A 43-year-old white woman with a history of chronic pancreatitis, diabetes, anxiety, recurrent urinary tract infections, and chronic lower-back and pelvic pain was noted to have normocytic anemia (hemoglobin concentration, 9 to 10 g/dL) since 1987. Serum iron, vitamin B (₁₂), and folate levels were normal.

Macrocytic anemia with 20% to 30% reticulocytes and a peripheral smear consistent with intravascular hemolysis Figure 1 were first noted in October 1991. Heinz body preparations and urine hemosiderin results were positive. Diagnostic evaluations for erythrocyte enzyme deficiencies, Wilson disease, autoimmune hemolytic anemias, heavy metal toxicity, and paroxysmal nocturnal hemoglobinuria were negative. Examination of a bone marrow specimen showed erythroid hyperplasia. Tests to determine the presence of an unstable hemoglobin concentration produced inconclusive results. During repeated questioning by at least two physicians, the patient vehemently denied the intake of oxidant drugs.

View larger version (127K): [in this window] [in a new window]   Figure 1. Bite cells on peripheral blood smear (original magnification, x 1000; oil).

In September 1992 and January 1993, the patient presented with near syncope, orthostatic hypotension, and a hemoglobin concentration in the range of 5 to 6 g/dL. Physical examination showed marked pallor and a distinct lemon-yellow color of the skin and conjunctiva (although the serum bilirubin concentration never exceeded 4 mg/dL; conjugated, 0.3 mg/dL), a 3-cm smooth and nontender liver, glove-and-stocking distribution hypoesthesia, and a 2/6 basal systolic murmur. Lymphadenopathy, skin ulcers, and splenomegaly were absent.

The critical diagnostic clue was provided by the patient's clinical pharmacist, who in February 1993 notified the physician that the patient or her husband had purchased, over the counter, the daily dosing equivalent of 4 to 5 g of phenazopyridine (Pyridium; Parke Davis, Morris Plains, New Jersey) for several months. He called the last physician who had prescribed medication, a neurologist, who promptly called one of the authors. When confronted with this evidence, the patient initially denied and then admitted to the abuse. After stopping the drug, her hemoglobin concentration peaked at 14.3 g/dL during the next 6 months and has since remained stable. The peripheral smear no longer shows evidence of intravascular hemolysis.

The reported toxic spectrum of phenazopyridine does not include chronic intravascular hemolysis [1, 2]. This patient's yellow skin color in the absence of marked hyperbilirubinemia was a potential diagnostic clue [3].