Association between CFTR modulators and changes in iron deficiency markers in cystic fibrosis

Abstract

Background:

Iron deficiency (ID) is a common extrapulmonary manifestation in cystic fibrosis (CF). CF transmembrane conductance regulator (CFTR) modulator therapies, particularly highly-effective modulator therapy (HEMT), have drastically improved health status in a majority of people with CF. We hypothesize that CFTR modulator use is associated with improved markers of ID.

Methods:

In a multicenter retrospective cohort study across 4 United States CF centers 2012–2022, the association between modulator therapies and ID laboratory outcomes was estimated using multivariable linear mixed effects models overall and by key subgroups. Summary statistics describe the prevalence and trends of ID, defined a priori as TSAT <20% or serum iron <60μg/dL (<10.7μmol/L).

Results:

A total of 568 patients with 2571 person-years of follow-up were included in analyses. Compared to off modulator therapy, HEMT was associated with +8.4% TSAT (95% confidence interval [CI], +6.3–10.6%; p<0.0001) and +34.4μg/dL serum iron (95% CI, +26.7–42.1μg/dL; p<0.0001) overall; +5.4% TSAT (95% CI, +2.8–8.0%; p=0.0001) and +22.1μg/dL serum iron (95% CI, +13.5–30.8μg/dL; p<0.0001) in females; and +11.4% TSAT (95% CI, +7.9–14.8%; p<0.0001) and +46.0μg/dL serum iron (95% CI, +33.3–58.8μg/dL; p<0.0001) in males. Ferritin was not different in those taking modulator therapy relative to off modulator therapy. Hemoglobin was overall higher with use of modulator therapy. The prevalence of ID was high throughout the study period (32.8% in those treated with HEMT).

Conclusions:

ID remains a prevalent comorbidity in CF, despite availability of HEMT. Modulator use, particularly of HEMT, is associated with improved markers for ID (TSAT, serum iron) and anemia (hemoglobin).

1. Introduction

Extrapulmonary manifestations are a major source of morbidity and poor quality of life for people with cystic fibrosis (pwCF), and include pancreatic exocrine and endocrine dysfunction, intestinal obstruction, chronic liver disease, reproductive dysfunction, and significant nutritional deficiencies.^1,2 Small, single center studies have reported that iron deficiency (ID) is a common extrapulmonary comorbidity in pwCF, with a reported prevalence of 40–80%.^3–5 The presence of ID increases with advancing age and has been associated with markers for poor outcomes, including shorter time to pulmonary exacerbations, lower lung function, and increased sputum Pseudomonas aeruginosa (PsA).^4–7

ID in other inflammatory chronic diseases such as congestive heart failure, chronic kidney disease (CKD), and inflammatory bowel disease is strongly associated with increased morbidity and mortality, and likely stems from a dysregulation of key iron transporters in the inflamed state.^8,9 The prominent feature of chronic inflammation in cystic fibrosis (CF) raises concern for the clinical impact of ID in CF.^4,10,11 Better characterization of this highly prevalent problem in CF through a systematic approach is required to inform clinical practice and future research directions.

Since the availability of CF transmembrane conductance (CFTR) modulator drugs, the burden of lung disease and overall life expectancy have improved substantially for those with highly-responsive genetic defects.^2,12 However, the effectiveness of modulator therapies is dependent on the specific drug in relation to the individual CFTR mutation.¹² So far, the clinical impact of modulators on lung disease is well-characterized, while the extrapulmonary effects of modulator therapy continue to be reported through real world observational studies; but information specific to ID is lacking.^1,12,13

In this study, we established a multicenter retrospective database including relevant ID variables, which are not available in CF patient registries.^2,14 We hypothesize that use of modulators is associated with improved markers of ID in pwCF. Here, we present original data on the estimated impact of modulator use on ID, stratified by modulator effectiveness relative to CFTR variant. Furthermore, we describe this association in key subgroups, selected based on expected impact on iron homeostasis or important CF clinical outcomes.¹⁵ Lastly, we provide descriptive data on the prevalence of ID in pwCF with respect to modulator use status.

2. Methods

2.1. Study Population:

This retrospective cohort study included adults age ≥18 years with CF from 4 United States (US) CF centers 6/1/2012–2/1/2022 for analyses. Individuals were included in the study if TSAT (%), serum iron (μg/dL), ferritin (μg/L), or hemoglobin (g/dL) laboratory data was available. Data from individuals with any organ transplant were excluded. Repeated measurements were included for each patient for the duration of follow-up, which was restricted to full clinical datasets (i.e., establishment of care until death or loss to follow-up). The study was conducted in accordance with Institutional Review Boards approvals at University of Michigan, University Hospitals Cleveland Medical Center, Dartmouth College, and Maine Medical Center.

2.2. Outcomes and exposures of interest:

The primary outcome of interest was the estimated association between modulator use and laboratory markers: TSAT, serum iron, ferritin, and hemoglobin. Due to the varying effectiveness of modulator drugs in CF, modulator use status was characterized as a 3-level variable: Off, moderately-effective modulator therapy (MEMT), and highly-effective modulator therapy (HEMT). For this study, we limited the definition of HEMT to CFTR variants expected to have a robust modulator response: elexacaftor-tezacaftor-ivacaftor (ETI) in patients with at least 1 copy of F508del, and any modulator in patients with a gating mutation (G551D, G178R, S549N, S549R, G551S, G970R, G1244E, S1251N, S1255P, G1349D) since all formulations contain ivacaftor.¹² MEMT included lumacaftor-ivacaftor or tezacaftor-ivacaftor in patients with 2 copies of F508del. Other clinical situations of CFTR-mutation/modulator therapy combinations were excluded from the analyses to maximize interpretability of the results. Iron deficient status was defined a priori as TSAT <20% (reference range: 20–50%) or serum iron <60 μg/dL (<10.7 μmol/L, reference range: 60–170 μg/dL).^4,5

2.3. Statistical Analysis:

Baseline characteristics and descriptive ID values of the study population are provided using medians and interquartile range (IQR) or percentages as appropriate overall, by subgroups, and by modulator status.

Linear mixed effect models were used to analyze repeated measurements of TSAT, serum iron, ferritin, and hemoglobin values over time; these models assumed a random intercept to account for correlation of measures within the same patient. Time-dependent modulator exposure was modeled as a 3-level variable, with off modulator as the reference. These models were estimated within the overall population and within subgroups with adjustments, where appropriate, for age (years), sex at birth, Global Lung Initiative % predicted of forced expiratory volume in 1 second (ppFEV1), body mass index (BMI, kg/m²), CF-related diabetes (CFRD), pancreatic insufficiency (PI), modulator use duration (months), baseline values of model-specific outcomes at start of follow-up, recent intravenous (IV) or oral (PO) iron supplementation within 6 months, pregnancy, cancer, CKD, methicillin-resistant Staphylococcus aureus (MRSA) or PsA in sputum, and bleeding events classified as minor for estimated blood loss (EBL) of 30–500 mL and major for EBL ≥500 mL.¹⁶

All analyses were performed using R version 4.3.1. A two-sided P-value <0.05 was considered statistically significant.

3. Results

3.1. Baseline characteristics of study population:

Of the total 568 patients, the numbers of individuals contributing data to each ID marker and to each modulator status group are shown in Figure 1. Table 1 provides summary statistics of the study population for the 10-year study duration, showing clinical characteristics similar to the general US CF population.² Total follow-up duration during which repeated measurements for ID were included in statistical analyses were 2571 patient-years from 568 individuals in the overall study, 1639 patient-years from 494 individuals in the off modulator group, 307 patient-years from 149 individuals in the MEMT group, and 567 patient-years from 326 individuals in the HEMT group (Table 2). Individuals could contribute data to more than 1 modulator group throughout the duration of the study period.

Figure 1:

STROBE diagram of the study population, including numbers of patients excluded due to age or organ transplant and numbers of patients contributing data in iron laboratory marker of interest in the overall study and by modulator status group. *Some patients contribute data to several modulator status groups throughout the study period. MEMT = moderately-effective modulator therapy; HEMT = highly-effective modulator therapy.

Table 1:

Baseline characteristic of patients contributing data during the study period.

Study Population Characteristics (n = 568)	Median (IQR) or n (%)
Age (years)*	26.6 (20.8–36.7)
Sex at birth (F)	291 (51.2)
ppFEV1*	69.0 (46.0–89.0)
BMI (kg/m2)*	22.4 (20.0–25.1)
Homozygous F508del	255 (44.9)
Heterozygous F508del	236 (41.5)
Off modulator†	494 (87.0)
On MEMT†	149 (26.2)
On HEMT†	326 (57.4)
CFRD†	183 (32.2)
PI†	507 (89.2)
PsA in sputum†	441 (77.6)
MRSA in sputum†	277 (48.8)
Bleeding event†	113 (19.9)
Pregnancy†	36 (6.4)
Cancer†	9 (1.6)
CKD†	14 (2.5)

Definition of abbreviations: BMI = body mass index; CFRD = cystic fibrosis–related diabetes; CKD = chronic kidney disease; HEMT = highly-effective modulator therapy; MEMT = moderately-effective modulator therapy; MRSA = methicillin-resistant Staphylococcus aureus; ppFEV1 = Global Lung Initiative % predicted forced expiratory volume in 1 second; PI = pancreatic insufficiency; PsA = Pseudomonas aeruginosa.

^*Baseline data of patients at the beginning of individual follow-up periods.

^†Sum of patients with an occurrence at any point during the study period.

Table 2:

Descriptive statistics of study population and study dataset.

	Study Group by Modulator Use Status
	Off Modulator (n = 494)	On MEMT (n = 149)	On HEMT (n = 326)	Total Study|| (n = 568)
TSAT, % (median, IQR)*	12.0 (7.0, 20.3)	14.0 (8.0, 20.0)	21.0 (12.0, 30.0)	15.8 (8.0, 25.0)
Iron, serum, μg/dL (median, IQR)*	38.0 (22.0, 67.3)	45.5 (25.8, 69.3)	73.0 (42.5, 106.5)	48.0 (26.0, 80.3)
Ferritin, µg/L (median, IQR)*	50.7 (21.0, 103.8)	45.0 (22.0, 96.8)	48.0 (31.7, 99.7)	48.1 (24.3, 101.1)
Hemoglobin, g/dL (median, IQR)*	11.7 (10.1, 13.2)	12.1 (10.8, 13.3)	12.3 (10.7, 13.8)	11.9 (10.3, 13.3)
Iron deficient† patient (n, %)‡	150 (30.4)	47 (31.5)	107 (32.8)	222 (39.1)
Iron deficient† lab result (count, %)§	408 (89.3)	96 (77.4)	136 (48.7)	586 (67.1)
IV iron treatment (n, %)‡	25 (5.1)	10 (6.7)	11 (3.4)	42 (7.4)
IV iron treatments (count)	95	31	26	154
IV iron treatment / 100 follow-up years	5.9	10.1	4.6	6.0
PO iron treatment (n, %)‡	80 (16.2)	5 (3.4)	31 (9.5)	144 (25.4)
PO iron treatment (years) / treated person¶	2.1	1.7	1.6	2.1
Follow-up (person-years)	1614.4	307.7	567.2	2571.0
Assessment for iron deficiency§	408	124	279	873
Assessment / follow-up year	0.283	0.403	0.492	0.340

Definition of abbreviations: HEMT = highly-effective modulator therapy; IQR = interquartile range; IV = intravenous; MEMT = moderately-effective modulator therapy; TSAT = transferrin saturation.

^*Reference ranges: TSAT 20–50%; serum iron 60–170 μg/dL; ferritin 15–200 μg/L; hemoglobin 12.3–15.3 (female) and 14.0–17.5 (male) g/dL.

^†Iron deficient is defined as TSAT < 20% or serum iron < 60 μg/dL (< 10.7 μmol/L).

^‡Event rates are presented as any occurrence of iron deficiency or IV iron treatment per patient, throughout the duration of modulator exposure status and in the total study, out of total patients in each respective study group. Not all patients in each study group had assessments for iron deficiency.

^§Assessment for iron deficiency is defined as any laboratory testing for TSAT or serum iron. Event rate is presented as an occurrence of iron deficiency lab results, out of total lab assessments for iron deficiency in each respective study group.

^¶PO iron treatment exposure is presented in years, normalized to number of persons treated with PO iron, during the period of each respective study group and total study.

^||Total study summary statistics includes some data points excluded due to unknown modulator status; thus, data is not summative.

3.2. Extent of iron deficiency in study population:

The scope of ID and practice patterns over the 10-year study period are presented in Table 2. In these descriptive statistics, laboratory markers indicate a substantial prevalence of ID throughout all modulator status groups. These data suggest a trend of improving ID markers with escalating effectiveness of modulator use, which we evaluated further using multivariable models as presented below. Despite this, in the HEMT group: the 25^th percentile of TSAT values was 12.0%; 32.8% of patients were iron deficient at least once during the follow-up period, inclusive of those without ID laboratory assessments; and 48.7% of total laboratory assessments were consistent with ID. Assessments per patient-year increased over time, with highest ratio in the HEMT group of 0.492 assessments/person-year. Similar trends were seen in hemoglobin, with the 25^th percentile=10.7 g/dL in patients using HEMT. IV iron treatment events, a potential indicator for severe ID, were slightly less frequent in the HEMT group. PO iron treatments were high in the HEMT compared to those on MEMT. Patient level ID status changes between modulator groups are summarized in Figure S1 (Supplementary Material).

3.3. Changes in ID laboratory markers associated with modulator therapy:

Forest plots in Figure 2 display estimated differences in laboratory ID markers (A. TSAT, B. serum iron, C. ferritin, D. hemoglobin) for HEMT and MEMT relative to off modulators. To aid in interpretation of the results, summary statistics for laboratory ID markers in the off-modulator reference group are shown in Table S1 (Supplementary Material).

Figure 2:

(A-D) Forest plots displaying the estimated differences in laboratory markers (A. transferrin saturation [TSAT], B. serum iron, C. ferritin, D. hemoglobin) for highly-effective modulator therapy (orange, HEMT) and moderately-effective modulator therapy (blue, MEMT), with dot and bar representing parameter estimate and 95% confidence interval; reference group of off modulator therapy indicated with vertical dashed line (black). Estimated differences are shown for overall study population and by subgroups with adjustments, where appropriate, for age (years), sex at birth, Global Lung Initiative % predicted of forced expiratory volume in 1 second (ppFEV1), body mass index (BMI, kg/m²), CF-related diabetes (CFRD), pancreatic insufficiency, modulator use duration (months), baseline values of model-specific outcomes at start of follow-up, recent intravenous or oral iron supplementation within 6 months, pregnancy, cancer, chronic kidney disease, Methicillin-resistant Staphylococcus aureus and/or Pseudomonas aeruginosa (PsA) in sputum, and bleeding events classified as minor for estimated blood loss (EBL) of 30–500 mL and major for EBL ≥500 mL.

Compared with off modulators, HEMT was associated with higher TSAT (+8.4%; 95% confidence interval [CI], +6.3–10.6%; p<0.0001) and serum iron (+34.4 μg/dL; 95% CI, +26.7–42.1 μg/dL; p<0.0001); while no statistically significant difference was seen in those using MEMT in TSAT (+1.2%; 95% CI, −2.3 to +4.7%; p=0.51) or serum iron (5.9 μg/dL; 95% CI, −6.8 to +18.6 μg/dL; p=0.36) (Figure 2A–B). No significant differences in ferritin levels were observed with either HEMT (−8.9 μg/L; 95% CI, −32.5 to +14.7 μg/L; p=0.46) or MEMT (−2.9 μg/L; 95% CI, −42.3 to +36.5 μg/L; p=0.89) in the overall study population, or in any subgroup (Figure 2C). Hemoglobin values were observed to be higher depending on effectiveness of the modulator, with differences in the overall study population as follows: HEMT (+0.66 g/dL; 95% CI, +0.60–0.76 g/dL; p<0.0001) and MEMT (+0.34 g/dL; 95% CI, +0.22–0.46 g/dL; p<0.0001) (Figure 2D).

The magnitude of the increases in ID markers, TSAT and serum iron, associated with HEMT were subgroup specific (Figure 2A–B). HEMT use in males was associated with higher differences in TSAT (+11.4%; 95% CI, +7.9–14.8%; p<0.0001) and serum iron (+46.0 μg/dL; 95% CI, +33.3–58.8 μg/dL; p<0.0001) compared to the modest differences observed in females (TSAT: +5.4%; 95% CI, +2.8–8.0%; p=0.0001) (serum iron: +22.1 μg/dL; 95% CI, +13.5–30.8 μg/dL; p<0.0001). Modulator use of any type was associated with statistically significantly higher hemoglobin in males (HEMT: +0.87 g/dL; 95% CI, +0.72–1.01 g/dL; p<0.0001) (MEMT: +0.67 g/dL; 95% CI, +0.47–0.87 g/dL; p<0.0001) (Figure 2D). However, only HEMT was associated with statistically significantly higher hemoglobin in females (HEMT: +0.50 g/dL; 95% CI, +0.38–0.62 g/dL; p<0.0001) (MEMT: +0.13 g/dL; 95% CI, −0.01 to +0.27 g/dL; p=0.068). In the off-modulator reference group, summary statistics indicate that the median (IQR) of TSAT, serum iron, and hemoglobin are lower in females compared to males (Table S1).

HEMT was associated with greater differences in TSAT and serum iron in subgroups with lower ppFEV1 (Figure 2A–B). Patients with ppFEV1 <40 had observed increases of +13.2% in TSAT (95% CI, +9.9–16.6%; p<0.0001) and +51.3 μg/dL in serum iron (95% CI, +39.6–62.9 μg/dL; p<0.0001), whereas patients with ppFEV1 ≥40 to <70 had observed increases of +10.2% in TSAT (95% CI, +6.5–13.9; p<0.0001) and +40.0 μg/dL in serum iron (95% CI, +26.5–53.6; p<0.0001). Patients with ppFEV1 ≥70 did not have a statistically significant improvement in TSAT (+2.0%; 95% CI, −1.7 to 5.6%; p=0.29) associated with HEMT relative to the reference. However, the observed improvement in serum iron met statistical significance (+14.2 μg/dL; 95% CI, +1.0–27.3 μg/dL; p=0.037) in this group with preserved lung function. Similar trends were observed in the hemoglobin subgroup analyses, with HEMT associated with the highest observed improvement in those with ppFEV1 <40 (+1.08 g/dL; 95% CI, +0.86–1.30 g/dL; p<0.0001) (Figure 2D). Patients with ppFEV1 ≥70 in the off-modulator group had higher medians (IQRs) of TSAT, serum iron, and hemoglobin relative to those with lower ppFEV1 (Table S1).

Other subgroup analyses showed similar trends of HEMT associated with higher TSAT and serum iron levels, with consistency across the majority of subgroups (Figure 2A–B). The observed associations between modulator use and increased hemoglobin in the BMI, age, CFRD, and sputum PsA subgroups was dependent on the effectiveness of the modulator therapy, with the largest observed improvements associated with HEMT in individuals with BMI <18.5 kg/m² (hemoglobin: +1.58 g/dL; 95% CI, +1.25–1.92 g/dL; p<0.0001) (Figure 2D).

4. Discussion

ID is a major extrapulmonary manifestation with a high prevalence in pwCF,^4,5 and has been associated with an increased all-cause mortality in non-CF populations.^8,9,17 However, little published data characterizes ID, with available literature being predominantly single center descriptive reports.^{3–5,7,11,18–21} The lack of ID markers in currently captured CF patient registries has precluded robust multicenter analyses to date.^2,14 By creating a comprehensive patient database spanning 10 years across 4 clinical institutions, we reduce the problems inherent in single center studies and offset fluctuating practice pattern variations. We also take advantage of information obtained before and after large scale use of HEMT and provide critical data specific to pwCF without access to or intolerant of present-day CFTR modulator therapies.

We found that HEMT was associated with higher TSAT and serum iron, compared to off modulator therapy, both overall and in most subgroups (Figure 2A–B). The magnitude of TSAT increase associated with HEMT is clinically meaningful (8.4% higher overall; 13.2% in those with ppFEV1 <40), considering the narrow clinical reference range (20–50%)²² and the relatively low values observed in the off modulator group of this overall cohort (12.0%) (Table S1). This magnitude of difference is similar to what would be considered therapeutic success as in historic clinical trials of IV iron repletion resulting in improvements of TSAT ranging ~8–11%.²³ In other inflammatory chronic diseases, ID has been associated with increased morbidity and mortality and timely treatment leads to improved outcomes,^9,24 suggesting a potential clinical impact of HEMT due to improvements in iron status.^8,19

MEMT in our cohort was not associated with statistically significant or clinically meaningful differences in TSAT or serum iron values in any study group. This lesser impact of MEMT compared to HEMT on ID markers is commensurate with its smaller effect on pulmonary disease seen in prior clinical studies.¹²

A key subgroup difference emerged in the sex-related magnitude of differences in TSAT and serum iron associated with HEMT (Figure 2A–B). There has been considerable interest in sex differences in the CF population, given the previously-described higher morbidity and mortality in women with CF.^25–27 In our study, ID markers in females were modestly higher while on HEMT compared to male counterparts (Figure 2A–B), but were also lower in the reference (off modulator) female subgroup compared to males (Table S1); thus HEMT further exaggerated the separation in iron status between sexes in our cohort. In the general population without CF, ID is significantly more prevalent in females due to risk factors such as menorrhagia and pregnancy, gastrointestinal bleeding, and nutritional deficiencies, with reported prevalence of up to 40% in some cohorts even in the developed world.^15,17 A speculation for our study results may be that the true iron homeostasis differences seen in the women in general population remain apparent in those taking HEMT, with ID persisting as a major problem. Conversely, newer real-world data suggest that pulmonary exacerbations, a marker for illness in CF, remain higher in women compared to men despite HEMT,²⁷ which raises a concern for persistent sex differences in CF outcomes such as we observed in our cohort.

The improvements in TSAT and serum iron associated with HEMT were more evident in those with lower ppFEV1 (Figure 2A–B). We suspect the lack of difference seen in the subgroup with ppFEV1 ≥70 is likely due to a ceiling effect since we noted that the baseline TSAT and serum iron values in this subgroup were near normal (Table S1), providing less opportunity for improvement. Conversely, those with a low ppFEV1 <40 had lowest baseline TSAT and serum iron across subgroups, providing more room for improvement to normal ranges. A lower iron status in patients with lower lung function was previously observed³ and potentially speaks to the importance of lung disease burden in the pathophysiology of ID in CF,^8,20 even when compared to BMI as a marker for overall nutritional status (Table S1). As such, modulators therapies known to substantially improve lung function may be expected to improve ID.

While ferritin is used as the primary marker for ID in the general population (World Health Organization: <15 µg/L),²⁸ it is also an acute phase reactant expected to increase in the setting of chronic diseases or acute exacerbations of infections.^8,29 Not surprisingly, estimated ferritin was not significantly different with modulator therapy of any type (Figure 2C), likely due to concurrent improvements in overall inflammation and iron stores. In non-CF chronic inflammatory conditions, exceedingly higher thresholds of ferritin have traditionally been used for the diagnosis of ID (<100 µg/L).^8,9 However, recent scrutiny suggests it may be a poor marker for true ID.³⁰ Our results show that trends in ferritin are highest in subgroups with lowest TSAT, serum iron, and hemoglobin (Table S1), supporting that ferritin correlates poorly with other ID markers and may rather indicate pulmonary infections.^4,19

Anemia in CF has historically been the result of multifactorial pathologies, with ID and chronic inflammation as primary etiologies.^6,8 We present hemoglobin laboratory data as a marker for anemia and noted improved hemoglobin associated with modulator use, an observation similar to a US CF registry study (Figure 2D).¹³ In our study, the improvements observed varied respective to modulator effectiveness. Estimated differences in sex- and ppFEV1-based subgroup analyses followed trends like TSAT and serum iron outcomes, with HEMT use associated with highest differences. However, it is important to note that the magnitude of this difference is relatively small; only 2 subgroups in our analyses were observed with hemoglobin increases ≥1 g/dL,^23,24 associated with HEMT alone (ppFEV1 <40: 1.08 g/dL; BMI <18.5 kg/m²: 1.58 g/dL). The magnitude of hemoglobin improvement associated with HEMT in those with low BMI is noteworthy, potentially signaling the impact of a catabolic state on red blood cell production. The estimated impact of HEMT on hemoglobin is more substantial than on markers for ID, pointing to the presence of multifactorial mechanisms for anemia in CF, including non-iron micronutrient deficiencies such as vitamin A/D and folate/B12 on anemia in CF.^1,31,32

In a descriptive survey of ID prevalence specific to modulator usage, several indicators of a high burden of ID emerged (Table 2). Trends show improved laboratory markers with HEMT use, consistent with multivariable model results (Figure 2). However, the proportion of patients who were ID (TSAT <20% or serum iron <60 μg/dL)^4,5,9 remained high, likely due to increased ID testing in the HEMT group. The wide IQR of TSAT and serum iron in the HEMT indicates substantial portions of laboratory results remained low (25^th percentile TSAT: 12.0%; 25^th percentile serum iron: 42.5 μg/dL). Considering subgroup mixed-model results, these unadjusted data may represent a higher proportion of women (Figures 2A–B). PO and IV iron treatment events are dependent on the number of patients as well as follow-up duration for each study group, but provides important treatment pattern information throughout the study period. In patients with available laboratory data, more patients who were ID off modulators were no longer ID on HEMT, compared to MEMT (Figure S1). These results are limited to individuals with either TSAT or serum iron both off and on modulator therapy. However, this observation is complementary to the larger estimated differences in TSAT and serum iron seen with HEMT in adjusted models (Figure 2).

A limitation of this work is the retrospective nature of the study. The results must be interpreted considering the fluctuating clinical practice patterns of diagnostic and treatment approaches for ID in CF. For example, the more contemporary practice of ID screening and liberalized iron therapy likely coincides with the clinical availability of more effective modulator therapies. Furthermore, the lack of standardized routine screening for ID in CF potentially introduces bias towards ID diagnosis by indication of testing. To offset sampling bias, we provide study data availability information to assist with interpretability of these results. More consistent sampling in parallel modulator groups would require a prospective study with predetermined schedules of laboratory data acquisition. Fewer data points available in the MEMT group is largely due to the short time pwCF were prescribed these dual-combination modulators before the clinical availability of ETI.^2,12 Thus the possibility of type 2 error in MEMT group cannot be excluded. In descriptive statistics of the study population (Table 2, Figure S1), we defined iron deficient status based on a critical assessment of prior literature.^4,5,8,9 However, the incidence of ID in CF is highly dependent on specific laboratory thresholds,⁴ suggesting that a clinically meaningful definition should be systematically evaluated in future studies. Lastly, this study excludes post-transplant and pediatric^11,33 patients due to the differing pathophysiology of ID in these populations.

In this study, we found that HEMT was associated with improved markers for ID in patient with CF even after adjusting for confounders that may independently affect iron biology. However, the problem of ID in CF remains substantial even in the era of HEMT and particularly for those not on modulators due to lack of a qualifying mutation, intolerance to treatment, or lack of availability. Well-designed prospective studies are needed to better characterize the clinical impact of ID in CF in the modern era.