Abstract
Most children with cystic fibrosis (CF) in current times are likely to be picked up through newborn screening programs and confirmed by the finding of two CF-causing mutations and a positive sweat test. They may be clinically well at the time of detection but likely to develop CF symptoms in time and will require prompt, effective intervention. Some children will be missed through the screening program, and pediatricians must remain vigilant to this possibility. No racial group is exempt from the disorder, and children of ethnic minorities or mixed heritage are at greatest risk of a delayed or a missed diagnosis. The clinical features of recurrent chest infections, malabsorption with pancreatic insufficiency in many but not all, salt losing syndromes, or an infant presenting with meconium ileus or rectal prolapse requires investigation. Recent consensus statements have provided useful guidance on the diagnostic criteria for CF.
One of the unintended consequences of CF screening programs has been the detection of children in whom the full diagnostic criteria are not met. This group is referred to by the acronym CFSPID/CRMS. These children may require monitoring over time and possibly reevaluation depending on their clinical course. The range of clinical features for individuals with 2 CFTR mutations is wide and some may not develop symptoms at all or have an incomplete phenotype—often with single organ disease. Currently, such individuals often present clinically as adults and are not necessarily given the full cystic fibrosis label but considered to have a “CFTR related disorder.”
Keywords
cystic fibrosis, diagnosis, sweat test, CFTR mutation analysis, CFTR function, CFSPID/CRMS, CFTR related disorders
A diagnosis of cystic fibrosis (CF) has lifelong implications for affected individuals, their families, and their acquaintances. This important step needs to be taken accurately and early on. A late diagnosis is often preceded by a catalogue of doctor’s visits, family anguish, and anger associated with a delay in the initiation of treatment that may have an impact on long-term outcome. Equally disturbing is a small but increasingly documented experience of children, diagnosed as having CF, whom on review—often years later—are found not to fulfill diagnostic criteria.
Many countries have now introduced newborn screening programs for the diagnosis of CF. This has changed the experience of diagnosing CF for both parents and pediatricians. Appropriate monitoring of the child’s condition and early therapy are introduced, often before the emergence of signs and symptoms traditionally associated with CF. Apart from the advantages of early detection through newborn screening, new challenges have arisen from the screening program, including:
- 1.
Uncertainties about the diagnostic label in some infants who, following a positive screening result, do not fulfill all the criteria for a CF diagnosis ;
- 2.
The recognition of an ever-widening phenotype for an individual with two cystic fibrosis transmembrane conductance regulator (CFTR) mutations, ranging from a traditional CF phenotype, a spectrum of atypical disease forms, and even individuals with no discernible CFTR dysfunction at all and no evidence of end organ disease at the time of assessment ;
- 3.
The recognition that a CF phenotype can emerge and change over time, presenting de novo in adulthood or moving from “atypical” forms to a more classical CF ;
- 4.
The need for pediatricians and adult physicians to remain vigilant in order to diagnose those individuals who may have been missed by the screening process or did not undergo screening.
In this chapter, we will describe the diagnostic criteria for CF, the diagnostic techniques available for assessing the CFTR function, the screening process, and the need to identify and label atypical cases—both screened and clinically detected.
Diagnostic Criteria for Cystic Fibrosis
In more than 90% of cases, the diagnosis of CF in an unscreened patient arises from a clinical suspicion supported by a test of CFTR function, the sweat test, with genetic testing for CFTR mutation used as confirmatory testing. A Cystic Fibrosis Foundation (CFF) Consensus Panel (United States) synthesized diagnostic criteria for CF. The key features are summarized in Box 50.1 . The basic premise of the consensus statement is that CF is a clinical and not a genetic diagnosis, although acknowledging that genetic testing may have a role in sorting out atypical clinical situations. Any such document must be considered a work in progress to accommodate new developments and acknowledged shortcomings. The latest version of the consensus guidelines from the CFF is now available online ( https://www-ncbi-nlm-nih-gov.easyaccess2.lib.cuhk.edu.hk/pubmed/28129811 ).
One or more characteristic phenotypic features consistent with cystic fibrosis (CF):
Chronic sinopulmonary disease
Gastrointestinal and nutritional abnormalities
Salt loss syndromes
Male urogenital abnormalities resulting in obstructive azoospermia
OR
A history of CF in a sibling
OR
A positive newborn screening test result
AND
an increased sweat chloride concentration
or identification of two disease causing CF mutations
or demonstration of abnormal nasal epithelial ion transport.
Making the Diagnosis of Cystic Fibrosis
A clinician may need to confirm a diagnosis of CF in various settings such as when
- 1.
A patient has one or more suspicious clinical features;
- 2.
A newborn screening program has identified a child at risk for CF;
- 3.
An individual is being examined after the diagnosis of CF in a family member; or
- 4.
Postnatal confirmation is required when an antenatal test has proven suspicious for CF.
Clinical Suspicion
The majority of unscreened children with CF present with a history of bulky fatty stools, failure to thrive, and recurrent chest infections. Shortly after birth, 10%–15% will present with intestinal obstruction as meconium ileus. However, there is a wide range of less common presenting features ( Box 50.2 ). Any children or adults presenting with suspicious features of CF should be investigated further, even if they have undergone newborn screening. A possible diagnosis of CF in a child with suggestive clinical findings should not be discounted just because the child appears too well or is thought to be too old. Although the diagnosis is established in most children by the age of 1 year—and earlier if screened—in approximately 10% of children the diagnosis is delayed until after 7 years of age. Patients with pancreatic sufficiency and non-Caucasian patients are particularly vulnerable to delays in diagnosis.
0–2 Years
Failure to thrive
Steatorrhea
Recurrent chest infections including bronchiolitis/bronchitis
Meconium ileus
Rectal prolapse
Edema/hypoproteinemia/“kwashiorkor” skin changes
Severe pneumonia/empyema
Salt depletion syndrome
Prolonged neonatal jaundice
Vitamin K deficiency with bleeding diathesis
3–16 Years
Recurrent chest infections or “asthma”
Clubbing and “idiopathic” bronchiectasis
Steatorrhea
Nasal polyps and sinusitis
Chronic intestinal obstruction, intussusception
Heat exhaustion with hyponatremia
CF diagnosis in a sibling
Adulthood (Often Considered CFTR-Related Disorder)
Azoospermia/congenital absence of the vas deferens
Bronchiectasis
Chronic sinusitis
Acute or chronic pancreatitis
Allergic bronchopulmonary aspergillosis
Focal biliary cirrhosis
Abnormal glucose tolerance
Portal hypertension
Cholestasis/gall stones
CF, Cystic fibrosis.
The Sweat Test
The sweat test was first described in 1959 and remains a gold standard for the diagnosis of CF. In the appropriate clinical setting, whether the child has entered the diagnostic algorithm via clinical suspicion or newborn screening, a positive sweat chloride test is diagnostic of CF. There are a number of other rare conditions (sometimes single case reports) that have been associated with a positive sweat test, but these are usually clearly distinguishable by their clinical features ( Box 50.3 ).
Adrenal insufficiency or stress
Anorexia nervosa
Ectodermal dysplasia
Eczema
Fucosidosis
G6PD deficiency
Glycogen storage disease type 1
Human immunodeficiency virus infection
Hypoparathyroidism
Hypothyroidism
Malnutrition from various causes
Nephrogenic diabetes insipidus
Pseudohypoaldosteronism
Chronic arsenic exposure
The standard sweat test (Gibson and Cooke technique) requires skill and care, and should be undertaken by accredited laboratories. Localized sweating is stimulated by the iontophoresis of pilocarpine into the skin. Sweat is collected on filter paper, gauze, or in microbore tubing over a controlled period of time to ensure that the rate of sweating and the total sweat collected are sufficient and standardized. Guidelines for sweat testing procedures and precautions are available in published and electronic format ( http://www.acb.org.uk/docs/sweat.pdf ).
Chloride is the analyte of choice. The sodium levels and osmolality of sweat are less reliable and should never be used in isolation. A sweat chloride concentration of more than 60 mmol/L is considered positive, and levels below 30 mmol/L are likely to be in the normal range. This lower limit of 30 mmol/L was previously set at 40 mmol/L in children older than 6 months, but recent consensus guidelines have revised to a chloride level of less than 30 mmol/L for all age groups. Results between 30 and 60 mmol/L have traditionally been considered intermediate and require further evaluation. A high proportion of patients with chloride concentrations in this intermediate range will have two CFTR mutations—usually one at least that is not defined as disease causing, a situation found with increasing frequency with the advent of more detailed CFTR mutation testing.
Following the introduction of newborn screening (NBS) for CF, sweat tests are increasingly performed on infants. It is recommended that sweat chloride testing in asymptomatic newborns with a positive NBS test be performed when the infant is older than 2 weeks of age and more than 2 kg. The vast majority of affected infants will have a chloride level greater than 60 mmol/L. A sweat chloride value between 30 and 59 mmol/L should be considered intermediate and trigger further patient evaluation, including repeat testing when older. Certain CFTR mutations (such as R117H, R334W, 3849+10kbC>T) associated with a milder CF phenotype may also demonstrate a normal sweat chloride.
Some normal adolescents and adults can have sweat chloride values in the intermediate range, and sweat chloride levels alone may be insufficient to diagnose CF in the older adolescent. A false-positive sweat test in the severely malnourished child or the critically ill child in intensive care needs cautious interpretation and follow-up.
Research continues to explore the development of an easier sweat test. Collecting systems such as the Macroduct and Nanoduct simplify collection and are now in routine practice. The role of sweat conductivity as a diagnostic tool in CF is debated. In one large trial, the best conductivity cutoff value to diagnose CF was ≥90 mmol/L, and the best conductivity cutoff value to exclude CF was less than 75 mmol/L. However, most clinicians and laboratories will chose to confirm a positive sweat conductivity result with a formal measurement of chloride concentration.
Mutation Analysis
The 1989 identification of the CF gene and the characterization of its protein product (CFTR) held the promise that the diagnostic dilemmas for the condition were over. If you had two CFTR mutations, you had CF—if not, you did not. Unfortunately it has not worked out that simply. Even though the presence of two mutations is very supportive of the diagnosis of CF in the appropriate clinical setting, two alterations in the gene that encodes CFTR does not necessarily mean that you will develop classic CF disease, as outlined later.
There are more than 2000 different CF gene mutations documented, although the clinical impact of these mutations is frequently uncharacterized. A valuable online resource lists identified mutations at http://www.genet.sickkids.on.ca/app .
Clinically available techniques do not routinely allow a full screen of the entire CF genome, and most laboratories will only search for the most common mutations within their geographical region—focusing on those known to be disease causing. Examples of disease-causing CFTR mutations are listed in Table 50.1 , highlighting the dominance of the ΔF508 mutation (present in >70% of CF alleles in Caucasian populations). Customizing mutation panels to match the patient’s ethnic background and clinical presentation can enhance the sensitivity of DNA testing in CF.
Traditional Name | HGVS b Nomenclature | Frequency (%) |
---|---|---|
ΔF508 | p.Phe3508del | 75.0 |
G551D | p.Gly551Asp | 3.4 |
G542X | p.gly542X | 1.8 |
R117H | p.arg117His | 1.3 |
621+1G>T | c.489+1G>T | 1.3 |
ΔI507 | p.Ile507del | 0.5 |
N1303K | p.Asn1303Lys | 0.5 |
R560T | p.arg560Thr | 0.4 |
Q493X | c.1477C>T | 0.3 |
R1162X | p.Arg1162X | 0.3 |
R533X | p.Arg553X | 0.3 |
W1282X | p.Trp1282X | 0.3 |
3659delC | c.3527_3528delC (p.Lys1177Serfs) | 0.3 |
1154insTC | c.1021_1022dup (p.Phe342HisfsX28) | 0.3 |
E60X | p.Glu60X | 0.2 |
G85E | p.Gly85Glu | 0.2 |
P67L | c.200C>T (p.Pro67Leu) | 0.2 |
R347P | p.Arg347Pro | 0.2 |
V520F | p.Val520Phe | 0.2 |
1078delT | c.946_947delT (p.Phe316Leufs) | 0.1 |
2184delA | c.2052_2053delA (p.Lys684Asnfs) | 0.1 |
A455E | p.Ala455Glu | 0.1 |
R334W | p.Arg334Trp | 0.1 |
S549N | p.Ser549Asn | 0.1 |
2789+5G>A | c.2657+5G>A | 0.1 |
3849+10kbC>T | c.3717+10kbC>T | 0.1 |
711+1G>T | c.579+1G>T | 0.1 |
1717-1G>T | c.1585-1G>A | 0.6 |
1898+1G>T | c.1766+1G>A | 0.6 |
a In Caucasian populations—variations in frequency occur between different ethnic groups and geographic regions.
Second, there is a range of mutation types that have been classified into classes according to the functional impact on the CFTR. The final protein product may be incomplete, complete but incorrectly packaged and processed, or a final CFTR molecule that is unstable or incapable of reaching the cell surface in sufficient quantity to be physiologically effective. This is discussed further in Chapter 49 .
In addition to the “disease-causing” mutations, there are also recognizable polymorphisms that do not necessarily result in a clinical phenotype but may influence the structure of the final protein product when associated with another mild mutation. The thymidine run in intron 8 is a well-described example where the 5T allele leads to a substantial reduction in functional protein compared with the 9T allele; the 7T allele is intermediate in its effect. The clinical phenotype associated with two CFTR mutations is far broader than previously anticipated ( Box 50.4 ). Caregivers should avoid making prognostic predictions based on genotype alone. Examples of the potential clinical impact of selected CFTR mutations is shown in Table 50.2 and can also be researched on the CFTR2 website ( http://www.cftr2.org/ ), which also regularly updates the growing list of disease causing mutations.
“Classical CF” with pancreatic insufficiency
Sinopulmonary disease, pancreatic sufficiency and positive sweat test
Sinopulmonary disease and male infertility with a normal sweat test
Severe sinusitis and congenital bilateral absence of the vas deferens
Male infertility only
Chronic pancreatitis only
Allergic bronchopulmonary aspergillosis
Sclerosing cholangitis
Positive sweat test only
No clinical features including normal sweat chloride.