A 40-Year-Old Man With an Unexpected Drop in Hemoglobin Level
Author:
Ronald N. Rubin, MD—Series Editor
Citation:
Rubin RN. A 40-year-old man with an unexpected drop in hemoglobin level. Consultant. 2019;59(3):75-76,96.
A 40-year-old African American man presented to urgent care with renal colic. He is a “stone former” and had had multiple episodes over the past decade. He knows to remain hydrated and to use analgesia as needed. He also knows to seek medical care in the presence of fever, which was now present and had prompted his presentation. Overnight and on the morning of presentation, he had self-administered several doses of trimethoprim-sulfamethoxazole from his medicine cabinet. He was taking no other medications.
Physical examination revealed an obviously uncomfortable man who was experiencing significant renal colic on the left side. His temperature was 38.3°C, pulse was 108 beats/min, respiratory rate was 16 breaths/min, and blood pressure was 135/80 mm Hg. The mucosae were moist. Significant baseline data included a low hemoglobin level of 12.9 g/dL with normal mean corpuscular volume; an elevated white blood cell (WBC) count of 16,000/µL; normal basic metabolic panel findings, including a blood urea nitrogen (BUN) level of 19 mg/dL and a creatinine level of 1.1 mg/dL; and an elevated lactate dehydrogenase (LDH) level of 290 U/L. The rest of the laboratory test results were within normal limits. He had hematuria, with the urine a rose wine color.
A computed tomography scan revealed a degree of obstructive uropathy with mild dilatation of the left ureter originating at the ureterovesical junction. He was admitted with plans to administer hydration and parenteral antibiotics pending urologic surgical removal if the stone were to not pass.
On the morning of hospital day 2, laboratory test results now revealed that the hemoglobin level had dropped even lower to 10.1 g/dL. The WBC count was now 11,500/µL, with a normal platelet count. The creatinine level essentially had remained stable at 1.2 mg/dL. His urine remained rose wine in color, and overnight output measured 1100 mL. His temperature was 38.1°C, and his blood pressure was 130/80 mm Hg. When he was informed that he was anemic, he recalled having experienced a similar episode some years ago.
Answer: C, initiate serial complete blood cell count monitoring and obtain haptoglobin, bilirubin, LDH, and reticulocyte levels.
The salient clinical findings in this patient’s case are, of course, the urolithiasis syndrome that prompted his presentation and the relatively large and unexpected drop in hemoglobin that accompanied it.
When an unexpected drop in hemoglobin occurs, we round up the usual suspects. Initial thoughts include some form of blood loss, which may be hemodynamically significant and thus life-threatening. Gastrointestinal tract bleeding is the most common and should be excluded by way of serial stool guaiac testing. GU tract bleeding is far less common as a cause, and arithmetic explains why. With hematuria, useful clinical saws are that rose wine (pink) urine has a hematocrit of 1% to 2%, while port wine or darker wine urine has a hematocrit of 2% to 5%. Given our patient’s overnight urine output of 1100 mL, that would equate to a GU blood loss of approximately 10 to 20 mL, which is nowhere near enough to cause the hemoglobin decrease seen here, thus making Answer B extremely unlikely and incorrect here.
Another often-considered etiology of decreased hemoglobin is hydration. Indeed, vigorous hydration in a dry patient can lead to changes in the hemoglobin level. However, other parameters such as BUN and creatinine will also “hydrate.” Our patient was not clinically dehydrated (as indicated by his normal blood pressure and creatinine levels) and did not change with intravenous therapy.
When blood loss and dehydration have been excluded, it is time to consider everyone’s favorite diagnosis—hemolysis, when enhanced destruction of red blood cells (RBCs) exceeds the normal marrow’s significant reserve capacity to make more RBCs. Sophomore medical students and surgical intensive care unit staff favor this diagnosis at a rate that is out of proportion with its true incidence as a cause of anemia. Happily, once hemolysis makes its way into the differential diagnosis, a battery of relatively simple laboratory tests with quite rapid turnaround are reasonably specific and sensitive to exclude or confirm its presence. These tests rely on the fact that RBC destruction, whether it is extracellular or intracellular, will result in the release of hemoglobin and RBC cellular contents, with catabolism of the hemoglobin molecules into identifiable products—specifically, LDH (concentration in RBCs) and bilirubin, as well as the consumption and subsequent lowering of the protective protein haptoglobin, which binds to the released hemoglobin. The lowering of the hemoglobin with erythropoietin release and (by a yet poorly understood mechanism) hemolysis itself trigger a marrow response to form new RBCs such that the reticulocyte count rises. This process may take a week or so and may not be evident early on.
Although we must think and act in a coordinated way in the clinic and thus often use a “shotgun” approach, at least cerebrally I would ask that we confirm or exclude the presence of hemolysis as a cause for hemoglobin decline using those 4 tests—levels of LDH, bilirubin, and haptoglobin and reticulocyte count—and then proceed to define the reason for the decrease. In this patient’s case, therefore, Answer C is correct—obtain those key tests for hemolysis while monitoring the hemoglobin level for further declines that might require transfusion while sorting out specific causation.
Once hemolysis has been confirmed, a myriad of etiologies must be considered. Two clinically useful schemes have evolved to aid in sorting the large number of possible causes into more manageable subcategories. Hemolysis may result from a hereditary condition or an acquired one. Or, hemolysis may be caused by an intracellular abnormality within RBCs, making them liable to premature destruction, or it may be caused by an extracellular abnormality that is capable of destroying otherwise normal RBCs. Both of these schemes are more useful than categorizing hemolysis as intracellular or extracellular in this author’s opinion and that of others.1 Finally, there is great overlap of extracellular causes being predominantly acquired in nature, while intracellular causes are more often inherited.
CONTINUED
In the acquired/extracellular category, lesions such as autoantibody-induced hemolysis by the reticuloendothelial system and microangiopathic lesions with fragmentation destruction hemolysis as found in vascular heart disease and disseminated intravascular coagulation are examples. In the inherited/intracellular category, once can work from the membrane, such that (1) membrane defects in the spectrin membrane sodium-potassium pump can result in hereditary spherocytosis and osmotic lysis; (2) cytoplasmic defects such as inherited abnormal glucose-6-phosphate dehydrogenase (G6PD) renders the RBCs liable to oxidative denaturation and destruction2,3; and (3) genetically altered hemoglobin molecules, such as hemoglobin S, form abnormal tactoids within RBCs, with the subsequent shape change causing RBC destruction. A valuable clue to help initial probing into hemolysis causation is that the inherited group usually manifests at a younger age, and hemolysis in an older adult makes an acquired form more likely. Finally, a professionally examined peripheral blood smear is extremely useful in pinpointing a cause. In fact, smear examination can identify spherocytes, schistocytes, and sickle cells, thus confirming hemolysis and its cause very quickly despite its monetary value of a scant 2 relative value units.
Regarding Answer A, iron deficiency simply cannot cause this acute change so quickly, and thus this answer is incorrect. Answer D, renal insufficiency with inadequate erythropoietin, is also incorrect given the speed of the decline on one hand and the minimal degree of renal insufficiency on the other.
PATIENT FOLLOW-UP
Broad-spectrum parenteral antibiotics were administered along with analgesics. The patient passed the kidney stone on the second day, with prompt resolution of fever and other symptoms. His hemoglobin remained stable in the 9 to 10 mg/dL range. Results of laboratory tests on day 2 revealed a haptoglobin level of less than 6 mg/dL, an LDH level of 356 U/L (reference value, <250 U/L), a bilirubin level of 2.2 mg/dL, and a reticulocyte count of 31 × 103/µL (reference range, 10-25 × 103/µL), confirming hemolysis. A detailed review of the patient’s records revealed occurrence of these episodes over time in association with urinary tract infection and sulfa antibiotics. Results of a quantitative G6PD assay subsequently confirmed deficiency. At 3 months, his hemoglobin level had totally normalized.
Ronald N. Rubin, MD, is a professor of medicine at the Lewis Katz School of Medicine at Temple University and is chief of clinical hematology in the Department of Medicine at Temple University Hospital in Philadelphia, Pennsylvania.
References:
- Puig A, Dighe AS. Case 20-2013: a 29-year-old man with anemia and jaundice. N Engl J Med. 2013;368(26):2502-250
- Frank JE. Diagnosis and management of G6PD deficiency. Am Fam Physician. 2005;72(7):1277-1282.
- Cappellini MD, Fiorelli G. Glucose-6-phosphate dehydrogenase deficiency. Lancet. 2008;371(9606):64-74.
TAKE-HOME MESSAGE
Although hemolytic anemia is a general condition with a copious number of specific causes, it is a rather uncommon cause of anemia. The initial aspect when entertaining hemolysis as the cause of anemia in a patient is to confirm its presence. The key tests are measurements of haptoglobin, LDH, bilirubin, and reticulocytes. Once confirmed, specific causes of hemolysis can be sought. Convenient classification schemes such as hereditary vs acquired and extracellular vs intracellular causes can help isolate a specific diagnosis. A frequently overlooked cause is G6PD deficiency in African American patients. Episodic mild episodes related to drugs and infections suggest the diagnosis.