Cancer-associated Mutations in Congenital Pulmonary Malformations: A Prospective Cohort

Abstract

To the Editor:Cancer-associated mutations are identified in some congenital pulmonary malformations (CPMs), including mutations of the V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) gene in particular (1, 2).They were described in cystic malformations, with or without mucinous islets (1, 2).The theoretical risk of a link between CPMs and cancer risk is fueling the debate about whether preventive surgery should be performed for asymptomatic CPM.We used the large French prospective population-based MALFPULM cohort to evaluate the actual prevalence of cancerassociated mutations in CPMs and to correlate CPM phenotype with the presence of at least one oncogenic mutation.This cohort has been described in detail elsewhere (3,4).Briefly, all pregnant women whose fetus had a prenatal diagnosis of CPM were invited to participate in the study, whatever the ultrasonic phenotype of the CPM: cystic, purely hyperechoic, or mixed.This study received institutional review board approval (Comit e de Protection des Personnes Ile-de-France IV, U.S. Department of Health and Human Services Agreement no.00003835).Follow-up of this cohort was performed up to 2 years of age and had no impact on the decision as to whether or not to perform elective surgery.When compared with children with elective surgery, nonoperated children had fewer cystic malformations but as many malformations with systemic vascularization or emphysematous images.Postnatal thoracic computed tomography description and final histological diagnosis (5) were prospectively collected in a online database.Tumor DNA was extracted from macrodissected tissue samples from the area of the malformation.The 22 gene Ion AmpliSeq Colon and Lung Cancer Research Panel v2 (Life Technologies-Thermo Fisher Scientific) was used.Mutations in DNA were detected at allele frequencies as low as 0.001 for insertions or deletions .2bp and 0.003 for single-nucleotide variants (6).For CPMs with mutation, tissue samples from apparently normal lung parenchyma, collected at some distance from the malformation sample, were also tested, when available.A blind histological review of slides from CPMs with mutation, together with randomly selected slides from 15 tissues without mutation, was performed using a standardized grid that was created by the review panel of four pathologists from three centers.Univariate analysis compared the following parameters between malformations with mutations and those without mutations: pre-and postnatal imaging appearance of the malformation; maximum value during pregnancy of the ratio of CPM volume to head circumference (CVRmax); severe fetal compression; and neonatal respiratory distress (3, 4).x 2 tests, Fisher's exact tests, t tests, and univariate logistic regression were used.Of the 426 live-born infants in the MALFPULM cohort, 285 underwent surgery before the age of 2 years, and 195 of them were genotyped.The main reason for the absence of genomic analysis was the omission of the additional consent collection.The characteristics of CPMs sent for analysis did not differ from the ones not sent.A KRAS mutation was identified in 17 cases and an fibroblast growth factor receptor 2 (FGFR2) mutation in one case (Table 1).Seven CPMs with mutation had a concomitant analysis of histologically healthy tissue.In four of them, the same mutation was identified as in the area of the malformation, with similar or lower allelic ratios.CPMs with mutation demonstrated higher prenatal CVRmax values (odds ratio, 1.08; 95% confidence interval, 1.02-1.14for every 0.1 cm 2 increase in CVRmax; P = 0.007), higher rate of prenatal compressive complications (P = 0.037), higher rate of neonatal respiratory distress (P = 0.024), higher rate of surgery in the first week of life (P , 0.001), and higher rate of surgery because of symptoms (P , 0.001) (Table 2).All CPMs with mutation were cystic in appearance on postnatal imaging (P = 0.001), with no difference in postnatal mean diameter of the largest cyst between CPMs with or without mutation.In 3 of the 18 CPMs with mutation, the cysts were not detected on prenatal imaging.The histological diagnosis was congenital pulmonary airway malformation (CPAM) for all CPMs with mutations (P = 0.01).Of the 66 CPMs with postnatal systemic vascularization, including 17 who were also cystic, none had a mutation (P = 0.002).Microscopically, a mucinous component was identified only in tissues with mutation (P , 0.005).However, 56% of the mutated tissues had no identifiable mucinous component on the slides reviewed.Inflammation was noted in more than half the tissues, with no association with the presence of mutations.Microscopy findings were consistent with the diagnosis of type 2 CPAM in five cases with mutations.In our prospective cohort, we demonstrate that mutations of cancer-associated genes are identified only in CPMs with a cystic component on postnatal imaging, with a prevalence of 15% within these cystic CPMs.All cystic CPMs are at risk of mutation, regardless of cyst size, type of CPAM, or identifiable mucinous epithelial cells by microscopy.This is concordant with the previous description of KRAS mutations away from mucinous islets of type 1 CPAMs or in type 2 CPAMs (1).Among the 22 genes tested on our panel, only KRAS and FGFR2 were identified with mutations.The FGFR2 mutation in one case was the first to be described in a CPM.As onethird of tissues with mutation were obtained from neonates, the results suggest that these mutations are prenatally acquired,

Topics & Concepts