Increasing life expectancy in cystic fibrosis: Advances and challenges

Based on data from 2019, the Cystic Fibrosis Foundation Registry Report from the United States calculated the predicted median survival age of a child born that year with cystic fibrosis to be 48.4 years¹, a remarkable achievement. Summarized data from the registries of the United Kingdom, Canada, Belgium, Europe, Australia, France and Ireland show a range of median survival age from 44 to 53 years, and median age at death ranging from 29 to 35.6 years². The greater longevity for patients with CF is not only a therapeutic triumph, but also a therapeutic challenge, as diseases and complications of aging now must be factored into the care of patients with CF. This article and this volume (in much greater depth than this article can provide) will address how the CF field achieved this greatly improved survival expectation, which factors have promoted it and which have not yet been appropriately addressed, and what the rapidly developing therapeutic strategies hold for the future. Since there is variability around the world in the timing and application of various advances, in this article we will concentrate on milestones in the United States.

Early days – initiation of aggressive care

When CF was first described in 1938³, it was a diagnosis made postmortem on infants who were less than 18 months of age. After World War II, the first antibiotics were available to the general public, and treated some of the lung infections in infants with CF, but with only modest consequences for survival. The heat wave in New York City in 1948 brought the discovery of the sweat defect in CF⁴, and as the sweat test came into use as a diagnostic tool, milder cases of CF were identified during life. Still, in 1954, the median survival age for patients with CF was estimated at 4-5 years. About that time, the first CF Centers were formed. The collection of many patients under the care of a few physicians provided them with the experience needed to manage the disease better. At the Cleveland Center at Rainbow Babies and Children’s Hospital, institution of a Comprehensive Care Program resulted in a remarkable increase in median survival age, to the mid-20s –simply by aggressively treating each complication as it arose^5,6. The concept of comprehensive care programs at Centers as a pillar of the clinical care of CF took root. By 1978 or so, the estimated median survival nationwide had climbed to 11 years.

As understanding of the pathobiology grew, therapeutic strategies were refined. Lung disease was managed by vigorous airway clearance and antibiotic treatment of pulmonary exacerbations. Improved nutrition was achieved not by avoidance of fatty foods (as was the early approach to pancreatic insufficiency, to avoid cramps), but by increasing enzyme doses to allow consumption of high-calorie, high-fat diets⁷. The fat-soluble vitamins A, D, E, and K were provided in excess and in water-miscible form⁸. Subsequent improvements in the formulation of pancreatic enzyme supplements with microencapsulation to allow them to escape destruction in the stomach also helped in digestion and absorption⁹. Some Centers improved the pH in the gut by suppressing stomach acid production, allowing the pancreatic enzyme supplements to work at closer to their optimal physiologic pH¹⁰. All of these interventions led to better nutrition and probably improved immune defense.

Improvements in pulmonary therapy and other complications

To manage the lung disease, institution of routine cultures of the sputum or the deep throat allowed tailoring of the antibiotic regimen to the specific organisms harbored by that patient. New antibiotics were being developed, particularly drugs active against gram-negative bacteria. This was especially important, because Pseudomonas aeruginosa emerged as the dominant organism recovered in the CF lung. Tobramycin was approved in 1975 by the FDA, and many other injectable antibiotics came to market, including semisynthetic penicillins and cephalosporins. The strategy of inhaled antibiotics was gradually more widely adopted, and some antibiotics were specifically formulated for aerosolization, which allowed delivery of high concentrations in the airways with greatly reduced systemic toxicity¹¹. The first of these, inhaled tobramycin, was approved in 1998. A major advance occurred with the development of oral antibiotics that were truly effective against P. aeruginosa, such as ciprofloxacin in 2004¹². Prevention of viral infection, especially immunization against measles and influenza, and the availability of anti-influenza drugs, was also protective for patients with CF.

Treatment of the excessive inflammatory response in patients 5-17 years of age with CF by high dose ibuprofen slowed the rate of decline of pulmonary function and improved life expectancy ^13,14,15. Airway clearance techniques improved so that a partner was not required as was true for classical postural drainage and percussion. Such mechanical devices as the high frequency chest wall oscillation (vest) devices and the positive pressure oscillatory exhalation devices (e.g. Flutter ™) which produced airway vibrations to shake loose adherent secretions greatly aided the ability to clear sputum¹⁶. Recombinant human DNase helped to liquefy the sputum to facilitate clearance¹⁷. Exercise was increasingly recognized as beneficial but only now is a multinational comprehensive study being undertaken¹⁸. Major complications of CF that can be fatal, such as massive hemoptysis and pneumothorax have seen better and more proactive treatment in recent years, partly due to the development and implementation of consensus practice guidelines¹⁹.

The leading cause of death in cystic fibrosis is respiratory failure. Early reports suggested that intubation and mechanical ventilation usually failed²⁰. However, more recently, over half of intubated and ventilated patients survived to discharge from the ICU^21,22 . In addition, a growing body of literature describes successful respiratory support of patients with non-invasive ventilation, which has been shown to stabilize lung function²³ and can serve as a bridge to transplantation. Organ transplantation provided another survival boost. Though the lung is among the most difficult organs to preserve for transplantation, techniques soon improved and survival following transplantation has improved to over 10 years today in some reports²⁴, though the complications of transplantation stand as the second leading cause of death in CF¹.

Liver disease is the third leading specified cause of death in CF in the US but only accounts for 3.2% of deaths¹. However, there are still no agreed-upon diagnostic criteria for this complication, and the only treatment is ursodeoxycholic acid, which has not been definitively shown to prolong life²⁵. Portal hypertension and esophageal varices can be lethal. Liver transplantation, sometimes in combination with lung transplantation, may be required.

Although women with CF are less fertile than other women of comparable age, pregnancies are common, and the outcomes are generally good. One recent series of 149 pregnancies in women with CF found no stillbirths or infant deaths, and no deaths or lung transplants among the mothers in the two years following delivery, even for women who began pregnancy with FEV1 less than 50% predicted²⁶. However, infants born to CF mothers with initial FEV1 less than 50% predicted were more likely to be premature (41.9% vs 28.7%) and had lower birth weight (average 2705 g vs 3044 g). Drugs which are teratogenic are not administered and those for which the data are equivocal are avoided All the infants will have one CF allele from the mother. If the father is positive for a CF allele, one in two of the offspring will be affected. While some series indicate that patients with CF lose pulmonary capacity following pregnancy, others suggest that the pulmonary function of women with CF who have had children actually improves following pregnancy²⁷. Attention to the details of pulmonary care and nutrition has improved the maternal outcomes of pregnancy and contributed to the overall improved survival.

With all of these adjustments, plus widespread adoption of the aggressive comprehensive care program, the median survival age crept steadily upwards, into the thirties and the forties¹. For the five year period 2010-2014, which centers on the year the first CFTR modulator was approved (2012), predicted median survival age for a child born with CF was 40 years. Even before drugs directed at CFTR were available, the quality of life had steadily improved. Oral anti-pseudomonas antibiotics allowed patients to stay out of the hospital more, as did regular immunization against respiratory pathogens. Patients were liberated from prior constraints such as restrictive diets and airway clearance techniques that required a human partner. Of course, patients still had to spend time on airway clearance daily, and they still took dozens of pills each day, but in general they could feel well enough to attend college and/or work, marry, and start a family. In 2012, the impact of therapies directed at the basic defect was still to come.

Ongoing challenges to CF survival

There were some substantial challenges to survival as well. In the early 1980’s a new organism, Burkholderia cepacia (first called Pseudomonas cepacia), was seen in CF populations³⁰. Its acquisition resulted in marked acceleration of the downhill course in CF, and, for a time, halted the upward trend of median survival age. However, conventional infection control measures, including separating patients known to be infected (Thomassen, 1986) and later avoiding contact among CF patients entirely, helped control the spread of this very difficult infection. Strategies for antibiotic therapy evolved, and the course became milder. This organism is still found in the population, in low prevalence ¹. More recently nontuberculous mycobacteria and fungi have been observed in the CF lung³¹. Fortunately these are rare, but difficult to treat. B. cepacia, mycobacteria, and fungi, however, are relative contraindications to some life-extending treatments such as transplantation in many centers and are negative prognostic signs. Low socioeconomic status is associated with poorer outcomes in most diseases, and CF is no exception. Patients supported by Medicaid have about four times the age adjusted death rate as those covered by other insurance, commensurately worse pulmonary function, and worse nutritional status³². This disparity occurs despite comparable number of physician visits. Another study which divided patients by household income showed that those in households with the lowest income had about 50% higher age adjusted mortality than those in the highest income category, despite apparently comparable prescriptions for common CF drugs³³. A later study used several measures of socioeconomic status (zip code, maternal education, state insurance coverage), all of which were inversely associated with severe disease, and the disparities were not explained by utilization of health services or prescription for chronic care³⁴. The effects of poverty are profound and pervasive.

CF is most prevalent in non-Hispanic whites but the prevalence of CF in minority racial and ethnic groups (Black, Hispanic whites, Asian, American Indian, and others) is increasing³⁵. After adjustment for socioeconomic status, minority racial and ethnic groups with CF suffer worse health outcomes. For example, Hispanic white and Black people with CF have poorer pulmonary function than their non-Hispanic white counterparts³⁶. In the Hispanic population specifically, a pulmonary function gap and higher mortality is present despite a higher body mass index (BMI), higher likelihood of pancreatic sufficiency, and the presence of more mild CFTR mutations than in non-Hispanic Whites³⁶. In California, patients who self-identify as Hispanic white have 2.8 times the age-adjusted mortality of those who identify as non-Hispanic white³⁷, and another national study found that Hispanic white patients have 1.27 times the mortality of non-Hispanic whites after adjusting for age, sex, genetic mutations, bacterial cultures, lung function, body mass index, use of CF respiratory therapies, low socioeconomic status, pancreatic enzyme use, and CF-related diabetes in all regions except the South³⁸. The explanation for the worse health outcomes in minorities with CF is unclear, but later diagnosis due to the insensitivity of the newborn screen to certain alleles, a lack of representation in pharmaceutical clinical trials, less eligibility for modulators, and racial bias in the healthcare system most likely play a part³⁹ . It will be important to identify and mitigate the factors responsible for racial disparity in healthcare outcomes.

Early diagnosis using newborn screening

A randomized study of newborns in Wisconsin provided the earliest, clearest evidence for the benefit of neonatal screening⁴⁰. Today, although the specific testing strategy varies state to state, all fifty states in in the U.S. offer newborn screening to identify infants with CF in the neonatal period. In the majority of states this screen combines a blood spot analysis for elevated levels of immunoreactive trypsin (IRT), a product released from a diseased pancreas, and reflexive screening for one or more mutant CF alleles. The diagnosis is then confirmed by a sweat test at approximately one month of age⁴¹. More than 6000 infants were followed in the U.S. CFF registry during the first 9 years of universal newborn screening, with the age at first evaluation for CF being <30 days for the majority of them. Few preventable complications of CF were reported at the first clinical encounter, although about 40% of infants had a weight for age <10%tile at that visit⁴². Patients identified at an early age can begin treatment: pancreatic enzyme and vitamin supplements to optimize nutrition, meticulous attention to treatment of the lung disease, and monitoring for liver disease and other complications, but risk infectious exposure in early visits to clinic. The cohort of patients identified by neonatal screening has improved nutritional status and growth compared to patients who are identified by symptoms later in life⁴³, and many studies, though not all, indicate improved pulmonary function in the screened group^,44. Median age at diagnosis is now under one month of age¹, and it is expected that earlier treatment results in improved survival.

Although newborn screening has brought improvements in outcomes for patients with CF, for every affected newborn, ten newborns are identified as carriers of a mutant CF allele⁴⁵. For each of these children and their parents, the false positive result leads to additional testing, anxiety, and in some cases, an incorrect or inconclusive diagnosis⁴⁶.

Newborn screening for CF has been implemented around the world in a fragmented fashion with variable protocols. The protocols most used in the U.S., testing for the 23 to 40 CFTR mutations most commonly responsible for CF in Caucasian and Ashkenazi Jews, do not include some of the alleles more common in infants of African, Hispanic, Asian, and Middle Eastern ancestry⁴⁷. CF is being recognized more frequently among diverse ethnic groups, with the spectrum and frequency of CFTR mutations varying by ethnic group and geographic location^35,48. A study of the US CF registry demonstrated that those with Hispanic, Black, or Asian ancestry were less likely to have two identified CFTR variants, and were more likely to carry no mutations on the commonly used 23 mutation carrier screening⁴⁸. Moreover, a study comparing the genotypes of people included in the US CF registry to the most commonly used newborn screening panels demonstrated that 13-16% of Hispanic, 15-18% of Black, 23-28 % of Asian, and 7- 8% of Native Americans would be missed compared to only 3% of non- Hispanic Whites⁴⁷. The delay in diagnosis caused by failure to identify CF on the newborn screen could be an important prognostic factor, as children diagnosed due to symptoms rather than by NBS are diagnosed later (14.5 months of age as opposed to 2 weeks), and have increased complications including hospitalization, pseudomonas colonization, and deceased lung function^49,50 . To reduce the racial bias in screening protocols based on genetic testing, a strategy has been proposed that couples IRT with next generation sequencing (NGS), allowing for rapid simultaneous detection of a larger number of known pathogenic mutations, though still not all^47,51. A recent study provided evidence of the feasibility of NGS, as well as the increased sensitivity, specificity and positive predictive value of screening with this method⁵².

In summary, newborn screening has improved the outcome of infants diagnosed with cystic fibrosis by allowing early intervention, particularly nutritional support. Unfortunately, the most common newborn screening algorithms introduce racial inequality in diagnosis, preventing every infant from experiencing the benefit of early treatment. This inequality may contribute in part to the disparate outcomes noted previously. Recognition of the limitations of current screening methods and the introduction of NGS are important first steps to improving the sensitivity of newborn screening.

Uncertain CF diagnoses

Among the more than one thousand mutations in the CF gene that can cause disease, there is great variation in the amount of residual CFTR function they entrain. This variation creates a continuum of disease expression and severity, which may complicate the diagnosis of CF disease and skew survival data. For example, men with isolated congenital bilateral absence of the vas deferens may have no or mild pulmonary disease and abnormal sweat chloride, but otherwise might not have been diagnosed with CF based on symptoms. This variability in phenotypic expression can further complicate newborn screening as well. Infants may be identified as having an intermediate sweat chloride, suggesting inadequate CFTR function, but only have one identified CFTR mutation. Alternatively, an infant may be identified by newborn screening with two CFTR mutations, only one of which is known to cause disease, but have a sweat chloride of <30mm/L. Both pose diagnostic challenges and are therefore classified as CFTR-Related Metabolic Syndrome (CRMS) or, in Europe,• Cystic Fibrosis Screen Positive, Indeterminate Diagnosis” (CFSPID), and are followed closely to monitor for evolving signs and symptoms of CF⁵³. Inclusion of patients with mild forms of CF or borderline diagnoses among the CF population will improve apparent CF survival, whereas their exclusion will result in apparently poorer survival. Understanding the inclusion and exclusion criteria defining CF is essential for interpretation of CF survival statistics.

The advent of CFTR modulator therapy

An enormous step toward achieving normalcy for people with CF has been the development of small molecule therapies directed at the abnormal CFTR protein and referred to as modulators. Modulators target restoring or improving CFTR function and are targeted at different functional classification of the CFTR mutation. This therapeutic strategy was a significant advance from prior symptomatic therapies; rather it addressed the actual defect in chloride transport. The first modulator, the CFTR potentiator ivacaftor (IVA), was introduced in 2012 for patients with at least one copy of the G551D or other Class III gating mutations, and has now been extended for use for over 90 CFTR mutations. The initial 48 week phase III trial showed significant sustained effects of IVA treatment including improved lung function (mean absolute change in percentage of predicted FEV₁ of 10.6 percentage points), body weight increase of 2.7 kg, a decrease in sweat chloride of 48.1 mmol/L), respiratory quality of life measured by questionnaire was better, and the frequency of pulmonary exacerbations fell by 55% compared to placebo⁵⁴. However, only 6% of people with CF have at least one copy of a gating mutation.

The most common mutation in the CF population is F508del, a processing mutation in which the vast majority of the abnormal CFTR protein is destroyed prior to reaching the epithelial cell surface, and the channel protein that does reach the cell surface fails to open appropriately. Therefore, modulators developed for F508del combine a potentiator such as IVA with corrector(s) of CFTR folding and transport lumicaftor (LUM) or tezacaftor (TEZ). These combination drugs made CFTR modulation available to an additional 45% of patients with CF homozygous for F508del, though their clinical impact was not as profound as ivacaftor on patients with gating mutations, with a mean absolute improvement in percentage of predicted FEV₁ compared to placebo of only 2.8 percentage points (LUM/IVA) ⁵⁵ and 4.0 percentage points (TEZ/IVA)⁵⁶ . Most recently, the triple combination of CFTR potentiator IVA with two CFTR correctors TEZ and elexacaftor (ELX) (which work at different steps in the protein folding process) in people with a single F508del allele was found to significantly improve CFTR function as measured by sweat chloride, and resulted in significant improvement in percentage of predicted FEV₁ of 14.3 percentage points, BMI and patient reported quality of life, as well as a 63% lower rate of pulmonary exacerbations relative to placebo⁵⁷. Even patients with advanced lung disease, including candidates for lung transplantation, have benefited from the introduction of this latest modulator therapy⁵⁸. Elexacaftor-tezacaftor-ivacaftor can provide a highly effective modulator treatment regimen for more than 85% of the people living with cystic fibrosis world- wide.

The full impact of modulators on the health and longevity of those with cystic fibrosis is unclear due to the limited long term data at this time. The acute increase in FEV₁ and BMI, as well as symptom improvement, are important short term outcomes, but the rate of decline of lung function over time is more closely associated with survival ^59,60. Ivacaftor-treated G551D patients, compared to propensity-matched F508del patients displayed reduced rate of FEV₁ decline by nearly half⁶¹. A more recent study of G551D patients (from the GOAL-e2 study) also indicated that there was ongoing, albeit slower, disease progression⁶². Trials of LUM/IVA (PROGRESS) and TEZ/IVA (EXTEND) demonstrated that despite only modest acute increases in FEV₁, both modulator combinations also significantly reduced FEV₁ rate of decline compared to matched controls^63,64. A recent study examined the effect of ivacaftor on survival (among other outcomes) using data from the US and UK cystic fibrosis registries. Ivafactor-treated patients not only had had better preserved lung function, but a significantly lower risk of death and lung transplantation when compared to an untreated matched cohort⁶⁵. More time is needed to evaluate the ultimate benefit of elexacaftor-tezacaftor-ivacaftor on lung disease progression and survival.

Although CFTR modulators decrease the rate of pulmonary function decline, they do not arrest it. Disease progression continues, in part, due to ongoing infection and inflammation, as well as pulmonary exacerbations, superimposed on structural damage in the lung. Although pulmonary exacerbations decrease with modulator therapy, they are not eliminated. Ivacaftor treatment reduces the frequency of pulmonary exacerbation, but does not improve the rate of complete lung function recovery after pulmonary exacerbation when compared with placebo⁶⁶. For patients who have already sustained structural damage in the airways, many with frank bronchiectasis, alveolar simplification and cysts, structural changes do not reverse and the predisposition to infections remains. It remains to be seen whether modulator treatment in young children with healthy lungs will alter the pattern of infection and pulmonary exacerbation and prevent structural damage.

Extra-pulmonary comorbidities of CF and modulator therapy

In addition to the effects of modulators on pulmonary function, pulmonary exacerbations, and quality of life, there is growing evidence that CFTR modulators affect other complications of cystic fibrosis. A recent review summarizes evidence of the beneficial effect of IVA on weight and growth, pancreatic function, gastrointestinal and hepatobiliary systems, sinus disease, bone disease, exercise tolerance, fertility, mental health, and immunity⁶⁷. However, the author points out that the impact of modulator therapy on in distal intestinal obstructive syndrome (DIOS), nephrolithiasis, CF related arthropathy, and the post-transplant course have yet to be described.

The improvement in exocrine pancreatic function described in children age 1–5 years^68–70 after IVA treatment, as well as several case reports describing the reversal of documented pancreatic insufficiency⁷¹, including in older children^72,73 taking IVA, suggest that pancreatic damage can be reversible. Furthermore, a recent case report describes a fetus homozygous for F508del exposed to ELX/TEZ/IVA during the entire pregnancy and subsequently born with a normal fecal pancreatic elastase and a normal newborn screen⁷⁴. The birth of this child supports the theory that in utero exposure to mutation-appropriate modulators may be able to restore CFTR function in order to prevent the pancreatic damage usually present at birth. This has led to speculation that in utero exposure to a highly effective modulator might also preserve the vas deferens in males with CF, preventing infertility.

Limitations of modulator therapies

Although the modulation of CFTR has been shown to improve pulmonary function, BMI, patient-reported respiratory signs and symptoms, and to decrease pulmonary exacerbations, they have not proven to be the hoped-for cure. For most patients, the disease continues to progress, albeit much more slowly. Drawbacks to lifelong medical treatment with CFTR modulators include adverse effects, significant monetary cost, and the concern about possible consequences of medication withdrawal. Adverse effects from modulators include hepatic dysfunction, abdominal pain, headache, rash, and neuropsychiatric events, among others⁷⁵. Some of them are sufficiently severe to warrant drug discontinuation. Sudden discontinuation of modulator therapy may result in unusually severe pulmonary exacerbation, and besides adverse effects, factors such as high cost, complexity of navigating specialty drug pharmacies, and lapses in insurance coverage may result in interruption of treatment⁷⁶. Moreover, modulators are mutation specific: patients with premature stop codons who produce no CFTR are not benefitted. Only an estimated 85-90% of patients with CF in the US are both eligible for modulators and able to tolerate them, and the proportion may be even less in many other countries where the mutation profile is different. Even in the United States, there are racial and ethnic disparities in eligibility for CFTR modulators, with 92.4% of non-Hispanic White patients, 69.7% of Black/African American patients, 75.6% of Hispanic patients, and 80.5% of other race patients eligible for CFTR modulators⁷⁷. These differences in formal eligibility may delay approval of payment by insurance carriers, but many can likely be overcome by in vitro testing of the mutant or analogy to approved mutants. Nevertheless, mutation-agnostic therapies are urgently needed for those who do not carry a modulator responsive mutation, or who cannot tolerate modulator therapy due to side effects. Although these new therapies have changed the lives of many living with CF, the quest for additional definitive therapies must continue. A better understanding of factors leading to continued decline in lung function and the side effects that limit modulator use, and identification of mutation agnostic therapies are critically important.

Looking forward

Great progress has been made in improving the longevity and quality of life of patients with CF over the eighty years since the disease was first described. Aggressive, comprehensive plans of care, team-based care, development of specific drugs and approaches such as exercise to treat the pulmonary disease of CF and improve nutritional status, newborn screening for early diagnosis to allow maximum impact of therapies early in life, better management of complications including respiratory failure, better management of pregnancy, improvements in lung transplantation, and engagement of families in the care plan all contributed to the remarkable improvement in CF longevity to about 40 years, even before the release of drugs aimed at the basic defect. Still, inequities that impact outcomes are seen for racial and ethnic minorities and also by income, which must be addressed in CF as they must for other aspects of health care in the US. All of these issues will be discussed later in this volume. Now, with CFTR modulators available, the last decade has seen acceleration in these improvements. Though the use of some of these drugs has not yet been extended to children or infants, who might be predicted to be the most likely to benefit from them long term, there has still been substantial improvement in quality of life and increase in projected longevity of about 8 years⁷⁸. It is likely that the less damage sustained by the organs prior to starting treatment, the better the preservation of function. Therefore, beginning CFTR modulator therapy earlier, and even considering in utero intervention, might well provide even greater improvements for patients with CF.

However, existing drugs may not be sufficient. First, there are mutation classes that are not affected by any of the current drugs on the market, especially those with “stop” codons. Second, some patients cannot tolerate current modulator therapy. Third, adverse effects and/or cost may become prohibitive with lifelong therapy. And fourth, the current state of CFTR modulators is that they do not obviate the need for other aspects of CF care, and the disease continues to progress, albeit at a slower rate. For all those reasons many investigators are exploring additional pharmacotherapies, improved methods for gene therapy that would be allele-agnostic, and gene correction using CRISPR-CAS. There is already a report of successful in vivo treatment of Transthyretin Amyloidosis⁷⁹ using CRISPR-CAS. If the entire defective CF gene could be replaced at a site shortly after the initiation codon, all mutations would be corrected. If this could be done early in life, the chances of physiologic normalcy would greatly increase. The field has come a long way since CF was first identified, first by assiduous treatment of complications and now CFTR modulator therapy, but we still have a road ahead to address the basic defect completely and restore normalcy to our patients.