Erythrocytosis in Chronic Mountain Sickness (CMS) in Andeans

Introduction: Chronic mountain sickness (CMS) was first described as erythrocytosis of high altitude among Andean high altitude dwellers. Later description of this condition quickly implicated increasing age and increasing hematocrit as the main drivers for many physiological changes. Now CMS is a full clinical syndrome, and the severity is based on scoring of symptoms, physical examination signs, and hemoglobin (Hb) levels. Protective alleles in EPAS1 and EGLN1 have been associated with less erythrocytosis in Tibetans. Recently, SNPs in ANP32D and SENP1 leading to increases in their expression have been reported in CMS due to excessive hypoxia response. Fetal hemoglobin (HbF), having higher oxygen affinity than adult Hb, has been tested in small number of participants. There was no difference in those with or without CMS. Thus we tested these genes and laboratory parameters in a larger Andean cohort.

Methods: Healthy 153 adults and 32 children age ≥5 years old, born and lived in Cerro de Pasco, Peru (4338 m), enrolled in the study. Native residency was defined as living, eating, and sleeping in households in the study area, with total absences <1 year at lower altitudes (<3000 m), and being > 6 months at high altitude before the study entry. Blood sample for blood counts, hemoglobin electrophoresis, plasma erythropoietin (EPO), and brain natriuretic peptide (pro BNP) were tested. SNP identification was carried out by PCR and Sanger sequencing of the 4 candidate genes (EPAS1, EGLN1, ANP32D, and SENP1). Testing for JAK2 V617F was also performed.

Results: 58 adults and no children had CMS. 52% of the participants belonged in the healthy group (CMS score ≤5), 24% in mild CMS (score 6-10), 12% moderate CMS (score 11-15), and 2.2% severe (score >15). The mean Hb level was lower in the healthy group, compared to the CMS groups (P<0.001). The entire group (n= 184) had normal levels of HbF, median value of 0.4% (range 0.2-1.9). There were no differences (P=0.9) between healthy and CMS participants, adults and children, or men and women.

In the participants who had sufficient quantity of plasma, EPO and proBNP determinations were performed. Mean plasma EPO levels correlated negatively with increasing CMS score, but none was statistically significant (P=0.24). In contrast, proBNP levels were the lowest levels in healthy group (27.5±3.1 pg/mL) and highest in the mild CMS group (66.5±14.1 pg/mL).

* P<0.001, compared to healthy group by one way ANOVA

** P=0.005, compared to healthy group by one way ANOVA

Previous studies showed that high altitude dwellers in Tibet harbor protective SNPs in the EPAS1 (or HIF 2a) and EGLN1 (or PHD2). Thus we chose 26 men with the highest Hb and 13 individuals with normal Hb, and performed SNP sequencing for these 2 genes. There was no difference between the 2 groups in the EPAS1 gene; all had CC allele. The T content in EGLN1 was 69% in CMS, and 85% in the control group (odds ratio for C vs T, 2.4, P =0.14).

Another recent report showed that other candidate genes may be associated with CMS. Sequencing of ANP32D (rs72644851) revealed that G content in 25 CMS participants was 64%, and 96% in 13 controls (odds ratio for G vs A, 14.06, P=0.002). The percent of G content in SENP1 (rs7963934) in CMS was 84%, and 100% in controls (odds ratio for G vs C, 10.6, P=0.045). Additionally, 15 of these participants with CMS had JAK2 V617F mutation analysis; all showed the wildtype allele.

Conclusion: Compared to healthy adults, CMS participants had similar and normal HbF levels and statistically insignificant lower plasma EPO levels. Pro BNP levels were higher only in the mild CMS group. We confirmed the association between SNPs in ANP32D and SENP1 and CMS in Andeans. Furthermore, the protective SNPs present in the Tibetan population, EPAS1 or EGLN1, are not present in this population of Andeans, potentially explaining the increased prevalence in Andeans at high altitude.