|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departements de Pneumologie Pediatrique-INSERM U515 (Drs. Desmarquest, Tamalat, Boule, Fauroux, Tournier, and Clement), et de Biochimie (Dr. Feldmann), Hopital Trousseau AP-HP, Universite Paris VI, Paris, France.
Correspondence to: Annick Clement, MD, PhD, Departement de Pneumologie Pediatrique, Hopital Trousseau, 26 av Dr. Netter, 75012 Paris, France; e-mail: annick.clement{at}trs.ap-hop-paris.fr
| Abstract |
|---|
|
|
|---|
Patients and results: We
focused on children referred to the pulmonary department for various
types of pulmonary disease and who had several sweat chloride test
results with median values in the range of 40 to 60 mEq/L. Twenty-four
patients over a 10-year period were enrolled (mean age, 4.8 years).
Respiratory manifestations at initial evaluation included recurrent
bronchitis, wheezing, chronic cough, and pneumonia. The duration of the
follow-up ranged from 0.5 to 10.5 years. Sputum cultures revealed the
presence of Haemophilus influenzae (10 children),
Staphylococcus aureus (4 children), and
Pseudomonas aeruginosa (3 children). Pancreatic
insufficiency was found in two patients. Analysis of the entire coding
sequence allowed identification of 16 known mutations in
CFTR gene. Fifteen chromosomes (31.2%) carried a
mutation in CFTR gene and one allele carried two
mutations. Three patients were homozygous or double heterozygous
(
F508/
F508,
F508/3849 + 10 kb C
T, S1235R/G551D).
The 5-thymidine allele was identified in four children.
Conclusion: These results indicate an higher frequency of CFTR gene mutations in patients with borderline sweat chloride test results, compared to data reported in the general population. They lead to the recommendations for complete pulmonary and GI investigations in this group of patients, as well as assiduous care and medical follow-up.
Key Words: children cystic fibrosis genotype sweat chloride tests
| Introduction |
|---|
|
|
|---|
For many years, elevated sweat chloride levels were the "gold standard" for diagnosis of CF.14 Abnormal values were assumed for chloride values > 60 mEq/L, the 60 mEq/L cutoff discriminating between the populations with CF and without CF.14 15 However, patients have been reported with characteristic manifestations of CF and chloride levels < 60 mEq/L.16 17 Moreover, in several cases, genetic analyses have documented two mutated alleles in the CFTR gene, and it is likely that such observations will increase in the next years with the identification of more mutations.18 19 Some of these mutations will possibly appear to correlate with a phenotype of mild diseases. Sweat chloride test values < 60 mEq/L can be separated into two groups: a group of values in the intermediate range of 40 to 60 mEq/L (borderline test results), and a group of values < 40 mEq/L.20 From the growing number of observations reported in the literature, there is some evidence that patients with borderline test results share more similarities with the CF patients (eg, patients with chloride values > 60 mEq/L) in terms of clinical presentations and disease progression than the group of patients with chloride values < 40 mEq/L.
The crucial issue of definition of CF raised by the number of observations that do not fulfill the classical criteria explain the efforts that need to be made to provide more information on both the genotype and the phenotype of patients with chloride levels < 60 mEq/L.21 In the present work, we focused on children who were referred to our pulmonary pediatric department for various pulmonary diseases and whose sweat chloride tests were in the intermediate range of 40 to 60 mEq/L. The aim of the study was to analyze the genotype and the phenotypic characteristics of these patients.
| Materials and Methods |
|---|
|
|
|---|
Clinical Evaluation
Personal and familial history, mode of presentation, and age at
first diagnosis were recorded for each patient. Sweat chloride levels
were measured by the standard protocol of Gibson and
Cooke.14
The quantitative pilocarpine iontophoresis tests
were performed with measurements of sweat weight and chloride
concentrations in duplicate, with sweat specimens being collected
concurrently from the right and left arm. A minimum sweat weight of 100
mg was required for analysis. At least four tests were performed for
each patient. Height and weight were recorded at each visit, and the
weight-to-height ratio representing the actual weight expressed as a
percentage of the ideal weight for height was calculated to determine
the nutritional status. Pulmonary disease was evaluated by physical
examination, chest radiograph, blood gas analysis at rest, and
pulmonary function tests. Before 6 years of age, pulmonary function
tests included measurements of functional residual capacity, dynamic
lung compliance (CLdyn), and total lung resistance. After 6
years of age, FVC and FEV1 were determined.
Results of these tests were expressed as a percentage of the predicted
values for height-matched children.22
23
Respiratory
bacterial infections were evaluated using quantitative sputum cultures,
or cultures of tracheal aspirates when sputum samples were not
available. Cultures were considered positive when microorganisms were
present at a concentration > 106 cfu/mL.
Pancreatic status was determined by the fat content in stool samples
collected over 3 days. Patients showing normal results (fecal fat < 5
g/d) and currently not treated by pancreatic enzyme replacement were
defined as pancreatic sufficient, and the remaining were defined as
pancreatic insufficient. The levels of blood vitamin A and E were also
evaluated.24
DNA Analysis
Peripheral blood samples were collected from the children and
genomic DNA was extracted by standard methods. For each patient, all 27
exons of the CFTR gene were analyzed by denaturing gradient
gel electrophoresis with polymerase chain reaction primers, as
described previously.25
Polymerase chain reaction products
that displayed an altered behavior in the gel were subsequently
sequenced. Mutations were identified by direct DNA sequencing using an
automatic DNA sequencer 373A (Applied Biosystems; Foster City,
CA). The cryptic splice mutation 3849 + 10 kb C
T was
detected by restriction analysis according to the instruction of
Highsmith et al.18
The 5-thymidine (5T) variant
of the intron 8 polythymidine tract was also documented by direct DNA
sequencing.
Statistical Analysis
Results of sweat chloride tests were reported as median (range).
Differences between proportions were tested by the
2 statistic. All p values were two-tailed, and
probabilities of < 0.05 were considered significant.
| Results |
|---|
|
|
|---|
|
|
|
F508, and two patients
carried compound heterozygous (
F508/3849 + 10 kb C
T,
S1235R/G551D). Patient 3 was heterozygous for the mutations R75X and
D1270H; however, the familial analysis revealed that the mutations R75X
and D1270H were both carried by the paternal allele. Eight patients
were heterozygous for a CFTR mutations previously described
in CF patients. It should be pointed out that patient 6, who was
heterozygous
F508, had the same genotype identified in his
brother, and familial analysis confirmed that the two children carried
the same alleles. As 12 of the 24 patients (50%) with intermediate
sweat chloride test results carried at least one CFTR
mutation, the frequency of CFTR mutations in the studied
population was significantly higher than in the general population
(4%), with t = 37.5 (p < 0.001).8
The
5T allele variant of intron 8 was identified in four
patients. Of the four patients, two patients carried a CFTR
mutation on one chromosome and the 5T on the other
chromosome (patients 5 and 22). Patient 9 carried the 5T
allele and the mild mutation S1235R on the same chromosome. The
frequency of the 5T allele in the population of patients
with intermediate sweat chloride test results was not statistically
different from the frequency reported in the general population
(frequency of the 5T allele in the general population,
5.2%).26
|
| Discussion |
|---|
|
|
|---|
Analysis of the genotype included study of the 27 exons of the
CFTR gene, as well as the surrounding intronic sequence
3849 + 10 kb C
T and the 5T allele. Results indicated
that 15 (of 48) chromosomes had a known mutation in CFTR
gene, with 1 chromosome bearing two mutations (R75X and D1270H). It is
interesting to point out that two patients were compound heterozygous
(
F508/3849 + 10 kb C
T, S1235R/G551D), and one was homozygous
F508. This genotype is usually observed in patients with high values
of sweat chloride concentrations.
Our results documenting the presence of various CFTR
mutations extend the conclusions of several reports in the literature
that have analyzed the levels of sweat chloride tests in different
groups of CF patients. The hypothesis that sweat chloride
concentrations may be directly reflective of CF transmembrane
conductance regulator activity and that different functional classes of
CFTR mutations would display differences in epithelial
chloride conductance and in sweat chloride values is no longer
sustained. In a study reviewing 455 patients who had two identified CF
mutations, Wilschanski et al27
analyzed the concentrations
of sweat chloride levels in CF in relation with the different classes
of mutations. They found no differences between patients bearing class
I mutations (85 patients), class II mutations (294 patients), and class
III mutations (48 patients), all of these patients except 1 in each
group having a phenotype of pancreatic insufficiency. They also found
similar sweat chloride levels in the group of class V mutations (11
patients), with all the patients included in this group having a
pancreatic sufficient status. The only statistical difference they
could document was in the group of patients bearing class IV mutations.
The mean sweat chloride level in this group was 95 mmol/L, compared to
104 mmol/L in the group of
F508/
F508 patients. This difference
should be interpreted with caution, because of the smaller number of
patients in the group of class IV mutations (17 patients). Indeed, one
would have expected class IV and class V mutations to give similar
sweat chloride concentrations: these classes of mutations produce
proteins that reach the apical membrane and generate cAMP-regulated
apical membrane chloride current, with the only observed difference
being a reduction in the amount of current for class IV mutations and a
reduced amount of normal protein for class V mutations.28
The 60 mEq/L value of sweat chloride concentrations has been used for a
long time to discriminate between the populations of patients with CF
and without CF. As for the relation between sweat chloride
concentration and CF transmembrane conductance regulator activity, the
60 mEq/L cutoff is questionable. Indeed, from several reports in the
literature, it is admitted that normal sweat chloride values do not
exclude the diagnosis of CF.18
19
In the present study,
one patient was
F508/
F508. Similar observations have already been
reported by other investigators for various mutations. Stewart et
al17
described two patients from two families with the
compound heterozygotic CF mutations
F508/3849 + 10 kb C
T.
Phenotypic description indicated that these patients had a mild
expression of the disease with sweat chloride concentrations of 28
mEq/L and 46 mEq/L, respectively. Interestingly, investigation of the
siblings of one patient revealed sweat test levels of 49 mEq/L and 45
mEq/L, and sputum culture positive for P aeruginosa. In the
population included in the present study, a patient also carried the
genotype
F508/3849 + 10 kb C
T, with a sweat chloride
concentration of 50 mEq/L. The 3849 + 10 kb C
T mutation was
identified by Highsmith et al18
and corresponds to a point
mutation in intron 19. This mutation seemed to be associated with
milder disease.29
A report by Augarten et
al30
of sweat chloride concentrations in patients with
this mutation indicated various results less or more than 60
mEq/L. Another interesting finding was the result of the 5T.
In our study population, the 5T allele was present in four
patients. The 5T allele corresponds to a DNA variant of the
intron 8 polypyrimidine tract at the branch acceptor site of exon 9.
This variant gives rise to a normal transcript and to a transcript with
an in-frame deletion of exon 9, leading to a protein without
cAMP-activated chloride conductance activity.26
31
32
33
Kerem et al26
32
showed that the 5T allele can
be associated with various clinical presentation and sweat chloride
levels ranging from normal to elevated.
Several studies have attempted to correlate genotype and phenotype in
CFTR mutations.11
12
13
34
35
Although it is well
admitted that some mutations, such as
F508, are associated with a
more severe clinical presentation, expression of pulmonary disease
among the patients carrying these mutations varies considerably. The
same observation applies to the pancreatic status as well as to the
levels of sweat chloride. This extreme variability supports the concept
that disease expression may be the result of CFTR mutations
in association with additional genetic and/or environmental
factors.6
36
Hull and Thomson36
have
addressed the question of the contribution of genes other than
CFTR to disease severity in a group of CF patients. Based on
the current concept of the role of inflammation in CF pathophysiology,
they speculated that the proinflammatory cytokine tumor necrosis
factor-
and the detoxifying enzyme glutathione S-transferase M1
could influence disease severity. Interestingly, they provided data
suggesting some relation between alteration in pulmonary function and
tumor necrosis factor-
-308 promoter polymorphism, as well as
homozygosity for the null allele of glutathione S-transferase M1.
The report of the consensus conference initiated by the CF Foundation in the United States stated that the criteria for the diagnosis of CF should include the following: (1) one or more characteristic phenotypic features, or a history of CF in a sibling, or positive newborn screening test results; and (2) an elevated sweat chloride concentration by pilocarpine iontophoresis (> 60 mmol/L) on two or more occasions, or identification of two CF mutations, or demonstration of abnormal nasal epithelial ion transport.21 According to these criteria, two of our patients fulfilled the two groups of criteria. For one patient, there was a diagnostic dilemma. To follow the recommendations of the authors of the consensus report, it is clear that there is a need to define more precisely the spectrum of CF phenotypic features, as well as to redefine the guidelines for sweat chloride test interpretation. This is sustained by studies indicating a higher incidence of CFTR gene mutation in patients with diffuse bronchiectasis.37 In addition, the growing number of described mutations and the complexity of exploring CFTR gene expression indicates that the genetic analysis of the CFTR gene for a given patient may never be complete.38 Stewart et al17 discussed the value of nasal transepithelial voltage measurements; however, such measurements are not validated in young children. This leads to the key question: How can we progress to confirm the diagnosis of CF? Considering the difficulties in answering this question, several authors39 suggest introducing the terms of CFTR-associated disease or CFTR-related disease to cover the various expressions of the disease. With similar concern, Farrell and Koscik20 discussed the levels of sweat chloride concentrations that should be considered as normal. They concluded that any sweat chloride value > 40 mEq/L had a low probability of being a true-normal and, therefore, that a diagnosis of CF is likely for levels in the range of 40 to 60 mEq/L.
Data reported in the present work provide support to the conclusion of Farrell and Koscik.20 From the current understanding of the role of CFTR in airway disease, assiduous care and medical follow-up of the patients with intermediate sweat chloride test results should be recommended. These patients should undergo a treatment program according to protocols designed for CF patients, with an aggressive management of pulmonary exacerbations.40 41 42 43 44
| Acknowledgements |
|---|
| Footnotes |
|---|
This work was supported by Association Française de Lutte contre la Mucoviscidose.
Received for publication August 25, 1999. Accepted for publication June 20, 2000.
| References |
|---|
|
|
|---|
F508). N Engl J Med 323,1517-1522[Abstract]
This article has been cited by other articles:
![]() |
K De Boeck, M Wilschanski, C Castellani, C Taylor, H Cuppens, J Dodge, M Sinaasappel, and on behalf of the Diagnostic Working Group Cystic fibrosis: terminology and diagnostic algorithms Thorax, July 1, 2006; 61(7): 627 - 635. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. LEBECQUE, T. LEAL, C. DE BOECK, M. JASPERS, H. CUPPENS, and J.-J. CASSIMAN Mutations of the Cystic Fibrosis Gene and Intermediate Sweat Chloride Levels in Children Am. J. Respir. Crit. Care Med., March 15, 2002; 165(6): 757 - 761. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |