Background: Glucose 6-phosphate dehydrogenase (G6PD) deficiency is a genetic disorder that occurs in approximately 400 million people worldwide. It is characterized by varying degrees of hemolysis most often triggered by infections, drugs, or certain foods. However, most people with G6PD deficiency do not have a clinically significant phenotype. Some medications thought to trigger hemolytic crises in patients with G6PD deficiency can be life-saving (e.g. rasburicase in severe tumor lysis syndrome). Thus, the ability to stratify patients according to risk of hemolysis may have significant clinical implications. In December 2014, our reference lab increased the lower limit of normal of the G6PD enzymatic assay from 7 units/g Hb to 9.9 units/g Hb, which increased the number of patients diagnosed with G6PD deficiency. However, it remains unknown if patients with intermediate levels (between 7 and 11 units/g Hb) experience hemolysis when exposed to triggers compared to patients with levels below 7 units/g Hb.

Methods: We conducted a retrospective cohort analysis of patients with G6PD enzymatic levels between 2014 and 2016 to determine risk of hemolysis in patients with low or intermediate levels upon exposure to hemolysis-triggering medications (rasburicase, dapsone, nitrofurantoin, primaquine, cotrimoxazole, methylene blue, probenecid, phenazopyridine, moxifloxacin and aspirin). Hemolysis was defined by clinical documentation and further characterized by need for transfusion after drug exposure as well as a decrease in pre-exposure hemoglobin after exposure. Inclusion criteria were patients 0-21 years of age at time of G6PD assay, G6PD assay level less than 11 units/g Hb, exposure to one of the listed medications, and those with peri-exposure hemoglobin and transfusion data available. Patients were excluded if G6PD assay level was ≥ 11 units/g Hb, they were over 21 years of age at time of assay, or were not exposed to the listed medications.

Results: We identified 704 patients who had available G6PD assays during the study period. Assay levels in this cohort ranged from 0.3 to 270 units/g Hb with 291 having G6PD assay levels less than 11 units/g Hb. Of these patients, 39 had both qualifying G6PD levels and documented exposure to one or more of the triggering medications; 5 patients had low G6PD levels (<7 units/g Hb), 14 patients had intermediate levels (7-10 units/g Hb) and 20 patients had high intermediate G6PD levels (10-11 units/g Hb). The distribution of hemolysis-triggering medication exposure was as follows: cotrimoxazole (n=35), rasburicase (n=5), dapsone (n=1), primaquine (n=1), methylene blue (n=1), moxifloxacin (n=1), aspirin (n=1), nitrofurantoin (n=0), probenecid (n=0), and phenazopyridine (n=0). Of the triggering medications we identified, rasburicase is the only medication with a known black box warning against its use in patients with G6PD deficiency.

There were two patients amongst those in the study who had significant hemolytic events associated with anemia warranting blood transfusion. One patient (a male) with a low G6PD assay level (2.4 units/g Hb) had clinically significant hemolysis in the setting of rasburicase exposure. The other patient (a female) with a high intermediate G6PD assay (10.8 units/g Hb) had clinically significant hemolysis in the setting of cotrimoxazole exposure, but had a complicated clinical course making it difficult to attribute causality of hemolysis definitively to cotrimoxazole. Of note, no patients with intermediate G6PD assay levels (between 7 and 10 units/g Hb) had significant hemolysis despite exposure to a triggering medication.

Conclusions: Our results indicate that there was no clear association between G6PD assay levels above 7 units/g Hb and the incidence of hemolysis. Despite a general avoidance of cotrimoxazole exposure in patients with known G6PD deficiency, our study did not reveal a significant correlation between drug exposure and hemolytic events in this population. Importantly, the only patient with significant hemolysis in the setting of rasburicase exposure had a very low G6PD level. Further correlation with genotype and G6PD assay level may have significant clinical implications to predict one's risk of clinically significant hemolysis.

Disclosures

Smith-Whitley:Pfizer: Membership on an entity's Board of Directors or advisory committees. Lambert:Novartis: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract

Sign in via your Institution