| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Service de Pneumologie (Drs. Loubières, Grenet, and Stern), Hôpital Foch, Suresnes; Centre d’Etudes de Biologie Prénatale (Dr. Simon-Bouy), SESEP, Université de Versailles, Versailles; Laboratoire de Biostatistique (Drs. Medioni and Landais) Hôpital Necker-Enfants-Malades, Paris; and Laboratoire de Génétique Moléculaire et d’Histocompatibilité (Dr. Férec), INSERM, Brest, France.
Correspondence to: Marc Stern, MD, Service de Pneumologie, Hôpital Foch, 40 rue Worth, 92151 Suresnes cedex, France; e-mail: m.stern{at}hopital-foch.org
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Design: Retrospective study.
Setting: A respiratory unit of a teaching hospital.
Patients: We studied 51 adult CF patients for whom current data and genotype were available. Thirty-seven patients carried two severe mutations associated with pancreatic insufficiency phenotype (group S). Fourteen patients carried at least one mild (and dominant) mutation associated with pancreatic sufficiency phenotype (group M).
Measurements: We compared the course of the disease between the two groups, looking for a genotype/phenotype association for lung disease.
Results: The mean age of the population was 30 years. Patients with two severe mutations presented more severe disease with earlier onset (1.7 years vs 7.9 years, p = 0.0001). They presented with a more severe respiratory impairment, with a lower mean FEV1 (29% of predictive value vs 58% of predictive value, p < 0.001); a higher Pseudomonas colonization rate (97% vs 57%, p < 0.01); a more frequent end-stage respiratory insufficiency, defined by a FEV1 < 30% (73% vs 29%, p < 0.05); and a more marked yearly decline of FEV1 (3% vs 1.4%, p < 0.001). By multivariate logistic regression analysis, carrying two severe mutations was the only independent predictor of a terminal respiratory insufficiency (relative risk, 6.75; 95% confidence interval, 1.79 to 26.50; p = 0.003).
Conclusion: This study suggests that pulmonary disease appears to be associated with the severity of CF transmembrane regulator mutations.
Key Words: adults • cystic fibrosis • genotype • phenotype
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
F508) and accounts for 60 to 80% of allelic mutations in various
populations.2
3
At the molecular level, the mutations
include nonsense, missense, frameshift, in-frame deletions, and
splicing mutations. The CFTR mutants are classified into six classes
(from class I to class VI by order of decreasing severity) based on the
primary mechanism of reduction of CFTR function.4
5
6
Defects in the CFTR gene cause abnormal chloride concentration across
the apical membrane of epithelial cells and mainly result in
progressive lung disease, pancreatic dysfunction, elevated sweat
chloride level, and male infertility.
CF is characterized by wide variations in clinical
expression, partly explained by the number of mutations approaching
1,000 within the gene, although the majority of them are limited to a
single patient (http://www.genet.sickkids.on.ca/cftr/).
Consequently, various modes of presentation are observed
from birth to adulthood, with a wide range of severity and
rate of disease progression. Although most CF patients receive their
diagnosis during their first years of life in a context of lung disease
and pancreatic insufficiency (pi), an increasing number of
patients have CF detected later in adulthood with less typical disease.
Studying a large cohort of CF patients, Kerem et al7
demonstrated that the pancreatic status was genetically determined: two
severe mutations concerning pancreatic status (such as the F508
mutation) confer the pi phenotype, whereas a severe mutation
and a mild mutation, or two mild mutations confer the pancreatic
sufficiency (ps) status.7
Mild mutations
predominate over severe mutations in terms of pancreatic phenotype.
Despite extensive research into genotype/phenotype relationships and
because of the heterogeneity of CF, no association with lung disease
has been established,8
9
apart from rare mutations that
seem to be associated with a milder disease.10
11
We
retrospectively report the clinical course of a cohort of adult CF
patients according to their genotype and discuss the various factors
influencing phenotype, with particular emphasis on lung disease.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Diagnosis Criteria
All patients had to present characteristic symptoms of CF and
had to meet at least two of the diagnostic criteria described by Kerem
and Kerem.12
At least two sweat tests were performed in
each patient using the standard protocol. A sweat chloride
concentration > 70 mEq/L was considered to be positive for the
diagnosis of CF.13
The diagnosis was confirmed by
identification of two mutations of the CFTR gene, either at the time of
diagnosis, or later in the course of the disease. DNA analysis was
performed by standard procedures on peripheral blood. The most frequent
CF mutations usually found in the French population (F508, I507,
1717–1G
A, G542X, G551D, R553X, W1282X, N1303K) were analyzed by
polymerase chain reaction and allele-specific oligonucleotide with the
INNO-LIPA CF2 kit (Innogenetics; Zwijnaarde, Belgium). Almost
80% of CF cases among French children are accounted for by these eight
mutations.14
When a CF mutation was not found, complete
screening of the 27 exons of the gene was performed by denaturing
gradient gel electrophoresis, using a previously described
procedure.15
For each mutation, we specified its frequency
in our population and other characteristics, such as its class
according to the classification of Welsh and Smith4
and
its site on the CF gene (Table 1
).
|
Parameters Collected at Inclusion Period
The following data were noted for each patient: sweat chloride
level, age at first symptoms, type of first symptoms (respiratory:
cough, purulent sputum or digestive: diarrhea), monosymptomatic
beginning, age at diagnosis of CF (calculation of the time interval
between onset of first symptoms and diagnosis), and age at first
isolation of Pseudomonas aeruginosa in sputum.
Clinical Assessment and Follow-up
The following parameters routinely collected in CF patients in
our institution were recorded during every 6-month visit until the end
of the study.
General Status:
Karnofsky index, body mass index (weight
[kilograms]/height [meters squared]), and occupation.
Respiratory Status on Stable Clinical Situation:
Dyspnea
according to New York Heart Association classification; arterial blood
gas analysis on room air; FVC and FEV1; yearly decline of
FEV1 calculated between birth and end of the study
([100 - percent predictive FEV1]/age)10
;
6-min walk test; airways colonization with P aeruginosa;
annual number of courses of IV antibiotics; bronchial hyperreactivity,
defined as improvement of FEV1 
20% after
bronchodilators; episodes of pneumothorax and hemoptysis; and end-stage
respiratory insufficiency, defined as FEV1 < 30%.
GI Manifestations:
pi, defined clinically by
the necessity of pancreatic enzyme therapy; distal intestinal
obstruction; and cholestasis, defined as alkaline phosphatase > 1.2
times the normal value.
Other Manifestations:
Diabetes mellitus, nasal polyps,
sinusitis, and genital defect.
Other Conditions:
Lung transplantation or death.
Statistical Analysis
All data were entered into a computerized database. The various
parameters were compared between the two groups in order to assess the
genotype/phenotype association. We used a Student’s t test
to compare continuous variables (mean ± SD) or a
2 test to compare categorical variables
(percent) in a univariate analysis. For small patient samples, the
adjusted Yates 2 test was used. We completed
the univariate analysis using multivariate analysis. Variables with
p < 0.1 after the univariate analysis were entered into a logistic
regression model to evaluate their independent prognostic roles in the
poor respiratory outcome defined by a FEV1
< 30% of predictive value. The level of significance was set a
priori at 5% (p < 0.05).
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Genotype Analysis and Group Description
One hundred two mutations were isolated in these 51 patients
(Table 1) . The frequency of F508 mutation in this population was
71.5%. The severity for each mutation, insofar as presence or absence
of pi was concerned, was determined from a review of the
medical literature. Fifteen mutations (14.7%) were considered to be
mild mutations, and 87 mutations (85%) were considered to be severe
mutations. Twenty-seven patients (52.9%) were homozygotes for F508.
The incidence and characteristics of other mutations are shown in Table 1
.
Group S and group M comprised 37 patients and 14 patients, respectively. Only one patient in group M presented two mild mutations (Table 2 ). All severe mutations belong to class I or II of the classification of Welsh and Smith4 and would be expected to produce little or no functional CFTR, whereas mild mutations belong to class III, IV, and V and are associated with a partial CFTR function (Table 1) . All mechanisms of mutation (nonsense, missense, frameshift, deletion, and splicing) can lead to a severe mutation, while only missense and splicing mutations can lead to a mild mutation according to pancreatic status.
|
|
|
|
Respiratory disease appeared to be more frequent and more severe in group S patients than in group M patients (Table 4 ). Group S patients had a faster yearly decline of FEV1 (3%/yr) than group M patients, with a yearly loss of FEV1 of 1.4% (p < 0.001). Group S patients should therefore theoretically reach the threshold value of 30% of predicted FEV1 earlier than group M patients, at a mean age of 23 years and 50 years, respectively. The monthly rate of lung function decline (FEV1) during the study period was not significantly different between the two groups. Bronchial colonization by P aeruginosa was more frequent in group S, and these patients required more frequent courses of IV antibiotics (Table 4) . P aeruginosa colonization tended to occur earlier in group S (p = NS). The incidence of pneumothorax, hemoptysis, and bronchial hyperreactivity was similar in the two groups. End-stage respiratory disease was more frequent in group S (73%) than in group M (29%; p < 0.05; Table 3 ). Thirteen lung transplantations were performed in group S patients vs only 1 lung transplantation was performed in group M (p = 0.04). Seven patients died without transplantation in group S, and two patients died without transplantation in group M.
|
The incidence of GI and respiratory diseases increased in group S between the time of diagnosis of CF and the end of the surveillance study, contrary to the group M, where the rate of progression of the disease seems to be stable (Table 3) . Using multivariate logistic regression analysis, belonging to group S (two severe mutations) was the only independent predictor of a terminal respiratory insufficiency (relative risk, 6.75; 95% confidence interval, 1.79 to 26.50; p = 0.003).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
F508 mutation and the distribution between severe and
mild allele mutations are similar to those reported in the CF
consortium cohort.9
Most severe mutations were class I
nonsense mutations and most mild mutations were class IV missense
mutations. Twenty-seven percent of the patients (group M) had two mild
mutations or one mild mutation and one severe mutation, indicating
ps. This is much higher than generally reported (10
to 15%), and might indicate that the pi patients had
significantly shorter survival and thus are underrepresented in the
study. Most of our patients presented a typical form of CF
(pi, severe respiratory disease, positive sweat test result,
and early diagnosis during their childhood) as in the majority of
published series of adult CF patients.17
The patients with
CF diagnosed during adulthood, many years after the onset of their
first symptoms, are a more interesting subgroup. Seven patients
received a diagnosis after the age of 25 years, which has been rarely
reported, although it tends to be more frequent than it once
was.18
Mild and misleading presenting symptoms and
a negative sweat test result were the main causes for delayed
diagnosis, as were actions by parents, responses by physicians, and
variations in the health-care delivery systems. This emphasizes the
importance of the following elements for the diagnosis of CF: (1) a
negative sweat test result does not exclude the
diagnosis19
20
(six of our patients [11%] had a
negative sweat test result and a sweat chloride level in the normal or
intermediate range); (2) an accurate questionnaire is needed,
especially looking for unsuspected infertility due to bilateral absence
of vas deferens in male patients (In our experience, the diagnosis of
CF was suggested in some male adults with previously undetected or
misunderstood symptoms following demonstration of this genital
defect.); (3) the current genotype analysis is needed to ascertain the
diagnosis of atypical and mild forms of CF with negative sweat test
results. Our study showed that not only pancreatic function, as shown in previous reports,9 21 but also sweat chloride concentrations and pulmonary phenotype are correlated with the severity of the mutations. In fact, group S patients had earlier onset and more severe digestive and respiratory disease, and all but one of them had a positive sweat test result. Their mean sweat chloride level was higher than in group M patients. This agrees with the hypothesis that sweat gland function and chloride conductance may directly reflect the severity of the mutation,5 and that the genotype is by far the most important factor influencing the sweat gland function.22 The higher residual secretion resulting from a partially functional CFTR protein (type IV and type V mutations) might explain the more delayed diagnosis and the milder clinical impairment of group M patients.23 Gaskin et al24 also showed that sweat chloride levels were lower in ps patients than in pi patients.
Several mutations are associated with mild lung
disease.10
11
Gan et al10
and Hubert et
al16
suggested a genotype/phenotype association concerning
the severity of lung disease. In our study, all patients with two
severe mutations experienced respiratory symptoms. Their respiratory
function was more severely impaired with a more rapid yearly decline
than that of patients with at least one mild mutation (group M). Group
S patients also had a higher incidence of Pseudomonas colonization and
required more frequent courses of IV antibiotics. If, as indicated by
Davis et al,25
pulmonary function parameters are
considered to be normal at birth, group S patients should reach an
FEV1 equal to 30% of predictive value much
earlier than group M patients, at the age of 23 years and 50 years,
respectively. This FEV1, 30% of the
predicted value, is an important threshold associated with a median
survival of 2 years.26
Our findings that group S patients
are likely to develop more rapid decline of their respiratory
functional status, leading more frequently to lung transplantation or
death, are concordant with the demonstration that
FEV1 decline is a more powerful predictive index
than the simple cutoff value of 30% of
FEV1.27
Our results concerning a
genotype/phenotype association are in agreement with those reported by:
(1) Kerem and Kerem,12
Gaskin et al,24
and
Corey et al,28
who found less severe respiratory disease
in the case of ps; and (2) Johansen et al29
and
Kubesch et al,30
who found a higher incidence of
Pseudomonas colonization in F508 homozygous pi patients
and a lower risk for P aeruginosa acquisition in patients
with mild mutations and ps phenotype; and (3) Ferec et
al,31
who found milder respiratory disease, lower risk for
P aeruginosa acquisition, and delayed pi in
patients bearing mild alleles. Our results differ from the
analysis9
of a large cohort of CF patients in which no
genotype/phenotype association for pulmonary status was established.
This conflicting finding might be mainly explained by a difference in
the mean age of the CF populations. As our study only concerned adult
patients, the difference of yearly FEV1 decline
between group S and group M resulted in a statistically significant
difference of respiratory function that might not have been observed in
much younger patients, as in the study by Kerem et al.7
Only in the long run could a statistically significant difference be
emphasized, as pulmonary function parameters are normal at
birth.25
The different course of respiratory function
observed between group S and group M might not be exclusively related
to genotype, as the better pulmonary status in group M might also
reflect a better nutritional status.32
The severity of
respiratory disease has been assumed to be a combination of various
mutations influenced by others genetic and environmental
factors.12
We confirm that pancreatic function and its rate of progression are related to genotype. In our study, malabsorption may have been incorrectly estimated due to the absence of quantitative determination of steatorrhea, but for most clinicians, symptoms of malabsorption and the response to pancreatic enzyme treatment constitute sufficient evidence of pancreatic exocrine insufficiency.33 The incidence of the pancreatic disease increased between the time of diagnosis and the end of the study only in group S patients. This progression of GI disease with time, despite pancreatic enzyme therapy, which does not influence pancreas destruction, has already been observed in other studies.12 24 30 The relative old age of group M patients may explain the high proportion of insufficient pancreatic patients in this group (29%) and its tendency to progression during the study period. We did not observe any episode of ileus in the two groups probably because of the efficacy of pancreatic enzyme therapy. By contrast, appropriate respiratory care may reduce progression of lung disease and prevent end-stage respiratory failure, especially in group M patients. Pancreatic function, like sweat gland function, can therefore be considered to directly reflect the genotype, while the severity of lung disease partly depends on the genotype, but also on other factors such as early treatment, compliance with treatment, and pollution or tobacco exposure.34
The genotype/phenotype association is complex. The clinical course of
some patients was not consistent with the classification, confirming
that the genotype is not the only predictor of the severity of the
disease: one group S patient with a F508/F508 genotype received a
diagnosis at the age of 26 years and presented with mild disease, while
some group M patients presented with severe disease and one of them
even required lung transplantation. Another patient presented a G85E
mutation, which has been associated with both ps and
pi phenotypes.35
In our study, this mutation
was considered to be a mild mutation, although this patient had severe
disease with a sweat chloride concentration of 157 mEq/L,
pi, and severe respiratory disease
(FEV1 at 26% of predictive value and Pseudomonas
colonization). The decision to classify this patient in group M
therefore lowered the significance of the observed difference of the
results of this study. Moreover, discrepancies between the phenotype of
siblings or patients with the same genotype36
argue in
favor of the role of other genetic or nongenetic (environmental)
factors to explain the wide spectrum of clinical manifestations of the
disease. Kerem and Kerem12
reported substantial variations
of respiratory impairment among children presenting the same mutations.
Several mechanisms may explain phenotype variations among patients with
the same genotype: some allelic differences in genes associated with
the traffic of the mutant protein,12
a second mutation on
the same CFTR allele that attenuates the effect of the main
mutation,37
unidentified genes outside the CF
locus,38
a DNA variant in a noncoding region
(eg, 5T, 7T or 9T in intron 8) of CFTR,39
40
and tissue-specific alternative splicing of messenger RNA may also
occur.38
The phenotype at the organ level depends on three components: the patient’s specific pair of mutations, the rest of the patient’s genome, and the environment, which is particularly relevant to respiratory status.38 Moreover, other genes modulating the inflammatory response and influencing the pathogenesis of lung disease may also alter progression of respiratory tract involvement.38 The sensitivity to the CFTR defect appears to differ from one tissue to another and seems to be directly related to the inability to maintain luminal hydration of ductal macromolecules and a high flow rate.41 This variable sensitivity explains the varying frequencies of involvement of different organs; the vas deferens is the most vulnerable to functional CFTR defect, followed by the sweat glands, lungs, and pancreas.25 Pancreatic function therefore appears to be a good marker of residual CFTR activity.5
|
Conclusion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
|
Footnotes |
|---|
Received for publication April 27, 1999. Accepted for publication April 3, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
F 508). N Engl J Med 323,1517-1522[Abstract]
F508 mutation. Lancet 337,631-634[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  J de Gracia, F Mata, A Alvarez, T Casals, S Gatner, M Vendrell, D de la Rosa, L Guarner, and E Hermosilla Genotype-phenotype correlation for pulmonary function in cystic fibrosis Thorax, July 1, 2005; 60(7): 558 - 563. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
