‘More work to do’ to determine cost-effectiveness of universal Duffy testing
Key takeaways:
- Universal testing for Duffy null phenotype may be cost effective depending on the type of neutropenia workup used.
- Universal testing could lead to more equitable health care.
SAN DIEGO — Lauren E. Merz, MD, MSc, can still feel both the relief and anger two parents expressed during her first year of medical school at University of Michigan.
The parents had a daughter about 4 years old who had been diagnosed with neutropenia.
They had visited numerous clinicians looking for an answer. Most assumed she had an immunodeficiency but, after extensive testing including a bone marrow biopsy, they could not find one.
Eventually, the parents took her from their home in Iowa to Michigan to visit a pediatric hematologist with whom Merz worked with at the time.
Merz’s mentor tested the child for the Duffy null phenotype, a genetic variant that can cause low circulating neutrophil counts without an increased infection risk and has a high prevalence among people with African or Middle Eastern ancestry.
The test — which costs about $10 — showed the girl, who was Black, had the Duffy null phenotype.
“It had been months of their lives,” Merz, now a hematology fellow at Dana-Farber Cancer Institute and Mass General Brigham, told Healio. “Their child had been through the wringer, and then someone just tells them that they’re normal. They’re fine. Everything’s going to be OK. [The reaction was] simultaneously ‘Thank God!’ and ‘Are you serious?’ We paid how many thousands of dollars to take time off work and to travel across state lines for this?”
Proponents have suggested universal Duffy testing could prevent other families from experiencing similar hardships; however, data to support universal testing had been limited.
Merz and colleagues evaluated the cost-effectiveness of universal testing for children aged 9 to 12 months.
The conclusion of the findings they presented in December at ASH Annual Meeting and Exposition can be summarized in a word: Maybe.
“The purpose of this is continuing the conversation on how to be better in health care and making things equitable for everyone,” Merz said. “We’re so white-centric in the way that we do our health care system.”
Background
Approximately 10% of individuals in the U.S. have the Duffy null phenotype, according to study background.
Less than 1% of individuals who identify as white or Asian have the Duffy null phenotype, compared with 66% who identify as either Black, African American, North African or Middle Eastern.
“That was actually an advantage, because it’s partially protective against a form of malaria called Plasmodium vivax,” Merz said. “That was probably a great selective advantage. That explains why we see the geography of the Duffy null phenotype in those areas.”
However, neutrophils are distributed differently among people with a Duffy null phenotype.
For most individuals, between 1% and 5% of neutrophils circulate in the blood, Merz said. Among those with Duffy null phenotype, a greater proportion of neutrophils reside in the bone marrow and spleen.
“Everyone has approximately the same amount of neutrophils if they’re healthy,” Merz said. “If you are Duffy null, they just don't like to live in your peripheral blood as much as they like to live in the spleen. When we draw blood, it looks like people with the Duffy null phenotype have lower neutrophils, but that’s not an accurate reflection of what’s going on in the rest of their body. That helps us understand why we don’t see things like an increased risk for infection and we’re not seeing any clinical differences in people by Duffy status.”
In 2023, Merz and colleagues published a report in JAMA with neutrophil ranges adjusted for Duffy status. Merz said the lower limit of normal for an adult is 1,900/L at her institution, but the lower limit for people with the Duffy null phenotype is 1,200/L.
“We built white blood cell and neutrophil reference ranges based on the Duffy status that 99% of people of European genetic ancestry will have,” Merz said. “We’ve completely neglected the most common variant — the Duffy null phenotype — that 80% to 100% of people from West Africa will have.
“This study is that next step of, if we are going to have these reference intervals, how do we use them if people don’t know their Duffy status?” she added.
Methods
In their most recent study, Merz and colleagues designed a Markov cohort model to evaluate the cost-effectiveness of incorporating universal Duffy testing into the American Academy of Pediatrics-recommended anemia and lead screening for children aged 9 to 12 months.
They used the 2022 American birth cohort of 3.67 million children as a study population.
They determined that 23.8% of people with Duffy null phenotype would have a count less than 2,000 — based on previous findings from adult data — compared with 2.5% for those without Duffy null phenotype, based on pathology standards that aim to capture the central 95% of values within a population.
Investigators compared the monetary benefits if patients had a basic pediatric neutropenia workup or a full evaluation. The basic workup included a referral to hematology with a follow-up visit, complete blood count differential, anti-neutrophil antibodies, Duffy testing, immunoglobulins, and fecal elastase or pancreatic amylase.
The full workup added evaluation of nutrition, inflammation, autoimmunity, infection, congenital neutropenia genetic testing, bone marrow biopsy, flow cytometry and cytogenetics.
A Duffy test costs $12.70, the basic work-up cost $386.35 and the full evaluation cost $2,284.04.
Incremental net monetary benefit for universal testing vs. no testing served as the primary endpoint. A threshold sensitivity analysis of minimum Duffy prevalence for testing to be cost effective served as the secondary endpoint.
Key findings
If every patient with an absolute neutrophil count below 2000/uL received a basic evaluation, universal Duffy testing would have an incremental net monetary benefit of negative $15 million (95% CI, 5-24). Universal testing is cost ineffective in this model.
If every patient received a full workup, universal Duffy testing would have an incremental net monetary benefit of $133 million (95% CI, 96-174). Universal testing is cost effective in this model.
This pattern was consistent in 100% of iterations in a probabilistic and deterministic sensitivity analysis.
Merz and colleagues determined 14.7% of the population would have to have the Duffy null phenotype for universal testing to be cost effective with the basic workup. However, only 2.6% of the population would have to have Duffy null phenotype for universal testing to be cost effective with the full evaluation.
“Is there any world in which this is going to be cost effective?” Merz asked. “We came away with ‘maybe.’ It would have been really nice to be like, ‘Yes, definitely,’ or ‘no, there’s no world,’ but this is kind of what we expected. The answer is maybe. We have more work to do to think about which populations should get Duffy testing or in which scenarios.”
Every child should know
Because most individuals who have a Duffy-null phenotype identify as Black or Middle Eastern, race seemingly would be a place to start.
Merz agreed it is an “important triage point,” but she has concerns making race the primary reason for testing.
“I’m not saying that we should be colorblind,” she said. “We tried that in medicine. We tried that in our society. That negates the experiences of people, and it also takes away an important triage tool. . Most people who have sickle cell disease are Black. It’s OK to look at someone and ask, ‘Do you have any genetic ancestry from sub-Saharan Africa?’ And if so, recognizing they are at higher risk for sickle cell trait or sickle cell disease. I worry in our medical field and in our society that we are stopping at race. We’re not looking at the biologic driver.
“With the Duffy null phenotype, I worry that people are going to say, ‘You’re Black, so you’re probably going to be on that lower neutrophil range, so we’re just going to set you there,’” Merz added. “That leads to risks for things like missing diagnoses of some cancers or some immunodeficiencies.”
Another potential problem by using race would include individuals who have one parent who identifies as Black and another who identifies as white.
“It’s really hard to practically implement a lot of these recommendations if we hinge our recommendations on a social construct,” Merz said.
Merz emphasized the need for more research.
She and her co-investigators want to get a better grasp of patterns of actual neutropenia diagnostic testing and determine the cost-effectiveness of that.
“Those [examples from the study] are binary,” she said. “That is not how people actually practice. We want to quantify how people practice. We’ll have a more tiered assessment.”
Merz also wants to evaluate how universal Duffy testing could impact quality of life.
“When you Google ‘neutropenia,’ it’s going to tell you that you have leukemia,” Merz said. “If you have neutropenia, you are often referred to a cancer center. There’s a quality of life impact we can financially quantify to avoid this referral.”
Additionally, neutrophil levels can disqualify individuals from clinical trials, and they are used for medication dosing.
“We know that people who identify as Black have significantly worse outcomes in cancer care than white people,” Merz said. “We haven’t really been able to figure out the why of it. My hypothesis is there’s at least partial blame on not understanding the Duffy null phenotype, inappropriately withholding chemotherapy due to neutropenia — even if it’s someone’s baseline counts — which then leads to worse overall survival.”
Merz said future research could make universal testing the norm for all children.
“I hope that in 10 to 20 years every child knows their Duffy status,” she said. “I hope that we have things integrated into the medical system that’ll help facilitate that. It’s automatically pulled in, which means that every single person — when they present for medical care — will know their Duffy status, that Duffy status will automatically trigger the appropriate reference interval, which means that we have a more personalized medical system.”
References:
- Merz LE, et al. Abstract 162. Presented at: ASH Annual Meeting and Exposition; Dec. 7-10, 2024; San Diego.
- Merz LE, et al. JAMA. 2023;doi:10.1001/jama.2023.7467.
For more information:
Lauren E. Merz, MD, MSc, can be reached at lauren_merz@dfci.harvard.edu.
Collapse
Merz LE, et al. Abstract 162. Presented at: ASH Annual Meeting and Exposition; Dec. 7-10, 2024; San Diego.