Genotype-phenotype associations in children with a positive cystic fibrosis newborn screening

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 1

Genotype-Phenotype Associations in Children with a Positive Cystic Fibrosis
Newborn Screening

Danieli Barino Salinas

Pediatric Pulmonology Department, Children’s Hospital Los Angeles

Keck School of Medicine

University of Southern California

Author Note: This paper was created for the appointment of Master in Science in the program of
Clinical and Biomedical Investigations.
Program director: Stanley Azen, Ph.D.
Date of submission June 24
th
, 2015

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 2
Table of Contents
1. Table of contents…………………………………………….…………………………..2
2. Abstract………….………………………………………………….……………….......3
3. Introduction………………………………………………………………………….…..4
4. Methods…...…………………………………………………………………………..5-8
5. Results…….……………………………………………….…………………………9-11
6. Discussion…………...……………………………………………………………...12-16
7. Acknowledgements…….………………………………………………………..……..16
8. References.……………………………………………………………….…………17-20
9. Tables and Figures……………………………………………………….…………21-27

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 3
Abstract
Cystic Fibrosis (CF) is an autosomal recessive disorder caused by a defective trans-epithelial
channel called CF Transmembrane Conductance Regulator (CFTR). Of the nearly 2000 CFTR
variants described only 201 have been formally assigned to three categories of disease liability:
CF-causing, varying clinical consequence, and non-CF causing. California’s racially diverse
population, along with exome sequencing as part of a newborn screening model, provides the
opportunity to examine the phenotypes of young children with uncategorized variants to help
inform disease liability and penetrance. Methods: Retrospective cohort study of children
screened from 2007 to 2011 and followed for two to six years. Newborns that screened positive
were divided into three genotype groups: those with two CF-causing mutations (CF-C); those
with one varying clinical consequence (VCC); and those with one variant of unknown disease
liability (Unknown). Clinical characteristics used as diagnosis criteria were compared including
sweat chloride tests, pancreatic sufficiency status, and Pseudomonas aeruginosa colonization.
Individual genotype subgroups were also analyzed. Results: Compared to children with two CF-
causing variants (and therefore meeting genetic diagnosis criteria) who presented with a classical
CF phenotype (n=234), 5% of VCC (4/78) and 11% of Unknown (27/244) met diagnostic criteria
for CF. Children carrying unknown variants 1525-42G>A, L320V, L967S, and R170H had a
particularly benign clinical presentation, suggesting these are non-CF causing. Conclusions:
These findings can help clinicians, researchers, and parents understand the risk of CF and CF-
related symptoms for children with variants of uncertain clinical significance and impact clinical
care protocols.
Keywords: cystic fibrosis newborn screening, CFTR mutations, genotype-phenotype
associations.
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 4
Introduction
Cystic Fibrosis (CF) is a lethal disorder caused by a defective trans-epithelial channel
called Cystic Fibrosis Transmembrane Conductance Regulator (CFTR). The defective CFTR
causes mucus secretions to become thick and viscous, obstructing the lungs and pancreas and
consequently causing life-threatening lung infections and malnutrition. Newborn Screening
(NBS) for cystic fibrosis (CF) is performed throughout the U.S. and other countries based on
known long-term benefits from early nutritional interventions (Grosse, 2004; Farrell, 1997).
Although different algorithms are used, most CF NBS programs involve DNA analysis
(Wagener, 2012). The challenge of diagnosis confirmation after a positive newborn screen result
is that there are nearly 2000 distinct variants in the CF Transmembrane Conductance Regulator
(CFTR) gene

(http://www.genet.sickkids.on.ca/cftr/Home.html), and most of them are rare and of
unknown clinical significance (Rowe, 2005; Watson, 2004). The Clinical and Functional
TRanslation of CFTR (CFTR2) project (www.CFTR2.org) has classified disease liability based
on the high prevalence of variants among CF patients. CFTR2 has so far classified 201
mutations (178 as CF causing, 12 as Non CF-causing, 11 as Varying Clinical Consequence)
(Sosnay, 2013). However, there remain over 1700 mutations of low frequency with uncertain
disease liability.
California started CF NBS in 2007. In order to address the challenges of screening in an
ethnically diverse state and also to minimize the number of false positives and children needing
further diagnostics such as sweat testing, a 3-step model was adopted that includes CFTR
sequencing. Sequencing opened the opportunity to explore the full scope of the CF spectrum by
understanding the penetrance of CFTR variants of uncertain clinical significance.
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 5
We analyzed data from the California CF NBS database of all children who were screen
positive born in the first 4 years of the program, comparing outcomes for 3 distinct genotype
groups based on CFTR2 classifications: those with two CF-causing variants (CF-C), one CF-
causing and one variant of varying clinical consequence (VCC), and one CF-causing and one
variant of unknown disease liability (Unknown).

Methods
Study Design and Population
We conducted a retrospective cohort study of all newborns with a positive California CF
NBS from July 2007 to July 2011, with two to six years of follow up (data collected until August
2013).
The California CF screening model starts with an elevated immunoreactive trypsinogen
(fixed cut off at IRT ≥ 62 ng/ml, 98.4
th
percentile) followed by a 40 CFTR-mutation panel
previously identified as prevalent among CF patients in California (Table 1). All but three of the
variants on the panel (F311del, 935delA and Q98R) have been evaluated by CFTR2 and
determined to be CF-causing. Newborns with only one panel mutation had DNA from the
original filter paper blood spot used for CFTR sequencing, which was done by direct CFTR DNA
Sanger sequencing at Stanford Molecular Pathology Laboratory. Sequencing was performed by a
semi-automated process adapted from protocols previously described (Strom, 2003). The assay
covered 983 bases of 5’ untranslated region, 27 exons, 20 bases into the 5’ and 3’ ends of all
introns, the CFTR poly T status and TG tract, intron 19 surrounding the 3849+10kbC>T variant,
and intron 11 surrounding the 1811+1634A>G variant. Analysis for known and novel variants in
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 6
these areas of the gene was performed first by unidirectional sequencing, confirming all
suspicious and positive variants with sequencing in the opposite direction. Children with two or
more variants identified were considered screen positive and referred to a pediatric CF care
center for follow up. Children identified with only one CFTR variant were considered carriers
and parents were offered telephone genetic counseling.
Seventeen CF centers in California received referrals from the NBS program and
followed children with CF and those with so-called CFTR-related metabolic syndrome (CRMS)
(Borowitz, 2008). CRMS are individuals with two CFTR variants identified by screening, but no
clinical criteria for CF diagnosis, in other words they are asymptomatic, well nourished, and
have sweat test < 60 mmol/L. Clinical and laboratory data were entered into the state’s web-
based screening information system in the initial assessment visits (quarters 1 to 3) until the final
diagnosis of CF or CRMS was established and yearly thereafter. After personal identifiers were
removed the data on pre-selected variables was sent to the Children’s Hospital Los Angeles
(CHLA) research team for analysis. The CHLA-Institutional Review Board and the California
Health and Human Services Agency Committee for the Protection of Human Subjects approved
the study and waived informed consent for participating families.
Genotype Groups
Three genotype groups including all screen-positive children were created. The CF-causing
group (CF-C) consisted of individuals with two CF-causing mutations (either from the California
40 mutation panel or from CFTR sequencing classified as CF-causing by CFTR2). The varying
clinical consequence group (VCC) consisted of those with one CF-causing mutation and a
variant identified by sequencing that is characterized as having varying clinical consequences by
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 7
CFTR2. Lastly, the unknown disease liability group (Unknown) consisted of those with one CF-
causing mutation and a variant identified by sequencing that is characterized as unknown or has
not yet been studied by CFTR2.
Outcome Measures
In addition to genetic data, outcome variables included standard tests and parameters
used to confirm the diagnosis of CF, and to evaluate disease severity such as: sweat chloride
concentration, pancreatic sufficiency status, growth parameters, rate of first acquisition of
Pseudomonas aeruginosa in the first year of life, and persistent colonization with P. aeruginosa.
Sweat chloride tests to measure CFTR’s activity in the sweat gland has long been used to make a
clinical diagnosis (Davis, 1984). Sweat tests were performed according to a standardized
protocol and results categorized as positive, intermediate, and negative (Farrell, 2008). Those
with initial negative or intermediate results had repeat testing at ~ 6, 12, and 24 months, and the
highest value used for analysis. Pancreatic status was assigned based on the most recent fecal
elastase (FE) value to pancreatic insufficient (PI: <200mcg/g) or pancreatic sufficient (PS: ≥200
mcg/g) (Farrell, 2008; O’Sullivan, 2013). In 20% of subjects missing FE values, pancreatic
status was assigned based on use of pancreatic replacement enzyme therapy. Growth analysis
was examined in subjects with at least three weight and height measurements. As P. aeruginosa
is a common and disease severity-defining pathogen in CF (Emerson, 2002), all subjects with at
least one positive respiratory culture for this organism were reviewed separately with their
respective centers to determine persistent colonization status. Cultures were obtained during
clinic visits by deep-throat swabs as recommended by CF and CRMS practice guidelines
(Borowitz, 2009a; Borowitz, 2009b). Persistent colonization was determined according to a
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 8
modified LEEDS criteria (Lee, 2003; Pressler, 2011) as follows: a) mucoid-type, ever; and b) at
least two positive cultures within twelve months. Other categories included: P. aeruginosa-yes-
ever for subjects with positive cultures who did not meet the criteria for persistent; negative for
those who never had a positive culture; and unknown for others with missing data (fewer than
two collected cultures per year and/or fewer than two years of follow up). “Lost to follow up”
was defined as no medical visit for ≥ eighteen months.
In order to further characterize outcomes from variants found in subjects in VCC and
Unknown groups, we subdivided them by genotype and selected only those with at least 6 or
more subjects for further analysis.
Statistical analysis
Comparisons across the three genotype groups and additional subgroups were made with
ANOVA or Wilcoxon rank sum tests for continuous variables and Chi-square or Fisher’s exact
test for categorical variables. Longitudinal weight-for-height Z-scores (based on CDC growth
curves from year 2000) during the first two years of life were analyzed using quadratic repeated
measures mixed models, with CF-C further divided into PI and PS. Outlier growth data points
were removed if not biologically plausible. Dunnett’s post hoc comparisons of least squares
means were made with CF-C-PS children as the reference group. Time-to first positive
respiratory culture for P. aeruginosa during the first year of life was analyzed with the Kaplan-
Meier method and compared across groups with the log-rank test. Comparisons between groups
were also made adjusting for ethnicity (Hispanic vs. all others) with Cochran- Mantel–Haenszel
test for categorical variables, 2-way ANOVA for continuous variables, and Cox regression for
time to first P. aeruginosa culture. Statistical analysis was performed with SAS/STAT® v9.2
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 9
software. Statistical tests were 2-sided with statistical significance set at p<0.05.
Results
Among 2,172,002 births during the 4-year study period, 32,818 had further testing due to
IRT levels above the fixed cutoff of ≥ 62 ng/ml. Of those, 174 who had two CFTR variants
identified from the California 40 panel and 1,855 with detection of only one panel mutation
underwent CFTR sequencing (Figure 1). Of 1,855 individuals sequenced, a second mutation was
found in 674, while the 1,181 with no variants found were labeled as screen negative carriers and
offered genetic counseling. Among the 674 sequence-positives, 292 total were excluded from
this analysis: 27 had mutations in cis (= mutations on the same side of the chromosome), 57 had
known non-CF-causing mutations; and 208 had intron 9 variants (poly-T and TG track) known to
affect exon 10 splicing that will be reported separately.
There were 234 subjects in the CF-C group, in which 60 different variants were
represented. In addition to the 40 from the panel, there were 20 established CF-causing variants
detected on sequencing (seen in 60/234 subjects). The VCC group contained 9 sequenced
variants found in 78 subjects, and the Unknown group contained 158 variants found in 244
subjects.
There was a lower median IRT, higher mean birth weight, and no cases of meconium
ileus among VCC and Unknown subjects compared to the CF-C group (Table 2). Individuals in
the VCC and Unknown groups were more likely to be female. There were 6 deaths in the CF-C
group. Fifty-four percent of those in the Unknown and 37% in the VCC groups were lost to
follow up, compared to 18% in the CF-C group (p<0.0001; Table 2).
Initial sweat tests were performed at a median age of 54 days (interquartile range (IQR):
41-75) days. The overall quantity not sufficient (QNS) rate was 11%. Testing was repeated at
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 10
least once in 70% of those with an initial result <60mmol/L or QNS. One subject from the VCC
(genotype: F508del and R117H/7T-9T, highest sweat chloride=63 mmol/L) and 27 from the
Unknown groups met sweat chloride diagnostic criteria for CF (≥60 mmol/L), while 39 and 33
subjects, in the respective groups, had values in the borderline range (30 to <60 mmol/L) (Table
2). The genotypes among those with diagnostic sweat chloride levels and unknown liability
mutations are shown in Table 3. The sweat chloride conversion rate from <60 mmol/L
(indeterminate) to ≥60 mmol/L (diagnostic for CF) happened in 7 subjects from the CF-C group,
2 from the VCC (genotypes: F508del with R117H/7T-9T [n=1] and F508del with 1874insT and
R117H/5T [n=1]); and 4 from the Unknown group (genotypes: F508del with L32M, S1159P,
T1076P [n=3], and 663delT with I105N [n=1]). The mean (±SD) conversion age was 16±11
months.
Pancreatic insufficiency was observed in 4% of VCC and 7.7% of Unknown groups,
compared to 79% of CF-C (p<0.0001; Table 2). VCC and PI occurred in individuals carrying
F508del and the following mutations: R1070W and 1525-42G>A (n=1), D1152H (n=1), and
R117H/7T-9T (n=1). Genotypes of those with PI and mutations of unknown liability are
described in Table 3. Conversion from PS to PI was seen in 4 subjects: one from the CF-C group
(homozygous 663delT), 2 from the VCC group (Genotypes: F508del and D1152H [n=1]; and
F508del and R117H/7T-9T [n=1]) and one from the Unknown group (Genotype S549N and
1949del84), setting the conversion rate at 1.3% considering 315 total subjects initially labeled
PS. The mean ±sd age of conversion was 15±11 months. We also observed 4 subjects whose
initial FE values were consistent with PI (between 50 and <200 mcg/g), who later converted to
PS (>200 mcg/g).
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 11
Repeated measures analyses of growth data through 24 months for 492 subjects grouped
by genotype and PS/PI status found that PS subjects (CF-C, VCC, and Unknown) had
significantly better growth compared to CF-C-PI subjects based on Weight for Height Z scores
(WHZ) (p values respectively: 0.0247, 0.0013, and <0.0001; Figure 2).
Though subjects from all groups cultured P. aeruginosa, individuals in CF-C did so more
frequently and earlier. In the first 12 months of life P. aeruginosa-free probability was lower in
CF-C subjects (p<0.0088). The acquisition rate was 18% for CF-C, 11% in VCC, and 8% in
Unknown groups. For a large proportion of subjects in VCC and Unknown (52% and 61%
respectively) P. aeruginosa persistent colonization status was not able to be determined, largely
due to lost to follow up. No statistical testing was done with this variable considering the large
number of missing data. Still, among those with sufficient data, 9% in VCC and 6% in Unknown
groups met criteria for persistent colonization (Table 2).
The mutations found in at least six subjects in VCC and Unknown groups allow better
insight into possible genotype-phenotype correlations (Table 4). Sweat chloride levels were
borderline or beneath the diagnostic threshold in the VCC and Unknown groups except for one
subject (with R117H/7T-9T) who reached diagnostic level at 63 mmol/L. A high percentage of
subjects with D1152H and R117H/7T-9T (70% and 64% respectively) had sweat chloride values
in the intermediate range of 30 and 59 mmol/L suggesting some level of CFTR defect. One
subject with D1152H and another with R117H/7T-9T had PI, while 5/24 (21%) in VCC and 1/17
(6%) in the Unknown groups had persistent P. aeruginosa colonization. Among the frequently
seen unknown liability variants, 1525-42G>A, L320V, L967S, and 296+28A>G, when not part
of a complex allele and in combination with a CF-C mutation, had sweat chloride results ≤ 33
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 12
mmol/L, no cases of pancreatic insufficiency, and only one isolated case of persistent P.
aeruginosa.

Discussion
The California CF-NBS protocol uses DNA sequencing to search for a second variant in
individuals with elevated IRT and one CF-panel mutation. Our study is unique in that it
describes the genotype and phenotype variability in a large and ethnically diverse population
identified by CF newborn screening that applies exon sequencing. We showed that children with
a CF-C genotype had classical disease phenotype (88% had sweat chloride ≥60 mmol/L, 80%
were PI, 18% acquired P. aeruginosa in the first year, and 22% met criteria for persistent
colonization), while those with VCC and Unknown liability mutations had greater phenotype
variability (respectively, 1.3% and 12% had sweat chloride ≥60 mmol/L, 4% and 8% were PI,
11% and 8% acquired P. aeruginosa in the first year, and 9% and 6% met criteria for persistent
colonization). Five percent of VCC and 11% of Unknown group subjects met diagnostic criteria
of CF either by sweat chloride ≥60 mmol/L and/or pancreatic insufficiency status.
Out of a total of 556 subjects studied, there were 290 subjects (52%) with CFTR variants
identified (located on opposite chromosomes to the panel CF-causing mutation), who did not
meet definitive CF diagnostic criteria at this time. A portion of these patients may never develop
disease; while others will go on to develop CF or single organ system problems such as CFTR-
related disorder or isolated infertility due to Congenital Bilateral Absence of the Vas Deferens
(CBAVD) (Bombieri, 2011). However, it is currently impossible to predict the overall
percentage or which individual patient will develop disease in the future.
CF diagnosis in individuals with a negative sweat test. Perhaps most at risk in the
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 13
uncertain categories are those with borderline sweat chloride values (51% of VCC and 14% of
Unknown). Nearly 14% of individuals diagnosed with CF as adults in the US Cystic Fibrosis
Foundation patient registry had sweat values <60 mmol/L, with the CF diagnosis established by
genotype or clearly recognized phenotype (Farrell, 2008) These individuals are most worthy of
longer follow-up, when pulmonary function testing and other objective measures of disease can
be used to establish a CF diagnosis.
Pancreatic status predicted growth pattern. Pancreatic status was an independent
predictor for growth. Considering that PI is strongly associated with genotyping, we can state
that genotype predicted growth patterns, mediated by pancreatic insufficiency status. PI subjects
had a remarkable growth recovery in the first 12 months. As suggested by Farrell et al, this is
one measurement of success of a CF-NBS program, as it reflects early and proper nutritional
management for those at high risk of malnutrition (Farrell, 2001).
Fecal elastase should be re-measured beyond the first visit. As previously reported by
O’Sullivan (2013), there was fluctuation in the levels of FE. Conversion to PI and to PS
occurred beyond the first 3 months of life in children from our study (n=4). Therefore, if the
initial FE value is equivocal, repeat FE at 12 months should be considered.
P. aeruginosa colonization in subjects with variants of varying or unknown
significance was higher than what is known for healthy controls and lower than classical
CF. The presence of P. aeruginosa is considered pathognomonic for CF and represents an
important contributor to lung function decline (Emerson, 2002; Rosenfeld, 2001). The reported
incidence of first acquisition in the first year of life in CF patients is approximately 25% (Hayes,
2010), compared to 1-3% in healthy controls (Rosenfeld, 2012; Carlson, 2009). In our study,
CFTR dysfunction in VCC and Unknown subjects could explain the P. aeruginosa colonization.
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 14
Alternatively, many of these children may have transient colonization similar to normal controls.
However, there is also an often-discussed concern of increased exposure due to frequent
surveillance and frequent contact with CF patients, which may increase their risk of cross
contamination and even influence the final diagnosis of CF (Hayes, 2010; Massie, 2014).
Classical CF as well as benign phenotypes are seen with VCC and Unknown
genotypes. Though the majority of subjects identified in VCC and Unknown groups had a
benign presentation, the full spectrum of phenotype associated with CFTR allelic heterogeneity
was present. Twenty-nine subjects (of 322 in the VCC and Unknown groups combined) met
either sweat chloride or PI criteria for CF diagnosis after screening. From what it was observed
in the genotype variability among CF individuals identified by sequenced (Table 2), there is a
clear need to understand complex alleles, as several patients in this group had a total of 3 CFTR
variants identified through screening. This is not so problematic for this particular group of
patients as they have a clear diagnosis of CF through their phenotype, which lead to the
categorization of these mutations as likely pathogenic. However, individuals carrying mutations
of unknown significance and negative phenotype in the early years will not know their risk of
developing disease unless phasing is done and their families are willing to continue to follow and
monitor for early symptoms.
Consistent with previously published reports, patients with one CF-causing and variants
known to be associated with variable penetrance for CF and CFTR-related disorders such as
D1152H and R117H were identified frequently (Cuppens, 1998; Burgel, 2010; Pont-Kingdon,
2004). A study of CF patients with one copy of D1152H and another CF-causing mutation has
shown low incidence of symptoms in childhood, but as many as 70% of adults have
bronchiectasis (Burgel, 2010). Though this may be biased towards severe cases referred to CF
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 15
centers, it illustrates the potential these variants have to produce a pathologic phenotype. The
children identified as part of this study may better represent the spectrum of penetrance of these
genotypes, as the sweat chloride values are lower than published reports of CF patient registry
data (Sosnay, 2013). Alternatively, this study is limited to evaluate the full penetrance of some of
these VCC and Unknown variants as it only analyses the first few years of life.
The relative short-term follow up is a limitation of our study. When our findings are
compared to what is described in CFTR2, older patients with VCC variants described there have
a more severe profile. Specifically, in 194 patients with D1152H and a deleterious variant,
average age 30 years, 32% were PI, 42% were colonized with P. aeruginosa, and the mean sweat
test was 46 mmol/L. We had 23 subjects with this genotype, 5% were PI, and 21% were
colonized with P. aeruginosa, and the mean sweat chloride was 32 mmol/L. This is not an
accurate comparison as the CFTR2 study subjects were pre-selected as having CF and DNA
sequencing was not required, which means that another hidden variant may be the actual cause of
the more severe phenotype. It may be that our newborn screen children will eventually reach a
higher value of sweat values; some may convert to PI, and/or become more colonized with
pathogenic bacteria, as they get older. The advantage of the current study and this model of
screening is the knowledge of the numerator and denominator to determine the penetrance of
these VCC and Unknown variants. As we follow these positive screen children over time, we
also have the opportunity to study the gene-environment interactions that may influence the
progression to diagnostic thresholds.
Our study has also brought attention to variants that have low penetrance for CF. Variant
1525-42G>A, a single change of G>A, 42 nucleotides into intron 9 (Schrijver, 2005), so far
rarely described in the literature, was present in 19 children in this cohort. All had a negative
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 16
sweat chloride and were pancreatic sufficient. Its frequency of 0.4% in the general population
and 2.2% in Hispanics (www.1000genome.org) explains the high prevalence in our population,
suggesting this is a non-CF causing variant or benign polymorphism. Other frequently identified
variants with also benign presentations included the missense mutations L320V, L967S, and
R170H. These are rare variants described in case reports of individuals with negative sweat
chloride and CBAVD or pancreatitis, but are not known to cause CF (Pelletier, 2010; Claustres,
1993; LaRusch, 2014).
This study helps us understand the penetrance of CFTR variants and the full spectrum of
CFTR- related disease that may begin in early childhood. These data can help clinicians,
researchers, public health experts, and parents understand the risk of CF for children with
variants of uncertain clinical significance and impact clinical care protocols. More reliable in
vivo CFTR function assays or other discriminating biomarkers are needed to address the issue of
uncertain prognosis in CRMS subjects. Long-term follow up of this cohort are required to better
understand how CFTR variants contribute to disease beyond our current understanding of
classical CF.

Acknowledgements
Author Contributions: Dr. Danieli Salinas and Colleen Azen had full access to all the
data in the study and take responsibility for the integrity of the data and the accuracy of the data
analysis. Study concept and design: Salinas, Chin, Azen, Keens, Sosnay, and Kharrazi. Data
acquisition and analysis: Salinas, Chin, Azen, Young, and Kharrazi. Interpretation: Salinas,
Keens, Azen, Young, Raraigh, Sosnay, and Kharrazi. Drafting of the manuscript: Salinas,
Sosnay, Raraigh, and Kharrazi. Statistical analysis: Azen.
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 17
References
Bombieri, C. et al. Recommendations for the classification of diseases as CFTR-related
disorders. Journal of cystic fibrosis: official journal of the European Cystic Fibrosis
Society 10 Suppl 2, S86-102 (2011).
Borowitz, D. et al. Cystic Fibrosis Foundation evidence-based guidelines for management of
infants with cystic fibrosis. The Journal of pediatrics 155, S73-93 (2009).
Borowitz, D. et al. Cystic Fibrosis Foundation practice guidelines for the management of infants
with cystic fibrosis transmembrane conductance regulator-related metabolic syndrome
during the first two years of life and beyond. The Journal of pediatrics 155, S106-16
(2009).
Burgel, P.R. et al. Non-classic cystic fibrosis associated with D1152H CFTR mutation. Clinical
genetics 77, 355-64 (2010).
Carlson, D. et al. Oropharyngeal flora in healthy infants: observations and implications for cystic
fibrosis care. Pediatric pulmonology 44, 497-502 (2009).
Claustres, M. et al. Analysis of the 27 exons and flanking regions of the cystic fibrosis gene: 40
different mutations account for 91.2% of the mutant alleles in southern France. Human
molecular genetics 2, 1209-13 (1993).
Cuppens, H. et al. Polyvariant mutant cystic fibrosis transmembrane conductance regulator
genes. The polymorphic (Tg)m locus explains the partial penetrance of the T5
polymorphism as a disease mutation. The Journal of clinical investigation 101, 487-96
(1998).
Davis, P.B. & di Sant'Agnese, P.A. Diagnosis and treatment of cystic fibrosis. An update.
Chest 85, 802-9 (1984).
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 18
Emerson, J., Rosenfeld, M., McNamara, S., Ramsey, B. & Gibson, R.L. Pseudomonas
aeruginosa and other predictors of mortality and morbidity in young children with cystic
fibrosis. Pediatric pulmonology 34, 91-100 (2002).
Farrell, P.M. et al. Nutritional benefits of neonatal screening for cystic fibrosis. Wisconsin
Cystic Fibrosis Neonatal Screening Study Group. The New England journal of medicine
337, 963-9 (1997).
Farrell, P.M. et al. Early diagnosis of cystic fibrosis through neonatal screening prevents severe
malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal
Screening Study Group. Pediatrics 107, 1-13 (2001).
Farrell, P.M. et al. Guidelines for diagnosis of cystic fibrosis in newborns through older adults:
Cystic Fibrosis Foundation consensus report. The Journal of pediatrics 153, S4-S14
(2008).
Grosse, S.D. et al. Newborn screening for cystic fibrosis: evaluation of benefits and risks and
recommendations for state newborn screening programs. MMWR. Recommendations
and reports : Morbidity and mortality weekly report. Recommendations and reports /
Centers for Disease Control 53, 1-36 (2004).
Hayes, D., Jr. et al. Pseudomonas aeruginosa in children with cystic fibrosis diagnosed through
newborn screening: assessment of clinic exposures and microbial genotypes. Pediatric
pulmonology 45, 708-16 (2010).
LaRusch, J. et al. Mechanisms of CFTR functional variants that impair regulated bicarbonate
permeation and increase risk for pancreatitis but not for cystic fibrosis. PLoS genetics 10,
e1004376 (2014).
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 19
Lee, T.W., Brownlee, K.G., Conway, S.P., Denton, M. & Littlewood, J.M. Evaluation of a new
definition for chronic Pseudomonas aeruginosa infection in cystic fibrosis patients.
Journal of cystic fibrosis : official journal of the European Cystic Fibrosis Society 2, 29-
34 (2003).
Massie, J. & Gillam, L. Uncertain diagnosis after newborn screening for cystic fibrosis: An
ethics-based approach to a clinical dilemma. Pediatric pulmonology 49, 1-7 (2014).
O'Sullivan, B.P. et al. Evolution of pancreatic function during the first year in infants with
cystic fibrosis. The Journal of pediatrics 162, 808-812 e1 (2013).
Pelletier, A.L. et al. CFTR gene mutation in patients with apparently idiopathic pancreatitis:
lack of phenotype-genotype correlation. Pancreatology : official journal of the
International Association of Pancreatology 10, 158-64 (2010).
Pont-Kingdon, G., Jama, M., Miller, C., Millson, A. & Lyon, E. Long-range (17.7 kb) allele-
specific polymerase chain reaction method for direct haplotyping of R117H and IVS-8
mutations of the cystic fibrosis transmembrane regulator gene. The Journal of molecular
diagnostics : JMD 6, 264-70 (2004).
Pressler, T. et al. Chronic Pseudomonas aeruginosa infection definition: EuroCareCF Working
Group report. Journal of cystic fibrosis : official journal of the European Cystic Fibrosis
Society 10 Suppl 2, S75-8 (2011).
Rosenfeld, M. et al. Early pulmonary infection, inflammation, and clinical outcomes in infants
with cystic fibrosis. Pediatric pulmonology 32, 356-66 (2001).
Rosenfeld, M. et al. Prevalence of cystic fibrosis pathogens in the oropharynx of healthy
children and implications for cystic fibrosis care. Journal of cystic fibrosis : official
journal of the European Cystic Fibrosis Society 11, 456-7 (2012).
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 20
Rowe, S.M., Miller, S. & Sorscher, E.J. Cystic fibrosis. The New England journal of
medicine 352, 1992-2001 (2005).
Schrijver, I. et al. Diagnostic testing by CFTR gene mutation analysis in a large group of
Hispanics: novel mutations and assessment of a population-specific mutation spectrum.
The Journal of molecular diagnostics : JMD 7, 289-99 (2005).
Sosnay, P.R. et al. Defining the disease liability of variants in the cystic fibrosis transmembrane
conductance regulator gene. Nature genetics 45, 1160-7 (2013).
Strom, C.M. et al. Extensive sequencing of the cystic fibrosis transmembrane regulator gene:
assay validation and unexpected benefits of developing a comprehensive test. Genetics
in medicine : official journal of the American College of Medical Genetics 5, 9-14
(2003).
Wagener, J.S., Zemanick, E.T. & Sontag, M.K. Newborn screening for cystic fibrosis. Current
opinion in pediatrics 24, 329-35 (2012).
Watson, M.S. et al. Cystic fibrosis population carrier screening: 2004 revision of American
College of Medical Genetics mutation panel. Genetics in medicine : official journal of
the American College of Medical Genetics 6, 387-91 (2004).

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 21
Tables and Figures
Table 1: CFTR variant-panel used by the California Newborn Screening Program.

Date added Mutations added or deleted – cDNA name (legacy name) No.
July 16, 2007
c.164+2T>A (296+2T>A)
28
c.254G>A (G85E)
c.274-1G>A (406-1G>A)
c.489+1G>T (621+1G>T)
c.579+1G>T (711+1G>T)
c.595C>T (H199Y)
c.933_935delCTT (delF311)
c.1000C>T (R334W)
c.1519_1521delATC (delI507)
c.1521_1523delCTT (delF508)
c.1585-1G>A (1717-1G>A)
c.1624G>T (G542X)
c.1646G>A (S549N)
c.1652G>A (G551D)
c.1657C>T (R553X)
c.1675G>A (A559T)
c.1680-1G>A (1812-1G>A)
c.1973-1985del13insAGAAA (2105-2117del13insAGAAA)
c.2175_2176insA (2307insA)
c.2988+1G>A (3120+1G>A)
c.3196C>T (R1066C)
c.3266G>A (W1089X)
c.3485G>T (R1162X)
c.3611G>A (W1204X [3743G>A])
c.3717+12191C>T (3849+10kbC>T)
c.3744delA (3876delA)
c.3846G>A (W1282X)
c.3909C>G (N1303K)
October 4, 2007 c.1153_1154insAT (1288insTA) 29
December 12, 2007
c.1475C>T (S492F)
38
c.1923_1931del9insA (2055del9>A)
c.223C>T (R75X)
c.293A>G (Q98R)
c.3140-26A>G (3272-26A>G)
c.531delT (663delT)
c.54-5940_273+10250del21kb (CFTRdele2,3(21kb))
c.613C>T (P205S)
c.803delA (935delA)
August 12, 2008
c.3659delC (3791delC)
40
c.3612G>A (W1204X [3744G>A])
c.988G>T (G330X)
Removed c.164+2T>A (296+2T>A) on August 12, 2008

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 22
Table 2: Description of the Study Population

Population Characteristics CF-C
a
(n=234) VCC
b
(n=78)
Unknown
c
(n=244)
p value
d

Age at last follow up,
mean±SD, months

49±14

48 ±13

49±13

0.91
Female gender, n (%) 99 (42.3) 45 (57.7) 144 (59.0) 0.0007
Race/ethnicity, n (%)
Whites
Hispanics
Non-Hispanics Blacks
Multiple and others
Missing

113 (48.3)
93 (39.7)
12 (5.1)
15 (6.4)
1(0.4)

29 (37.2)
40 (51.3)
2 (2.6)
7 (9.0)
0(0)

87 (35.7)
110 (45.1)
17 (7.0)
30 (12.3)
0 (0)

0.17
e

Birth weight, mean±SD, kg 3.11±0.61 3.34±0.54 3.30±0.58 0.0005
IRT, median (IQR)
f
, ng/mL 162 (109,235) 89 (74,116) 80 (69,106) <0.0001
Meconium Ileus, n (%) 36 (15.4) 0 (0) 0 (0) <0.0001
Maximum Sweat Chloride,
Median (IQR), mmol/L

Sweat Chloride distribution,
n (%), mmol/L
<30
30- 59
≥ 60
n=215
94 (83,103)

4 (1.9)
21 (9.8)
190 (88.4)
n=77
30 (20, 40)

37 (48.1)
39 (50.7)
1(1.3)
n=228
18.5 (12, 30)

168 (73.7)
33 (14.5)
27 (11.8)
<0.0001

<0.0001

Pancreatic Insufficiency, n (%) 181/228 (79.4) 3/75 (4.0) 16/208 (7.7) <0.0001
P. aeruginosa during 1
st
year,
Probability (S.E.)

P. aeruginosa Colonization, n (%)
Unknown
Never
Yes-ever
Persistent
n=231
0.18 (0.03)

40 (17.1)
90 (38.5)
53 (22.6)
51 (21.8)
n=78
0.11 (0.04)

41 (52.5)
23 (29.5)
7 (9.0)
7 (9.0)
n=243
0.08 (0.02)

150 (61.5)
57 (23.4)
23 (9.4)
14 (5.7)

0.0088

n/a

Lost to follow up, n (%) 42 (18.0) 29 (37.2) 131 (53.7) <0.0001
Deaths, n (%) 6
g
0 0 0.0183

a,b,c
Denominators are the total of subjects per group unless otherwise specified
a
Two CF-causing variants (from the 40-panel or one from the panel plus sequenced and classified as CF-
causing by CFTR2)
b,c
Children identified with one CF causing variant from the 40-panel and one or more sequenced
classified as varying clinical consequence by CFTR2
b
or unknown disease liability
c
. Unknown disease
liability includes the 7 variants in CFTR2 classified as unknown as well as those not yet studied.
d
p values indicate, unless otherwise specified, the comparison of all 3 groups. If p<0.05 means that there
is a difference between at least 2 groups.
e
p value for Hispanics compared to all other races and ethnicities.
f
IRT=Immunoreactive Trypsinogen, IQR=Interquartile Range.

g
The causes of death were: complications from prematurity including one confirmed case of acute
respiratory distress syndrome (n=3), complications from Biotinidase deficiency (n=1), sepsis post-
intestinal perforation repair (n=1), and acute respiratory failure secondary to CF-related pneumonia and
liver disease (n=1).
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 23
Table 3: Subjects carrying one CF-Causing and one or more mutations of unknown disease
liability who met diagnostic criteria for CF based on sweat chloride level and/or pancreatic
insufficiency status.

Unknown disease liability group: Individual genotypes
n
CF-C
mutation
Sequenced Mutation 1
Legacy name (cDNA)
Sequenced Mutation 2
Legacy name (cDNA)
Mutation type
Mutation 1, mutation 2
Sweat chloride 30-59 mmol/L and PI
1 F508del
2789+2insA
(c.2657+2_2657+3insA)
- Non-canonical splice
Sweat Chloride ≥60 mmol/L and PI
1
F508del

1138insG
(c.1006_1007insG)
- Frameshift
1 F1016S (c.3047T>C) L102R (c.305T>G) Missense, missense
1 1343delG (c.1211delG) - Frameshift
1
296+28A>G;2686-2687insT
(c.164+28A>G;
c.2554-2555insT)
-
Non-canonical splice;
frameshift
1
2481_2482insT
(c.2349_2350insT)
- Frameshift
2
2215insG
(c.2083_2084insG)
D836Y (c.2506G>T) Frameshift, missense
1 I1005R (c.3014T>T) - Missense
1
3199del6
(c.3067_3072delATAGTG)
- In-frame deletion
1 -816C>T (c.-887C>T) F1107L (c.3319T>C) Promoter, missense
1 1410delC (c.1278delC) I556V (c.1666A>G) Frameshift, missense
1
3015_3018dupGTCA
(c.2883_2886dupGTCA)
- Frameshift
1 S549N
1949del84
(c.1817_1900del84)
- In-frame deletion
1 W1089X
1811+1G>A
(c.1679+1G>A)
- Canonical splice
Sweat Chloride ≥60 mmol/L and PS
1
F508del

S1159P (c.3475T>C) - Missense
1 T1076P (c.3226A>C) Missense
1 L32M (c.94C>A) - Missense
1 T1036N (c.3107C>A) - Missense
1 c.-152G>C (c.-152G>C) - Promoter
1 G126D (c.377G>A) - Missense
1 Y917C (c.2750A>G) - Missense
1 R75X T1036N (c.3107C>A) - Missense
1 P205S
K114del
(c.472_474delAAG)
- In-frame deletion
1 N1303K K162E (c.484A>G) - Missense

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 24
Table 4: Subgroup analysis of children with one CF causing variant identified by CA-40 panel
and second identified by sequencing and classified as varying clinical consequence and unknown
disease liability, with frequency ≥ 6. PI=pancreatic insufficiency; PSA-P=Pseudomonas
aeruginosa persistent colonization according to a modified Leeds criteria.

Legacy name c. DNA n
Highest Sweat
Chloride in
mmol/L:
Median, range
(Included)
PI: n/total
included
PSA – P:
n/total
included
Sequenced CFTR mutation - varying clinical consequence
D1152H c.3454G>C 23 32,16-56 (23) 1/21 3/14
R117H/7T-9T c.350G>A 22 32,17-63 (22) 1/21 2/10
F1052V c.3154T>G 11 18,11-54 (10) 0/11 0/5
Sequenced CFTR mutations – unknown disease liability
1525-42G>A c.1393-42G>A 19 12, 8-33 (19) 0/19 1/6
L320V c.958T>G 13 14, 10-26 (12) 0/12 0/4
L967S c.2900T>C 9 20, 12-27 (7) 0/6 0/2
R170H c.509G>A 9 20, 7-30 (8) 0 /6 0/3
296+28A>G c.164+28A>G 6 14, 9-23 (6) 0 /4 0/2

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 25
Figure 1: Flowchart of total number of newborns screened from July 16
th
, 2007 to July 31, 2011.
IRT = immunoreactive trypsinogen, CF-causing group = cystic fibrosis-causing, VCC group =
variants of varying clinical consequence, and Unknown liability group.

a
The original total number of 2-panel positives was 172; two additional patients were selected to
be part of this group as their sequenced mutations were added to the panel: 1288insTA in
October 2007 and R75X in December 2007.
b
Panel variants only: n=174; panel plus sequenced and CF-causing: n=60 (total 234).

!
Total births
n=2,172,002
Valid IRT results
n=2,124,050
≥62 ng/ml
n=32,818
<62 ng/ml
n=2,091,232
Screen negative
≥2
n=174
a

1
n=1,855
California CFTR
mutation panel
None
n=30,789
Sequencing
No second mutation
identified
n=1,181
Screen negative,
carrier
Second mutation
identified
n=674
Excluded, n= 292: cis (n=27),
non CF-causing mutations
(n=57), and (TG)-5T (n=208;
unless part of complex allele)
CF-causing
n=234
b

Unknown
liability
n= 244
VCC
n= 78
GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 26
Figure 2: Growth for different subgroups: CF-C-PI = CF-causing group, pancreatic insufficient;
CF-C-PS = CF-causing group, pancreatic sufficient; VCC = mutations of varying clinical
consequence group; and Unk = mutation of unknown liability group. Pancreatic insufficient
subjects were removed from VCC (n=3) and Unk (n=16) groups for this analysis. There was a
statistically significant difference between CF-C-PI and CF-C-PS (p=0.0247), VCC (p=0.0013),
and Unk (p<0.0001). There is no statistically significance difference between CF-C-PS and
VCC and Unk (respectively p=0.9767 and p=0.9922).
Weight-for-Height
Predicted from Quadratic Model
Age in Months
0 2 4 6 8 10 12 14
WHZ
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
CFC-PI
CFC-PS
VCC
Unk

GENOTYPE-PHENOTYPE ASSOCIATIONS IN CYSTIC FIBROSIS 27
Figure 3: Pseudomonas aeruginosa first acquisition in the first 12 months of life. PSA =
Pseudomonas aeruginosa, CF-C = CF-causing group (n=231), VCC = mutation of varying
clinical consequence group (n=78), Unk = mutation of unknown liability group (n=243). The
probability of acquiring PSA was different among the 3 groups (p=0.0088 Log-Rank test). CF-C
group had a probability of acquisition rate of 0.18 (s.e 0.03), which was not statistically
significantly different than the VCC group (0.11 s.e 0.04, p=0.0916), but statistically higher than
the Unk group (0.08 s.e. 0.02, p=0.0034) according to pairwise comparisons.

0.70
0.75
0.80
0.85
0.90
0.95
1.00
0 90 180 270 360 450
PSA During First Year
Days
Probability PSA-free
Group
CF-C
VCC
Unk

Asset Metadata

Creator Salinas, Danieli Barino (author)

Core Title Genotype-phenotype associations in children with a positive cystic fibrosis newborn screening

Contributor Electronically uploaded by the author (provenance)

School Keck School of Medicine

Degree Master of Science

Degree Program Clinical and Biomedical Investigations

Publication Date 07/17/2015

Defense Date 07/16/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag CFTR mutations,cystic fibrosis newborn screening,genotype-phenotype associations,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Advisor Azen, Stanley P. (committee chair), Keens, Thomas (committee member), Patino Sutton, Cecilia M. (committee member), Ward, Sally L. (committee member)

Creator Email dsalinas@chla.usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-597264

Unique identifier UC11298990

Identifier etd-SalinasDan-3626.pdf (filename),usctheses-c3-597264 (legacy record id)

Legacy Identifier etd-SalinasDan-3626.pdf

Dmrecord 597264

Document Type Thesis

Format application/pdf (imt)

Rights Salinas, Danieli Barino

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract Cystic Fibrosis (CF) is an autosomal recessive disorder caused by a defective trans-epithelial channel called CF Transmembrane Conductance Regulator (CFTR). Of the nearly 2000 CFTR variants described only 201 have been formally assigned to three categories of disease liability: CF-causing, varying clinical consequence, and non-CF causing. California’s racially diverse population, along with exome sequencing as part of a newborn screening model, provides the opportunity to examine the phenotypes of young children with uncategorized variants to help inform disease liability and penetrance. Methods: Retrospective cohort study of children screened from 2007 to 2011 and followed for two to six years. Newborns that screened positive were divided into three genotype groups: those with two CF-causing mutations (CF-C)