Revised Prevalence Estimate of Possible Hereditary Xerocytosis as Derived from a Large U.S. Laboratory Database Harvey W. Kaufman1 , Justin K. Niles1, Denis R. Gallagher1, Alicia Rivera2, Seth L. Alper2 , Carlo Brugnara3 , L. Michael Snyder1. 1. Quest Diagnostics, Marlborough, MA 01752 2. Division of Nephrology and Vascular Biology Research Center, Beth Israel Deaconess Medical Center, and Department of Medicine, Harvard Medical School, Boston, MA 02215 3. Department of Laboratory Medicine, Boston Children's Hospital and Department of Pathology, Harvard Medical School, Boston, MA 02115 This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process which may lead to differences between this version and the Version of Record. Please cite this article as an ‘Accepted Article’, doi: 10.1002/ajh.24923 This article is protected by copyright. All rights reserved. American Journal of Hematology To the Editor: The rare autosomal dominant hemolytic anemia known as hereditary xerocytosis (HX) is characterized by disease-causing mutations in one of two ion-channel proteins of the red cell membrane. The mechanosensitive cation channel PIEZO1 is mutated in the majority of cases, while a smaller subset of patients harbor mutations in the Ca2+-gated K+ channel, KCNN4 (Gardos channel)  . When normal circulating erythrocytes undergo reversible distention and distortion during their navigation through narrow tissue capillaries and hepatosplenic sinuses, the mechanical stimuli activate PIEZO1, elevating intracellular [Ca2+]. The elevated intracellular [Ca2+] in turn activates KCNN4 to allow K+ efflux and cell shrinkage. These two channels thus constitute sequential steps in an erythrocyte Ca2+ signaling pathway [2, 3]. The heterozygous missense mutations in either PIEZO1 or KCNN4 that cause autosomal dominant HX appear to confer “gain-of-function,” manifested as abnormally increased channel activity that is usually, but not uniformly, secondary to prolonged channel inactivation time [2-4]. Increased KCNN4 channel activity leads directly to erythrocyte dehydration via loss of K+, Cl-, and water. In the case of PIEZO1 mutations, the dysfunction increases Ca2+ entry into erythrocytes, in turn increasing KCNN4 activation and cellular dehydration [2-4]. The hemolytic anemia associated with HX varies from fully compensated to severe, and is accompanied by chronic fatigue, muscle weakness, abdominal pain from splenic enlargement, and pathological iron accumulation in body tissues . However, while splenectomy can relieve abdominal pain, it may be of no benefit for correcting anemia and carries an exceptionally high risk of thromboembolic complications (at least in the context of PIEZO1 mutations) . The central role of KCNN4 in the red blood cell dehydration of HX has led to the proposed use of the 2 John Wiley & Sons This article is protected by copyright. All rights reserved. Page 2 of 11 Page 3 of 11 American Journal of Hematology KCNN4 (Gardos channel) blocker, senicapoc, to prevent dehydration and subsequent hemolysis of red blood cells . Evidence-based estimates of the number of HX patients would be of benefit in planning clinical tests of senicapoc as a possible treatment for HX and in developing molecular diagnostics for those patients. Classical estimates of HX prevalence of ~1:50,000  have been based on small numbers of reported patients. However, the study of HX cohorts in which hemolytic disease occasionally cosegregates with disease variants (or disease-associated variants) of higher allele frequency [6, 8] suggests that HX is more common than previously believed. As the sometimes extreme red blood cell dehydration of HX is often revealed in increased values of mean corpuscular hemoglobin concentration (MCHC) , we mined a large commercial lab database for complete blood counts (CBCs) results. All CBCs performed between January 2014 and December 2016 that included numerical results for MCHC, hemoglobin, and mean corpuscular volume (MCV) were retrieved from the Quest Diagnostics Health Trends™ Database. Individual patients were identified either by an enterprise master patient identifier or by a combination of patient gender, date of birth, ZIP Code®, and insurance provider. Patients from all 50 states and District of Columbia were represented. The dataset was edited to include only the first CBC performed within the study period for each individual patient of identified sex between the ages of 18 and 99 years. Statistical significance testing of group proportions was conducted using the chi-square test. Data were analyzed using SAS Studio (SAS Institute, Cary, N.C.). This Quest Diagnostics Health Trends™ report was deemed exempt by the Western Institutional Review Board. Median age was 51 years, and 59.1% of subjects were female. Hemoglobin and MCHC values grouped by reference range (RR) are shown in Table 1. Of the 48,403,254 patients 3 John Wiley & Sons This article is protected by copyright. All rights reserved. American Journal of Hematology included in this analysis, 27,108 (0.056%) had MCHC values above the RR, of whom 4,029 (0.0083% of overall population) also had hemoglobin levels below the RR. This latter, smaller group includes anemic patients with possible HX, referred to as "candidate HX (anemia)." Among subjects less than 50 years of age, the criteria for "candidate HX (anemia)" were more commonly present in females than in males (0.0059% vs. 0.0036%; p<0.01). In contrast, those criteria in subjects 50 or older were more common in males than in females (0.0129% vs 0.0100%; p<0.01). Table 2 shows concurrent CBC values and indicators of hemolysis for subjects with high MCHC in whose samples cold agglutinins had been ruled out. Among subjects with in-range or elevated hemoglobin values, 52.5% had elevated reticulocyte counts, 10.5% had elevated total bilirubin concentrations, and 12.9% had elevated serum levels of lactate dehydrogenase (LD). In addition, 15.9% of patients with high MCHC had elevated serum ferritin levels, consistent with the systemic iron overload of HX, exceeding that expected from the degree of hemolysis. Also notable among these patients were elevated MCH values in 46.8%, consistent with the macrocytosis commonly noted in HX patients. Patients with combinations of elevated MCHC, in-range or elevated hemoglobin, and elevated total bilirubin, ferritin, LD, or reticulocyte count were defined as having possible HX with compensated hemolysis, and referred to as "candidate HX (compensated hemolysis)." For subjects of all ages, elevated MCHC was more frequent in males than in females (0.0831% vs. 0.0372%; p<0.01), as was also true for the group "candidate HX (compensated hemolysis)" (0.0103% vs. 0.0019%; p<0.01). These data suggest that the number of subjects with hematological indices consistent with HX is higher than would be predicted by previous estimates of HX prevalence based on sporadic clinical presentation . Extrapolation of this analysis to the 2010 US Census population 4 John Wiley & Sons This article is protected by copyright. All rights reserved. Page 4 of 11 Page 5 of 11 American Journal of Hematology between the ages of 18-99 (243,226,274; https://www.census.gov/2010census/) suggests that approximately 146,000 US residents have elevated MCHC values (Table 3). Among these individuals, 18,718 exhibit additional hematological parameters consistent with "candidate HX (anemia)", and 15,531 have indices consistent with "candidate HX (compensated hemolysis)" (Table 3). Thus, an extrapolated 34,249 people in the US exhibit elevated MCHC accompanied by compensated, mild, or severe anemia, representing ~1:7,000 U.S. adults aged 18-99 years. We therefore suggest that patients with unexplained anemia or compensated hemolysis accompanied by elevated MCHC, as well as symptomatic patients with elevated MCHC in the absence of anemia, may, be candidates for genetic screening for mutations associated with HX. However, elevated MCHC values cannot be attributed exclusively to HX. MCHC can be elevated by hyperlipidemia, hyperbilirubinemia, hyponatremia, and hemolysis, none of which factors have been systematically controlled for in our analysis . Fewer than 4% of hereditary spherocytosis (HS) patients have elevated MCHC . This is equivalent to the 3.67% of our subjects with both elevated MCHC and anemia, who also have low MCV (Table 2), representing ~750 patients within our extrapolated US population estimate. The ~10% or more of sickling hemoglobinopathy patients (SS and SC) with elevated MCHC  may represent as many as 1,850 patients among our extrapolated US population estimate, assuming a US sickle cell disease population of 100,000. Moreover, homozygous HbC disease affects 0.02% of the approximately 29,000,000 American adults of self-described African descent aged 18-99 (2010 US Census), and accounts for 5,800 additional patients with elevated MCHC. However, the "candidate HX" groups capture only that ~15% of HbC disease patients with MCV within or above RR , numbering ~900 patients. Thus, subtracting the above ~3,500 patients with HS or with hemoglobinopathies SS, SC, or CC from the extrapolated 34,249 US subjects with elevated 5 John Wiley & Sons This article is protected by copyright. All rights reserved. American Journal of Hematology MCHC and either anemia or compensated hemolysis leaves ~31,000 subjects from our extrapolated study population in the category of "candidate HX." This value suggests a possible HX prevalence in the US population (the "anemia" group plus the "compensated hemolysis" group) of 1:8,000 US adults, or ~6-fold greater than classical estimates based on clinical presentation. However, even this estimated prevalence is likely a conservative one, since a not insubstantial minority of HX patients have MCHC within RR. In addition, several PIEZO1 variants genetically linked to HX [6, 8] have minor allele frequencies of 0.0033 or higher in the general population (exac.broadinstitute.org) The main strength of this study is its evaluation of nearly 20% of American adults, providing a broad window into the laboratory findings associated with anemia and compensated hemolysis accompanied by elevated MCHC. The main weakness of the study is that subject ascertainment reflects the variations in degree of clinical suspicion of anemia and other medical conditions by healthcare providers, as reflected in ordering of CBC testing. Thus, the study is limited by the large proportion of patients who did not undergo tests of the additional parameters contributing to our definition of "presumed HX (compensated hemolysis)," likely rendering these estimates somewhat conservative. Another limitation is extrapolating prevalence from a population who received medical care to the entire U.S. population. In contrast, possible interference with MCHC measurement by concomitant hyperbilirubinemia, hyperlipidemia, hyponatremia, or intravascular or post-collection hemolysis would contribute to overestimation of the candidate HX population. Although no specific treatment for HX is currently available, the KCNN4 inhibitor senicapoc, has been well tolerated in previous phase III clinical trials  and has been proposed as a targeted therapy for HX [2, 6, 8]. Future studies of senicapoc and of other nascent 6 John Wiley & Sons This article is protected by copyright. All rights reserved. Page 6 of 11 Page 7 of 11 American Journal of Hematology therapeutic approaches for patients with elevated MCHC and anemia or compensated hemolysis , will evaluate the utility of PIEZO1 and KCNN4 genetic screening in defining eligibility for treatment with senicapoc or other drugs. These studies may define patient subsets who may most benefit from genetic testing. Acknowledgements: We thank Dr. Jay Wohlgemuth and Jeff Radcliff for support and critical comment. Conflict of interest disclosure: Drs. Alper, Brugnara, and Snyder are co-investigators in a clinical project for the treatment of HX (dehydrated stomatocytosis) with Springworks Therapeutics. Dr. Brugnara and Boston Children’s Hospital hold patents potentially related to senicapoc. Dr. Alper has received research funding from Quest Diagnostics. Dr. Kaufman, Mr. Niles, Mr. Gallagher, and Dr. Snyder are employees of Quest Diagnostics. 7 John Wiley & Sons This article is protected by copyright. All rights reserved. American Journal of Hematology References 1. Andolfo I, Russo R, Gambale A, et al. New insights on hereditary erythrocyte membrane defects. Haematologica 2016;101:1284-1294. 2. Andolfo I, Russo R, Manna F, et al. Novel Gardos channel mutations linked to dehydrated hereditary stomatocytosis (xerocytosis). American Journal of Hematology 2015;90:921-926. 3. Cahalan SM, Lukacs V, Ranade SS, et al. Piezo1 links mechanical forces to red blood cell volume. eLife 2015;4. 4. Alper SL. Genetic Diseases of PIEZO1 and PIEZO2 Dysfunction. Current Topics in Membranes 2017;79:97-134. 5. Iolascon A, Andolfo I, Barcellini W, et al. Recommendations regarding splenectomy in hereditary hemolytic anemias. Haematologica 2017;102:1304-1313. 6. Rapetti-Mauss R, Picard V, Guitton C, et al. Red blood cell Gardos channel (KCNN4):the essential determinant of erythrocyte dehydration in Hereditary Xerocytosis. Haematologica 2017. 7. Ataga KI, Reid M, Ballas SK, et al. Improvements in haemolysis and indicators of erythrocyte survival do not correlate with acute vaso-occlusive crises in patients with sickle cell disease: a phase III randomized, placebo-controlled, double-blind study of the Gardos channel blocker senicapoc (ICA-17043). British journal of haematology 2011;153:92-104. 8. Glogowska E, Schneider ER, Maksimova Y, et al. Novel mechanisms of PIEZO1 dysfunction in hereditary xerocytosis. Blood 2017. 9. Berda-Haddad Y, Faure C, Boubaya M, et al. Increased mean corpuscular haemoglobin concentration: artefact or pathological condition? International journal of laboratory hematology 2017;39:32-41. 10. Rocha S, Costa E, Rocha-Pereira P, et al. Complementary markers for the clinical severity classification of hereditary spherocytosis in unsplenectomized patients. Blood cells, molecules & diseases 2011;46:166-170. 11. Parrow NL, Tu H, Nichols J, et al. Measurements of red cell deformability and hydration reflect HbF and HbA2 in blood from patients with sickle cell anemia. Blood cells, molecules & diseases 2017;65:41-50. 12. Murphy JR. Hemoglobin CC disease: rheological properties or erythrocytes and abnormalities in cell water. The Journal of clinical investigation 1968;47:1483-1495. 8 John Wiley & Sons This article is protected by copyright. All rights reserved. Page 8 of 11 Page 9 of 11 American Journal of Hematology Table 1: Hemoglobin and MCHC Reference Range Groups MCHC Values (g/dL) Hemoglobin Values (g/dL) Male: >17.1 Female: >15.5 Male: 13.1 - 17.1 Female: 12.2 - 15.5 Male: <13.1 Female: <12.2 Total >36 2,605 (0.0054) 20,474 (0.0423) 4,029 (0.0083) 27,108 (0.0560) 32-36 807,177 (1.67) 35,640,342 (73.63) 5,917,868 (12.23) 42,365,387 (87.53) <32 57,626 (0.12) 3,456,431 (7.14) 2,496,702 (5.16) 6,010,759 (12.42) Total 867,408 (1.79) 39,117,247 (80.82) 8,418,599 (17.39) 48,403,254 (100) ______________________________________________________________________________ Parentheses indicate values in % of total population tested. 9 John Wiley & Sons This article is protected by copyright. All rights reserved. American Journal of Hematology Page 10 of 11 Table 2: Select Hematologic Indices in Patients with Elevated MCHC and with Hemoglobin Values either Below Reference Range or Within or Above Reference Range MCH MCV RBC RDW Total Bilirubin Ferritin LD Reticulocyte Count Hemoglobin Below Reference Range Below Above Reference Within Reference TOTAL Range, N Reference Range, N N (%) Range, N (%) (%) 1,301 2,691 4,029 37 (0.92) (32.29) (66.79) 2,924 4,029 148 (3.67) (72.57) 957 (23.75) 3,689 3,895 (94.71) 205 (5.26) 1 (0.03) 1,970 1,662 3,632 0 (0) (54.24) (45.76) 1,918 2,449 4 (0.16) (78.32) 527 (21.52) 306 24 (7.84) 189 (61.76) 93 (30.39) 121 5 (4.13) 77 (63.64) 39 (32.23) 88 6 (6.82) 33 (37.50) Hemoglobin Normal or Above Reference Range Below Above Reference Within Reference TOTAL Range, N Reference Range, N N (%) Range, N (%) (%) 12,238 10,800 23,079 41 (0.18) (53.03) (46.8) 21,071 23,079 306 (1.33) (91.30) 1,702 (7.37) 3,512 16,294 20,219 (17.37) (80.59) 413 (2.04) 16,083 17,577 5 (0.03) (91.50) 1489 (8.47) 15,704 1,838 17,553 11 (0.06) (89.47) (10.47) 774 15 (1.94) 636 (82.17) 123 (15.89) 518 13 (2.51) 438 (84.56) 67 (12.93) 49 (55.68) 59 5 (8.47) 23 (38.98) ______________________________________________________________________________ N, Number of patients in each category for whom laboratory value was available in the database. Parentheses indicate % of "TOTAL N". 10 John Wiley & Sons This article is protected by copyright. All rights reserved. 31 (52.54) Page 11 of 11 American Journal of Hematology Table 3: Extrapolated Population Prevalence Estimates by Age and Sex Estimated U.S. Population Prevalence Candidate HX (Compensated hemolysis) Candidate HX (anemia) Elevated MCHC N (%) Age Group (Years) Female Male Female Male 18-49 4,594 (0.0318) 7,696 (0.0888) 21,692 60,610 82,302 50-99 6,036 (0.0427) 8,782 (0.0787) 24,501 39,257 63,757 TOTAL 10,630 (0.0372) 16,478 (0.0831) 46,193 99,867 146,059 18-49 851 (0.0059) 316 (0.0036) 4,047 2,471 6,518 50-99 1,421 (0.0100) 1,441 (0.0129) 5,873 6,327 12,200 TOTAL 2,272 (0.0079) 1,757 (0.0089) 9,920 8,798 18,718 18-49 171 (0.0020) 879 (0.0124) 1,406 8,820 10,226 50-99 187 (0.0017) 772 (0.0087) 974 4,331 5,305 TOTAL 358 (0.0019) 1,651 (0.0103) 2,380 13,151 15,531 Total ______________________________________________________________________________ Total United States population estimates for the categories Elevated MCHC, "Candidate HX (anemia)", and "Candidate HX (compensated hemolysis)" were based on extrapolations from 2010 U.S. Census data (https://www.census.gov/2010census/) stratified by sex and age, Sexstratified prevalence data for the groups aged 18-49 and 50-99 were adjusted by corresponding U.S. population estimates. Parentheses indicate values in % of total population tested. 11 John Wiley & Sons This article is protected by copyright. All rights reserved.