Association Between Genetically Determined Pancreatic Status and Lung Disease in Adult Cystic Fibrosis Patients* Yann Loubie`res, MD; Dominique Grenet, MD; Brigitte Simon-Bouy, MD, PhD; Jacques Medioni, MD; Paul Landais, MD; Claude Fe´rec, MD, PhD; and Marc Stern, MD Study objectives: The association between genotype and phenotype in cystic fibrosis (CF) has been clearly established for pancreatic status, but not for lung disease. 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 ⴝ 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 ⴝ 0.003). Conclusion: This study suggests that pulmonary disease appears to be associated with the severity of CF transmembrane regulator mutations. (CHEST 2002; 121:73– 80) Key words: adults; cystic fibrosis; genotype; phenotype Abbreviations: CF ⫽ cystic fibrosis; CFTR ⫽ cystic fibrosis transmembrane regulator; NBF ⫽ nucleotide binding fold; NS ⫽ not significant; pi ⫽ pancreatic insufficiency; ps ⫽ pancreatic sufficiency; R ⫽ regulatory; TM ⫽ transmembrane
ystic fibrosis (CF) is an inherited autosomal C recessive disease of exocrine glands caused by mutations of a single gene. The CF gene, identified in 1989, encodes the CF transmembrane regulator (CFTR), a regulated chloride channel composed of five domains: two transmembrane (TM) domains
*From the Service de Pneumologie (Drs. Loubie`res, Grenet, and Stern), Hoˆpital Foch, Suresnes; Centre d’Etudes de Biologie Pre´natale (Dr. Simon-Bouy), SESEP, Universite´ de Versailles, Versailles; Laboratoire de Biostatistique (Drs. Medioni and Landais) Hoˆpital Necker-Enfants-Malades, Paris; and Laboratoire de Ge´ne´tique Mole´culaire et d’Histocompatibilite´ (Dr. Fe´rec), INSERM, Brest, France. Manuscript received April 27, 1999; revision accepted April 3, 2001. Correspondence to: Marc Stern, MD, Service de Pneumologie, Hoˆpital Foch, 40 rue Worth, 92151 Suresnes cedex, France; e-mail:
[email protected]
(TM1, TM2), two nucleotide binding fold (NBF) domains (NBF1, NBF2), and a regulatory (R).1,2 The most common allelic mutation deletes a phenylalanine at position 508 (⌬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 – 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. CHEST / 121 / 1 / JANUARY, 2002
73
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 Patient Selection The records of all patients referred for CF to our respiratory medicine unit from 1990 to 1999 were reviewed. Inclusion criteria were age ⬎ 18 years and complete data including identification of both mutations of the CF gene. 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, 1717–1G3 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 74
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). Patient Distribution According to Genotype Patients with two severe mutations concerning pancreatic status, who are supposed to express the pi phenotype, comprised the first group (group S). Patients with at least one mild mutation (two mild mutations, or one mild mutation and one severe mutation), who are supposed to express the ps phenotype, comprised the second group (group M). 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 ⱖ 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). Clinical Investigations
Table 1—Characteristics of the Mutations* No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Mutation
%
K 14 X G 85 E 621 ⫹ 1 G3T E 193 X L 206 W I 336 K R 347 H Q 493 X ⌬F 508 1717 ⫺ 1 G3A G 542 X R 553 X Y 569 C E 595 X R 668 C 2183 AA-G 2789 ⫹ 5 G3A G 1061 R R 1066 H D 1152 H D 1182 H 3849 ⫹ 4 A3G 3849 ⫹ 10 kb C3T S 1251 N N 1303 K
0.9 0.9 0.9 1.9 0.9 0.9 0.9 0.9 71.5 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 1.9 3.9
Severity Severe Mild Severe Severe Mild Mild Mild Severe Severe Severe Severe Severe Mild Severe Mild Severe Mild Mild Mild Mild Mild Mild Mild Mild Severe
Class
Site on the Gene
I IV I I IV IV IV I II I I I III I III I V IV IV IV IV V V IV II
Ex 1 Ex 3 Int 4 Ex 5 Ex 6a Ex 7 Ex 7 Ex 10 Ex 10 Int 10 Ex 11 Ex 11 Ex 12 Ex 13 Ex 13 Ex 13 Int 14b Ex 17b Ex 17b Ex 18 Ex 19 Int 19 Int 19 Ex 20 Ex 21
Site of CFTR TM1
TM1 TM1 TM1 NBF1 NBF1 NBF1 NBF1 NBF1 R R R TM2 TM2 TM2
NBF2 NBF2
Type of Mutation Nonsense Missense Splicing Nonsense Missense Missense Missense Nonsense Deletion Splicing Nonsense Nonsense Missense Nonsense Missense Frameshift Splicing Missense Missense Missense Missense Splicing Splicing Missense Missense
*Ex ⫽ exon; Int ⫽ intron.
Results Characteristics of the Population Sixty-one adult CF patients were referred to our center for CF during the 9-year study period. Ten patients were excluded from analysis because of at least one undetected mutation. Fifty-one patients (25 men and 26 women) were therefore included in this retrospective study. Their mean age at the end of the study was 30 ⫾ 6 years (28 ⫾ 4 years in group S and 34 ⫾ 7 years in group M, p ⫽ 0.003). Many patients presented severe respiratory disease (mean FEV1, 38% of predictive value; 86% were colonized by P aeruginosa colonization) and pancreatic insufficiency (88% of patients) at the end of the study. The mean follow-up duration by patient was 47 ⫾ 30 months.
tions. 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. Data at the Time of Inclusion
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 muta-
Forty-six percent of patients in group S were males vs 57% in group M (p ⫽ not significant [NS]). The first manifestations of the disease appeared earlier in group S (1.7 ⫾ 3 years) than in group M (7.9 ⫾ 5 years; p ⫽ 0.0001). Age at diagnosis was 3.4 ⫾ 5 years in group S and 20 ⫾ 10 years in group M (p ⫽ 0.0001). The time to diagnosis was sevenfold shorter in group S (1.8 ⫾ 5 years vs 12.5 ⫾ 10 years, p ⫽ 0.001). Monosymptomatic onset of disease ocCHEST / 121 / 1 / JANUARY, 2002
75
Table 2—Description of the Two Groups Genotypes Group S (n ⫽ 37) ⌬F508/⌬F508 ⌬F508/N 1303 K ⌬F508/621 ⫹ 1 G3T ⌬F508/1717 ⫺ 1 G3A ⌬F508/E 595 X ⌬F508/G 542 X ⌬F508/Q 493 X ⌬F508/R 553 X Group M (n ⫽ 14) ⌬F508/Y 569 C ⌬F508/3849 ⫹ 10 kb C3T ⌬F508/2789 ⫹ 5 G3A ⌬F508/D 1152 H ⌬F508/R 668 C ⌬F508/D 1182 H ⌬F508/I 336 K ⌬F508/3849 ⫹ 4 A3G ⌬F508/L 206 W E 193 X/S 1251 N K 14 X/G 1061 R 2183 AA-G/G 85 E R 347 H/R 1066 H
Class
No.
II/II II/II II/I II/I II/I II/I II/I II/I
27 4 1 1 1 1 1 1
II/III II/V II/V II/IV II/III II/IV II/IV II/V II/IV I/IV I/IV I/IV IV/IV
1 1 1 1 1 1 1 1 1 2* 1 1 1
*Brothers.
curred in 85% of group M patients vs 51% in group S patients (p ⫽ 0.02). Monosymptomatic respiratory onset of disease was much more frequent in group M (72% vs 24% in group S, p ⫽ 0.002). GI symptoms were more often the first symptoms of the disease in group S (76%) than in group M (29%; p ⫽ 0.002). Respiratory symptoms were present at the onset of the disease in a similar proportion of patients in the two groups (73% and 86%, respectively; p ⫽ NS), but the beginning of the disease was delayed in group M (Table 3). Diagnosis of CF was established during the first year of life in 49% of patients in group S vs none in group M (p ⫽ 0.001; Fig 1). The sweat test result was positive in 97% of group S patients (only one patient had a negative test result) but in only 64% of group M patients (five patients with negative sweat test results; p ⫽ 0.001), with a higher mean sweat chloride level in group S (117 ⫾ 28 mEq/L) than in group M (82 ⫾ 41 mEq/L; p ⫽ 0.01). Sweat test results were more
often negative as the age of diagnosis increased (Fig 2). Four of the six negative sweat test results were within the intermediate range of 40 to 70 mEq/L. Parameters Recorded at the Last Visit of the Study Period Group S patients had a poorer general status than group M patients, with a lower body mass index (15 ⫾ 2 kg/m2 vs 18 ⫾ 3 kg/m2, p ⬍ 0.001) and a lower Karnofsky index (58 ⫾ 17 vs 78 ⫾ 22, p ⫽ 0.01). They also worked less frequently (49% vs 79%, p ⫽ 0.01). 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. As expected, more patients presented with GI disease in group S than in group M: pi and cholestasis occurred in 97% compared to 64% (p ⫽ 0.05) and 43% vs 7% (p ⬍ 0.02) of patients, respectively. No difference was observed in the rate of ileus. Sixteen percent of group S patients were diabetic vs no patients in group M (p ⬍ 0.02). Of the 25 men of the study, 6 subjects had a documented genital defect.
Table 3—Rate of GI and Respiratory Disease Between the Onset of the Disease and the End of the Study Digestive Disease, % Groups
Symptomatic Disease Duration, yr
First Symptoms
End of Study
S M p Value
27 ⫾ 4 26 ⫾ 7 NS
76 29 ⱕ 0.001
97 64 ⱕ 0.001
76
Respiratory Disease, % p Value ⬍ 0.05 NS
First Symptoms
End of Study
73 86 NS
100 93 NS
p Value ⱕ 0.001 NS
Clinical Investigations
Figure 1. Percentage of cumulative diagnosis according to age at diagnosis.
All of them belong to group M because we particularly focused on the absence of vas deferens in misleading symptoms cases and not systematically for all patients. The frequency of nasal polyps and sinusitis was higher in group S (89%) than in group M (64%; p ⬍ 0.04). 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).
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,
Table 4 —Patient Characteristics According to Genotype*
Discussion
Characteristics
Overall, this adult CF cohort is similar to other clinical series of adult CF patients reported in the literature.16,17 The incidence of ⌬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
Follow-up, mo Age, yr New York Heart Association class 6-min walk test, m FVC, mL FVC, % predictive FEV1, mL FEV1, % predictive Annual loss of FEV1, % FEV1/FVC, % Pao2, mm Hg Paco2, mm Hg Pseudomonas colonization Age at first detection, yr Yearly IV antibiotic courses Terminal respiratory insufficiency Transplantation Death before transplantation
Figure 2. Sweat test results according to age at diagnosis. pts ⫽ patients.
Group S (n ⫽ 37)
Group M (n ⫽ 14)
42 ⫾ 30 28 ⫾ 4 3⫾1
58 ⫾ 26 34 ⫾ 7† 2 ⫾ 1†
395 ⫾ 160 2,100 ⫾ 1,100 50 ⫾ 19 1,100 ⫾ 700 29 ⫾ 15 3 ⫾ 0.9 49 ⫾ 11 65 ⫾ 12 43 ⫾ 9 36 (97) 17 ⫾ 7 4.5 ⫾ 2.9 27 (73)
564 ⫾ 105† 3,400 ⫾ 1,500† 72 ⫾ 26† 2,200 ⫾ 1,400‡ 58 ⫾ 32‡ 1.4 ⫾ 1.2‡ 63 ⫾ 19† 80 ⫾ 15† 38 ⫾ 4† 8 (57)‡ 21 ⫾ 10 1.9 ⫾ 1.8† 4 (29)†
13 (35) 7 (19)
1 (7)† 2 (14)
*Data are presented as mean ⫾ SD or No. (%). †p ⬍ 0.05. ‡p ⱕ 0.001. CHEST / 121 / 1 / JANUARY, 2002
77
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 78
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 Clinical Investigations
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 tissuespecific 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 This study demonstrates that not only pancreatic status but also pulmonary disease are correlated with the type of CFTR mutations. Since mortality in patients with CF depends on progression of lung disease, patients with at least one mild mutation have a better prognosis than patients with two severe mutations, which present a higher risk of terminal respiratory insufficiency. As proposed by Tsui,22 CF phenotypes can be classified into two categories: one category includes clinical manifestations directly related to the genotype, such as pancreatic function; and the other category, as seen with pulmonary status, is defined by the genotype as well as other factors. References 1 Riordan JR, Rommens JM, Kerem B, et al. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 1989; 245:1066 –1073 2 Kerem B, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245: 1073–1080 3 Tsui LC. Mutations and sequence variations detected in the cystic fibrosis transmembrane conductance regulator (CFTR) gene: a report from the Cystic Fibrosis Genetic Analysis Consortium. Hum Mutat 1992; 1:197–203 4 Welsh MJ, Smith AE. Molecular mechanisms of CFTR chloride channels dysfunction in cystic fibrosis. Cell 1993; 73:1251–1254 5 Wilschanski M, Zielenski J, Markiewicz D, et al. Correlation of sweat chloride concentration with classes of the cystic fibrosis transmembrane conductance regulator gene mutations. J Pediatr 1995; 127:705–710 6 Zielenski J. Genotype and phenotype in cystic fibrosis. Respiration 2000; 67:117–133 7 Kerem E, Corey M, Kerem BS, et al. The relation between genotype and phenotype in cystic fibrosis: analysis of the most common mutation (⌬F 508). N Engl J Med 1990; 323:1517– 1522 8 Burke W, Aitken ML, Chen SH, et al. Variable severity of CHEST / 121 / 1 / JANUARY, 2002
79
9 10 11
12 13 14
15 16 17 18 19 20 21 22 23
24
80
pulmonary disease in adults with identical cystic fibrosis mutations. Chest 1992; 102:506 –509 The Cystic Fibrosis Genotype-Phenotype Consortium. Correlation between genotype and phenotype in patients with cystic fibrosis. N Engl J Med 1993; 329:1308 –1313 Gan KH, Veeze HJ, Van Den Ouweland MW, et al. A cystic fibrosis mutation associated with mild lung disease. N Engl J Med 1995; 333:95–99 Highsmith WE, Burch LH, Zhou Z, et al. A novel mutation in the cystic fibrosis gene in patients with pulmonary disease but normal sweat chloride concentrations. N Engl J Med 1994; 331:974 –980 Kerem E, Kerem B. Genotype-phenotype correlations in cystic fibrosis. Pediatr Pulmonol 1996; 22:387–395 Hodson ME, Beldon I, Power R, et al. Sweat tests to diagnose cystic fibrosis in adults. BMJ 1983; 286:1381–1383 Simon-Bouy B, Mornet E, Serre JL, et al. Nine mutations in the cystic fibrosis (CF) gene account for 80% of the CF chromosomes in French patients. Clin Genet 1991; 40:218 – 224 Fe´ rec C, Audrezet MP, Mercier B, et al. Detection of over 98% cystic fibrosis mutations in a Celtic population. Nat Genet 1992; 1:188 –191 Hubert D, Bienvenu T, Desmazes-Dufeu N, et al. Genotypephenotype relationships in a cohort of adult cystic fibrosis patients. Eur Respir J 1996; 9:2207–2214 Di Sant’Agnese PA, Davis PB. Cystic fibrosis in adults: 75 cases and a review of 232 cases in the literature. Am J Med 1979; 66:121–132 Pescatore P, Marteau P, Lemie`re E, et al. Mucoviscidose reconnue apre`s l’aˆ ge de 25 ans. Gastroenterol Clin Bio 1994; 18:195–199 Strong TV, Smit LS, Turpin SV, et al. Cystic fibrosis gene mutation in two sisters with mild lung disease and normal sweat electrolyte levels. N Engl J Med 1991; 325:1630 –1634 Gan KH, Geus WP, Bakker WP, et al. Genetic and clinical features of patients with cystic fibrosis diagnosed after the age of 16 years. Thorax 1995; 50:1301–1304 Kristidis P, Bozon D, Corey M, et al. Genetic determination of exocrine pancreatic function in cystic fibrosis. Am J Hum Genet 1992; 50:1178 –1184 Tsui LC. The cystic fibrosis transmembrane conductance regulator gene. Am J Respir Crit Care Med 1995; 151:S47– S53 Veeze HJ, Halley DJJ, Bijman J, et al. Determinants of mild clinical symptoms in cystic fibrosis patients: residual chloride secretion measured in relation to the genotype. J Clin Invest 1994; 93:461– 466 Gaskin K, Gurwitz D, Durie P, et al. Improved respiratory prognosis in patients with cystic fibrosis with normal fat absorption. J Pediatr 1982; 100:857– 862
25 Davis PB, Drumm M, Konstan W. Cystic fibrosis: state of the art. Am J Respir Crit Care Med 1996; 154:1229 –1256 26 Kerem E, Reisman J, Corey M, et al. Prediction of mortality in patients with cystic fibrosis. N Engl J Med 1992; 326:1187– 1191 27 Milla CE, Warwick WJ. Risk of death in cystic fibrosis patients with severely compromised lung function. Chest 1998; 113:1230 –1234 28 Corey M, Gaskin K, Durie P, et al. Improved prognosis in CF patients with normal fat absorption. J Pediatr Gastroenterol Nutr 1984; 3:S99 –S105 29 Johansen HK, Nir M, Hoiby N, et al. Severity of cystic fibrosis in patients homozygous and heterozygous for ⌬F508 mutation. Lancet 1991; 337:631– 634 30 Kubesch P, Dork R, Wulbrand U, et al. Genetic determinants of airway’s colonization with Pseudomonas aeruginosa in cystic fibrosis. Lancet 1993; 341:189 –193 31 Ferec C, Verlingue C, Guillermit H, et al. Genotype analysis of adult cystic fibrosis patients. Hum Mol Genet 1993; 2:1557–1560 32 Tsui LC. The spectrum of cystic fibrosis mutations. Trends Genet 1992; 8:392–398 33 Stern RC. The diagnosis of cystic fibrosis. N Engl J Med 1997; 336:487– 491 34 Campbell PW III, Parker RA, Roberts BT, et al. Association of poor clinical status and heavy exposure to tobacco smoke in patients with cystic fibrosis who are homozygous for the F508 deletion. J Pediatr 1992; 120:261–264 35 Kerem E, Nissim-Rafinia M, Argaman A, et al. A missense cystic fibrosis transmembrane conductance regulator mutation with variable phenotype [abstract]. Pediatrics 1997; 100:E5 36 Ferrari M, Cremonesi L. Genotype-phenotype correlation in cystic fibrosis patients. Ann Biol Clin 1996; 54:235–241 37 Do¨ rk T, Wulbrand U, Richter T, et al. Cystic fibrosis with three mutations in the cystic fibrosis transmembrane conductance regulator gene. Hum Genet 1991; 87:441– 446 38 Bienvenu T. Les bases mole´ culaires de l’he´ te´ roge´ ne´ ite´ phe´ notypique dans la mucoviscidose. Ann Biol Clin 1997; 55: 113–121 39 Chu CS, Trapnell BC, Curristin S, et al. Genetic basis of variable exon 9 skipping in cystic fibrosis transmembrane conductance regulator mRNA. Nat Genet 1993; 3:151–156 40 Chillon M, Do¨ rk T, Casals T, et al. A novel donor splice site in intron of the CFTR gene, created by mutation 1811⫹1.6kbA-⬎G, produces a new exon: high frequency in Spanish cystic fibrosis chromosomes and association with severe phenotype. Am J Hum Genet 1995; 56:623– 629 41 Tizzano EF, Buchwald M. CFTR expression and organ damage in cystic fibrosis. Ann Intern Med 1995; 123:305–308
Clinical Investigations
