogenase (LDH) are also common hemolytic markers, but normal values do not rule out hemolysis. LDH increases significantly in acute hemolysis and is especially useful for monitoring microangiopathic hemolytic anemia. However, the level may be only slightly elevated or normal in patients with chronic hemolysis, and it may be artificially high if RBCs break in the needle during the blood draw, as sometimes occurs in young pediatric patients or during any difficult blood draw. Therefore, I prefer to focus on the reticulocyte count, which is extremely valuable when any question exists regarding the RBCs.
The pathophysiology of anemia is either decreased production or increased destruction of RBCs. Hemolysis, from the Greek hema (“blood”) and lysis (“breaking up, dissolving”), literally means “blood breaking”—that is, increased RBC destruction. An increased reticulocyte count confirms a healthy production of RBCs, pointing to the diagnosis of hemolysis. However, a normal or decreased reticulocyte count does not always rule out hemolytic anemia. For example, in autoimmune hemolytic anemia (AIHA), the reticulocyte count may not increase within the first 24 to 48 hours after onset and may remain low until treatment begins, when anti-RBC antibodies also target antigens expressed on the reticulocyte membrane. In infants aged 2 weeks to 2 months, the reticulocyte response to hemolytic anemia may be suboptimal because the erythropoietin-mediated stimulation of erythropoiesis is relatively blunted during this developmental period. Additionally, cases of mild chronic hemolytic anemia, such as mild hereditary spherocytosis (HS), may present for the first time during bone marrow suppression due to a viral infection, such as with parvovirus. In these cases, reticulocytosis may be severely suppressed.
Of note, hemolysis may occur with or without anemia. Examples of hereditary hemolysis without anemia include cases of mild spherocytosis and most cases of hereditary xerocytosis (HX), in which hemolysis is well compensated by reticulocytosis.
In the differential diagnosis, it is important to distinguish between AIHA and red cell membrane disorders. The RBCs in warm AIHA have morphologic and functional characteristics of acquired spherocytosis as a consequence of membrane loss due to immunoglobulin G (IgG) binding, appearing as spherocytes and microspherocytes on the blood smear. If these cells are sufficiently numerous, they may give increased values in the osmotic fragility assay or an HS-like curve on osmotic gradient ektacytometry.1
H&O Versions of your algorithm for hemolysis appear in Blood and Hematology.1,2 What other algorithms are in use?
TK The Figure shows a version of my diagnostic algorithm for the evaluation of hemolytic anemia in general. I would also recommend the diagnostic algorithm for AIHA by Barcellini and Fattizo,3 which includes and emphasizes the possibility that a hemolytic anemia is of autoimmune etiology even with a negative conventional DAT result. We have to remember that the result of a DAT performed by most of the local laboratories may be falsely negative in 5% to 10% of patients with warm AIHA because of technical limitations such as the following: (1) poor detection of relatively low levels of anti-RBC autoantibodies that still cause hemolysis; (2) low affinity of the antibodies, causing them to be removed easily from the RBC surface during the preparative washings in conventional DATs; and (3) sensitization of RBCs with IgA or a warm-reacting, monomeric IgM rather than IgG, without complement fixation.4
H&O What are the key clinical clues that should prompt a physician to suspect a red cell membrane disorder rather than other causes of hemolytic anemia?
TK Although signs of variable degrees of anemia and/or jaundice, splenomegaly, and possibly signs of cholecystitis are common findings in the physical examination of patients with a hemolytic anemia of any etiology, a good history and family history may provide more specific clues. A history of neonatal hyperbilirubinemia, especially one that started within the first 24 hours of life, increases the possibility of a hereditary hemolytic anemia, including red cell membrane disorders. A positive family history of hemolytic anemia, including a history of splenectomy or cholecystectomy at a young age in first-degree relatives, is also of significance. Red cell membrane disorders are caused by pathogenic variants in several genes coding for membrane skeleton proteins or transport channels on the RBC surface. As genetic diseases, they are hereditary, most of them with autosomal dominant inheritance, so a detailed family history is frequently illuminating. Approximately 5% to 10% of cases of HS are of autosomal recessive inheritance, in which the family history in parents and grandparents may be negative but the disease may affect siblings or cousins. Most cases of hereditary pyropoikilocytosis (HPP) also have a type of compound heterozygous inheritance, in which family members may have common, nonhemolytic hereditary elliptocytosis (HE) with a single allele defect (dominant inheritance). This condition is frequently undiagnosed in parents with HE because they have been asymptomatic and their blood smear has not been evaluated. Moreover, 10% to 15% of cases of HS—and likely a small percentage of the other red cell membranopathies—are caused by pathogenic genetic variants arising de novo. In these cases, the disease is not present in other family members but may be inherited by the patient’s offspring.5
The laboratory evaluation and especially the red cell morphology on the blood smear will provide more specific clues to support the possibility of a red cell membranopathy.
H&O What morphologic changes occur in RBCs in patients with these disorders?
TK A normal blood smear reveals minimal variation in the size (anisocytosis) and shape (poikilocytosis) of RBCs. Approximately 0.5% to 1.5% of the RBCs have a purplish hue because they are reticulocytes providing a normal level of polychromasia on the smear, and the RBCs have a central pallor approximately one-third their diameter in size.
When I review the smear of a patient with anemia, I first look for polychromasia; increased polychromasia means an increased number of reticulocytes, which indicates hemolysis. As previously discussed, hemolysis can rarely exist without increased reticulocytes. Then I look for anisopoikilocytosis, a variability in the sizes and shapes of the RBCs. I specifically look for spherocytes (ie, cells with absent or decreased central pallor), which characterize HS. Occasional acanthocytic spherocytes and mushroom-shaped cells may also be present. Elliptocytes are prominent in HE, and elliptocytes and fragmented cells may occur in HPP. In rare cases, we see hybrids of HS and HE or HPP, in which spherocytes, elliptocytes, and ovalocytes with decreased or absent central pallor and fragmented cells are present on the same smear. Stomatocytes (from the Greek stoma, “mouth” and kytos, “cell”) are RBCs with a slitlike central pallor that resembles a mouth; these may be acquired as a consequence of liver disease—which changes the lipid composition of the membrane, especially in adults—or may be caused by hereditary stomatocytosis (HSt). The HSt syndromes include the RBC hydration disorders, such as HX (also known as dehydrated stomatocytosis) and OHSt (overhydrated stomatocytosis). In HX, stomatocytes, target cells, and pyknocytes are present among many normal or mildly macrocytic RBCs, whereas in OHSt, macrocytic stomatocytes are the prominent morphology noted on the blood smear. Additional examples of HSt disorders are sitosterolemia, in which stomatocytes and spherocytes are seen along with macrothrombocytopenia, and cryohydrocytosis, in which the RBC potassium leak is exacerbated in cold storage conditions.
H&O What role does osmotic fragility testing play in today’s diagnostic workup?
TK Osmotic fragility testing is still used to detect a decrease in the surface-to-volume ratio, such as in HS, which causes an increase in osmotic fragility. However, not all red cell membrane disorders are characterized by increased osmotic fragility. HE and HPP typically have normal osmotic fragility, whereas HX has an increased membrane-to-volume ratio that leads to a decrease in osmotic fragility, which is not always recognized reliably in osmotic fragility testing.
Osmotic gradient ektacytometry, on the other hand, provides a curve that clearly demonstrates decreased, normal, or increased osmotic fragility. It also measures the deformability of the RBCs, a parameter affected by the membrane skeleton mechanics as the RBCs are exposed to an osmolality gradient. Additionally, the distant part of the ektacytometry curve depends on the intracellular viscosity of the erythrocytes, which may be affected by the hemoglobin concentration and the hydration status of the cell.6 The ektacytometry curve is characteristic for each type of red cell membranopathy, providing a differential diagnosis among HS, HX, HE/HPP, OHSt, and Southeast Asian ovalocytosis.1,2
H&O How would you compare flow cytometry testing for red cell membrane disorders with ektacytometry?
TK Flow cytometry for eosin-5′-maleimide (EMA) is a screening test that is readily available and has been used for more than 20 years. EMA flow cytometry can show membrane loss, which happens in HS and in a cell population fraction in HPP, but not in HE or HX. Although the EMA assay can be helpful to confirm suspicion for HS, it cannot differentiate reliably among the various red cell membrane disorders.7
One important reason to diagnose a red cell membrane disorder appropriately is to avoid recommending splenectomy for HX, a disease that was considered extremely rare in the past but is now recognized as the diagnosis of approximately 10% of patients who have hereditary hemolysis with or without anemia. Splenectomy is contraindicated in HX because it does not decrease hemolysis, which is mainly intravascular in HX. Additionally, patients with HX have a high risk for life-threatening thrombophilia if they have splenectomy. It is essential to diagnose the cause of hemolytic anemia accurately before treatment to uphold the principle of “first, do no harm.” It is also important to diagnose HX that is associated with increased risk of iron overload in young adults, to prevent liver and heart disease in these patients.
Osmotic gradient ektacytometry clearly differentiates between HS and HX and gives a signature curve for each of the different types of red cell membranopathies, allowing an accurate diagnosis in patients with such disorders if they have not recently received a transfusion (ie, in the previous 3-4 months). Ektacytometry has dramatically changed the diagnostic landscape in the field of red cell membrane disorders.
I want to disclose here that I am the co-director of the Erythrocyte Diagnostic Laboratory at Cincinnati Children’s Hospital, and we do offer osmotic gradient ektacytometry as a Clinical Laboratory Improvement Amendments (CLIA)-certified, orderable test. The Mayo Clinic Laboratories have also recently started offering this test within their hemolytic anemia workup.
Regardless of the test being used, expert interpretation is required. I recommend that the hematologist or hematopathologist interpreting the tests also examine the indices of the CBC count, including the mean corpuscular volume, mean corpuscular hemoglobin concentration, and red cell distribution width, along with the absolute reticulocyte count and blood smear. The patient’s personal and family history should also be factored in.
H&O What are the most common pitfalls in interpreting red cell membrane studies that lead to missed or delayed diagnoses?
TK We know that even gold standard tests are not 100% sensitive or specific. In addition, test results must be interpreted in the context of what happened before the blood specimen was obtained. For example, phenotypic evaluation of RBCs is difficult in a patient who has recently received blood transfusions because donor cells may be a significant fraction of the cells evaluated, especially if the patient required several units of blood before the blood specimen was obtained. If the patient has recently received a transfusion or receives them chronically, it is difficult to do phenotypic testing, but a smear evaluation is still useful. If the patient has not received transfusions, hemoglobin electrophoresis is the first phenotypic screening testing that I recommend. Hemoglobin disorders are very common, and even if they do not cause disease, they have the potential to modify other erythrocyte disorders and confound further phenotypic testing. If the patient has a chronic microcytic anemia not explained by iron deficiency, we need to consider the need for globin gene sequencing and deletion/duplication analysis because a basic hemoglobin electrophoresis test will not detect all the possible globin variants, including the pathogenic unstable hemoglobins before splenectomy—they are by definition unstable! If a hemoglobin disorder is unlikely, the next most frequent causes of chronic hereditary hemolytic anemia are membranopathies; therefore, we should move on to ektacytometry, osmotic fragility, or EMA flow cytometry. If the results of these tests are normal, then we should evaluate for RBC enzymopathies and also consider congenital dyserythropoietic anemias, a heterogeneous group of rare disorders characterized by ineffective erythropoiesis and iron overload.
The gene panels and subpanels that have become clinically available for the evaluation of hereditary hemolytic anemias over the last 10 to 15 years are a powerful addition to our diagnostic toolbox. Such workups are especially valuable in transfusion-dependent patients, in whom evaluation of their own RBCs is practically impossible outside research settings. These workups are also important before an irreversible treatment such as splenectomy is undertaken.
H&O When should clinicians consider splenectomy for patients with HS, and what are the current risk-benefit considerations?
TK As recently as 3 decades ago, the standard of care was to remove the spleen of a patient with a diagnosis of HS, even if the patient had mild HS with well-compensated hemolysis. However, the spleen is a useful organ of the immune system that also provides protection against intravascular hemolysis, thrombocytosis, and thrombophilia, so it is important to consider the risks and benefits of splenectomy in the context of the severity of the disease. A full workup, including a genetic evaluation, should be conducted to verify the diagnosis before splenectomy is recommended. We do not want a patient to have a splenectomy for presumed HS if the true diagnosis is HX or autoimmune hemolytic anemia with an underlying combined variable immunodeficiency or autoimmune lymphoproliferative syndrome (ALPS). Removal of the spleen can be detrimental in HX, predisposing to life-threatening thrombophilia.2 It can also be detrimental in ALPS, as it greatly increases the risk of overwhelming sepsis.4,8
If a patient has HS with transfusion dependency, splenectomy is a reasonable treatment because of the risk of iron overload with chronic transfusions. Partial rather than complete splenectomy can be a good option for patients with moderate spherocytosis. Patients with mild spherocytosis (defined by hemoglobin level of 11-15 g/dL with a reticulocyte count of 3%-8%)9 may be better off avoiding splenectomy. The decision to have total or subtotal splenectomy is individualized and must be made by the patient and family after a thorough diagnostic evaluation and consultation with the hematologist and the surgeon regarding the risks and benefits of the procedure.
Patients need to complete immunizations for encapsulated organisms at least 2 weeks before splenectomy. Patients who have had splenectomy require regular pneumococcal and meningococcal boosters per Centers for Disease Control recommendations, antibiotic prophylaxis, and prompt medical attention in case of fever to decrease the risk of life-threatening sepsis.
Disclosures
Dr Kalfa has received research funding from Forma Therapeutics/Novo Nordisk and Agios Pharmaceuticals and has done consulting for Forma Therapeutics/Novo Nordisk.
References
1. Risinger M, Kalfa TA. Red cell membrane disorders: structure meets function. Blood. 2020;136(11):1250-1261.
2. Kalfa TA. Diagnosis and clinical management of red cell membrane disorders. Hematology (Am Soc Hematol Educ Program). 2021(1):331-340.
3. Barcellini W, Fattizzo B. Strategies to overcome the diagnostic challenges of autoimmune hemolytic anemias. Expert Rev Hematol. 2023;16(7):515-524.
4. Kalfa TA. Warm antibody autoimmune hemolytic anemia. Hematology (Am Soc Hematol Educ Program). 2016;2016(1):690-697.
5. Mohandas N. Inherited hemolytic anemia: a possessive beginner’s guide. Hematology (Am Soc Hematol Educ Program). 2018;2018(1):377-381.
6. Mohandas N, Clark MR, Jacobs MS, Shohet SB. Analysis of factors regulating erythrocyte deformability. J Clin Invest. 1980;66(3):563-573.
7. Da Costa L, Suner L, Galimand J, et al; Society of Hematology and Pediatric Immunology (SHIP) group; French Society of Hematology (SFH). Diagnostic tool for red blood cell membrane disorders: assessment of a new generation ektacytometer. Blood Cells Mol Dis. 2016;56(1):9-22.
8. Rao VK. Approaches to managing autoimmune cytopenias in novel immunological disorders with genetic underpinnings like autoimmune lymphoproliferative syndrome. Front Pediatr. 2015;3:65.
9. Eber S, Lux SE. Hereditary spherocytosis—defects in proteins that connect the membrane skeleton to the lipid bilayer. Semin Hematol. 2004;41(2):118-141.