Exploring Turkey's mosaic of novel variants and complex alleles in cystic fibrosis genetics

Abstract

Backgrounds

Genetic variants in CF transmembrane conductance regulator (CFTR) gene causes cystic fibrosis (CF), a prevalent autosomal recessive disorder. More than 2000 variants in CFTR have been described as disease causative. This study aims to delineate the genotypic and phenotypic landscape of CF among people with CF (pwCF) followed at the largest CF center in Turkey.

Methods

We conducted a descriptive and retrospective analysis of 481 patients registered with the European CF Society Patient Registry and followed at Marmara University Selim Çöremen CF Center from 2015 onwards. Comprehensive CFTR analysis was utilized for genetic diagnosis. Besides the whole cohort, novel variants and complex alleles were also described.

Results

Our cohort exhibited a broad spectrum of CFTR variants, with 136 different variants detected, indicating substantial genetic diversity. The F508del variant was less prevalent in our cohort compared to US and European averages, which could reflect unique genetic and demographic characteristics of the Turkish population. Additionally, we identified nine novel variants in 12 alleles, which enhances the understanding of CF's genetic complexity in this region, and complex alleles in 32 pwCF.

Conclusion

Our research underscores the heterogeneity of CFTR variants in Turkey and highlights the necessity for extensive genetic profiling particularly for diverse populations to provide effective personalized treatment strategies. It is crucial to understand the full spectrum of CFTR variants with the advent of CFTR modulators.

1. INTRODUCTION

Cystic fibrosis (CF), the most common lethal genetic disorder in populations of Northern European descent, is an autosomal recessive disease that occurs as a result of variants in CF transmembrane conductance regulator (CFTR) gene and causes progressive multisystemic involvement, especially in the lung and pancreas. ¹ The gene is located on the long arm of the 7th chromosome (chr7:117,479,963‐117,668,665) and it spans a region of nearly 188,000 base pairs and includes 27 exons. ² Approximately 2000 different variants have been identified in CFTR to date. ³ These variants, including those that alter single codons, can be identified through comprehensive gene sequencing, while techniques like multiple ligation dependent probe amplification (MLPA) are utilized to detect larger gene deletions or duplications. ⁴ The wide range of CF characteristics has been found to be influenced not just by the type of CFTR variants but also by other genetic elements like modifier genes and environmental factors.

As CFTR modulators, (ivacaftor [I], lumacaftor [L], elexacaftor [E], and tezacaftor [T]), become increasingly common, with proven efficacy for patients with certain variants; the importance of genotyping has grown to make a conclusive CF diagnosis, foresee the prognosis and guide patients towards the most effective treatment options with these drugs.

It is noteworthy that the distribution and clustering of variants in CFTR are specific to some geographical regions due to migrations throughout history, racial origins, and inbreeding, which is especially common in isolated societies. Over the course of time, genetic drift and the founder effect can cause an increase in the prevalence of specific variants among the descendants, resulting in the development of unique or primarily local genetic traits. ⁵ This situation causes c.1521_1523del (F508del) variant, which is the most common around the globe, to be seen at lower frequency compared to the rest of the world, in Turkey, which is the second most populous country bordering Europe and holds a geopraphically unique position.

In Turkey, the newborn screening program started in January 2015, consisting of immunoreactive trypsinogen (IRT) testing twice; however, genotyping is not currently integrated into the program. Due to the formation of the newborn screening program, patients who are referred because of a positive screening result undergo whole gene sequencing of the CFTR if their sweat test is either suspicious or positive. If a heterozygous variant is identified, then MLPA analysis is also carried out. While using whole gene analysis instead of a panel allows genotypic diversity to be documented in more detail, this practice may lead to a lack of diagnosis because the cluster to which genetic analysis is applied is very specific.

In our study, we aimed to describe the detailed genotypic and clinical features of pwCF followed in our clinic, the largest CF center in Turkey, as well as to detect new variants and complex alleles in our population.

2. METHODS

2.1. Study population

This is a descriptive and retrospective study that evalautes data of 481 people with CF (pwCF) who have been registered in European CF Society (ECFS) Patient Registry since 2015 and have been following up at Marmara University Selim Çöremen CF Center.

Patients who were referred to our center due to a positive newborn screening test for CF, underwent a sweat test for definitive diagnosis. If the sweat test result was positive (Cl‐ > 60 mmol/L) or borderline (Cl‐ = 30−59 mmol/L), a genetic analysis was performed, which included CFTR gene sequence analysis and MLPA.

The pwCF who were being followed up before the newborn screening program began, were diagnosed with genetic analysis based on sweat test results after they presented with typical clinical findings (recurrent bronchopneumonia, chronic cough and sputum production, bronchiectasis, failure to thrive, meconium ileus, salt depletion, repeated episodes of pancreatitis).

Those who were followed up at our center with a diagnosis of CF, whose files did not have any missing information that would affect the study results, and who agreed to participate in the study were included in the study. Those of CF screen positive, inconclusive diagnosis, those with a follow‐up period of less than 1 year and who did not give consent were excluded. Deceased patients were also included in the study with their genetic results but they were excluded in the demographic and clinical follow‐up data, except for those who died within the last year.

2.2. Clinical status

Body mass index (BMI) Z scores at the last examination for growth evaluation, maximum forced expiratory volume in the first second (FEV₁) value (as percentage) of the last 12 months, pancreatic insufficiency (PI) status, presence of CF‐related diabetes mellitus (CFRD), status of liver disease, colonization of Pseudomonas Aeruginosa (PsA), need of noninvasive positive pressure ventilation (NIPPV) or nasal oxygen (O₂) support, number of acute exacerbations and hospitalization days in the last year were recorded. People with fecal elastase value below 200 mcg/g were considered as PI and a diagnosis of CFRD was made through an oral glucose tolerance test.

2.3. Genetic and clinical status

Variant analysis results (CFTR sequencing analysis and MLPA) were obtained from the files and while classifying the variants of the patients, first the variants in both alleles were examined separately and then they were classified according to certain characteristics of the variants. Patients with the presence of more than two nucleotide variants on one allele (complex alleles) were identified and evaluated separately.

To determine the characteristics and novelty of the variants https://varsome.com/, https://franklin.genoox.com/, https://www.ncbi.nlm.nih.gov/clinvar/, https://cftr2.org/, https://cftr.iurc.montp.inserm.fr/cftr online databases were used. Variants were classified as follows:

• 1.According to American College of Medical Genetics and Genomics (ACMG), the alleles were divided into groups of benign, likely benign, variant of uncertain significance (VUS), likely pathogenic or pathogenic. ⁶
• 2.The determination of missense, nonsense, frameshift, and noncoding variants was based on the mechanism that caused the change in the encoded protein. CFTR classification was also made based on these results (Class I variants cause a deficiency in the synthesis of CFTR protein. Class II variants result in modified protein maturation. Class III variants lead to faulty channel regulation. Class IV variants result in diminished channel conductance. Class V variants decrease CFTR protein synthesis. Class VI variants result in decreased CFTR stability.).
• 3.Alleles were classified according to variant type, specifically differentiating between large exon deletions, exonic, and intronic variants.
• 4.Point mutations were identified and recorded by subtype including base substitution, deletion, deletion‐insertion, duplication, and insertion.
• 5.Splice site variants were also specified separately for each allele.

2.4. Statistical analysis

We conducted all statistical analyses using the IBM SPSS 21.0 software (SPSS Inc.). To compare categorical variables across groups, we employed the Chi‐square test. Student's T‐test was used for normally distributed continuous variables, while the Mann−Whitney U test was applied to those not normally distributed. The distribution of continuous variables was evaluated using the Kolmogorov−Smirnov test and/or the Shapiro−Wilk test. Categorical variables were presented as numbers (percentages) and continuous variables as either mean ± standard deviation or median (range), depending on their distribution. A p Value of less than .05 was considered statistically significant.

2.5. Ethics

According to the Genetic Diseases Diagnosis Center workflow, information about testing was provided and written patient/parent informed content for data collection was obtained from the patients. The study protocol was approved by the ethical committee of Marmara University School of Medicine (Ethical approval number: 04.2024.353).

3. RESULTS

3.1. Demographic and clinical features

Among all patients involved to the study (n = 481, 51.3% males), 27 who died before the previous year were excluded in terms of clinical feature. The remaining group (n = 454, 52.3% males) had a median age at follow‐up of 11.8 years, with a range of 0.46−49.9 years (Table 1). The age at diagnosis varied widely, with a median of 3.6 years and a range from 0.96 to 523.2 months. Regarding neonatal screening, 117 of them (25.9%) had a positive result. Meconium ileus was diagnosed in 36 (8%) of the cases, and required surgical intervention, and a further 1.1% also suffered from this condition but did not undergo surgery. Sweat test results showed a median chloride level of 91 mmol/L (9−200 mmol/L). Nutritional status as assessed by standard Z scores indicated that the median BMI Z score had a median value of −0.2, ranging from −5.06 to 3.7. The median of FEV1 (%), to present pulmonary function, was 86.7% (21.3−126.4%). Twelve (2.7%) patients needed NIPPV and 68 (15.4%) of the group were colonized with PsA in respiratory cultures, while 136 (30.9%) had intermittent growth. Diabetes was present in 30 (6.6%) of the cases, and 15 (3.3%) of the patients had a diagnosis of cirrhosis. The median duration of hospital stay was 18 days (1−327).

Table 1.

Demographic and clinical features of the patient group.

	n (%) or median (minimum−maximum) or mean ± SD
Age at diagnosis (months)	3.6 (0.9−523.2)
Age at last visit (years)	11.8 (0.4−49.9)
Gender (female)	216 (44.8)
Neonatal screening
Positive result	117 (25.9)
Negative result	40 (8.9)
Not done	294 (65.2)
Meconium ileus
Yes, operated	36 (8)
Yes, not operated	5 (1.1)
No	410 (90.9)
Z Score
Weight	−0.4 (−6.9 to 3.2)
Height	−0.4 ± 1
BMI	−0.2 (−5 to 3.7)
Sweat test (mmol/L)	91 (9−200)
FEV1 (%)	86.7 (21.3−126.4)
Oxygen support	12 (2.7)
Pseudomonas aeruginosa growth
Colonized	68 (15.4)
Intermittant	136 (30.9)
No	237 (53.7)
Pancreatic insufficiency	338 (70.2)
CFRD
Yes, insulin treatment	27 (6.1)
Yes, no need for insulin treatment	3 (0.6)
No	411 (93.2)
Number of hospitalizations last year	0 (0−10)
Number of exacerbations last year	2 (0−10)

Abbreviations: BMI, body mass index, CFRD, cystic fibrosis related diabetes mellutis; FEV1, forced expiratory volume in 1 s.

3.2. Genotypes

In our study a total of 136 different variants (including 10 different large exon deletions) were detected in a cohort of 481 patients (962 alleles). The total number of variations detected was 1007 due to the presence of complex alleles. The 30 most frequently observed variants are shown in Table 2. Two patients had only one variant each. Although, the nationality of 469 patients in our cohort was Turkey; there were also patients from Syria (3, 0.6%), Azerbaijan (2, 0.4%), Iran (2, 0.4%), Kosovo (2, 0.4%), Palestine (1, 0.2%), Iraq (1, 0.2%) and Russia (1, 0.2%). At the time of the study, 33 of the patients included in our cohort were dead.

Table 2.

The 30 most frequently observed variants in the patient group.

Legacy name	c.DNA position	Number of alleles	CFTR2
F508del	c.1521_1523del	293	CF‐C
1677delTA	c.1545_1546del	63	CF‐C
	c.2657 + 5 G > A	44	CF‐C
N1303K	c.3909 C > G	42	CF‐C
E92K	c.274 G > A	36	CF‐C
CFTRdele2		33
K684Sfs*38	c.2051_2052delAAinsG	30	CF‐C
G542X	c.1624 G > T	25	CF‐C
D110H	c.328 G > C	22	CF‐C
R347P	c.1040 G > C	18	CF‐C
L467F	c.1399 C > T	17	VUSa
G85E	c.254 G > A	16	CF‐C
W1282X	c.3846 G > A	16	CF‐C
I1000Lfs*2	c.2998del	15	CF‐C
	c.3964‐3 C > G	14	CF‐Ca
	c.489 + 1 G > T	12	CF‐C
	c.1116 + 1 G > A	10	CF‐C
	c.1393‐1 G > A	10	CF‐C
R1158X	c.3472 C > T	10	CF‐C
S466X	c.1397 C > G	9	CF‐C
R1070Q	c.3209 G > A	9	VCC
	c.3718‐2477 C > T	9	CF‐C
G178R	c.532 G > A	9	CF‐C
CFTRdele4‐11		9
L732X	c.2195 T > G	8	CF‐C
E831X	c.2491 G > T	8	CF‐C
D1152H	c.3454 G > C	7	VCC
R347H	c.1040 G > A	6	CF‐C
Q353X	c.1057 C > T	5	CF‐C
	c.2988 + 1 G > A	5	CF‐C

Abbreviations: CF‐C, Cystic fibrosis causing; VCC, varying clinical consequence; VUS, variants of uncertain significance.

^aVariant could not be found in CFTR2, the result is from CFTR‐French database.

The most common genotype was nonF508del/nonF508del (274/481, 56.9%), followed by the F508del/nonF508del (123/481, 25.5%) and the F508del/F508del (84/481, 17.4%) genotypes. Among them, the most frequent variant was c.1521_1523del (F508del) (293 alleles, 29%), followed by c.1545_1546del (1677delTA) (63 alleles, %6.2), c.2657 + 5 G > A (44 alleles, 4.3%), c.3909 C > G (N1303K) (42 alleles, 4.1%) and c.274 G > A (E92K) (36 alleles, 3.5%). The most common large exon deletions were CFTRdele2 (33 alleles) and CFTRdele4‐11 (9 alleles). Three humdred forty patients (70.6%), 18 of whom were dead, regardless of age, had variants eligible for at least one of the modulator drugs and 114 (23.7%) were using.

Of all total alleles, 84.1% (n = 847) were pathogenic, 6.9% (n = 70) were likely pathogenic, and 5.5% (n = 56) were structural variants. The remaining 3.5% (n = 34) included VUS, likely benign and benign variants. Without including large exon deletions, 81.1% (n = 817) of the variations were exonic and 13.2% (n = 134) were intronic. When the variants were classified in terms of their effects on the protein produced, it was seen that there were 289 in‐frame, 279 missense and 162 nonsense type variations. While the number of alleles with noncoding type variants was 134, it was determined that variants in 82 alleles caused a shift in the reading frame. Splice site variants were found in 237 alleles. The distribution of the allelic relationships according to this type of classification is displayed in Figure 1.

Figure 1.

Variant classification according to protein coding.

3.3. Novel variants and phenotypes

We found nine novel variants in a total of 12 alleles of our nine patients (1.8%), three patients had their variants homozygous. According to ACMG, one of the novel variants was pathogenic and seven were likely pathogenic and one was VUS (Table 3). Five patients met the criteria for modulator drug eligibility based on their genotypes, yet none were using them, likely due to the continued lack of reimbursement coverage for CFTR modulators in Turkey.

Table 3.

Genetic features of the novel variants.

Novel variant	Legacy name	ACMG class	Exon	Variant type	Coding impact	Variant position	Other variant(s)
R1283G (c.3847 A > G)	R1283G	Likely pathogenic	23	Base subtitution	Missense	Exonic	F508del (c.1521_1523del)
P1372H (c.4115 C > A)	P1372H	Likely pathogenic	25	Base subtitution	Missense	Exonic	P1372H (c.4115 C > A)
R289Nfs*17 (c.865_869del)	R289Nfs*17	Pathogenic	7	Deletion	Frameshift	Exonic Splicing	R289Nfs*17 (c.865_869del)
G1208Afs*3 (c.3618del)	G1208Afs*3	Likely pathogenic	22	Deletion	Frameshift	Exonic	G1208Afs*3 (c.3618del)
G921R (c.2761 G > A)	G921R	Likely pathogenic	17	Base subtitution	Missense	Exonic	c.2657 + 5 G > A
P1175Lfs*17 (c.3524del)	P1175Lfs*17	Likely pathogenic	22	Deletion	Frameshift	Exonic	F508del (c.1521_1523del)
c.3963 + 1del	‐	Likely pathogenic	‐	Deletion	Splice junction loss	Intronic Splicing	W1282X (c.3846 G > A) D579 = (c.1737 T > C)
S1049I (c.3146 G > T)	S1049I	VUS	20	Base subtitution	Missense	Exonic Splicing	F311L (c.933 C > G)
E632Tfs*9 (c.1894_1895del)	E632Tfs*9	Likely pathogenic	14	Deletion	Frameshift	Exonic	F508del c.1521_1523del

Abbreviations: ACMG, American College of Medical Genetics and Genomics; VUS, variants of uncertain significance.

While one patient's drug was at the application stage, another patient was considered ineligible for modulators because he was younger than 2 years old, even if his genotype was compatible. The parents of the two patients were consanguineous and both of them were first‐degree cousins. Three of the patients did not have IRT results because they were born before the start of the newborn screening program (2015). None of them had hepatic involvement whereas seven were PI. Detailed genotypic and phenotypic characteristics of patients with novel variants are given in Tables 3 and 4.

Table 4.

Clinical features of the patients with novel variants.

Novel Variant	Age of diagnosis	Current Age	Cons.	Sex	IRT	Sweat test (mmol/L)	FEV1 (%)	Pseudomonas growth	Pancreatic status	BMI (Z score)	Number of hospitalizations	Modulator eligibility	Current modulator use
R1283G (c.3847 A > G)	40 days	1 year 10 month	−	M	+,+	107	−	No	Insufficient	N/A	0	ETI	No
P1372H (c.4115 C > A)	10 month	7 year 6 month	+	M	N/A,−	44	97.7	No	Insufficient	−0.19	0	None	‐
R289Nfs*17 (c.865_869del)	3.5 month			F	N/A	90	24.1	Chronically colonized	Insufficient	−6.26	‐	None	‐
G1208Afs*3 (c.3618del)	26 days	1 year 10 month	+	M	+,+	74	−	No	Insufficient	N/A	4	None	‐
G921R (c.2761 G > A)	9 month	9 year 4 month	−	F	N/A	72	91	No	Sufficient	0.83	0	TI, I	No
P1175Lfs*17 (c.3524del)	4 month	15 year 6 month	−	F	N/A	114	88.6	Intermittant	Insufficient	0.23	0	ETI	No
c.3963 + 1del	6 month	5 year 5 month	−	F	+,+	47	86.1	No	Insufficient	−0.55	0	None	‐
S1049I (c.3146 G > T)	12 month	7 year 2 month	−	M	+,−	67	97.1	No	Sufficient	−1.45	0	ETI, TI, I	No
E632Tfs*9 (c.1894_1895del)	4 month	3 year 9 month	−	M	N/A,+	77.6	−	Intermittant	Insufficient	−0.78	0	ETI	No

Abbreviations: BMI, body mass index; cons., consanguineous marriage; E, elexacaftor; F, female; I, ivacaftor; IRT, immunoreactive trypsinogen; M, male; m, months; T, tezacaftor; y, years.

3.3.1. c.3847 A > G (R1283G)

A point mutation on the 23rd exon of CFTR, where the adenine changes to guanine at the 3847th nucleotide, causing the codon to change and the amino acid to switch from arginine to glycine. This patient has c.1521_1523del (F508del) variant on the other allele, leading eligibility to ETI. He was diagnosed at the age of 40 days. His deep pharyngeal cultures haven't grown PsA so far and he only had mild respiratory symptoms.

3.3.2. c.4115 C > A (P1372H)

A point mutation on the 25th exon of CFTR, where cytosine changes to adenine at the 4115th nucleotide, causing the codon to change and the amino acid to switch from prolin to histamine. This patient is homozygous for this variant, and not eligible for modulator drugs. He was diagnosed at the age of 10 months. He has never had PsA in cultures, but intermittant MRSA growths were detected. He does not suffer from chronic respiratory symptoms.

3.3.3. c.865_869del (R289Nfs*17)

A deletion of five nucleotides (AGACA) on the splice site of the 7th exon of CFTR, which causes a shift in the triplet reading frame and makes an amino acid change from arginine to asparagine. This patient was homozygous for this variant, she was diagnosed when she was 3.5 months old and died at the age of 17 years. Her BMI Z score was −6.26 on her last visit and with a diagnosis of CFRDM, she was on parenteral insulin therapy. She suffered from chronic PsA and intermittent MRSA growths, experiencing persistent symptoms requiring daytime oxygen support and BiPAP during sleep, with an FEV₁ of 24.1%.

3.3.4. c.3618del (G1208Afs*3)

Homozygous deletion of an adenien on the 22nd exon of CFTR causing a frameshift mutatiton where glycine converts to alanine was detected in one patient. He was diagnosed on postnatal 26th day. The patient is fed with a nasogastric catheher due to retardation in weighing and frequent hyponatremia, and he has moderate respiratory symptoms and generally requires NIPPV during exacerbations.

3.3.5. c.2761 G > A (G921R)

A single nucleotide subtitution where guanine is replaced by an adenine, and results in glycine changing to arginine on the 17th exon of CFTR. This patient has an intronic c.2657 + 5 G > A variant on the other allele. Her variants are eligible for TI + I and I. But she is not on any modulator drugs. She occasionally experiences mild respiratory symptoms and does not have chronic PsA or MRSA.

3.3.6. c.3524del (P1175Lfs*17)

This single base deletion on the 22nd exon of CFTR causes a frameshift variant where proline changes to leucine. As the patient has c.1521_1523del (F508del) variant on the other allele, she is eligible for ETI but not on it. Her diagnosis was made when she was 4 months old. Her respiratory tract cultures have shown chronic PsA and intermittent B.cepacia growths and her best FEV₁ was 88.6% last year.

3.3.7. c.3963 + 1del

Variation refers to a deletion of a single guanine nucleotide at the splice donor site following exon 20 of the CFTR. It is in the intronic region, yet it is very close to the exon‐intron boundary, which is crucial for proper splicing of the mRNA. This patient has two other variations in CFTR, c.3846 G > A (W1282X) and c.1737 T > C (D579=), and and ineligible for all modulator drugs. She was diagnosed when she was 6 months old. She does not have chronic PsA growth and rarely has exacerbations.

3.3.8. c.3146 G > T (S1049I)

This variant involves a substitution of a guanine with a thymine at nucleotide position 3146 on the 20th exon of CFTR, and the codon change leads to an amino acid switch from serine to isoleucine. The patient has c.933 C > G variation on the other allele, which makes him eligible for ETI, TI + I and I. With mild respiratory complaints, he never has PsA or MRSA in his respiratory cultures.

3.3.9. c.1894_1895del (E632Tfs*9)

The variant involves the deletion of two nucleotides (AG) in the coding sequence on exon 14 of CFTR. Due to deletion of two nucleotides, it alters the reading frame of the gene resulting premature termination of the protein sequence. Since the patient carries c.1521_1523del (F508del) variant in the other allele, he is considered modulator eligible. He was diagnosed at 4 months of age and occasionally experiences moderate respiratory issues.

3.4. Complex alleles

We noticed that there were 32 pwCF in our group with complex alleles, three of whom were dead (Table). While 18 of these patients had one extra variation, 13 of them had two (Table 5). A total of six variations were detected in one of the patients. Nine of these patients were homozygous for F508del, on the other hand 13 of them did not have F508del variant at all. Sanger sequencing was performed on the parents of nine patients, while only the mother of one patient could be tested, the other parents did not consent to genetic testing. The most common genotype with the complex alleles was [F508del;L467F]/[F508del;L467F] and was detected in four patients. Within the group, five pwCF were ineligible for modulator drugs, whereas seven of them were currently using.

Table 5.

Complex alleles detected in our patient group.

[F508del;L467F]/[F508del;L467F]	4
[Y1424=;c.2989‐1 G > A]/[Y1424=; c.2989‐1 G > A]	2
[S466X;R1070Q]/[S466X;R1070Q]	3
[F508del;L467F]/[F508del]	3
[F508del;L467F]/[F311L]	2
[F508del;T760M]/[F508del;T760M]	1
[F508del;L467F]/[G1265R;M348K]	1
[I148T;c.3067_3072del]/[I148T;V754M]	1
[R1162Q;Y913C]/[R1162Q;Y913C]	1
[V754M;CFTRdele3‐11;CFTRdele16‐18]/[E831X]	1
[S466X;R1070Q]/[F508del]	1
[F508del;V754M]/[F508del]	1
[S466X;R1070Q]/[c.1393‐1 G > A]	1
[W1282X;D579=]/[c.3963+1del]	1
F508del (;) S737F (;) CFTRdele11	1
F508del (;) L467F (;) D1152H	1
F508del (;) S466X (;) R1070Q	1
F508del (;) I148T (;) c.3067_3072del	1
F508del (;) N1303K (;) L467F	1
c.2658‐1 G > C (;) N1303K (;) c.869+11 C > T	1
N1303K (;) c.2657 + 5 G > A (;) c.869+11 C > T	1
P5L (;) D110H (;) c.2620‐15 C > G	1
L88X (;) G1069R (;) c.2657+5 G > A	1

When patients with the complex allele were compared with those without, no statistically significant difference was detected in terms of clinical features (Table 6).

Table 6.

Comparison of clinical features of patients according to the presence of complex alleles.

	Patients with complex alleles (n = 32)	Patients without complex alleles (n = 449)	p Value
	Median (minimum−maximum) or n (%)	Median (minimum−maximum) or n (%)
Age (years)	8 (2−32)	12 (1−50)	.062
Gender (female)	15 (51.7)	201 (47.6)	.66
BMIa	−0.20 (−2.67 to 3.93)	−0.19 (−5.86 to 2.72)	.48
FEV1 b (%)	87 (38.1−123.6)	86.1 (19.9−129.8)	.83
Number of hospitalizations	0 (0−5)	0 (0−7)	.90
Number of exacerbations	2 (0−6)	2 (0−10)	.66
Pseudomonas aeruginosa colonization	4 (13.8)	81 (19.3)	.46
Pancreatic insufficiency	22 (75.9)	309 (73.6)	.78
CFRD	1 (3.4)	23 (5.5)	.63

Abbreviations: BMI, body mass index, CFRD, cystic fibrosis related diabetes mellutis; FEV1, forced expiratory volume in 1 s.

^aZ score.

^bBest in the last year.

4. DISCUSSION

In our study, we aimed to describe the detailed genotypic and clinical features of pwCF followed in our clinic, the largest CF center in Turkey. Additionally, our objective was to identify any novel variants or complex alleles within our population.

Turkey is a transcontinental country located at the crossroads of Europe and Asia, extending across the Anatolian peninsula in the west and Thrace in the Balkan region of Southeast Europe. The Anatolian population exhibits significant genetic diversity, as throughout history, this region has witnessed major migrations, contributing to its complex genetic composition. Although the range and prevalence of CFTR variants have been thoroughly researched in people of Northern European and North American Caucasian background, data regarding these variants remain limited for various other populations, indicating a need for more comprehensive studies. The spectrum of CFTR variants in Turkey could differ from those commonly observed in European or North American communities, potentially reflecting distinct genetic and demographic characteristics. The differences in the prevalence of CFTR variants across various regions can be explained by factors like the founder effect, genetic drift, population history, and consanguinity. For instance, the most frequent variant, F508del, varies in frequency from 87% in Northern Europe to 28% in Asia (Middle East), suggesting that the variant was carried by humans migrating into Europe, spreading as these populations expanded across the continent. ⁵

Although the most common CFTR variant globally is F508del, constituting approximately 80% of variants in Europe and 85.4% according to the latest CF Foundation report, over half of the individuals in our cohort were characterized by a genotype other than F508del/F508del, with the proportion of patients homozygous for the F508del constituting only 17%. In a study conducted by our center in 2018, the F508del rate was reported similarly. ⁷ , ⁸ In a study assessing the CFTR genotype throughout Africa and Asia, the F508del variant was found with a frequency of 18‐36% in Syria, our country's southern neighbor; showed comparable frequencies in Iran, our neighbor to the east. This finding supports the fact that F508del decreases in frequency from west to east, as shown before. ⁹ , ¹⁰ , ¹¹ , ¹²

Located in exon 10 near the F508del, 1677delTA, which is known to occur with a frequency of approximately 5% according the National CF Registry System 2022 data of Turkey, although not globally recognized as a common CFTR variant, was identified as the second most frequent variant in our cohort. ¹³ This finding is consistent with Erdogan et al.‘s research on other Turkish cohorts, which noted it as the fourth most prevalent. ¹⁴ Also, Bonyadi et al reported it with a frequency of 5.9% in Iranian Azerbaijani Turks. ¹⁵ This dispension suggests a possible genetic or historical link that has caused the variant to be more prevalent in this geography. This regional specificity has implications for treatment, as the variant is not targeted by current modulator therapies, which are designed based on the global prevalence of CFTR variants. The 2789 + 5 G > A (c.2657 + 5 G > A) variant, identified as the 5th most common in the ECFS Patient Registry Annual Data Report, was detected at 4.3% in our cohort, which seemed low, was actually high compared to the rest of the world. ¹⁶

In our cohort, N1303K, which was detected as the 3rd most common variant in Europe by ECFS and belongs to Class II, was seen as the 4th most common. This variant has been reported at a “significantly high” rate across 23 European countries, with Alonso et al. noting a 2.9% prevalence in the Spanish cohort and Paolis et al. documenting a 4.5% occurrence in Italy. ⁸ , ¹⁷ , ¹⁸ Despite not yet being deemed suitable for modulatory treatments by Vertex, N1303K stands out due to the proven effectiveness of modulator drugs in numerous in vitro studies involving intestinal organoids and nasal epithelial cells. ¹⁹ , ²⁰ , ²¹

Complex alleles, which means more than two nucleotide variants on a single allele, can make diagnosing the disease more difficult and alter the clinical manifestations of CF and treatment response. ²² This is because each variant in the complex allele can influence the functional activity of the CFTR protein, potentially increasing or decreasing its function. In our cohort we had 32 pwCF with complex alleles. In the study by Chevalier et al. published in France, the [F508del;L467F]/[F508del;L467F] complex allele, the most common among our patients, was ranked sixth in frequency, on the other hand in another study also from France did not show the same complex alleles with ours. ²³ , ²⁴ In the research by Raraigh et al., where they analyzed the genotypes of 5058 individuals, none of the complex alleles discovered were common to our cohort. ²⁵ Cases in which patients with similar variants have very different clinical features or response to treatment can be explained by the presence of complex alleles. ²⁴ In our study, when comparing patients with complex alleles to those without, in terms of clinical characteristics, we found no statistically significant differences. The absence of significant differences may be attributed to the relatively small number of patients with complex alleles compared to the whole group. Additionally, as reported in the literature, the impact of complex alleles on the response to CFTR modulators might be more pronounced. ²² This suggests that studies on this subject are needed, especially for diverse populations.

Accurately genotyping the pwCF is essential, as it has direct applications in genetic counseling, prenatal diagnosis, and establishing genotype‐phenotype correlations that enable assessing individual prognosis and determining appropriate patient management. Moreover, we think that this wide variant distribution is one of the biggest difficulties in utilizing treatment options for pwCF.

Our study emphasizes the crucial need for comprehensive genetic profiling across diverse populations. As we are standing on the edge of a new era of treatments, where gene therapies are promising for curing genetic disorders, understanding the distributions of variants and identification and reporting of novel CFTR variants becomes really important.

5. CONCLUSION

As CF is a genetically diverse disease, the discovery of new variants—particularly those unique to specific populations or regions—can provide valuable insights for contributing to the global knowledge. With the invention of CFTR modulators, understanding the full spectrum of CFTR variants is mandatory for the effective allocation of these therapies, ensuring that pwCF receive the most beneficial treatment based on their specific genotypes. In countries like Turkey, where a mosaic of novel variants and complex alleles may be present due to unique demographic histories, such reporting is even more crucial.

AUTHOR CONTRIBUTIONS

Ceren Ayça Yıldız: Formal analysis, investigation, software, writing—original draft. Merve Selçuk Balcı: Data curation, software. Şeyda Karabulut: Data curation, software. Zeynep Münteha Başer: Formal analysis, data curation, software. Mine Kalyoncu: Data curation, software. Neval Metin Çakar: Data curation, software. Müge Merve Akkitap Yiğit: Data curation, software. Eda Esra Baysal: Data curation, software. Fulya Özdemircioğlu: Data curation, software. Burcu Uzunoğlu: Data curation, software. Gamze Taştan: Data curation, software. Pınar Ergenekon: Methodology, review and editing. Yasemin Gökdemir: Methodology, review and editing. Ela Erdem Eralp: Methodology, review and editing. Fazilet Karakoç: Project administration, methodology, conceptualization, review and editing. Pınar Ata: Methodology, conceptualization, review and editing. Bülent Karadağ: Project administration, methodology, conceptualization, review and editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

ETHICS STATEMENT

According to the Genetic Diseases Diagnosis Center workflow, information about testing was provided and written patient/parent informed content for data collection was obtained from the patients. The study protocol was approved by the ethical committee of Marmara University School of Medicine.

Yıldız CA, Selçuk Balcı M, Karabulut Ş, et al. Exploring Turkey's mosaic of novel variants and complex alleles in cystic fibrosis genetics. Pediatr Pulmonol. 2024;59:3540‐3549. 10.1002/ppul.27249

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.