Voxelotor (GBT440) produces interference in measurements of hemoglobin S

Dear Editor:

A 35-y-old African-American male presented to the emergency department at Vanderbilt University Medical Center for treatment of sickle cell pain crisis. An emergent red blood cell (RBC) exchange was ordered and pre-exchange hemoglobin S (HbS) quantitation was performed. Our laboratory performs HbS quantitation by cation exchange high performance liquid chromatography (HPLC) using the BioRad Variant II β-thalassemia short program. Upon review of the HPLC chromatogram, the chemistry team observed a split peak spanning the hemoglobin D (HbD) and HbS windows (Fig. 1, Pre-transfusion). Repeat analysis on an alternate analyzer produced a similar profile, ruling out an issue with the analyzer. A new specimen was requested to rule out artifacts from pre-analytical sample handling, such as a mislabeled specimen. The chromatogram of the new sample was consistent with the results of the previous sample; however, due to the overlap of the peaks, HbS% could not be immediately reported. This subsequently led to a delay in the patien’s RBC exchange.

Fig. 1.

Cation–exchange HPLC and isoelectric focusing (IEF) gel analyses of samples drawn prior to the patient joining the Voxelotor study (Pre-study), when he presented at our emergency department for treatment of sickle cell pain crisis (Pre-transfusion), and 2 weeks following his exchange transfusion (2 weeks post-transfusion). Additional hemoglobin species, which we attribute to hemoglobin-Voxelotor complexes, are visible in the pre-transfusion and 2 weeks post-transfusion samples. Note that the prestudy HPLC and IEF analyses were performed on separate samples. Both are included to demonstrate that the additional hemoglobin species were not present prior to the patient joining the trial.

A detailed review of the patien’s medical record revealed a prior diagnosis of homozygous HbS (HbSS) sickle cell disease (SCD). The patient was receiving hydroxyurea, a standard treatment for SCD. There were no recent RBC transfusions or exchanges noted in the medical record and discussions with the clinical team revealed that they were not aware of any recent transfusions at an outside hospital, eliminating this potential source of an unexpected hemoglobin variant. An isoelectric focusing gel (IEF) was performed, which displayed the expected HbS band as well as additional bands that migrated on either side of HbS (Fig. 1, Pre-transfusion). Of note, the patient was recently enrolled in a phase 3 clinical trial (“Study to Evaluate the Effect of GBT440 Administered Orally to Patients with Sickle Cell Disease (GBT_HOPE)”) studying the efficacy of a new SCD medication called Voxelotor (previously known as GBT440). Further discussions with the clinical team and a review of his medical record established that the patient was taking Voxelotor.

To investigate Voxelotor as the source of the interference, we examined the patien’s previous hemoglobin analyses (Fig. 1, Pre-study), performed prior to joining the clinical trial. The chromatogram displayed a single HbS peak, along with hemoglobin F (HbF) and hemoglobin A₂ (HbA₂) peaks, consistent with the patien’s diagnosis of HbSS and his hydroxyurea therapy. The IEF gel was also consistent with the HbSS diagnosis (Fig. 1, Pre-study). The patient continued to take Voxelotor and a second sample collected 2.5 weeks after his RBC-exchange procedure showed a split peak in the HbS window and two additional unexpected peaks that eluted before hemoglobin A (HbA; Fig. 1, 2 weeks post-transfusion), which was not present in a sample obtained immediately after his exchange. A split peak in the HbS zone was also observed when this sample was analyzed using capillary zone electrophoresis (CZE, Sebia MiniCap), an alternate method for HbS quantitation (Fig. 2). IEF performed on this sample displayed the expected HbS band, an HbA band from his transfusion, as well as multiple additional bands that migrated on either side of HbS, and also near HbA (Figs. 1, 2 weeks post-transfusion). Voxelotor, which has received Fast Track, Orphan Drug and Rare Pediatric Disease designations from the U.S. Food and Drug Administration, has been shown to form complexes with the N-terminus of hemoglobin a chains [1–3]. Qualitative matrix-assisted laser desorption/ionization (MALDI) mass spectrometry experiments confirmed the presence of Voxelotor-hemoglobin alpha chain complexes in specimens collected from this patient (Fig. 2). We attribute the unexpected hemoglobin species present on the chromatograms and IEF gels to hemoglobin-Voxelotor complexes. As evidenced in Figs. 1 and 2, we observed multiple hemoglobin-Voxelotor complexes, and the ability to resolve these complexes from the unmodified hemoglobin species depended on the measurement technique that we used. These additional complexes may represent Voxelotor modification of minor hemoglobins like HbA₂ and HbF.

Fig. 2.

Results of capillary zone electrophoresis (CZE) and matrix-assisted laser desorption/ionization mass spectrometry (MALDI) performed on the sample drawn 2 weeks post-transfusion. In keeping with the HPLC results, a split peak was also noted in the HbS zone by CZE. An additional Hb-Voxelotor complex migrating near HbA2 was resolved by CZE. Voxelotor is known to interact with the alpha chain of hemoglobin, and a species consistent with the expected mass to charge ratio of a Voxelotor-alpha chain complex was detected qualitatively by MALDI.

A drug that can increase hemoglobin concentration and decrease hemolysis in patients with SCD could significantly improve their quality of life; however, our data suggest Voxelotor can interfere with laboratory tests that are routinely used to monitor these patients and guide their therapy [4,5]. For example, HbS% is used to calculate the number of RBC units to administer during exchange transfusions. According to recent College of American Pathologists (CAP) survey data, 60% of CAP-accredited laboratories performing hemoglobinopathy evaluations use either BioRad HPLC or Sebia CZE methodologies. Sites participating in clinical trials of Voxelotor should notify their laboratories of this interference, and it should be noted in the Voxelotor package insert. In this case, the interference caused a several day delay in the patient’s RBC exchange while the laboratory worked to identify the unexpected hemoglobin species. In order to avoid potential delays in patient care, we encourage laboratorians and test manufacturers to determine whether their methods are subject to the interference described here. Finally, there is not currently any guidance on how laboratories should report HbS% in the presence of complexed HbS, specifically, whether the HbS-Voxelotor complexes should be regarded as equivalent to HbS. The decision to report “total” HbS or HbS without HbS-Voxelotor complexes could lead to differences in the number of units of RBCs a patient receives in an exchange procedure. Additional studies into the structure-function relationship of the hemoglobin-Voxelotor complexes are necessary to understand how quantitative HbS results should be reported in the presence of this interference.

Abbreviations:

HbS

hemoglobin S

HbD

hemoglobin D

HbSS

homozygous hemoglobin S

SCD

sickle cell disease

IEF

isoelectric focusing

HbF

hemoglobin F

HbA₂

hemoglobin A₂

HbA

hemoglobin A

CZE

capillary zone electrophoresis

CAP

College of American Pathologists

RBC

red blood cell

HPLC

high performance liquid chromatography

MALDI

matrix-assisted laser desorption/ionization