INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Since the discovery of the genetic defect that causes cystic fibrosis (CF) by Riordan and colleagues (1), attempts have been made to correlate CF genotypes with clinical assessments (2). To date, the correlations between genotype and prognosis, mortality, and pulmonary function have been highly variable (3-5). Previous studies have demonstrated that patients homozygous for the F508 mutation have an earlier onset of disease, higher sweat chloride levels and a greater prevalence of pancreatic insufficiency (3, 6, 7). Despite these observations, patients with the F508 mutation have a wide range of severity of pulmonary disease, and severe pulmonary disease due to CF is not restricted to patients homozygous for the F508 mutation (8-10).

Molecular mechanisms by which genetic mutations disrupt the cystic fibrosis transmembrane regulator proteins (CFTR) have recently been proposed (11, 12). By categorizing these mechanisms into five different classes, it is possible to better understand the pathophysiology of the disease process and to develop therapeutic strategies for improvement in CFTR function (13) according to the class of the mutation. For class I mutations, there is a premature termination of the CFTR mRNA translation caused by either base substitutions that create stop codons or by mutations that shift the reading frame or nonsense mutations and result in defects in protein production. Examples of class I mutations are G542X and W1282X. The most common mutation, F508 and the N1303K mutation belong to class II mutations and result in defects in protein processing. In class II mutations, the CFTR protein is degraded in the endoplasmic reticulum and fails to reach its intended site of action at the plasma membrane. Class III mutations, of which G551D is an example, are regulatory mutations in which protein reaches the surface of the cell but fails to respond consistently to cyclic AMP activation signals. Class IV mutations also result in protein reaching the plasma membrane, but it has altered channel properties that results in defective protein conduction. The genes R117H and R347P are examples of class IV mutations. Class V mutations result in reduced amounts of RNA necessary for normal CFTR production.

As several new therapies have been based on the CFTR mutation class and exercise tolerance is a recommended outcome measure in interventional trials in CF, it is important to assess the relationship between CFTR mutation class and exercise tolerance in this population (14). The aim of this study was to investigate the relationship between fitness and nutrition with genotype in children with both CF and at least one copy of the F508 mutation. This is the first study that we are aware of that compares aerobic and anaerobic exercise capacities of patients on the basis of CFTR mutations.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Children aged 8 to 17 years that attended the CF clinic at The Children's Hospital at Westmead (Royal Alexandra Hospital for Children), New South Wales, Australia, were randomly selected for the study. The randomization was performed using medical record numbers and uniform random numbers generated by the Numerical Algorithms Group routines GO5CBF and GO5CAF (15). The subjects were excluded if there was a history of pulmonary exacerbation in the month preceding the test. For the purposes of this study, pulmonary exacerbation was defined as increased purulent sputum production, and not just those who received antibiotics or were formally diagnosed by their physicians as have had a pulmonary exacerbation. The study was approved by the Ethics Committee of the Royal Alexandra Hospital for Children, and written informed consent was obtained from all participants and their parents where applicable.

Parameters of Assessment

Lean body mass was calculated using skinfold thickness and the equations provided by Durnin and Rahaman (16). Height was measured using a Harpenden stadiometer (Seritex, Carlstadt, NJ), and weight by electronic scales (Acme, San Leandro, CA). The body mass index (BMI) was recorded for each subject.

The Shwachman score, a clinico-radiological measure of disease severity (17), was recorded at each child's annual progress check. This score, which has a range from 20 (most severe) to 100 (most mild), was used to compare the disease severity of the subjects.

Pulmonary function tests consisted of measuring forced expiratory volume in one second (FEV₁) and forced vital capacity (FVC) using a spirometer (Sensormedics 2000; Sensormedics, Yorba Linda, CA).

Peak aerobic capacity was assessed using a modified Bruce protocol (18) on an electronic treadmill (Cardiovit 100, Schiller, Dietikon, Switzerland). Breath-by-breath gas analysis was used to determine peak oxygen uptake (VO₂), minute ventilation (VE), and respiratory quotient (RQ). The results were expressed in terms of lean body weight. Anaerobic power was measured using the Wingate test (19) on an electronically braked cycle ergometer (Lode Ergometer, Groningen, the Netherlands). The results of the peak anaerobic power is expressed as per convention in terms of total body mass.

Data Analysis

The subjects were categorized into groups according to CFTR mutation. The five classes of CFTR mutation as described by Welsh and Smith (12) were adopted for this study. The values for the peak aerobic capacity, anaerobic power, lung function, Shwachman score, and body mass index for subjects with at least one copy of F508 mutation were compared according to the class of the child's second CFTR mutation (I-V). Duncan's Post Hoc test of analysis of variance (ANOVA) was used to assess differences among the groups for each of the key outcome variables and to detect the variable with the greatest difference between groups (20). The Pearsons correlation was used to assess the relationship between aerobic capacity and peak anaerobic power. Power calculations were performed using peak aerobic capacity, peak anaerobic power, lung function, body mass index, and Shwachman score. These calculations determined that, to detect a 0.5 standard deviation difference between the classes of mutations, with a significance of 0.05 and power of 80%, a minimum of 10 subjects in each group would be required. Statistical significance was assigned to p values less than 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Of the 101 randomly selected subjects, ninety-seven children with CF agreed to participate in the study. An upper limit of 40 was set for the number of children who were homozygous for the F508 mutation. Two patients, both of whom were homozygous for the F508 mutation, were excluded before testing because of a pulmonary exacerbation in the previous month. The mean age of the subjects was 14.1 years (range 8.4 to 16.8 years). There was a wide range of disease severity as measured by the Shwachman score with a mean of 60.0 (range 35 to 100). Twenty-three subjects who participated in the study were pancreatic sufficient (see Table 1). Pancreatic sufficiency was significantly more common in subjects with class IV and V second mutations compared with the other classes of mutations.

Those patients with a second mutation in class I had a mean peak aerobic capacity of 29.8 (SEM 4.2) ml/kg/min. This was not significantly different from the mean value of 32.1 (SEM 5.4) ml/kg/min in subjects with class II second mutations (Table 2). However, subjects with class III second mutations had a peak aerobic capacity that was 38.0% (p < 0.05) greater than those with class I second mutations. There were no significant differences in peak aerobic capacity between subjects with class IV and class V second CFTR mutations. However, subjects with class IV second mutations were 22.3% (p < 0.05) greater than those with class III second mutations.

Subjects with class III second CFTR mutations had 7.8% (p < 0.05) greater peak power than subjects with class I and II second CFTR mutations, but 10.9% (p < 0.05) lower power than subjects with class IV and V second CFTR mutations. Patients with class I and II second CFTR mutations had the lowest peak anaerobic power, individuals with class III mutations had intermediate power, and those with class IV and V CFTR mutations had the highest power of all groups.

Lung function, as indicated by FEV₁, was not significantly different among the groups and was thus unrelated to genotype (Table 2).

The overall correlation between peak anaerobic power and aerobic capacity for children with CF was very good (Pearsons r = 0.77, p < 0.01). The correlation coefficients for these parameters were not significantly different among the groups (r = 0.73, 0.80, 0.73, 0.76, 0.76 for subjects with class I, II, III, IV, and V second CFTR mutations, respectively).

Subjects were categorized into two groups based on BMI: those with second CFTR mutations belonging to classes I and II had a significantly lower (17.6%, p < 0.05) BMI than those with second CFTR mutations belonging to classes III, IV, and V. The Shwachman score of disease severity followed the same trend as BMI. Subjects with class I and class II second CFTR mutations had significantly worse disease scores than subjects with class III, IV, and V second CFTR mutations.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study has demonstrated that there is a significant relationship between the class of the second CFTR mutation and aerobic capacity, anaerobic power and BMI in patients with CF who were heterozygous for the F508 mutation. A second CFTR mutation, belonging to either class I or II, was associated with a lower peak aerobic capacity and anaerobic power, as well as a lower BMI and Shwachman score, while class III, IV, and V second mutations were associated with higher values for these parameters. The results of this study also confirm an absence of correlation between genotype and lung function.

The subjects in this study were randomly selected from a patient group aged 8 to 17 years that attended the CF clinic. Only four subjects declined to participate in the study, and a further two subjects were excluded because of a pulmonary exacerbation. Thus, selection bias was minimized. Before randomization, to study a wide range of genetic mutations and not just class II second mutations, an upper limit was set at 40 children with a genotype that was homozygous for F508 mutation.

All patients were recruited from a single center, and the sample size of this study was large compared with previously published studies of exercise capacity in children with CF (21). Certainly the sample size was sufficient to detect several important statistical differences in aerobic capacity and anaerobic power among the different classes of CFTR mutations. Power analysis, taking into account the different classes of CFTR mutations, suggested that type II errors were minimal. Nevertheless, the lack of statistically significant differences in some of the variables (such as lung function) may have been due to inadequate numbers of subjects. However, with ongoing suppuration intrinsic to CF pulmonary disease, there are clearly disadvantages of doing single measurements of exercise tolerance. This may have been a confounding variable in this study.

Kaplan and colleagues (22) demonstrated no differences in terms of nutrition and exercise tolerance between patients who were homozygous and heterozygous for the F508 gene mutation. The study was limited by the very small sample size. In addition, all heterozygous patients were combined into a single group for analysis in that study, failing to take into account the vast spectrum of disease within this patient population. By classifying patients according to the cellular physiological function conferred by class of the CFTR mutation, the present study has demonstrated significant differences in exercise tolerance, muscle power, and nutrition among different mutation classes. The correlation between aerobic capacity and peak anaerobic power was good. This is consistent with a pervasive pathophysiologic process that affects both aerobic and anaerobic muscle function.

Although CF genotype has been shown to be highly predictive of exocrine pancreatic function (23), to date there has been no proven association between genotype and pulmonary status (24). Most recent approaches toward understanding the genotype-phenotype correlation in cystic fibrosis are focused on in vitro studies of CFTR function. The impact of nutritional status on exercise ability in children with CF has previously been demonstrated (25). Malnutrition severely limits exercise tolerance in children with advanced CF. In this study, nutrition was directly assessed using measurement of lean body mass and BMI and, indirectly, using the Shwachman score. Despite the potential inaccuracies of using these methods of assessing nutrition, clinically, the nutritional status of children in this study was adequate. Therefore, the differences demonstrated in this study in aerobic and anaerobic exercise ability among the different classes of CFTR mutations cannot be completely explained purely on the basis of nutrition. Other mechanisms must be involved. Sheppard and colleagues (26) studied the expression of normal and F508 CFTR in epithelial cells and demonstrated that the F508 CFTR mutation produced an immature type of glycosylated core protein due to defective processing. Other mechanisms by which the different classes of mutations disrupt CFTR function have been proposed (11, 12). The classes of mutation as described by Welsh and Smith (12) were adopted for this study and the results indicate that the deficits conferred on epithelial function by the mutations may affect muscle function. CFTR is a single polypeptide comprising two similar halves. Each half has six transmembrane -helices followed by a cytoplasmic nucleotide binding domain (NBD) that are separated by an intracellular (R) domain containing numerous sites for phosphorylation by protein kinase A and protein kinase C (1). CFTR is a chloride ion pore (27) and mutations in the transmembrane -helices cause alterations in the characteristics of chloride channel permeation (28). Chloride conductance in skeletal muscle fibers, which is mediated by CFTR and other chloride channels, is the predominant parameter that contributes to the electrical stability of sarcolemma (29). In class I and II mutations, there is a complete absence of CFTR protein presented to the apical membrane and, hence, efficient chloride conductance is severely limited. Class IV and V mutations result in CFTR proteins that reach the apical membrane but have altered chloride channel properties and hence chloride conductance is present but at a reduced level. The varying degrees of electrical stability conferred on skeletal muscle fibers by the different classes of CFTR mutations may be a possible explanation for the findings in this study. CFTR chloride channels require ATP hydrolysis for normal function (30). In addition, CFTR possesses intrinsic ATPase activity (31). As energy is released with the degradation of ATP, it may be that depleted or defective CFTR protein results in inefficient energy production or utilization, which is clinically demonstrated by reduced aerobic capacity and peak anaerobic power.

This study has important implications for long-term prognosis as well as response to various therapies in interventional studies. On the basis of aerobic capacity (32), it may be that patients with class III, IV, or V CFTR mutations have a better overall prognosis. However, with new therapies based on the CFTR mutation class (33), exercise testing would be an appropriate method of assessing response.

The results indicate that the previous ambiguity in regard to the relationship between genotype and phenotype in CF may be clarified if interpreted in the light of the molecular mechanisms conferred by the particular mutation. Further studies on the effect of impaired CFTR on skeletal muscles are necessary.