Page 75 - 2019_10 resto del Mondo_web

Molecular diagnosis and clinical features of 117 Japanese FA patients
affect PALB2 interaction with RNF168 or BRCA2 which is mediated by the PALB2 C-terminal WD40 domain.35,36
The three FA-D1 patients (including the previous Japanese case), as well as the one FA-N patient, all devel- oped early-onset malignancies; this is in line with previous reports from Western countries (Table 4).34,37,38 Although it is important to note that the first clinical manifestation in such cases could be onset of malignancy without prior clinical problems, Cases 65 and 98 had severe physical anomalies as well. Their malformations did not fully meet VACTERL-H criteria (Table 2). Alter et al. had previously reported that FA-D1 and FA-N patients were characterized by frequent VACTERL-H association and early-onset tumors, such as Wilms tumor, or acute myeloid leukemia (AML), with a cumulative incidence of malignancy as high as 97% by the age of 5.2 years.34 Thus, Case 66 was highly unusual as a FA-D1 patient. He developed T-lymphoblas- tic lymphoma at 23 years of age, with a relatively short stature, and severe microcephaly (Online Supplementary Appendix). He received standard chemotherapy for the lymphoma, which caused prolonged pancytopenia. Then a mitomycin C-induced chromosome breakage test was performed, and he was diagnosed as FA. We list Case 66 as FA-D1, since he had biallelic, likely deleterious, BRCA2 variants but no other FA gene mutations. This case may expand the clinical spectrum of FA-D1. Alternatively, for the moment, the possibility that hidden FA gene variants caused his FA phenotype cannot be excluded.
Allele frequency of pathogenic variations in 22 Fanconi anemia genes in the Japanese population
To estimate the frequency of pathogenic FA gene varia- tions in the Japanese population, we analyzed WGS data for 22 FA genes from the 3.5KJPNv2 database. We identi- fied 66 deleterious genetic variations (nonsense, frameshifts, and splicing site mutations) in 19 FA genes (Table 1). In addition to the three common FANCA [c.2546elC (0.08%)] and FANCG mutations [c.307+1G>C (0.1%); c.1066C>T (0.04%)], carriers with FANCA c.2602- 2A>T, FANCD1 c.6952C>T, FANCG c.194delC, or FANCI c.157-2A>G mutations were detected at low percentages (0.01-0.08%), and these variants were identified as causative mutations in Japanese FA patients. Allele fre- quencies of FANCL c.170G>A (p.W57X) variants were rel- atively high (0.08%); however, no patients with these variants were identified in our FA collection.
Monoallelic mutations in some FA genes, such as BRCA1, BRCA2, BRIP1, PALB2 and RAD51C, cause adult- onset cancer predisposition39-41 and we identified 25 dele- terious variants in these genes (5 in BRCA1, 10 in BRCA2, 3 in BRIP1, 6 in PALB2, and 1 in RAD51C). BRCA1 c.188T>A (p.L63X) and BRCA2 c.6952C>T (p.R2318X) are well-known mutations in hereditary breast and ovarian cancer (HBOC) in Japan.42 The BRCA2 c.10150C>T (p.R3384X) was more prevalent than p.R2318X, but it has been classified as non-pathogenic because of its location near the 3'-end.43 The PALB2 c.2834+2T>C was recently identified in a Japanese female with bilateral breast cancer.44
From these analyses of allele frequency of FA-associated deleterious variants in 3,554 individuals, we estimated that approximately 2.6% of the Japanese could be consid- ered to be carriers of pathogenic variations in FA genes.
Discussion
In this study, we report the largest series of subtyped Japanese FA patients to date by updating our previously reported cases with an additional 13 new cases (Online Supplementary Table S2). We employed various methods, including PCR-direct sequencing and next generation sequencing. WES and targeted exome sequencing were extremely useful in identifying mutations, as reported pre- viously.8 However, approximately half of the cases were undiagnosed even after these procedures.8 When com- bined with the data generated by FANCA-MLPA, the diag- nosis rate was much enhanced, since FANCA deletion was frequent, and WES/target-seq is not necessarily effective in identifying deletions. We also noted that mutations affecting splicing, such as intronic or synonymous vari- ants, were difficult to detect by WES or targeted-seq. The former weak point was complemented by the use of aCGH, while RNA-seq was useful in detecting splicing abnormalities. We think the identification of two synony- mous mutations affecting splicing is of great significance, since this type of mutations could have been easily over- looked. Thus, our approach ultimately achieved molecular diagnosis in most of the cases, and many private and novel mutations were identified in 11 of the 22 known FA genes.
Given the present results, we suggest that a molecular work-up of Japanese FA patients should start with screen- ing for the three most-common mutations (FANCA c.2546delC, FANCG c.307+1G>C, and FANCG c.1066C>T) along with an MLPA assay for FANCA. As a next step, targeted-seq or WES analysis should be consid- ered. For the remaining unclassified cases, aCGH, WGS, and RNA-seq analysis may be useful to identify large indels or splicing defects. Through these combined and comprehensive efforts, correct genetic diagnosis may be obtained in more than 90% of the Japanese FA patients.
Aldehyde dehydrogenase 2 ( ALDH2) converts acetalde- hyde to acetate, and potentially catalyzes other aldehydes as well. In East Asian countries, including Japan, a signifi- cant fraction (approx. 50%) of the population carries ALDH2 variant ALDH2*504Lys which is encoded by the so-called A allele, and affects alcohol tolerance and some aspects of human health.45 We have previously described a subset of severe FA cases that were homozygous for the ALDH2*504Lys variant (the AA genotype), and who expe- rienced bone marrow failure and/or myelodysplastic syn- drome (MDS) immediately after birth.5,7 We also found several FA-B and FA-I cases that were accompanied by severe physical abnormalities, termed VACTERL-H. Two (Cases 18-1, 99-1) of the six previously reported FA cases carrying a homozygous ALDH2 AA genotype also dis- played these severe malformations7 (Table 2), but, interest- ingly, their siblings (Cases 18-2, 99-2) with ALDH2 GG genotype displayed relatively minor physical abnormali- ties (Online Supplementary Table S2). We note here that the FA-B or FA-I patients with VACTERL-H anomalies were carriers of the ALDH2 GG or the GA genotype. The impact of endogenous aldehyde catabolism on bone mar- row stem cells is very clear, and this effect also extends to the role of the FA proteins in preventing severe malforma- tions. It has been suggested that the extent of physical abnormalities and severity of hematologic defects tend to be correlated.46 In any event, FA-B and FA-I groups often exhibit severe malformations, as described previously47,48 and confirmed here in Japanese cases. Since many of our
haematologica | 2019; 104(10)
1971