- Email: [email protected]
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy Tyler Groves, MSc1,2, Paul Robinson, MBChB, FRACP, PhD2,3, Veronica Wiley, FHGSA, FFSc (RCPA), PhD3,4, and Dominic A. Fitzgerald, FRACP, PhD2,3 Objective To describe the clinical course of children who have intermediate sweat chloride values on initial screening for cystic fibrosis (CF).
Study design We performed a retrospective review of children with intermediate sweat chloride values (raised immunoreactive trypsinogen/1 copy of p.F508del CF mutation on newborn screening (NBS)/sweat chloride value of 30-59 mmol/L) presenting to The Children’s Hospital at Westmead over 15 years. Patients with an intermediate sweat chloride evolving to a formal diagnosis of CF (termed “delayed CF”) were matched (2:1) with NBS positive patients with CF (termed “NBS positive CF”). Clinical outcomes were compared. Results Fourteen of 29 (48%, 95% CI 0.3-0.66) patients with intermediate sweat chloride value evolved to a diagnosis of CF and were matched with 28 NBS positive patients with CF. Delayed CF had less pancreatic insufficiency (OR 0.06, 95% CI 0.01-0.44, P = .006), less colonization with nonmucoid Pseudomonas aeruginosa (OR 0.04, 95% CI 0.01-0.38, P = .005), milder obstructive lung disease (forced expiratory volume in 1 second/forced vital capacity ratio), and overall disease severity (Shwachman scores) at 10 years (mean difference 5.93, 95% CI 0.39-11.46, P = .04; mean difference 4.72, 95% CI 0.9-8.53, P = .015, respectively). Nutritional outcomes were better at 2 years for delayed CF but did not persist to later ages. Conclusions In this cohort, approximately one-half of infants with intermediate sweat chloride value were later diagnosed with CF. The clinical course of delayed CF was milder in some aspects compared with NBS positive CF. These results emphasize the importance of ongoing follow-up of infants with intermediate sweat chloride values. (J Pediatr 2015;166:1469-74). See editorial, p 1337
T
he introduction of newborn screening (NBS) for cystic fibrosis (CF) over the past 30 years has led to earlier diagnosis and treatment of CF with improved clinical outcomes.1-4 NBS has also created a cohort of infants and children who fail to fulfill formal CF diagnostic criteria for sweat chloride values,5 but lie outside the established normal range for sweat test results (defined as sweat chloride values <30 mmol/L). At present, a sweat test is indicated if an elevated immunoreactive trypsinogen (IRT) level is found on NBS with genetic testing showing only 1 copy of a known disease causing CF gene mutation (in our case p.F508del; ie, p.F508del heterozygote). Sweat chloride concentrations >60 mmol/L are confirmatory, whereas values between 30 and 59 mmol/L form a challenging diagnostic category.6,7 These patients who also lack 2 CF-causing mutations have historically been described as an intermediate sweat chloride values cohort, although more recently the term CF transmembrane conductance regulator-related metabolic syndrome (CRMS) has been introduced.8 Despite a reported prevalence of 3%-4% of infants falling into this intermediate category,7 data on long-term outcome is limited. Subsequent clinical course is a spectrum from no/minimal ongoing clinical concern to symptomatic CF transmembrane conductance regulator-related disorders (eg, bronchiectasis, pancreatitis, male infertility), and even later to a formal diagnosis of CF.8,9 A delayed diagnosis of CF is estimated to occur in 8%-15% of these cohorts,10,11 but the later clinical course of these infants is again unclear and may be characterized by misdiagnosis with other respiratory conditions, such as asthma.12 The group with a delayed diagnosis of CF, compared with patients with a classic diagnosis of CF at NBS, was presumed more likely to be pancreatic sufficient, have a milder phenotype, and more favorable clinical course. Ren et al included only 1 child who tracked from a suspected diagnosis of CF (CRMS) to a formal diagnosis of CF over a 6-year period with patients that had sweat chloride values undertaken at the discretion of clinicians.11 Consequently, clinicians have limited From the Sydney Medical School, University of Sydney; 1
2
CF CHW CRMS IRT NBS
Cystic fibrosis The Children’s Hospital at Westmead CF transmembrane conductance regulator-related metabolic syndrome Immunoreactive trypsinogen Newborn screening
Department of Respiratory Medicine, The Children’s Hospital at Westmead; 3Discipline of Pediatrics and Child Health, Sydney Medical School, University of Sydney; and 4New South Wales Newborn Screening Program, The Children’s Hospital at Westmead, New South Wales, Australia The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Crown Copyright ª 2015 Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.01.052
1469
THE JOURNAL OF PEDIATRICS
www.jpeds.com
information on which to base prognostic discussions about the likelihood of evolution to a formal diagnosis of CF with parents, or indeed the likely disease severity of infants with a “delayed CF” diagnosis. Moreover, there is little data available to determine frameworks for follow-up of infants with intermediate sweat chloride values. The aim of this study was to compare the clinical course of children with intermediate sweat chloride values identified at a tertiary pediatric center, The Children’s Hospital at Westmead (CHW), over the past 15 years (1996-2010) who progressed to a CF diagnosis in comparison with matched NBS positive infants with CF.
Methods Patients were identified by screening all sweat test results, from tests performed at CHW (Sydney, New South Wales, housing the New South Wales reference laboratory for NBS), from 1996-2010 inclusive, and subsequent CF mutation analyses in those with intermediate sweat test results. Institutional Ethics Committee approval was obtained for the study (QI Ethics No. QIE-2012-09-14, Activity No. 3982). On average, 100 000 infants are screened per year, with a CF incidence of approximately 1:3000 live births, resulting in 25-30 new CF cases per year in NSW. Our center, as 1 of 3 tertiary pediatric hospitals in NSW, typically accounted for approximately 10-15 of these new patients per year over the study period examined, of which 5%-10% are historically missed by NBS.13 During the 15-year study period, the established New South Wales NBS protocol was an IRT/DNA-based system: IRT levels >99th percentile (fixed cut-off) on the screening date had subsequent DNA mutation analysis performed for the p.F508del mutation (accounts for 70%-75% of mutations in the Australian population with CF with 94% of infants diagnosed on NBS having p.F508del mutation as at least 1 of their alleles).13-15 Patients fulfilling all the following criteria were eligible for inclusion: (1) an elevated NBS IRT level; (2) heterozygous for the p.F508del mutation; and (3) an intermediate initial sweat test (sweat chloride value 30-59 mmol/L).2,10 Sweat chloride levels were determined using the method established by Gibson and Cooke throughout the entire study period.16 A retrospective medical chart review was performed of all patients with an intermediate sweat chloride value who progressed to a delayed formal CF diagnosis, termed “delayed CF.” Delayed CF was defined as a diagnostic follow-up sweat test, evidence of pancreatic insufficiency, and/or a respiratory clinical course consistent with CF (ie, recurrent respiratory infections requiring regular follow-up and physiotherapy at the CHW CF Clinic). These patients with delayed CF were then matched with classical NBS-detected patients with CF, defined as sweat chloride value $60 mmol/L and/or p.F508del heterozygous or homozygous, and termed “NBS positive CF.” Patients with CF presenting with meconium ileus were excluded from the NBS positive CF cohort, based on these patients being recognized as a more severe CF 1470
Vol. 166, No. 6 phenotype.17 Matching was performed based on sex and age (within 6 months) in 2:1 of NBS positive CF cases to delayed CF cases. Medical data were collected for all treatment episodes at CHW prior to February 28, 2014. Outcomes assessed included IRT levels on NBS, extended genetic mutation analysis (Table I; available at www.jpeds. com), gastrointestinal tests and lung function tests performed, nutritional outcome (weight and height centiles), episodes of respiratory tract bacterial isolation, total clinic visits, hospital admissions, intravenous antibiotic treatment home courses, and annual Shwachman scores (a CF global severity assessment score).18 Technically acceptable spirometry results were included closest to 5 and 10 years. Lung function results were expressed as percent predicted values and z-scores using recent existing reference equations.19 Sputum cultures were obtained as clinically indicated prior to the formal CF diagnosis, but once incorporated into the CF clinic, they were performed at each review (typically every 3 months from formal CF diagnosis). For specific organisms, Staphylococcus aureus and Pseudomonas aeruginosa, time to first isolation, and time to colonization (defined as 3 consecutive positive sputum samples within any 12-month period, or commencement of regular antipseudomonal nebulized therapy) were recorded. Exocrine pancreatic sufficiency was assessed by 3-day fecal fat collection and was performed in all patients following a formal CF diagnosis or where clinically suspected. Statistical analyses were performed using SAS v 9.0 for Windows (SAS Institute, Cary, North Carolina). Continuous variables were expressed as mean and SD for parametric data and median and IQR for nonparametric data. Categorical variables were expressed as frequency and percentage. No adjustment was made for multiple statistical testing. To account for the matching of cases to controls, conditional logistic regression was used to estimate the effect of variables present at diagnosis on case/control status. For variables not present at diagnosis, a generalized estimating equation model was used assuming normal distribution for continuous variables, Poisson distributions for counts, and binomial distributions for binary variables. An unstructured correlation matrix was assumed to describe the correlation within case-control triplets. Only statistically significant results are reported, defined as P < .05.
Results Over the 15-year period, 29 patients with an intermediate sweat chloride value were identified, of which 14/29 (48%) became patients with delayed CF: diagnostic sweat chloride level on follow-up (2/14, 14%), proven pancreatic insufficiency (4/14, 29%) with a 3-day fecal fat collection, and/or experiencing recurrent pseudomonal or staphylococcal lower respiratory tract infections (8/14, 57%). Over the study period, 3/29 (10.3%) were lost to follow-up in our clinic, and all were from the “delayed CF” group. Two transferred to a different CF center, and 1 moved interstate. One further patient was lost to follow-up at the 2-year time Groves et al
ORIGINAL ARTICLES
June 2015 point but returned to the clinic subsequently (Figure 1; available at www.jpeds.com). The proportion providing follow-up data at age 2, 5, and 10 years was 13/14 (93%), 12/14 (87%), and 10/14 (72%), respectively. One child with delayed CF was aged between 5 and 10 years at the completion of data collection and so had no data available for the 10-year review. Demographic information for the delayed CF and matched NBS positive CF cohorts are shown in Table II. Significant differences were present between the cohorts: patients with delayed CF had lower initial sweat chloride values (as expected based on the criteria used for cohort identification) (mean difference 48.3, 95% CI 42.1-54.6, P < .0001) and a lower rate of pancreatic insufficiency (OR 0.06, 95% CI 0.01-0.44, P = .006). Individual clinical features for the delayed CF cohort are shown in Table III. The median age (range) of CF diagnosis in patients with delayed CF was 0.19 (0.11-4.76) years. Repeat sweat tests in patients with delayed CF were performed in 8/14 patients (57%), a median of 1.0 (1.0-3.0) occasion. Of the repeat sweat tests performed, 42% were completed by 6 months, 67% by 12 months, and 83% by 5 years of age. Increasing sweat chloride levels to the diagnostic range were observed in 4/8 patients, whereas 2 other patients showed increasing levels. Two sweat tests (16.7%) by 6 months of age were diagnostic of delayed CF, and 2 (16.7%) further at 1 and 13 years of age that supported the patient’s prior clinical delayed CF diagnosis. A similar number of sweat tests were performed in the remaining patients with an intermediate sweat chloride value (n = 15) who did not progress to delayed CF: repeated in 10/15 (67%) on a median of 1.5 (1.0-3.0) occasions. Of the repeat sweat tests performed, 56% were repeated by 6 months, 81% by 12 months, and 94% by 5 years of age. Longitudinally, all sweat chloride values decreased over time to below 30 mmol/L. All 15 patients were discharged from formal respiratory follow-up and had no further medical entries
Table II. Patient characteristics between patients with delayed CF and NBS positive CF control patients
Sex (male) IRT Genotype Homozygous p.F508del Heterozygous p.F508del Initial sweat chloride level (mmol/L) Age at CF diagnosis (y) Pancreatic sufficient at diagnosis Basis for CF diagnosis Diagnostic sweat chloride level Proven pancreatic insufficient Recurring respiratory symptoms
Delayed CF (n = 14)
NBS positive CF (n = 28)
7 (50) 147.6 114.3
14 (50) 175.5 75.9
0 (0) 14 (100) 45.5 8.1
17 (60.7) 11 (39.3) 92.4 14.6*
0.19 (0.10-4.76) 9 (64.3)
0.12 (0.07-0.31) 1 (3.5)†
2 (14.3) 4 (28.6) 8 (57.1)
28 (100) 0 (0) 0 (0)
Data are presented as mean SD, number of patients (percentage), or median (range). *Statistically significant difference between groups (mean difference 48.3, 95% CI 42.1-54.6, P < .0001). †Statistically significant difference between groups (OR 0.06, 95% CI 0.01-0.44, P = .006).
recorded in the system, nor had they presented to other CF clinics in Australia (according to the Australian National CF Registry). Extended mutation analysis was completed in all patients with delayed CF (Table III). Seven (50%) patients carried p.F508del/unknown mutations, 4 (29%) carried p.F508del/R117H mutations, and 1 (7%) subject each for p.F508del/S549N, p.F508del/621+1G>T, and p.F508del/ 1078dT mutations. At the time of diagnosis for the 4 patients with p.F508del/R117H, the additional testing for polyT status was not sought as these patients were diagnosed by recurrent respiratory tract infections. The specific mutations that have been screened for within the extended mutation panels in NSW over the screening period can be seen in Table I. The only difference in lung function detected between the cohorts at either 5 or 10 years of age was in the forced expiratory volume in 1 second/forced vital capacity ratio at 10 years: mean difference (95% CI) 5.9% (0.39-11.46) greater in patients with delayed CF (P = .04) (Figure 2; available at www.jpeds.com). Nutritional outcomes were also similar between cohorts at all time points, apart from greater body mass index and weight z-scores in delayed CF patients at 2 years of age: mean difference (95% CI) 0.43 (0.01-0.86, P = .047) and 0.54 (0.17-0.92, P = .005), respectively (Table IV; available at www.jpeds.com). This difference was not sustained at later time points. Sputum cultures were performed on fewer occasions in the delayed CF vs NBS positive CF cohort: mean (SD) 40.3 (30.1) vs 58.0 (23.1) occasions (P = .02). High isolation rates for Saureus were observed in both cohorts; however, a significantly lower incidence of nonmucoid P aeruginosa isolation (OR 0.04, 95% CI 0.01-0.38, P = .005) and colonization (OR 0.04, 95% CI 0.01-0.38, P = .005) occurred in the delayed CF vs NBS positive cohorts (Table V). No difference was seen for mucoid P aeruginosa isolation rates between groups. No difference in age of first isolation for any organism was found. No differences were observed for clinic visits, intravenous antibiotic courses, and total number of hospital admission days at any of the time points assessed. The degree of separation in Shwachman scores between the cohorts increased over time, and reached statistical significance at 10 years of age, with delayed CF a mean difference (95% CI) 4.7 (0.9-8.5) points greater vs NBS positive CF (P = .015) (Figure 3).
Discussion NBS has created an increasing population of infants with initial intermediate sweat chloride values, and evidencebased recommendations for management and follow-up of these patients is required. This study provides insight into the clinical course of these infants, and in particular the subsequent clinical course of those infants with a delayed CF diagnosis. Almost one-half of patients with intermediate sweat chloride values progressed to a formal CF diagnosis. This is
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1471
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 166, No. 6
Table III. Clinical features of patients with delayed CF Sex
IRT
Initial sweat chloride (mmol/L) (age, y)
Follow-up sweat (age, y)
Age at diagnosis (y)
Basis for CF diagnosis
1
Male
94
32 (0.17)
33 (0.75) 55 (2.33)
2.25
PI
p.F508del/unknown
1999
2 3 4
Male Male Male
107 89 64
52 (0.72) 57 (0.24) 35 (0.15)
0.72 0.24 4.76
Resp Resp Resp
p.F508del/R117H p.F508del/R117H p.F508del/unknown
1999 1999 2002
5 6 7 8 9 10 11 12 13 14
Male Male Male Male Female Female Female Female Female Female
81 160 48 106 105 232 120 216 144 500
54 (0.11) 41 (0.35) 38 (0.12) 41 (0.15) 43 (0.14) 43 (0.1) 45 (0.09) 45 (0.14) 58 (0.14) 53 (0.11)
0.11 0.38 0.12 0.24 0.14 0.1 0.09 0.14 0.61 0.11
SwCl SwCl Resp Resp Resp PI Resp PI Resp PI
p.F508del/unknown p.F508del/unknown p.F508del/unknown p.F508del/S549N p.F508del/R117H p.F508del/unknown p.F508del/R117H p.F508del/621+1G>T p.F508del/unknown p.F508del/1078dT
1999 2002 2002 1999 1999 1999 2002 2002 2002 2005
Patient
45 (0.26) 29 (4.27) 35 (14.1) 63 (0.15) 86 (0.38) 80 (1.14) 77 (12.68) 37 (0.36) 47 (0.35)
Gene mutation
Extended mutation panel screened*
PI, pancreatic insufficiency; Resp, recurrent respiratory tract infections; SwCl, diagnostic sweat chloride level. *See Table I for mutation panel details.
higher than the 2 previous studies, which reported rates of 8% and 13% in a similar population with intermediate sweat chloride values.10,11 The subsequent clinical course of these patients with delayed CF is similar to matched NBS positive patients with CF, but not in all outcomes. These patients with delayed CF had lower rates of pancreatic insufficiency, nonmucoid P aeruginosa isolation and colonization, less obstructive lung disease (based on percent predicted forced expiratory volume in 1 second/forced vital capacity ratios), and less severe disease (based on Shwachman scores) at 10 years of age. Benefits seen in nutritional outcomes at 2 years were not sustained at later time points. The CRMS classification is difficult to apply to this cohort as it specifies 2 subsequent intermediate sweat tests by 6 months of age.8 All patients with an intermediate sweat chloride value in this study were identified prior to that publication and received repeat sweat tests only when clinically
indicated and, thus, cannot strictly be classified as having CRMS. For this reason, the patients in our study have been termed patients with intermediate sweat chloride values. Ren et al described the clinical course of 12 patients with CRMS identified between 2002 and 2010, but on closer inspection, direct comparison with our data may be justified.11 Patients were retrospectively identified as CRMS using the Borowitz et al diagnostic criteria, despite patient screening preceding publication of the guidelines and follow-up sweat testing also performed at the discretion of the treating physician, rather than at repetitive early time points. Only 1 patient (8%) was subsequently diagnosed with CF over 6 years of
Table V. Sputum colonization and isolation for patients with delayed CF and NBS positive CF control patients Delayed CF, n = 14 NBS positive CF, n = 28 Staphylococcus aureus Isolation No isolation Age of first isolation (y) Chronic colonization P aeruginosa (nonmucoid) Isolation No isolation Age of first isolation (y) Chronic colonization P aeruginosa (mucoid) Isolation No isolation Age of first isolation (y) Chronic colonization
12 (85.7) 2 (14.3) 2.45 3.74 11 (78.6)
42 (100) 0 (0) 2.15 1.63 27 (96.4)
8 (57.1) 6 (42.9) 2.99 3.54 1 (7.1)
27 (96.4)* 1 (3.6) 4.51 3.54 10 (35.7)†
3 (21.4) 11 (78.6) 9.29 3.73 2 (14.3)
8 (28.6) 20 (71.4) 7.33 3.48 2 (7.1)
Data are presented as mean SD, or number of patients (percentage). *Statistically significant difference between groups (OR 0.04, 95% CI 0.01-0.38, P = .005). †Statistically significant difference between groups (OR 0.04, 95% CI 0.01-0.38, P = .005).
1472
Figure 3. Comparison of total Shwachman score between the delayed CF cohort (gray bars) and the NBS positive CF cohort (striped bars). Data are presented as mean SD. * Statistically significant difference between cohorts (mean difference 4.72, 95% CI 0.9-8.53, P = .015). P = NS for all comparisons. Groves et al
ORIGINAL ARTICLES
June 2015 follow-up. Massie and Clements described the clinical course of 27 infants with intermediate sweat chloride values from the Victorian NBS Program in Australia over a 10-year period (1989-1998).10 Only 4 (15%) patients had a delayed CF diagnosis, although the length of follow-up was not clearly described. Current literature, therefore, may underestimate the potential likelihood of formal CF diagnosis in the indeterminate sweat chloride value cohorts. The oldest formal CF diagnosis in our cohort occurred at 5 years of age, although 86% (12/14) occurred by 2 years of age. This emphasizes the importance of ongoing follow-up beyond the first year in these patients. One suggestion was for an annual follow-up for infants with intermediate sweat chloride values in infancy, but the length of follow-up was not specified.8 The data presented here may in fact underestimate true risk of delayed CF diagnosis, as complete follow-up data were not available. The lung function outcome, describing values within the low normal range at 5 and 10 years of age, is consistent with other studies.11,20,21 A lesser degree of obstructive lung disease at the end of follow-up was observed. The patients with delayed CF did not have sustained improved nutritional outcomes compared with those identified by NBS. This may reflect benefits of modern treatment in NBS positive patients with CF, with nutritional benefits reported in the existing literature.22 Ren et al reported normal nutritional indices for all patients11; however, abnormal height and weight z-scores (defined as < 2 SDs) were recorded in 2 (14%) and 1 (7%) patients with delayed CF in our cohort throughout follow-up, respectively. Determining the clinical relevance of these lung function and nutritional findings is difficult, and should not be over-interpreted given the relatively small cohort numbers and potential type II statistical error. Rates of nonmucoid P aeruginosa isolation and colonization in the delayed CF cohort were both lower than NBS positive patients with CF. This is clinically relevant as nonmucoid P aeruginosa is an important pathogen in CF with significant detrimental effects on morbidity and mortality.23 Rates of P aeruginosa colonization were similar to those reported by Ren et al (20% vs 25%, respectively).11 Isolation and colonization rates of mucoid P aeruginosa, an additional adverse prognostic factor,1,24-26 however, were similar between cohorts in our study at final follow-up at 10 years of age. One confounding factor is that NBS positive patients with CF received more frequent sputum cultures. Patients with delayed CF received sputum cultures when clinically indicated, or at minimum, each annual review. Finally, data suggested less severe overall CF disease with increasing difference in Shwachman scores, which became significant at the end of follow-up in the delayed CF cohort. The major limitation of this study is the relatively small cohort. This may have resulted in an analysis underpowered to detect clinically relevant differences and potential type II statistical error. Other limitations include the retrospective design and IRT/p.F508del screening approach. This approach is based on previous cost analysis and was used throughout the 15 years of screening in NSW.13 It has been reported that even with improved sensitivity of a larger mutation analysis, most infants
would still be detected using the screening program used throughout this study, and adding a test for the next most common mutation likely results in the detection of only 1 extra case every 3 years.13,27 Finally, 3 patients with delayed CF were lost to follow-up, and this may have led to biasing of the results. However, it is unclear if the clinical course of these patients differed from the rest of the cohort. In conclusion, this study provides important prognostic information for physicians managing infants with intermediate sweat chloride values. Our high rate of subsequent formal CF diagnosis reinforces the need for close follow-up in this population. n The authors are grateful to Liz Barnes (Biostatistician, The Children’s Hospital at Westmead) for her generous assistance with the statistical analyses. The authors are also grateful to the Department of Molecular Genetics (The Children’s Hospital at Westmead), including Bruce Bennetts, Gemma Jenkins, Edwina Middleton, and Felicity Collins, for their assistance with the extended genetic mutation panel and newborn screening at this institution. The annual checks performed for this study were a collaborative effort by all staff on the CF team at CHW. Submitted for publication Aug 8, 2014; last revision received Jan 15, 2015; accepted Jan 28, 2015. Reprint requests: Tyler Groves, MSc, Sydney Medical School, University of Sydney, 105/2 Dind St, Milsons Point, New South Wales, 2061, Australia. E-mail: [email protected]
References 1. Dijk FN, Fitzgerald DA. The impact of newborn screening and earlier intervention on the clinical course of cystic fibrosis. Paediatr Respir Rev 2012;13:220-5. 2. Farrell PM, Kosorok MR, Laxova A, Shen G, Koscik RE, Bruns WT, et al. Nutritional benefits of neonatal screening for cystic fibrosis. Wisconsin Cystic Fibrosis Neonatal Screening Study Group. N Engl J Med 1997; 337:963-9. 3. Farrell PM, Kosorok MR, Rock MJ, Laxova A, Zeng L, Lai HC, et al. Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal Screening Study Group. Pediatrics 2001;107:1-13. 4. Koscik RL, Farrell PM, Kosorok MR, Zaremba KM, Laxova A, Lai HC, et al. Cognitive function of children with cystic fibrosis: deleterious effect of early malnutrition. Pediatrics 2004;113:1549-58. 5. Dijk FN, McKay K, Barzi F, Gaskin KJ, Fitzgerald DA. Improved survival in cystic fibrosis patients diagnosed by newborn screening compared to a historical cohort from the same center. Arch Dis Child 2011;96:1118-23. 6. Moskowitz SM, Chmiel JF, Sternen DL, Cheng E, Gibson RL, Marshall SG, et al. Clinical practice and genetic counseling for cystic fibrosis and CFTR-related disorders. Genet Med 2008;10:851-68. 7. Parad RB, Comeau AM. Diagnostic dilemmas resulting from the immunoreactive trypsinogen/DNA cystic fibrosis newborn screening algorithm. J Pediatr 2005;147:S78-82. 8. Borowitz D, Parad RB, Sharp JK, Sabadosa KA, Robinson KA, Rock MJ, et al. Cystic Fibrosis Foundation practice guidelines for the management of infants with cystic fibrosis transmembrane conductance regulatorrelated metabolic syndrome during the first two years of life and beyond. J Pediatr 2009;155:S106-16. 9. Farrell PM, Rosenstein BJ, White TB, Accurso FJ, Castellani C, Cutting GR, et al. Guidelines for diagnosis of cystic fibrosis in newborns through older adults: Cystic Fibrosis Foundation consensus report. J Pediatr 2008;153:S4-14.
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1473
THE JOURNAL OF PEDIATRICS
www.jpeds.com
10. Massie J, Clements B. Diagnosis of cystic fibrosis after newborn screening: the Australasian experience—twenty years and five million babies later: a consensus statement from the Australasian Paediatric Respiratory Group. Pediatr Pulmonol 2005;39:440-6. 11. Ren CL, Desai H, Platt M, Dixon M. Clinical outcomes in infants with cystic fibrosis transmembrane conductance regulator (CFTR) related metabolic syndrome. Pediatr Pulmonol 2011;46:1079-84. 12. Fitzgerald D, Van Asperen P, Henry R, Waters D, Freelander M, Wilson M, et al. Delayed diagnosis of cystic fibrosis in children with a rare genotype (delta F508/R117H). J Paediatr Child Health 1995;31:168-71. 13. Wilcken B, Wiley V, Sherry G, Bayliss U. Neonatal screening for cystic fibrosis: a comparison of two strategies for case detection in 1.2 million babies. J Pediatr 1995;127:965-70. 14. Bobadilla JL, Macek M Jr, Fine JP, Farrell PM. Cystic fibrosis: a worldwide analysis of CFTR mutations—correlation with incidence data and application to screening. Hum Mutat 2002;19:575-606. 15. Clancy JP, Rowe SM, Accurso FJ, Aitken ML, Amin RS, Ashlock MA, et al. Results of a phase IIa study of VX-809, an investigational CFTR corrector compound, in subjects with cystic fibrosis homozygous for the F508del-CFTR mutation. Thorax 2012;67:12-8. 16. Gibson LE, Cooke RE. A test for concentration of electrolytes in sweat in cystic fibrosis of the pancreas utilizing pilocarpine by iontophoresis. Pediatrics 1959;23:545-9. 17. Evans AK, Fitzgerald DA, McKay KO. The impact of meconium ileus on the clinical course of children with cystic fibrosis. Eur Respir J 2001;18: 784-9. 18. Shwachman H, Kulczycki LL. Long-term study of one hundred five patients with cystic fibrosis; studies made over a five- to fourteen-year period. AMA J Dis Child 1958;96:6-15.
1474
Vol. 166, No. 6 19. Stanojevic S, Wade A, Stocks J, Hankinson J, Coates AL, Pan H, et al. Reference ranges for spirometry across all ages: a new approach. Am J Respir Crit Care Med 2008;177:253-60. 20. Desmarquest P, Feldmann D, Tamalat A, Boule M, Fauroux B, Tournier G, et al. Genotype analysis and phenotypic manifestations of children with intermediate sweat chloride test results. Chest 2000;118: 1591-7. 21. Sermet-Gaudelus I, Girodon E, Sands D, Stremmler N, Vavrova V, Deneuville E, et al. Clinical phenotype and genotype of children with borderline sweat test and abnormal nasal epithelial chloride transport. Am J Respir Crit Care Med 2010;182:929-36. 22. McKay KO, Waters DL, Gaskin KJ. The influence of newborn screening for cystic fibrosis on pulmonary outcomes in New South Wales. J Pediatr 2005;147:S47-50. 23. Davies JC. Pseudomonas aeruginosa in cystic fibrosis: pathogenesis and persistence. Paediatr Respir Rev 2002;3:128-34. 24. de Vrankrijker AM, Wolfs TF, van der Ent CK. Challenging and emerging pathogens in cystic fibrosis. Paediatr Respir Rev 2010;11:246-54. 25. Treggiari MM, Rosenfeld M, Retsch-Bogart G, Gibson R, Ramsey B. Approach to eradication of initial Pseudomonas aeruginosa infection in children with cystic fibrosis. Pediatr Pulmonol 2007;42:751-6. 26. West SE, Zeng L, Lee BL, Kosorok MR, Laxova A, Rock MJ, et al. Respiratory infections with Pseudomonas aeruginosa in children with cystic fibrosis: early detection by serology and assessment of risk factors. JAMA 2002;287:2958-67. 27. Massie J, Curnow L, Glazner J, Armstrong D, Francis I. Lessons learned from 20 years of newborn screening for cystic fibrosis. Med J Aust 2012; 196:67-70.
Groves et al
June 2015
ORIGINAL ARTICLES
Figure 1. Flow diagram of patients with intermediate sweat chloride, with follow-up of the delayed CF and matched NBS positive CF control cohorts over diagnosis, 2, 5, and 10 years of age. F/U, follow-up.
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1474.e1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 166, No. 6
Figure 2. Comparison of lung function parameters at A, 5 and B, 10 years of age between the delayed CF cohort (gray bars; n for A = 13, n for B = 13) and the NBS positive CF cohort (striped bars; n for A = 28, n for B = 28). FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; FEF25-75, forced expiratory flow rate in the middle half of FVC. Values are presented as mean SD.*Statistically significant difference between cohorts (mean difference 5.93, 95% CI 0.39-11.46, P = .04). P = not significant for all comparisons.
1474.e2
Groves et al
ORIGINAL ARTICLES
June 2015
Table I. Extended genetic mutation panel used by clinical genetics at CHW Year
Mutations tested
1999 2002
Added:
2005
Added: Removed: Added:
2013
F508del V520F R560T 2183AA>G S549R 3659delC 3876delA Y122X F508C A559T S1255X I148T
Removed:
I507del G542X S549N 2789+5G>A R347P 3905insT 3120+1G>A Q493X I507V 1898+5G>T
621+1G>T G551D W1282X 1898+1G>A R334W Y122X 394delTT 3849+4G>A I506V 2307insA
R117H R553X 1717-1G>A 711+1G>T 1078delTT Q493X 2184delA
A455E N1303K 3849+10kbC>T G85E R1162X 3849+4G>A I148T
Y122X Y1092X
R347H M1101K
Table IV. Nutritional parameter Z-scores for patients with delayed CF and NBS positive CF control patients Weight Delayed CF Age (y)
n
Diagnosis 2 5 10
13 13 13 13
z-score 0.02 1.13 0.52 0.65* 0.15 0.83 0.08 1.09
Height NBS positive CF
n 28 28 28 28
z-score 0.47 0.95 0.02 0.91* 0.13 0.83 0.56 0.92
Delayed CF n 11 13 13 13
z-score 0.09 1.45 0.11 1.08 0.35 0.98 0.04 1.18
Body mass index NBS positive CF
n 28 28 28 28
z-score 0.4 1.01 0.13 0.92 0.08 0.89 0.45 1.03
Delayed CF n N/A 13 13 13
z-score 0.66 0.91† 0.1 0.71 0.03 0.93
NBS positive CF n 28 28 28
z-score 0.28 0.66† 0.06 0.75 0.35 0.81
N/A, not applicable. Data are presented as mean SD. *Statistically significant difference between cohorts (mean difference 0.54, 95% CI 0.17-0.92, P = .005). †Statistically significant difference between cohorts (mean difference 0.43, 95% CI 0.01-0.86, P = .047).
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1474.e3
Study design We performed a retrospective review of children with intermediate sweat chloride values (raised immunoreactive trypsinogen/1 copy of p.F508del CF mutation on newborn screening (NBS)/sweat chloride value of 30-59 mmol/L) presenting to The Children’s Hospital at Westmead over 15 years. Patients with an intermediate sweat chloride evolving to a formal diagnosis of CF (termed “delayed CF”) were matched (2:1) with NBS positive patients with CF (termed “NBS positive CF”). Clinical outcomes were compared. Results Fourteen of 29 (48%, 95% CI 0.3-0.66) patients with intermediate sweat chloride value evolved to a diagnosis of CF and were matched with 28 NBS positive patients with CF. Delayed CF had less pancreatic insufficiency (OR 0.06, 95% CI 0.01-0.44, P = .006), less colonization with nonmucoid Pseudomonas aeruginosa (OR 0.04, 95% CI 0.01-0.38, P = .005), milder obstructive lung disease (forced expiratory volume in 1 second/forced vital capacity ratio), and overall disease severity (Shwachman scores) at 10 years (mean difference 5.93, 95% CI 0.39-11.46, P = .04; mean difference 4.72, 95% CI 0.9-8.53, P = .015, respectively). Nutritional outcomes were better at 2 years for delayed CF but did not persist to later ages. Conclusions In this cohort, approximately one-half of infants with intermediate sweat chloride value were later diagnosed with CF. The clinical course of delayed CF was milder in some aspects compared with NBS positive CF. These results emphasize the importance of ongoing follow-up of infants with intermediate sweat chloride values. (J Pediatr 2015;166:1469-74). See editorial, p 1337
T
he introduction of newborn screening (NBS) for cystic fibrosis (CF) over the past 30 years has led to earlier diagnosis and treatment of CF with improved clinical outcomes.1-4 NBS has also created a cohort of infants and children who fail to fulfill formal CF diagnostic criteria for sweat chloride values,5 but lie outside the established normal range for sweat test results (defined as sweat chloride values <30 mmol/L). At present, a sweat test is indicated if an elevated immunoreactive trypsinogen (IRT) level is found on NBS with genetic testing showing only 1 copy of a known disease causing CF gene mutation (in our case p.F508del; ie, p.F508del heterozygote). Sweat chloride concentrations >60 mmol/L are confirmatory, whereas values between 30 and 59 mmol/L form a challenging diagnostic category.6,7 These patients who also lack 2 CF-causing mutations have historically been described as an intermediate sweat chloride values cohort, although more recently the term CF transmembrane conductance regulator-related metabolic syndrome (CRMS) has been introduced.8 Despite a reported prevalence of 3%-4% of infants falling into this intermediate category,7 data on long-term outcome is limited. Subsequent clinical course is a spectrum from no/minimal ongoing clinical concern to symptomatic CF transmembrane conductance regulator-related disorders (eg, bronchiectasis, pancreatitis, male infertility), and even later to a formal diagnosis of CF.8,9 A delayed diagnosis of CF is estimated to occur in 8%-15% of these cohorts,10,11 but the later clinical course of these infants is again unclear and may be characterized by misdiagnosis with other respiratory conditions, such as asthma.12 The group with a delayed diagnosis of CF, compared with patients with a classic diagnosis of CF at NBS, was presumed more likely to be pancreatic sufficient, have a milder phenotype, and more favorable clinical course. Ren et al included only 1 child who tracked from a suspected diagnosis of CF (CRMS) to a formal diagnosis of CF over a 6-year period with patients that had sweat chloride values undertaken at the discretion of clinicians.11 Consequently, clinicians have limited From the Sydney Medical School, University of Sydney; 1
2
CF CHW CRMS IRT NBS
Cystic fibrosis The Children’s Hospital at Westmead CF transmembrane conductance regulator-related metabolic syndrome Immunoreactive trypsinogen Newborn screening
Department of Respiratory Medicine, The Children’s Hospital at Westmead; 3Discipline of Pediatrics and Child Health, Sydney Medical School, University of Sydney; and 4New South Wales Newborn Screening Program, The Children’s Hospital at Westmead, New South Wales, Australia The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Crown Copyright ª 2015 Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.01.052
1469
THE JOURNAL OF PEDIATRICS
www.jpeds.com
information on which to base prognostic discussions about the likelihood of evolution to a formal diagnosis of CF with parents, or indeed the likely disease severity of infants with a “delayed CF” diagnosis. Moreover, there is little data available to determine frameworks for follow-up of infants with intermediate sweat chloride values. The aim of this study was to compare the clinical course of children with intermediate sweat chloride values identified at a tertiary pediatric center, The Children’s Hospital at Westmead (CHW), over the past 15 years (1996-2010) who progressed to a CF diagnosis in comparison with matched NBS positive infants with CF.
Methods Patients were identified by screening all sweat test results, from tests performed at CHW (Sydney, New South Wales, housing the New South Wales reference laboratory for NBS), from 1996-2010 inclusive, and subsequent CF mutation analyses in those with intermediate sweat test results. Institutional Ethics Committee approval was obtained for the study (QI Ethics No. QIE-2012-09-14, Activity No. 3982). On average, 100 000 infants are screened per year, with a CF incidence of approximately 1:3000 live births, resulting in 25-30 new CF cases per year in NSW. Our center, as 1 of 3 tertiary pediatric hospitals in NSW, typically accounted for approximately 10-15 of these new patients per year over the study period examined, of which 5%-10% are historically missed by NBS.13 During the 15-year study period, the established New South Wales NBS protocol was an IRT/DNA-based system: IRT levels >99th percentile (fixed cut-off) on the screening date had subsequent DNA mutation analysis performed for the p.F508del mutation (accounts for 70%-75% of mutations in the Australian population with CF with 94% of infants diagnosed on NBS having p.F508del mutation as at least 1 of their alleles).13-15 Patients fulfilling all the following criteria were eligible for inclusion: (1) an elevated NBS IRT level; (2) heterozygous for the p.F508del mutation; and (3) an intermediate initial sweat test (sweat chloride value 30-59 mmol/L).2,10 Sweat chloride levels were determined using the method established by Gibson and Cooke throughout the entire study period.16 A retrospective medical chart review was performed of all patients with an intermediate sweat chloride value who progressed to a delayed formal CF diagnosis, termed “delayed CF.” Delayed CF was defined as a diagnostic follow-up sweat test, evidence of pancreatic insufficiency, and/or a respiratory clinical course consistent with CF (ie, recurrent respiratory infections requiring regular follow-up and physiotherapy at the CHW CF Clinic). These patients with delayed CF were then matched with classical NBS-detected patients with CF, defined as sweat chloride value $60 mmol/L and/or p.F508del heterozygous or homozygous, and termed “NBS positive CF.” Patients with CF presenting with meconium ileus were excluded from the NBS positive CF cohort, based on these patients being recognized as a more severe CF 1470
Vol. 166, No. 6 phenotype.17 Matching was performed based on sex and age (within 6 months) in 2:1 of NBS positive CF cases to delayed CF cases. Medical data were collected for all treatment episodes at CHW prior to February 28, 2014. Outcomes assessed included IRT levels on NBS, extended genetic mutation analysis (Table I; available at www.jpeds. com), gastrointestinal tests and lung function tests performed, nutritional outcome (weight and height centiles), episodes of respiratory tract bacterial isolation, total clinic visits, hospital admissions, intravenous antibiotic treatment home courses, and annual Shwachman scores (a CF global severity assessment score).18 Technically acceptable spirometry results were included closest to 5 and 10 years. Lung function results were expressed as percent predicted values and z-scores using recent existing reference equations.19 Sputum cultures were obtained as clinically indicated prior to the formal CF diagnosis, but once incorporated into the CF clinic, they were performed at each review (typically every 3 months from formal CF diagnosis). For specific organisms, Staphylococcus aureus and Pseudomonas aeruginosa, time to first isolation, and time to colonization (defined as 3 consecutive positive sputum samples within any 12-month period, or commencement of regular antipseudomonal nebulized therapy) were recorded. Exocrine pancreatic sufficiency was assessed by 3-day fecal fat collection and was performed in all patients following a formal CF diagnosis or where clinically suspected. Statistical analyses were performed using SAS v 9.0 for Windows (SAS Institute, Cary, North Carolina). Continuous variables were expressed as mean and SD for parametric data and median and IQR for nonparametric data. Categorical variables were expressed as frequency and percentage. No adjustment was made for multiple statistical testing. To account for the matching of cases to controls, conditional logistic regression was used to estimate the effect of variables present at diagnosis on case/control status. For variables not present at diagnosis, a generalized estimating equation model was used assuming normal distribution for continuous variables, Poisson distributions for counts, and binomial distributions for binary variables. An unstructured correlation matrix was assumed to describe the correlation within case-control triplets. Only statistically significant results are reported, defined as P < .05.
Results Over the 15-year period, 29 patients with an intermediate sweat chloride value were identified, of which 14/29 (48%) became patients with delayed CF: diagnostic sweat chloride level on follow-up (2/14, 14%), proven pancreatic insufficiency (4/14, 29%) with a 3-day fecal fat collection, and/or experiencing recurrent pseudomonal or staphylococcal lower respiratory tract infections (8/14, 57%). Over the study period, 3/29 (10.3%) were lost to follow-up in our clinic, and all were from the “delayed CF” group. Two transferred to a different CF center, and 1 moved interstate. One further patient was lost to follow-up at the 2-year time Groves et al
ORIGINAL ARTICLES
June 2015 point but returned to the clinic subsequently (Figure 1; available at www.jpeds.com). The proportion providing follow-up data at age 2, 5, and 10 years was 13/14 (93%), 12/14 (87%), and 10/14 (72%), respectively. One child with delayed CF was aged between 5 and 10 years at the completion of data collection and so had no data available for the 10-year review. Demographic information for the delayed CF and matched NBS positive CF cohorts are shown in Table II. Significant differences were present between the cohorts: patients with delayed CF had lower initial sweat chloride values (as expected based on the criteria used for cohort identification) (mean difference 48.3, 95% CI 42.1-54.6, P < .0001) and a lower rate of pancreatic insufficiency (OR 0.06, 95% CI 0.01-0.44, P = .006). Individual clinical features for the delayed CF cohort are shown in Table III. The median age (range) of CF diagnosis in patients with delayed CF was 0.19 (0.11-4.76) years. Repeat sweat tests in patients with delayed CF were performed in 8/14 patients (57%), a median of 1.0 (1.0-3.0) occasion. Of the repeat sweat tests performed, 42% were completed by 6 months, 67% by 12 months, and 83% by 5 years of age. Increasing sweat chloride levels to the diagnostic range were observed in 4/8 patients, whereas 2 other patients showed increasing levels. Two sweat tests (16.7%) by 6 months of age were diagnostic of delayed CF, and 2 (16.7%) further at 1 and 13 years of age that supported the patient’s prior clinical delayed CF diagnosis. A similar number of sweat tests were performed in the remaining patients with an intermediate sweat chloride value (n = 15) who did not progress to delayed CF: repeated in 10/15 (67%) on a median of 1.5 (1.0-3.0) occasions. Of the repeat sweat tests performed, 56% were repeated by 6 months, 81% by 12 months, and 94% by 5 years of age. Longitudinally, all sweat chloride values decreased over time to below 30 mmol/L. All 15 patients were discharged from formal respiratory follow-up and had no further medical entries
Table II. Patient characteristics between patients with delayed CF and NBS positive CF control patients
Sex (male) IRT Genotype Homozygous p.F508del Heterozygous p.F508del Initial sweat chloride level (mmol/L) Age at CF diagnosis (y) Pancreatic sufficient at diagnosis Basis for CF diagnosis Diagnostic sweat chloride level Proven pancreatic insufficient Recurring respiratory symptoms
Delayed CF (n = 14)
NBS positive CF (n = 28)
7 (50) 147.6 114.3
14 (50) 175.5 75.9
0 (0) 14 (100) 45.5 8.1
17 (60.7) 11 (39.3) 92.4 14.6*
0.19 (0.10-4.76) 9 (64.3)
0.12 (0.07-0.31) 1 (3.5)†
2 (14.3) 4 (28.6) 8 (57.1)
28 (100) 0 (0) 0 (0)
Data are presented as mean SD, number of patients (percentage), or median (range). *Statistically significant difference between groups (mean difference 48.3, 95% CI 42.1-54.6, P < .0001). †Statistically significant difference between groups (OR 0.06, 95% CI 0.01-0.44, P = .006).
recorded in the system, nor had they presented to other CF clinics in Australia (according to the Australian National CF Registry). Extended mutation analysis was completed in all patients with delayed CF (Table III). Seven (50%) patients carried p.F508del/unknown mutations, 4 (29%) carried p.F508del/R117H mutations, and 1 (7%) subject each for p.F508del/S549N, p.F508del/621+1G>T, and p.F508del/ 1078dT mutations. At the time of diagnosis for the 4 patients with p.F508del/R117H, the additional testing for polyT status was not sought as these patients were diagnosed by recurrent respiratory tract infections. The specific mutations that have been screened for within the extended mutation panels in NSW over the screening period can be seen in Table I. The only difference in lung function detected between the cohorts at either 5 or 10 years of age was in the forced expiratory volume in 1 second/forced vital capacity ratio at 10 years: mean difference (95% CI) 5.9% (0.39-11.46) greater in patients with delayed CF (P = .04) (Figure 2; available at www.jpeds.com). Nutritional outcomes were also similar between cohorts at all time points, apart from greater body mass index and weight z-scores in delayed CF patients at 2 years of age: mean difference (95% CI) 0.43 (0.01-0.86, P = .047) and 0.54 (0.17-0.92, P = .005), respectively (Table IV; available at www.jpeds.com). This difference was not sustained at later time points. Sputum cultures were performed on fewer occasions in the delayed CF vs NBS positive CF cohort: mean (SD) 40.3 (30.1) vs 58.0 (23.1) occasions (P = .02). High isolation rates for Saureus were observed in both cohorts; however, a significantly lower incidence of nonmucoid P aeruginosa isolation (OR 0.04, 95% CI 0.01-0.38, P = .005) and colonization (OR 0.04, 95% CI 0.01-0.38, P = .005) occurred in the delayed CF vs NBS positive cohorts (Table V). No difference was seen for mucoid P aeruginosa isolation rates between groups. No difference in age of first isolation for any organism was found. No differences were observed for clinic visits, intravenous antibiotic courses, and total number of hospital admission days at any of the time points assessed. The degree of separation in Shwachman scores between the cohorts increased over time, and reached statistical significance at 10 years of age, with delayed CF a mean difference (95% CI) 4.7 (0.9-8.5) points greater vs NBS positive CF (P = .015) (Figure 3).
Discussion NBS has created an increasing population of infants with initial intermediate sweat chloride values, and evidencebased recommendations for management and follow-up of these patients is required. This study provides insight into the clinical course of these infants, and in particular the subsequent clinical course of those infants with a delayed CF diagnosis. Almost one-half of patients with intermediate sweat chloride values progressed to a formal CF diagnosis. This is
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1471
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 166, No. 6
Table III. Clinical features of patients with delayed CF Sex
IRT
Initial sweat chloride (mmol/L) (age, y)
Follow-up sweat (age, y)
Age at diagnosis (y)
Basis for CF diagnosis
1
Male
94
32 (0.17)
33 (0.75) 55 (2.33)
2.25
PI
p.F508del/unknown
1999
2 3 4
Male Male Male
107 89 64
52 (0.72) 57 (0.24) 35 (0.15)
0.72 0.24 4.76
Resp Resp Resp
p.F508del/R117H p.F508del/R117H p.F508del/unknown
1999 1999 2002
5 6 7 8 9 10 11 12 13 14
Male Male Male Male Female Female Female Female Female Female
81 160 48 106 105 232 120 216 144 500
54 (0.11) 41 (0.35) 38 (0.12) 41 (0.15) 43 (0.14) 43 (0.1) 45 (0.09) 45 (0.14) 58 (0.14) 53 (0.11)
0.11 0.38 0.12 0.24 0.14 0.1 0.09 0.14 0.61 0.11
SwCl SwCl Resp Resp Resp PI Resp PI Resp PI
p.F508del/unknown p.F508del/unknown p.F508del/unknown p.F508del/S549N p.F508del/R117H p.F508del/unknown p.F508del/R117H p.F508del/621+1G>T p.F508del/unknown p.F508del/1078dT
1999 2002 2002 1999 1999 1999 2002 2002 2002 2005
Patient
45 (0.26) 29 (4.27) 35 (14.1) 63 (0.15) 86 (0.38) 80 (1.14) 77 (12.68) 37 (0.36) 47 (0.35)
Gene mutation
Extended mutation panel screened*
PI, pancreatic insufficiency; Resp, recurrent respiratory tract infections; SwCl, diagnostic sweat chloride level. *See Table I for mutation panel details.
higher than the 2 previous studies, which reported rates of 8% and 13% in a similar population with intermediate sweat chloride values.10,11 The subsequent clinical course of these patients with delayed CF is similar to matched NBS positive patients with CF, but not in all outcomes. These patients with delayed CF had lower rates of pancreatic insufficiency, nonmucoid P aeruginosa isolation and colonization, less obstructive lung disease (based on percent predicted forced expiratory volume in 1 second/forced vital capacity ratios), and less severe disease (based on Shwachman scores) at 10 years of age. Benefits seen in nutritional outcomes at 2 years were not sustained at later time points. The CRMS classification is difficult to apply to this cohort as it specifies 2 subsequent intermediate sweat tests by 6 months of age.8 All patients with an intermediate sweat chloride value in this study were identified prior to that publication and received repeat sweat tests only when clinically
indicated and, thus, cannot strictly be classified as having CRMS. For this reason, the patients in our study have been termed patients with intermediate sweat chloride values. Ren et al described the clinical course of 12 patients with CRMS identified between 2002 and 2010, but on closer inspection, direct comparison with our data may be justified.11 Patients were retrospectively identified as CRMS using the Borowitz et al diagnostic criteria, despite patient screening preceding publication of the guidelines and follow-up sweat testing also performed at the discretion of the treating physician, rather than at repetitive early time points. Only 1 patient (8%) was subsequently diagnosed with CF over 6 years of
Table V. Sputum colonization and isolation for patients with delayed CF and NBS positive CF control patients Delayed CF, n = 14 NBS positive CF, n = 28 Staphylococcus aureus Isolation No isolation Age of first isolation (y) Chronic colonization P aeruginosa (nonmucoid) Isolation No isolation Age of first isolation (y) Chronic colonization P aeruginosa (mucoid) Isolation No isolation Age of first isolation (y) Chronic colonization
12 (85.7) 2 (14.3) 2.45 3.74 11 (78.6)
42 (100) 0 (0) 2.15 1.63 27 (96.4)
8 (57.1) 6 (42.9) 2.99 3.54 1 (7.1)
27 (96.4)* 1 (3.6) 4.51 3.54 10 (35.7)†
3 (21.4) 11 (78.6) 9.29 3.73 2 (14.3)
8 (28.6) 20 (71.4) 7.33 3.48 2 (7.1)
Data are presented as mean SD, or number of patients (percentage). *Statistically significant difference between groups (OR 0.04, 95% CI 0.01-0.38, P = .005). †Statistically significant difference between groups (OR 0.04, 95% CI 0.01-0.38, P = .005).
1472
Figure 3. Comparison of total Shwachman score between the delayed CF cohort (gray bars) and the NBS positive CF cohort (striped bars). Data are presented as mean SD. * Statistically significant difference between cohorts (mean difference 4.72, 95% CI 0.9-8.53, P = .015). P = NS for all comparisons. Groves et al
ORIGINAL ARTICLES
June 2015 follow-up. Massie and Clements described the clinical course of 27 infants with intermediate sweat chloride values from the Victorian NBS Program in Australia over a 10-year period (1989-1998).10 Only 4 (15%) patients had a delayed CF diagnosis, although the length of follow-up was not clearly described. Current literature, therefore, may underestimate the potential likelihood of formal CF diagnosis in the indeterminate sweat chloride value cohorts. The oldest formal CF diagnosis in our cohort occurred at 5 years of age, although 86% (12/14) occurred by 2 years of age. This emphasizes the importance of ongoing follow-up beyond the first year in these patients. One suggestion was for an annual follow-up for infants with intermediate sweat chloride values in infancy, but the length of follow-up was not specified.8 The data presented here may in fact underestimate true risk of delayed CF diagnosis, as complete follow-up data were not available. The lung function outcome, describing values within the low normal range at 5 and 10 years of age, is consistent with other studies.11,20,21 A lesser degree of obstructive lung disease at the end of follow-up was observed. The patients with delayed CF did not have sustained improved nutritional outcomes compared with those identified by NBS. This may reflect benefits of modern treatment in NBS positive patients with CF, with nutritional benefits reported in the existing literature.22 Ren et al reported normal nutritional indices for all patients11; however, abnormal height and weight z-scores (defined as < 2 SDs) were recorded in 2 (14%) and 1 (7%) patients with delayed CF in our cohort throughout follow-up, respectively. Determining the clinical relevance of these lung function and nutritional findings is difficult, and should not be over-interpreted given the relatively small cohort numbers and potential type II statistical error. Rates of nonmucoid P aeruginosa isolation and colonization in the delayed CF cohort were both lower than NBS positive patients with CF. This is clinically relevant as nonmucoid P aeruginosa is an important pathogen in CF with significant detrimental effects on morbidity and mortality.23 Rates of P aeruginosa colonization were similar to those reported by Ren et al (20% vs 25%, respectively).11 Isolation and colonization rates of mucoid P aeruginosa, an additional adverse prognostic factor,1,24-26 however, were similar between cohorts in our study at final follow-up at 10 years of age. One confounding factor is that NBS positive patients with CF received more frequent sputum cultures. Patients with delayed CF received sputum cultures when clinically indicated, or at minimum, each annual review. Finally, data suggested less severe overall CF disease with increasing difference in Shwachman scores, which became significant at the end of follow-up in the delayed CF cohort. The major limitation of this study is the relatively small cohort. This may have resulted in an analysis underpowered to detect clinically relevant differences and potential type II statistical error. Other limitations include the retrospective design and IRT/p.F508del screening approach. This approach is based on previous cost analysis and was used throughout the 15 years of screening in NSW.13 It has been reported that even with improved sensitivity of a larger mutation analysis, most infants
would still be detected using the screening program used throughout this study, and adding a test for the next most common mutation likely results in the detection of only 1 extra case every 3 years.13,27 Finally, 3 patients with delayed CF were lost to follow-up, and this may have led to biasing of the results. However, it is unclear if the clinical course of these patients differed from the rest of the cohort. In conclusion, this study provides important prognostic information for physicians managing infants with intermediate sweat chloride values. Our high rate of subsequent formal CF diagnosis reinforces the need for close follow-up in this population. n The authors are grateful to Liz Barnes (Biostatistician, The Children’s Hospital at Westmead) for her generous assistance with the statistical analyses. The authors are also grateful to the Department of Molecular Genetics (The Children’s Hospital at Westmead), including Bruce Bennetts, Gemma Jenkins, Edwina Middleton, and Felicity Collins, for their assistance with the extended genetic mutation panel and newborn screening at this institution. The annual checks performed for this study were a collaborative effort by all staff on the CF team at CHW. Submitted for publication Aug 8, 2014; last revision received Jan 15, 2015; accepted Jan 28, 2015. Reprint requests: Tyler Groves, MSc, Sydney Medical School, University of Sydney, 105/2 Dind St, Milsons Point, New South Wales, 2061, Australia. E-mail: [email protected]
References 1. Dijk FN, Fitzgerald DA. The impact of newborn screening and earlier intervention on the clinical course of cystic fibrosis. Paediatr Respir Rev 2012;13:220-5. 2. Farrell PM, Kosorok MR, Laxova A, Shen G, Koscik RE, Bruns WT, et al. Nutritional benefits of neonatal screening for cystic fibrosis. Wisconsin Cystic Fibrosis Neonatal Screening Study Group. N Engl J Med 1997; 337:963-9. 3. Farrell PM, Kosorok MR, Rock MJ, Laxova A, Zeng L, Lai HC, et al. Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal Screening Study Group. Pediatrics 2001;107:1-13. 4. Koscik RL, Farrell PM, Kosorok MR, Zaremba KM, Laxova A, Lai HC, et al. Cognitive function of children with cystic fibrosis: deleterious effect of early malnutrition. Pediatrics 2004;113:1549-58. 5. Dijk FN, McKay K, Barzi F, Gaskin KJ, Fitzgerald DA. Improved survival in cystic fibrosis patients diagnosed by newborn screening compared to a historical cohort from the same center. Arch Dis Child 2011;96:1118-23. 6. Moskowitz SM, Chmiel JF, Sternen DL, Cheng E, Gibson RL, Marshall SG, et al. Clinical practice and genetic counseling for cystic fibrosis and CFTR-related disorders. Genet Med 2008;10:851-68. 7. Parad RB, Comeau AM. Diagnostic dilemmas resulting from the immunoreactive trypsinogen/DNA cystic fibrosis newborn screening algorithm. J Pediatr 2005;147:S78-82. 8. Borowitz D, Parad RB, Sharp JK, Sabadosa KA, Robinson KA, Rock MJ, et al. Cystic Fibrosis Foundation practice guidelines for the management of infants with cystic fibrosis transmembrane conductance regulatorrelated metabolic syndrome during the first two years of life and beyond. J Pediatr 2009;155:S106-16. 9. Farrell PM, Rosenstein BJ, White TB, Accurso FJ, Castellani C, Cutting GR, et al. Guidelines for diagnosis of cystic fibrosis in newborns through older adults: Cystic Fibrosis Foundation consensus report. J Pediatr 2008;153:S4-14.
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1473
THE JOURNAL OF PEDIATRICS
www.jpeds.com
10. Massie J, Clements B. Diagnosis of cystic fibrosis after newborn screening: the Australasian experience—twenty years and five million babies later: a consensus statement from the Australasian Paediatric Respiratory Group. Pediatr Pulmonol 2005;39:440-6. 11. Ren CL, Desai H, Platt M, Dixon M. Clinical outcomes in infants with cystic fibrosis transmembrane conductance regulator (CFTR) related metabolic syndrome. Pediatr Pulmonol 2011;46:1079-84. 12. Fitzgerald D, Van Asperen P, Henry R, Waters D, Freelander M, Wilson M, et al. Delayed diagnosis of cystic fibrosis in children with a rare genotype (delta F508/R117H). J Paediatr Child Health 1995;31:168-71. 13. Wilcken B, Wiley V, Sherry G, Bayliss U. Neonatal screening for cystic fibrosis: a comparison of two strategies for case detection in 1.2 million babies. J Pediatr 1995;127:965-70. 14. Bobadilla JL, Macek M Jr, Fine JP, Farrell PM. Cystic fibrosis: a worldwide analysis of CFTR mutations—correlation with incidence data and application to screening. Hum Mutat 2002;19:575-606. 15. Clancy JP, Rowe SM, Accurso FJ, Aitken ML, Amin RS, Ashlock MA, et al. Results of a phase IIa study of VX-809, an investigational CFTR corrector compound, in subjects with cystic fibrosis homozygous for the F508del-CFTR mutation. Thorax 2012;67:12-8. 16. Gibson LE, Cooke RE. A test for concentration of electrolytes in sweat in cystic fibrosis of the pancreas utilizing pilocarpine by iontophoresis. Pediatrics 1959;23:545-9. 17. Evans AK, Fitzgerald DA, McKay KO. The impact of meconium ileus on the clinical course of children with cystic fibrosis. Eur Respir J 2001;18: 784-9. 18. Shwachman H, Kulczycki LL. Long-term study of one hundred five patients with cystic fibrosis; studies made over a five- to fourteen-year period. AMA J Dis Child 1958;96:6-15.
1474
Vol. 166, No. 6 19. Stanojevic S, Wade A, Stocks J, Hankinson J, Coates AL, Pan H, et al. Reference ranges for spirometry across all ages: a new approach. Am J Respir Crit Care Med 2008;177:253-60. 20. Desmarquest P, Feldmann D, Tamalat A, Boule M, Fauroux B, Tournier G, et al. Genotype analysis and phenotypic manifestations of children with intermediate sweat chloride test results. Chest 2000;118: 1591-7. 21. Sermet-Gaudelus I, Girodon E, Sands D, Stremmler N, Vavrova V, Deneuville E, et al. Clinical phenotype and genotype of children with borderline sweat test and abnormal nasal epithelial chloride transport. Am J Respir Crit Care Med 2010;182:929-36. 22. McKay KO, Waters DL, Gaskin KJ. The influence of newborn screening for cystic fibrosis on pulmonary outcomes in New South Wales. J Pediatr 2005;147:S47-50. 23. Davies JC. Pseudomonas aeruginosa in cystic fibrosis: pathogenesis and persistence. Paediatr Respir Rev 2002;3:128-34. 24. de Vrankrijker AM, Wolfs TF, van der Ent CK. Challenging and emerging pathogens in cystic fibrosis. Paediatr Respir Rev 2010;11:246-54. 25. Treggiari MM, Rosenfeld M, Retsch-Bogart G, Gibson R, Ramsey B. Approach to eradication of initial Pseudomonas aeruginosa infection in children with cystic fibrosis. Pediatr Pulmonol 2007;42:751-6. 26. West SE, Zeng L, Lee BL, Kosorok MR, Laxova A, Rock MJ, et al. Respiratory infections with Pseudomonas aeruginosa in children with cystic fibrosis: early detection by serology and assessment of risk factors. JAMA 2002;287:2958-67. 27. Massie J, Curnow L, Glazner J, Armstrong D, Francis I. Lessons learned from 20 years of newborn screening for cystic fibrosis. Med J Aust 2012; 196:67-70.
Groves et al
June 2015
ORIGINAL ARTICLES
Figure 1. Flow diagram of patients with intermediate sweat chloride, with follow-up of the delayed CF and matched NBS positive CF control cohorts over diagnosis, 2, 5, and 10 years of age. F/U, follow-up.
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1474.e1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 166, No. 6
Figure 2. Comparison of lung function parameters at A, 5 and B, 10 years of age between the delayed CF cohort (gray bars; n for A = 13, n for B = 13) and the NBS positive CF cohort (striped bars; n for A = 28, n for B = 28). FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; FEF25-75, forced expiratory flow rate in the middle half of FVC. Values are presented as mean SD.*Statistically significant difference between cohorts (mean difference 5.93, 95% CI 0.39-11.46, P = .04). P = not significant for all comparisons.
1474.e2
Groves et al
ORIGINAL ARTICLES
June 2015
Table I. Extended genetic mutation panel used by clinical genetics at CHW Year
Mutations tested
1999 2002
Added:
2005
Added: Removed: Added:
2013
F508del V520F R560T 2183AA>G S549R 3659delC 3876delA Y122X F508C A559T S1255X I148T
Removed:
I507del G542X S549N 2789+5G>A R347P 3905insT 3120+1G>A Q493X I507V 1898+5G>T
621+1G>T G551D W1282X 1898+1G>A R334W Y122X 394delTT 3849+4G>A I506V 2307insA
R117H R553X 1717-1G>A 711+1G>T 1078delTT Q493X 2184delA
A455E N1303K 3849+10kbC>T G85E R1162X 3849+4G>A I148T
Y122X Y1092X
R347H M1101K
Table IV. Nutritional parameter Z-scores for patients with delayed CF and NBS positive CF control patients Weight Delayed CF Age (y)
n
Diagnosis 2 5 10
13 13 13 13
z-score 0.02 1.13 0.52 0.65* 0.15 0.83 0.08 1.09
Height NBS positive CF
n 28 28 28 28
z-score 0.47 0.95 0.02 0.91* 0.13 0.83 0.56 0.92
Delayed CF n 11 13 13 13
z-score 0.09 1.45 0.11 1.08 0.35 0.98 0.04 1.18
Body mass index NBS positive CF
n 28 28 28 28
z-score 0.4 1.01 0.13 0.92 0.08 0.89 0.45 1.03
Delayed CF n N/A 13 13 13
z-score 0.66 0.91† 0.1 0.71 0.03 0.93
NBS positive CF n 28 28 28
z-score 0.28 0.66† 0.06 0.75 0.35 0.81
N/A, not applicable. Data are presented as mean SD. *Statistically significant difference between cohorts (mean difference 0.54, 95% CI 0.17-0.92, P = .005). †Statistically significant difference between cohorts (mean difference 0.43, 95% CI 0.01-0.86, P = .047).
Long-Term Outcomes of Children with Intermediate Sweat Chloride Values in Infancy
1474.e3