Genotype-Phenotype Correlation and Molecular Heterogeneity in Pyruvate Kinase Deficiency: Data from the PKD Natural History Study

Background: Pyruvate kinase (PK) deficiency is the most common enzyme defect of the glycolytic pathway causing hereditary non-spherocytic hemolytic anemia. PK deficiency is autosomal recessive, caused by both homozygous and compound heterozygous mutations in the PKLR gene with >300 mutations reported.

Aim: To identify PKLR genotypes and determine the genotype-phenotype correlation in patients with PK deficiency.

Methods: Patients (pts) enrolled on the PKD Natural History Study, a prospective international study, at 30 sites in North America and Europe. In pts without prior PKLR gene sequencing results, the PKLR gene was analyzed by Sanger sequencing. All new missense mutations affected conserved residues in multiple domains of the PKLR gene, were not detected in 1000 Genomes and HGMD databases, and were considered pathogenic by several mutation algorithms. To evaluate genotype-phenotype associations (Fisher's exact or Kruskal-Wallis test), pts were grouped according to genotype (two missense mutations (M/M), one missense/one non-missense (M/NM), or two non-missense mutations (NM/NM)); non-missense included nonsense, frameshift, inframe small indel, large deletions, and splicing mutations. Pts with three pathogenic variants or promoter mutations (n=6) were excluded from the genotype-phenotype analysis, and 2 pts were excluded for insufficient phenotype data. Pts from the Amish community (n=55, homozygous for the splicing mutation R479H) were separately analyzed. Regular transfusions were defined as ≥6 transfusions within 12 months. The Holm-Bonferroni method was used to adjust for multiple testing (significance level p<0.0055).

Results: Genotype information was available on 255 enrolled pts. Molecular characterization confirmed the wide heterogeneity of PK deficiency with 123 different mutations identified: 79 missense, 36 non-missense mutations (12 splicing, 13 frameshift, 7 stop codons, and 4 large deletions), 5 inframe indel, and 3 promoter variants. Most pts had at least one missense mutation (85%). Fifty mutations, all affecting PK structural domains, have not been reported previously (31 missense, 3 stop codon, 7 frameshift, 3 inframe indel, 5 splicing mutation, 1 promoter mutation); 22% (56/255) carried at least one new molecular variant.

The genotype-phenotype analysis was stratified as: 111 with M/M (58%), 52 with M/NM (27%), and 29 with NM/NM (15%) (n=192, Table). There is a trend for pts with more severe mutations to be diagnosed at a younger age (p=0.049). The NM/NM group had the highest rate of splenectomy, 72%, vs. 50% in the M/NM and 44% in the M/M group (p=0.024). In those splenectomized and not on regular transfusions, the hemoglobin (Hb) levels were significantly different between the groups (p=0.003). There were significant differences in the transfusion status; 86% of the NM/NM group were previously or currently regularly transfused, versus 50% of the M/NM and 42% of the M/M group (p<0.0001). There were also significant differences in the total number of transfusions (p=0.0013); 96% in the NM/NM group received at least one lifetime transfusion compared to 81% and 75% of the M/NM and M/M groups, respectively. The NM/NM group had the highest maximum ferritin (p<0.0001) and was more likely to have iron overload, defined as ferritin >1000 ng/ml or having received chelation in the year prior to enrollment (p=0.0013). Notably, there was no evidence of an association of PK enzyme activity with the genotype.

The median age of diagnosis in the homozygous R479H group (n=55) was at birth (range 0-10 y). Compared to the other genotypes, the rate of splenectomy was highest in this group, 93%, whereas the rate of cholecystectomy was the lowest, 27%. In this cohort, 67% had previously been on regular transfusions, but 18% had never been transfused. The maximum ferritin level was 615 ng/ml (range 126-3258) with 33% having iron overload. The median Hb of those splenectomized and not regularly transfused was 9.4 g/dl (range: 8-11.2), similar to the M/M group.

Conclusions: Genotype-phenotype associations were observed in a large international cohort of pts with PK deficiency. Pursuing molecular testing may be useful to discuss prognosis and to establish a monitoring plan in pts based on genotyping results. Fifty new mutations were identified, thus confirming the wide heterogeneity of the molecular genotype and diagnostic complexities in PK deficiency.

View largeDownload slide

View largeDownload slide

Close modal