Iron deficiency in pulmonary arterial hypertension associated with congenital heart disease

Abstract

Objectives. We aimed to investigate the prevalence of iron deficiency (ID) in congenital heart disease associated with pulmonary arterial hypertension (CHD-PAH) and to explore the influence of ID on CHD-PAH patients. What was associated with ID in these patients was also looked into. Design. One hundred and fifty-three patients who were newly diagnosed with CHD-PAH were enrolled. Patients were divided into iron-deficient and iron-replete groups according to the following criteria. ID was defined as transferrin saturation <20% in male and transferrin saturation <25% in female. Clinical data of all participants were collected and compared. Logistic regression was performed to explore factors associated with ID in CHD-PAH. Results. Thirty-nine percent of 153 CHD-PAH patients were founded with ID. Iron-deficient group had greater proportion of female patients, shorter six minutes walking distance (6-MWD), higher N-terminal pro-brain natriuretic peptide levels, lower creatinine levels, greater ratio of diastolic right ventricle diameter to left ventricle diameter. Female (OR = 15.44, 95%CI 4.91–48.54, p < .01), 6-MWD (OR = 0.99, 95%CI 0.98–1.00, p = .02) and mean right atrial pressure (OR = 1.13, 95%CI 1.02–1.26, p = .02) were independently associated with ID in the overall CHD-PAH patients. Menstruation was independently associated with ID in the female subgroup (OR = 3.88, 95%CI 1.09–13.84, p = .04). Conclusions. ID was highly prevalent in CHD-PAH patients. Worse exercise tolerance and right heart function were observed in iron-deficient patients with CHD-PAH. Female, 6-MWD, mean right atrial pressure and menstruation are important variables indicating the presence of ID in CHD-PAH.

Keywords:

• Congenital heart disease
• Pulmonary arterial hypertension
• Transferrin saturation
• Iron deficiency

Introduction

Congenital heart disease associated with pulmonary arterial hypertension (CHD-PAH), which belongs to a subtype of group 1 pulmonary hypertension, is one of the reasons for high morbidity and mortality in CHD. Previous studies reveal that the prevalence of PAH in CHD ranges from 4% to 28% [1,2]. According to the European Society of Cardiology/European Respiratory Society guidelines [3], CHD-PAH is divided into the following clinical classifications: 1) Eisenmenger’s syndrome; 2) PAH associated with prevalent systemic-to-pulmonary shunts; 3) PAH with small/coincidental defects; 4) PAH after defect correction. As CHD-PAH progresses, hyperkinetic PAH will be converted to organic one, which is usually accompanied by bi-directional or reverse shunting [4]. Venous blood without adequate oxygenation directly flows into systemic circulation and causes hypoxemia [5]. Secondary erythrocytosis triggered by hypoxemia could increase iron consumption in CHD-PAH patients. Thus, the maintenance of iron homeostasis might be broken.

Iron, an essential trace element, plays key roles in normal physiological processes including the synthesis of hemoglobin and myoglobin, oxygen delivery and energy metabolism [6]. Since 2011, numerous studies have focused on iron deficiency (ID) in PAH, demonstrating that ID is highly prevalent and capable of influencing exercise tolerance and hemodynamics [7,8]. Furthermore, iron supplementation could improve life quality and prognosis of PAH patients accompanied by ID [9]. However, participants in previous studies were mainly idiopathic PAH. Whether these results are also true for CHD-PAH remains unknown. Our study aimed to investigate the prevalence of ID in CHD-PAH and to explore the influence of ID on these patients. In addition, potential factors associated with ID in CHD-PAH patients were also determined.

Materials and methods

This single-center study was conducted in Fuwai Hospital, National Center for Cardiovascular Diseases in Beijing, China. The study was carried out with the approval of the Fuwai Hospital Ethics Committee. Written informed consent of all participants was obtained.

Study sample

Patients who were newly diagnosed with CHD-PAH between September 2015 and December 2017 were retrospectively enrolled. Clinical materials of all patients were collected and assessed. The diagnosis of “CHD-PAH” was established as follows [3]: 1) mean pulmonary arterial pressure (mPAP) ≥25 mm Hg with pulmonary capillary wedge pressure ≤15 mm Hg at rest measured by right heart catheterization; 2) cardiac defects or defect correction confirmed by echocardiography; 3) no other forms of pulmonary hypertension being present. Patients having one of the following conditions were excluded [10]: 1) age less than eighteen years old; 2) pathological conditions affecting iron status like massive hemoptysis and hypermenorrhea; 3) major surgery within a year; 4) hematological tumor; 5) recent iron supplementation therapy; 6) severe liver or kidney dysfunction. Transferrin saturation is a more favorable marker of iron status than serum iron and ferritin [11,12]. In this study ID was defined as transferrin saturation <20% in male and transferrin saturation <25% in female. Patients were divided into iron-deficient and iron-replete groups according to the above criteria.

Measurements

Baseline characteristics including age, gender, body mass index, six minutes walking distance (6-MWD), and World Health Organization functional class were collected. All participants underwent echocardiographic examination on the first day of admission. Several important echocardiographic variables were measured and calculated including the ratio of diastolic right ventricle diameter to left ventricle diameter (RVD/LVD), systolic pulmonary arterial pressure, the existence of pericardial effusion and abnormal cardiac structures. Besides, CHD-PAH patients were divided into different clinical classifications based on echocardiographic measurements. Iron parameters, hemoglobin, hematocrit, mean corpuscular volume, red blood cell width, N-terminal pro-brain natriuretic peptide (NT-proBNP), uric acid, creatinine and high sensitive C-reactive protein were also regularly measured using fasting venous blood samples on the second day of admission. To confirm the diagnosis of PAH, right heart catheterization was conducted by experienced pulmonary vascular clinicians. Mean right atrial pressure (mRAP), mPAP, cardiac index and pulmonary vascular resistance (PVR) were measured and calculated.

Statistical analysis

Continuous variables are presented as mean ± standard deviation or median (interquartile range). Categorical variables are given as counts or percentages. In order to compare baseline parameters between two groups, Student’s t-test was used for normally distributed continuous variables and Mann-Whitney U test was used for non-normally distributed continuous variables, categorical variables were compared using χ2 test. Potential factors associated with ID in CHD-PAH were determined using logistic regression analysis. Those factors with P values less than 0.1 univariably were put into multivariate regression analysis with the "Enter" procedure. P value < 0.05 was considered as statistically significant. Data analysis was performed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA).

Results

There were 168 patients diagnosed with CHD-PAH. Among them, 7 patients were less than eighteen years old; 1 patient had abnormal renal function; massive hemoptysis occurred to 3 patients; hypermenorrhea was demonstrated in 4 female patients. Finally, 153 patients with CHD-PAH were included to analyze.

Demographic materials and baseline characteristics of patients

The characteristics of the patients were summarized in Table 1. Among the whole population, 60 patients (39%) were founded with ID. There was a difference in female sex between the two groups (92% vs 46%, p < .01). The prevalence of ID in the male and female subgroup patients were 9% and 56% respectively. In the female subgroup, the prevalence of ID in premenopausal (n = 81) and postmenopausal (n = 17) patients were also compared. The results showed that the prevalence of ID in premenopausal and postmenopausal women were 60% and 35% respectively (p = .06). In addition, 6-MWD was shorter in CHD-PAH patients with ID [(389 ± 99) m vs (448 ± 88) m, p < .01]. There was no difference in World Health Organization functional class, age, body mass index and clinical classifications of CHD-PAH between the two groups.

Table 1. Comparison of baseline characteristics between the two groups.

	Iron-deficient (n = 60)	Iron-replete (n = 93)	p-value
Clinical parameters
Female sex [n(%)]	55 (92)	43 (46)	<.01
Age (years)	35 ± 12	37 ± 13	.37
BMI (kg/m^2)	20.5 ± 3.8	21.0 ± 3.7	.28
WHO FC I/II/III/IV (n)	3/35/15/7	10/51/29/3	.11
6-MWD (m)	389 ± 99	448 ± 88	<.01
CHD-PAH classifications [n(%)]			.91
Eisenmenger’s syndrome	28 (47)	48 (52)
PAH associated with prevalent systemic-to-pulmonary shunts	14 (23)	21 (23)
PAH with small/coincidental defects	8 (13)	12 (13)
PAH after defect correction	10 (17)	12 (13)
Laboratory parameters
NT-proBNP (pg/ml)	517.8 (198.9–1784.5)	318.4 (94.0–860.0)	.01
Uric acid (μmol/L)	388 ± 144	402 ± 122	.14
Creatinine (μmol/L)	70.2 ± 12.9	75.8 ± 14.2	<.01
EGFR (mL/min/1.73 m^2)	100.9 ± 16.5	101.6 ± 16.6	.79
Hs-CRP (mg/L)	1.0 (0.6–2.5)	0.9 (0.5–1.7)	.36
Echocardiographic parameters
RVD/LVD	0.89 ± 0.33	0.78 ± 0.32	.03
SPAP (mmHg)	98 ± 27	94 ± 28	.37
Pericardial effusion [n(%)]	7 (12)	6 (7)	.26
Hemodynamic parameters
MRAP (mmHg)	7 ± 4	6 ± 4	.06
MPAP (mmHg)	60 ± 20	63 ± 20	.41
Cardiac index (L/min/m^2)	3.7 ± 1.2	3.7 ± 1.3	.91
PVR (dyn·s/cm^5)	936.6 ± 681.7	780.4 ± 528.7	.33

BMI: body mass index; WHO FC: World Health Organization functional class; 6-MWD: 6 minutes walking distance; CHD: congenital heart disease; PAH: pulmonary arterial hypertension; NT-proBNP: N-terminal pro-brain natriuretic peptide; EGFR: estimated glomerular filtration rate; Hs-CRP: high sensitive C-reactive protein; RVD/LVD: the ratio of diastolic right ventricle diameter to left ventricle diameter; SPAP: systolic pulmonary artery pressure; MRAP: mean right atrial pressure; MPAP: mean pulmonary arterial pressure; PVR: pulmonary vascular resistance.

In terms of laboratory testing, iron–deficient group had higher NT-proBNP levels [517.8 (198.9–1784.5) pg/ml vs 318.4 (94.0–860.0) pg/ml, p = .01] and lower creatinine levels[(70.2 ± 12.9) μmol/L vs (75.8 ± 14.2) μmol/L, p < .01].

Echocardiographic measurements showed that RVD/LVD was higher (0.89 ± 0.33 vs 0.78 ± 0.32, p = .03) in iron-deficient group compared with iron-replete group. There was a trend toward higher values of mRAP in CHD-PAH patients with ID, however it did not reach a difference (p = .06). mPAP, cardiac index and PVR did not make differences between two groups either.

As shown in Table 2, parameters associated with iron homeostasis including serum iron, total iron binding capacity, unsaturated iron binding capacity, transferrin saturation, ferritin, transferrin, hemoglobin, hematocrit, mean corpuscular volume and red blood cell width were different between the two groups (all p < .01).

Table 2. Comparison of the iron parameters between the two groups.

	Iron-deficient (n = 60)	Iron-replete (n = 93)	p-value
Serum iron (μmol/L)	9.4 ± 4.5	24.8 ± 7.8	<.01
TIBC (μmol/L)	70.9 ± 12.4	61.6 ± 9.3	<.01
UIBC (μmol/L)	61.5 ± 13.6	36.8 ± 10.6	<.01
Transferrin saturation (%)	14 ± 7	41 ± 14	<.01
Ferritin (μg/L)	17.5 (8.3–42.2)	114.6 (48.3–212.5)	<.01
Transferrin (g/L)	3.1 ± 0.6	2.6 ± 0.5	<.01
Hemoglobin (g/L)	149.0 ± 30.0	171.9 ± 24.7	<.01
Hematocrit (L/L)	0.46 ± 0.09	0.51 ± 0.07	<.01
MCV (fL)	84.9 ± 9.8	89.0 ± 5.5	<.01
Red blood cell width (%)	15 ± 3	13 ± 1	<.01

TIBC: total iron binding capacity; UIBC: unsaturated iron binding capacity; MCV: mean corpuscular volume.

Possible factors associated with ID in CHD-PAH

Factors potentially associated with ID in CHD-PAH including age, female, body mass index, 6-MWD, RVD/LVD, NT-proBNP, estimated glomerular filtration rate, mRAP, mPAP, PVR and cardiac index were put into univariate logistic regression analysis, which demonstrated that female (OR = 12.79, 95%CI 4.70–34.84, p < .01), 6-MWD (OR = 0.98, 95%CI 0.97–0.99, p < .01), RVD/LVD (OR = 2.78, 95%CI 1.01–7.65, p = .05) and mRAP (OR = 1.09, 95%CI 1.00–1.18, p = .04) were associated with ID in CHD-PAH. Above four factors with P values less than 0.1 were then enrolled into multivariate regression analysis, and the "Enter" procedure was employed. The results displayed that female (OR = 15.44, 95%CI 4.91–48.54, p < .01), 6-MWD (OR = 0.99, 95%CI 0.98–1.00, p = .02) and mRAP (OR = 1.13, 95%CI 1.02–1.26, p = .02) were independently associated with ID in CHD-PAH patients. The details were shown in Table 3.

Table 3. Result of logistic regression for the whole population.

	Univariate model			Multivariate model
variable	OR	95%CI	p-value	OR	95%CI	p-value
Age	0.99	0.96–1.01	.34
Female	12.79	4.70–34.84	<.01	15.44	4.91–48.54	<.01
BMI	0.96	0.88–1.05	.40
6-MWD	0.98	0.97–0.99	<.01	0.99	0.98–1.00	.02
RVD/LVD	2.78	1.01–7.65	.05	1.22	0.36–4.10	.75
NT-proBNP	1.01	0.99–1.02	.34
EGFR	1.00	0.98–1.02	.79
MRAP	1.09	1.00–1.18	.04	1.13	1.02–1.26	.02
MPAP	0.98	0.97–1.05	.41
PVR	0.99	0.98–1.01	.12
Cardiac index	1.03	0.83–1.27	.82

BMI: body mass index; 6-MWD: 6 minutes walking distance; RVD/LVD: the ratio of diastolic right ventricle diameter to left ventricle diameter; NT-proBNP: N-terminal pro-brain natriuretic peptide; EGFR: estimated glomerular filtration rate; MRAP: mean right atrial pressure; MPAP: mean pulmonary arterial pressure; PVR: pulmonary vascular resistance.

Meanwhile, we performed the same logistic regression procedure in the female subgroup. The results showed that menstruation was independently associated with ID (OR = 3.88, 95%CI 1.09–13.84, p = .04). The details were shown in Table 4.

Table 4. Result of logistic regression for the female group.

	Univariate model			Multivariate model
variable	OR	95%CI	p-value	OR	95%CI	p-value
Age	0.99	0.96–1.02	.49
Menstruation	2.81	0.94–8.35	.06	3.88	1.09–13.84	.04
BMI	1.03	0.92–1.16	.64
6-MWD	1.00	0.99–1.00	.04	1.00	0.99–1.00	.07
RVD/LVD	0.80	0.23–2.86	.73	1.23	0.36–4.18	.74
NT-proBNP	1.00	1.00–1.00	.14
EGFR	1.00	0.98–1.03	.81
MRAP	1.11	0.99–1.25	.07	1.12	0.98–1.27	.09
MPAP	1.01	0.99–1.03	.27
PVR	1.00	1.00–1.00	.16
Cardiac index	0.99	0.77–1.26	.91

BMI: body mass index; 6-MWD: 6 minutes walking distance; RVD/LVD: the ratio of diastolic right ventricle diameter to left ventricle diameter; NT-proBNP: N-terminal pro-brain natriuretic peptide; EGFR: estimated glomerular filtration rate; MRAP: mean right atrial pressure; MPAP: mean pulmonary arterial pressure; PVR: pulmonary vascular resistance.

Discussion

Previous studies barely focused on the relationship between ID and CHD-PAH. In this study, we found that the prevalence of ID was high in CHD-PAH patients. CHD-PAH patients with ID had greater proportion of female patients, shorter 6-MWD, higher NT-proBNP levels and RVD/LVD, which reflected their worse exercise tolerance and right heart function to some degree. Multivariate regression analysis demonstrated that female, 6-MWD, and mRAP were independently associated with ID in the overall CHD-PAH patients. Furthermore, menstruation was independently associated with ID in the female subgroup.

Looker AC et al. [13] found that less than 1% of young men and 9% to 11% of young women were iron deficient in the United States. Nevertheless, our study demonstrated that the prevalence of ID in male and female CHD-PAH patients were 9% and 56% respectively. Thus, CHD-PAH patients had much higher prevalence of ID than general population did.

We noticed that CHD-PAH patients with ID were mainly female (92%), which was in line with the research by Kaemmerer et al. [14]. Female was independently associated with ID in the overall CHD-PAH patients. Subanalysis of female patients showed a trend toward higher prevalence of ID in premenopausal patients (60% vs 35%, p = .06). After adjustment for 6-MWD, RVD/LVD and mRAP, menstruation was independently associated with ID in the female group. Thus, we thought premenopausal patients may be more susceptible to ID than men or postmenopausal women in the setting of CHD-PAH, due to their recurrent menstruation and less iron store [15].

Iron is a key component of hemoglobin and myoglobin. Therefore, ID could impair oxygen transport capacity of blood and disturb muscle oxygen hemostasis, leading to shortened 6-MWD in the iron-deficient group in our study, which was consistent with the results of Ruiter et al. [16,17]. Hence, shortened 6-MWD should be considered as the external manifestation of ID in these patients. This could be an explanation for why 6-MWD was independently associated with ID in CHD-PAH.

CHD-PAH patients with ID also had higher NT-proBNP levels and RVD/LVD. Of note, 6-MWD, NT-proBNP levels and RVD/LVD are associated with right heart function and prognosis of PAH patients. Thus, we speculated that ID might negatively affect the clicial status even the prognosis of CHD-PAH. However, this hypothesis needs to be confirmed in a prospective study.

There was a trend toward higher values of mRAP in the iron-deficient group (p = .06). Multivariate regression analysis demonstrated that mRAP was independently associated with ID in those patients. Previous studies revealed that mRAP was associated with the severity and prognosis of PAH [18]. This trend in the iron-deficient group may indicated that they had more severe PAH and bi-directional or reverse shunting, leading to more severe hypoxemia and significant secondary erythrocytosis. In this way, mRAP may be associated with ID in CHD-PAH.

There was another evidence supporting the hypothesis mentioned above. No difference was obseved in the clinical classifications of CHD-PAH between the two groups. In other words, the proportion of bi-directional or reverse shunting (i.e., Eisenmenger's syndrome) were comparable between the two groups. Hence, we speculated that ID was more likely to correlate with the severity of bi-directional or reverse shunting rather than the phenomenon of the shunting itself. The more severe bi-directional or reverse shunting was, the more likely ID appeared. Previous research performed by Ruiter et al. [17] showed that iron parameters were not associated with hemodynamics. Nevertheless, Van Empel et al. [7] found that ID was associated with hemodynamics and the explanation for this discrepancy might be the differences in the time interval between right heart catheterization and iron measurement. However, we failed to demonstrated the association between mPAP/PVR and ID. Further work is needed to test this hypothesis.

Tay et al. [19] found that administration of iron could improve exercise tolerance and life quality of CHD-PAH patients with ID. But for those who had relatively higher hemoglobin concentration, iron supplementation may aggravate blood viscosity by excessive erythropoietic response [20]. Since the prevalence of ID was high in CHD-PAH patients and those with ID demonstrated worse clinical status, detection of ID and individualized iron therapy in CHD-PAH patients are advised.

Limitations

The small number of participants in our study should be first noted. Second, as a single-center retrospective study, we did not follow up. Whether female, 6-MWD and mRAP are prognostic indicators for CHD-PAH patients remains unknown. Third, we only compared the ID prevalence in CHD-PAH patients with the epidemiological data of ID in general population, and no normal control group was set up. Finally, CHD has several subtypes including multiple abnormal structures. This makes CHD-PAH patients a heterogeneous group, which may affect the accuracy of our results to some degree. A multicenter, prospective study is further needed to identify the influence of ID on the development and prognosis of CHD-PAH.

Conclusion

The prevalence of ID was high in CHD-PAH patients, no matter it was examined in the whole population or the subgroups. CHD-PAH patients with ID demonstrated worse clinical status. Female, 6-MWD and mRAP were independently associated with ID in the overall CHD-PAH patients. Clinicians should pay close attention to the iron status of CHD-PAH patients, especially that of premenopausal ones. It is of great concern to recognize ID in CHD-PAH patients and give appropriate treatment.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by National Natural Science Foundation of China [81370326, 81641005], Beijing Municipal Science and Technology Project [Z181100001718200] and National Precision Medical Research Program of China [2016YFC0905602].