Second Diagnosis

This patient has macrocytic anemia, occurring in the context of pancytopenia. The most feared cause of macrocytic pancytopenia is aplastic anemia, which is characterized by low counts in all three blood cell lines. The cytopenia usually is quite severe. This patient's hemoglobin level was compatible with that diagnosis, but she had only mild leukopenia (3,900/mm3). Furthermore, the white cell differential count, which showed 48% neutrophils, was not consistent with aplastic anemia. While laboratories traditionally give the differential in percentages, it is more useful to think in absolute numbers. In this case, 48% of 3,900 is about 1,900, which is a normal neutrophil count. Similarly, the platelet count of 62,000/mm3 is considered to be fairly mild thrombocytopenia. In aplastic anemia, the platelet count often drops below 20,000/mm3, a level associated with a risk of bleeding.

The total bilirubin level (1.2 mg/dL) in this case also argues against aplastic anemia. Except in cases of liver disease, in which total as well as direct bilirubin levels are elevated because of impaired hepatic or biliary function, the bilirubin level (particularly the indirect level) reflects the rate at which hemoglobin is broken down and heme is metabolized. Since women normally have a lower red cell mass than do men, they usually have lower bilirubin levels. Thus, while this patient's bilirubin level is technically at the upper limit of normal, the fact that she is a woman suggests that for all practical purposes it is elevated. In aplastic anemia, the bilirubin level is decreased because red blood cell production and destruction are diminished. Typically, the total bilirubin level is below the normal range, often at 0.4 or 0.3 mg/dL. The bilirubin level may be elevated when the anemia is associated with accelerated destruction of red cells, either in the peripheral blood (hemolytic anemia) or in the bone marrow (ineffective erythropoiesis). In the latter condition, red cell precursors form but die in the bone marrow. Malignant or infectious infiltration of the bone marrow is another possibility in a patient with pancytopenia. Normal bone marrow also may be replaced in patients with acute leukemia. However, immature blast cells are seen in the peripheral blood of patients with acute leukemia, and this patient's white cell count clearly indicated that she did not have leukemia. It did not rule out, but did reduce, the probability of so-called aleukemic leukemia. A much more likely possibility is that this patient has one of the myelodysplastic syndromes, clonal disorders of abnormal blood cell precursor production that sometimes develop into acute leukemia.

Myelodysplastic syndromes are most common in the elderly and should be included in the differential diagnosis of elderly patients with pancytopenia, even if mild. Bone marrow aspiration and biopsy, with chromosome analysis, is required for diagnosis. This patient's CBC was consistent with myelodysplastic syndrome; however, bilirubin levels are usually normal in such cases.

Pancytopenia can also result from trapping or pooling of cells in an enlarged spleen, as is often seen in patients with chronic liver disease. This patient did not have splenomegaly, however. The most likely diagnosis in this case is megaloblastic anemia, which is a common cause of macrocytic pancytopenia and by far the most readily treatable cause. Megaloblastic anemia can result from vitamin B12 or folate deficiency. Initially, patients have only a mild anemia with normal white cell and platelet counts. By the time the anemia becomes severe, the white cell and platelet counts almost invariably are decreased. The decrease is usually mild, as in this case.
A high bilirubin level is an important clue to megaloblastic anemia. Pancytopenia in patients with megaloblastic anemia occurs because-even though bone marrow cell production is markedly increased-abnormal DNA synthesis causes precursor cells to die before they can be released into the blood stream. The red cell precursors have already produced hemoglobin, and after they die in the marrow and phagocytosis occurs, their heme is converted to bilirubin-just as in hemolytic anemia, in which the red cells are destroyed in the circulation.

It is true that in severe megaloblastic anemia, the MCV is usually elevated. When the hemoglobin level is 5.4 gm/dL, one might expect the MCV to be as high as 120 µm3.

In this patient, the MCV was only 103 µm3. Nevertheless, I would advise against being misled by an overly rigid view of the MCV. For the sake of convenience, physicians often define macrocytosis as an MCV greater than 100 µm3. The actual threshold is closer to 97 µm3, however, and any reading above that should be considered suspicious. In fact, the ideal way to gauge the MCV is to compare it to the patient's usual value. If the patient customarily has an MCV of 85 µm3 and presents with an MCV of 92 µm3, the condition can be regarded as macrocytosis, even though the reading is technically within the normal range. A final consideration in patients with macrocytosis is chronic alcoholism. Indeed, alcohol may be the most common cause of macrocytosis. Alcohol causes increased red cell size and often decreased platelet counts. However, except in the most severe cases, alcohol causes only mild anemia. Chronic alcoholism can therefore be dropped from consideration in this case.