Study population
A cross-sectional study was carried out from people with a confirmed diagnosis of CF registered in the Rare Diseases Information System of Murcia (SIER)32, whose disease had been detected until December 31, 2018. People with CFTR related disorders (CFTR-RDs), CF-screen positive inconclusive diagnosis (CF-SPID) and healthy carriers were excluded. The informed consent of the study population was not required as the SIER is subject to the personal data protection regulations and registered in the Spanish Data Protection Agency (nº 2101040243 of April 14, 2010)33. Even so, the study was presented to the Clinical Research Ethics Committee of the International Doctoral School of the University of Murcia (nº 3376/2021) and it was approved on May 6, 2021.
Rare Diseases Information System (SIER)
The SIER, existing since 2010, is a population registry of rare diseases (RDR) of the Region of Murcia, an Autonomous Community located in southeastern Spain with an estimated population of 1,493,898 inhabitants as of January 1, 2019, which constitutes 3.18% of the Spanish population. For the inclusion of people with some rare disease (RD), this system uses a list of selected codes from the International Classification of Diseases (ICD) and integrates information from different sources. Currently, the SIER has 47 different sources of information; administrative clinics such as the regional Minimum Basic Data Set (MBDS), pre-existing patient registries such as the renal disease registry, orphan or foreign drug dispensing database, databases of people with recognition of disability and dependency, notification of patient associations, or Clinical Hospital Units. For this study, the Regional CF Unit of the Virgen de la Arrixaca University Clinic Hospital (HCUVA) was the main source of information in the contribution of people with CF to the registry. The sources that incorporated some patients of the study are shown in Table 1.
Once possible cases of RD have been incorporated into the registry they undergo a validation process, confirming the evidence of the diagnosis once the electronic medical record of the patient has been reviewed32.
Regarding the codes for the detection of people with CF, the 277.0 [0-9] in its ninth revision, Clinical Modification of International Classification of Diseases (ICD-9-CM) was used until 2015, and the code E84 [0-9] in the tenth version of the Spanish Clinical Modification (ICD-10-ES) from 2016 to 2018.
Data collection
The data collected from each patient included:
Consultation of basic patient information: sex, age at diagnosis (<18 years or ≥ 18 years), age on December 31st, 2018, native country of parents, diagnosis by neonatal screening, death and transplant (yes/no).
Obtaining genetic information: Firstly, information was obtained about the variants of the CFTR gene. In addition, the project database CFTR28, Database of single nucleotide polymorphisms (dbSNP)34, Database of cystic fibrosis mutations7 and ClinVAr35 were consulted to include the type of alteration that patients presented in the gene sequence, along with the associated nucleotide change, amino acid change and its molecular and clinical consequence. Secondly, the patients were classified into 2 groups according to genotype; "high-risk" if there were 2 mutations of minimal function or class I, II, III, and VII and "low-risk" if at least one allele carried a residual mutation or class IV, V and VI, with some exceptions1, 17, 18.
Procurement of clinical manifestations: We collected up to December 31, 2018, the following clinical manifestations: respiratory and digestive symptoms, metabolic disturbances and others like bone alterations.
Respiratory symptoms: Evidence of at least one episode of Allergic Bronchopulmonary Aspergillosis (ABPA), one or more clinically relevant episodes of hemoptysis (> 200 ml), presence of nasal polyps, chronic respiratory colonizations by different microorganisms (Staphylococcus aureus, Burkholderia cepacia or Pseudomonas aeruginosa) and at least one documented acute infection with methicillin-resistant Staphylococcus aureus, Achromobacter xylosoxidans, or nontuberculous mycobacteria.
Lung function was evaluated using the best value of the forced expiratory volume in the first second (FEV1) recorded in 2018, normalized with respect to its theoretical using the Global Lung Function Initiative (GLI) tool and expressed as a percentage of the predicted value. The variable was dichotomized into ≤90% and> 90%, the cut-off point used in other studies36.
Digestive symptoms: presence of meconium ileus at birth, rectal prolapse, intussusception, distal intestinal obstruction syndrome (DIOS), pancreatic insufficiency, recurrent acute or chronic pancreatitis and CF-related liver disease (cirrhosis or liver disease with or without cirrhosis, including fatty liver).
Metabolic disturbances: insulin-dependent CF-related diabetes (CFRD) and at least one CF-related episode of dehydration requiring medical attention.
Others: bone disorders including low bone density, osteoporosis, and digital arthropathy.
For this work, the CF diagnosis of the studied population was contrasted with the responsible physician for the CF Regional Unit. In addition, the doctor provided the necessary information to collate and complete the different clinical manifestations and the genetic information of each of the people included.
Statistical analysis
We described the clinical and demographic variables in the two groups of genotypes established by hypothesis contrast test according to the type of variables and their normality. The normality test was carried out using the Kolmogorov-Smirnov test. The absolute and relative frequencies of the clinical and demographic variables were described. The allelic frequency of the CFTR gene variants in the studied population was also presented.
For the quantitative variables, Student's t-test was used if they were normally distributed and Mann Whitney's U-test if they were not. For qualitative variables, Chi-squared2 or Fisher test was used when applied.
Additionally, crude and adjusted odds (OR) and 95% confidence intervals (95%CI) were calculated using binary logistic regression analysis to examine associations between genotype and clinical manifestations of the participants. There was a significant statistically association between genotype and age at diagnosis and as of December 31, 2018 (P <0.01). Therefore, these variables were taken into account for the adjustment of the model.
In addition, a sensitivity analysis was performed to verify that the patients diagnosed by neonatal screening did not lead to biased.
All tests were two-tailed and the level of statistical significance was established at ≤0.05. Statistical analyzes were performed with the IBM SPSS 25.0 statistical package (IBM Corporation, Armonk, New York, USA).