The burden of radiation exposure in congenital heart disease: the Italian cohort profile and bioresource collection in HARMONIC project

As part of the ongoing HARMONIC project [12], we presented an Italian national cohort comprising patients with CHD who had not been diagnosed with cancer and had undergone at least one cardiac catheterization. In this single-center, observational, retrospective study, all patients referred for a cath procedure at the Pediatric Cardiology in Niguarda Hospital between 1 January 2015 and 30 October 2023 were screened for inclusion. Hospital medical records for 993 patients were obtained from the pediatric cardiology department. Patients were considered eligible for inclusion if they underwent at least one cardiac cath procedure at an age < 22 years. Based on the criteria set by HARMONIC [12], patients were excluded if they lacked follow-up data, or had received a cancer diagnosis before the first cath procedure. The flowchart in Fig. 1 shows the exclusion and inclusion of patients in this study. A total of 584 patients (261 females, 323 males) with 1181 cath procedures were included in the cohort.

The study was approved by the Institutional Ethical Committee (Study Protocol n. 605–112019).

We retrieved clinical, demographics and procedural information for each patient from the electronic hospital's archive.

Dose cath indicators, including Kerma Area Product (KAP, Gy·cm²) and Fluoroscopic time (FT, min) were recorded for each patient during the procedure by the system’s radiation dosimeter coupled to the angiography imaging system.

Until August 2010, cath procedures were performed using a Philips Integris H5000 system (Philips Healthcare). Subsequently, the hospital upgraded its equipment with a GE Innova 2100-IQ system (General Electric Healthcare) for diagnostic and interventional procedures and Philips Integris Allura, specifically for electrophysiology procedures.

In September 2012, a radiation dose-reduction program was also implemented in the catheterization laboratory. All medical and technical staff participated in a training course focused on optimizing doses in fluoroscopy-guided procedures, covering aspect such as fluoroscopy levels, frame rates and geometrical parameters.

Additionally, hospital medical records were also used to retrive information on other high dose procedures, including computed tomography (CT), nuclear medicine (NM) scans and whole body or cardiac positron emission tomography-computed tomography (PET-CT).

Radiation exposure was extracted from imaging reports by collecting Dose‐Length Product (DLP, mGy·cm) for CT and PET-CT, and administered activity (MBq, Megabecquerel) of the tracer for NM scans.

Biological samples were collected in a subgroup of these patients who underwent diagnostic and/or therapeutic cardiac catheterization procedures, at three-time points: immediately before, immediately after the procedure and approximately 1 year afterward. Informed consent was obtained from all subjects or their parents.

Briefly, a maximum of 5 mL of blood sample was collected in a vacutainer tube containing EDTA K2, one clot activator serum separation tube, and one BD Vacutainer® CPT™ tube for isolation of lymphocytes. Within two hours post collection, blood samples were centrifuged according to standard operating procedures to obtain plasma, serum, and lymphocytes. Aliquots of each fraction were prepared and immediately stored at − 80 °C for future biomarker analysis, as previously reported [13]. Saliva samples were also acquired and divided into aliquots.

Categorical variables are expressed as frequencies and percentages, while continuous variables as median and interquartile range (IQR): 25th and 75th percentiles. Nonparametric tests were chosen due to the skewed distribution. The comparison between two groups was assessed using the Mann–Whitney U test and correlations between continuous variables were tested by using Spearman’s rho value.

Patients were stratified into age categories at the procedures: newborn (0–30 days), infants (1–12 months), 1–5 years, 6–10 years, 11–15 years; and > 15 years.

The baseline and clinical characteristics of patients are detailed in Table 1. Among CHD patients, 1181 diagnostic and/or therapeutic cardiac catheterization procedures were conducted, averaging 2.02 procedures per person.

The median age at first exposure was 1 year (IQR: 0–8). Two hundred fifty-nine (44%) patients underwent their first procedure before reaching 1 year of age, and 109 patients (19%) underwent catheterization within hours or days after birth.

Additionally, 300 other high dose procedures (CT, PET-CT, and NM scan) were also performed in 148 CHD patients.

Dosimetric data were adequately collected for 94% of the total catheterization procedures.

The median KAP/weight was 0.09 Gy·cm²/kg (IQR: 0.03–0.20), and the median FT was 8.10 min (IQR: 4.00–16.25). KAP/weight showed a positive correlation with FT (Spearman’s rho = 0.679, p < 0.001).

Significant differences were observed in the KAP/weight and FT between diagnostic and therapeutic procedures (median KAP/weight 0.08 Gy·cm²/kg IQR: 0.03–0.17 versus 0.10 Gy·cm²/kg IQR: 0.05–0.25, p < 0.001, and median FT 6.00 min, IQR: 3.00–13.02 versus 11.22 min IQR: 6.32–23.09, p < 0.001, respectively).

Dosimetric parameters based on the individual procedure type are presented in Table 2.

When patients were categorized into age subgroups, a statistically significant difference (P < 0.001) was observed in the distribution between diagnostic and therapeutic procedures. The proportion of interventional cases decreased progressively with advancing years (Fig. 2). The radiation dose, expressed as KAP/weight, was significantly higher in therapeutic procedures compared to diagnostic ones in the age group > 15 years (P < 0.05 adjusted by multiple comparisons). Conversely, FT displayed significant differences between procedures in all age groups, except in the newborn group (Fig. 3).

The comparison of exposure data reveals a significant dose reduction with the upgraded technology system, transitioning from Philips Integris H5000 (N. of procedures = 86) to GE Innova 2100-IQ (N. of procedures = 1008) or Philips Integris ALLURA (N = 20). The KAP/weight exhibited a significant decrease in the upgraded equipment compared to the previous system (Fig. 4). In contrast, no significant difference was observed for FT parameters between equipments.

Moreover, we compared dose parameters before and after the training course carried out in September 2012. The analysis highlighted statistically significant differences in KAP/weight and FT between the two periods, as depicted in Fig. 5.

In our cohort, 148 patients underwent a total of 300 other high dose imaging procedures (272 CT and 28 NM scans), averaging 2.03 procedures per patient. DLP data for CT scans were available for only 219 exams, with a median exposure of 239 mGy·cm (IQR: 56–570). Data for NM scans were available for 25 patients, with a median of administered activity of 185 MBq (IQR: 119–217). The different types of CT and NM, with their median DLP or MBq values, are presented in Table 3. CT and NM scans accounted for only 21% of all high-dose procedures. The distribution of diagnostic cardiac catheterizations, interventional cardiac catheterizations, CTs and NM scans in the whole cohort is shown in Fig. 6.

A total of 60 patients were enrolled in the prospective study, comprising 38 males with a median age of 7 years (IQR: 5–12). Among them, 27 patients had no previous catheterization procedures, while 33 patients had at least one previous cardiac catheterization procedure. Twenty-six patients (20 males; median age 5 years, IQR: 5–11) underwent diagnostic procedures while 34 (19 males; median age 7 years, IQR: 8–12) therapeutic ones.

The median KAP/kg value was 0.07 Gy·cm²/kg (IQR: 0.03–0.16).

The Italian cohort collected a total of 1858 samples, including six different types of specimens: blood, plasma, serum, clot, cell pellet/lymphocytes and saliva. Figure 7 illustrates the study design for the collection of biological samples.