Introduction

Iodine-based contrast media (ICM) are widely used in clinical practice for various X-ray-based modalities, and can be classified, according to their osmolality, into hyperosmolar CM (HOCM), low-osmolar CM (LOCM), and iso-osmolar CM (IOCM) [1]. They can be further subdivided into ionic and non-ionic CM, which do not dissociate into ions in water and are therefore lower in osmolality [2].

Ioversol (Optiray®, Guerbet) is a non-ionic, monomeric LOCM, with an osmolality between 502 and 792 mOsm/kg, depending on iodine concentration (240, 300, 320, or 350 mg I/mL).

Despite the generally good safety profile of ICM, adverse drug reactions (ADRs) may occur and can be life threatening. Among these reactions, there are hypersensitivity reactions (HSRs) [3]. Immediate (acute) HSRs occur within 1 h after ICM administration and may include urticaria, angioedema, bronchospasm, laryngeal edema, and anaphylactic shock. Non-immediate (delayed) HSRs, with symptoms occurring between 1 h and several days after ICM administration, commonly manifest as delayed urticaria and maculopapular exanthema, and rarely as severe cutaneous adverse reactions (SCARs) [3].

Post-contrast acute kidney injury (PC-AKI) is a complication that might occur after intravascular exposure to ICM. PC-AKI has been associated with excess morbidity and mortality [4,5,6], and chronic kidney disease (CKD) is the most well-known risk factor [7]. The risk of PC-AKI could increase from 5% at an estimated glomerular filtration rate (eGFR) ≥ 60 to 30% at an eGFR < 30 mL/min/1.73 m2 [8]. Several definitions of PC-AKI, based on serum creatinine (SCr) concentration, have been proposed by different initiatives, the European Society of Urogenital Radiology (ESUR) [9], the Acute Kidney Injury Network (AKIN) [10], and the Kidney Disease Improving Global Outcomes (KDIGO) being the most recent [11].

As the causal relationship between ICM exposure and the occurrence of AKI is often confounded by several patient- and procedure-related factors, the term PC-AKI is preferred for AKI associated with CM administration for studies lacking a control population [9]. Only when the ICM is demonstrated as the causative factor is the term contrast-induced acute kidney injury (CI-AKI) or contrast-induced nephropathy (CIN) appropriate.

To support radiologists in their clinical practice, we sought to perform this systematic analysis of literature on the incidence of ADRs, HSRs, and PC-AKI after intravenous (IV) administration of ioversol and to position the safety profile of ioversol among the different ICM. Complications after intra-arterial administration will be discussed in a future review.

Materials and methods

This systematic literature review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [12].

Data sources and searches

A search of MEDLINE (PubMed) and EMBASE (Elsevier) references from January 1980 to May 2021 was performed using keywords related to adverse events usually associated with the use of ICM such as “allergic reaction,” “hypersensitivity,” “anaphylactic,” “nephrotoxicity,” and “kidney injury” (Appendix 1).

Study selection

Clinical studies documenting exposure to IV ioversol and the presence or absence of ADRs, and/or HSRs, and/or PC-AKI were included. Systematic or descriptive reviews, commentaries, letters, or case reports were excluded. Studies with less than 5 patients exposed to ioversol were excluded.

Study selection was conducted and reconciled between two independent authors. After a first screening step of all identified references, based on titles and abstracts, a full-text screening of potentially relevant publications was performed. Additional relevant publications were identified by cross-referencing.

Data extraction and study quality assessment

Key data extracted from selected articles were as follows: study design, patient characteristics, indication for which ioversol was used, number of patients exposed to ioversol and other ICM (if any) or number of administered doses, ICM dose, type of safety outcome and incidence, intensity [13] and seriousness if reported, and definition of PC-AKI (when applicable).

The methodological quality of the non-randomized studies was assessed using a modified Newcastle-Ottawa Scale (NOS) [14]. The score ranged from 0 to 8, based on 8 questions (one question excluded as not appropriate for safety outcomes) related to patient selection, comparability of cohorts, and outcomes assessment. Scores of 7–8 and 5–6 indicated high-quality and moderate-quality studies, respectively. The revised Cochrane Risk of Bias assessment tool for randomized trials (ROB 2) algorithm was used for randomized controlled trials (RCT) [15].

Results

Study selection

Among the 556 articles identified, 132 underwent a full-text screening and 4 articles were identified through citation tracking [16,17,18,19]. Finally, 31 articles were included: 16 related to the ioversol clinical development program [20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35] and 15 from other studies (Fig. 1). Twenty-five studies had a prospective design and 11 were RCT [20,21,22,23,24,25, 28, 29, 31, 36, 37]. Four studies were on pediatric patients [27, 30, 38, 39].

Fig. 1
figure 1

Flow diagram of the search strategy and study selection

Full size image

The NOS was applied to all non-RCT and one RCT (randomized for patient hydration and not for ICM allocation) [36], indicating high quality for 4 studies and medium quality for 18 studies. All RCTs had a low risk of bias, except one [37] where some concerns linked to a potential performance bias were raised as the study was not double blinded.

Twenty-nine studies indicated the number of patients exposed to ioversol (total of 57,837 patients, including 13,484 pediatric patients) while two studies indicated the number of administered doses of ioversol, with more than 1.5 million in An et al [17] and 20,958 doses in Morales et al [40] (Table 1).

Table 1 Description of all selected studies
Full size table

In adult studies conducted during the clinical development of ioversol, the mean administered dose ranged between 50 and 176 mL, while sparse information was retrieved from the other adult studies. In pediatric patients, the injected dose was 1–3 mL/kg [27, 30, 38, 39].

Among the selected studies, 26 [17, 18, 20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38, 40, 42,43,44,45] documented the incidence of all ADRs or specifically HSRs (56,502 patients and 1,613,481 doses) and 5 studies [16, 19, 39, 46, 47] reported the incidence of PC-AKI (1335 patients). Contrast-enhanced CT was the main indication for which ioversol was used, followed by venography and urography. The mean age was 28–78 years old in adult studies and 5–10 years old in pediatric studies.

Twelve publications reported information on intensity of reactions (Table 2), with detailed information on the methodology of classification in 4 of them (Table 3). In addition, 4 publications reported information on seriousness of reactions (Table 2).

Table 2 Incidence of ADRs/HSRs after intravenous administration of ioversol
Full size table
Table 3 Event classification by intensity
Full size table

Adverse drug reactions and hypersensitivity reactions

The overall incidence of ADRs in adults was reported in 15 studies [17, 20,21,22,23,24,25,26, 28, 29, 32, 33, 35, 42, 43] with a median of 0.23%. In two studies where heat sensation was assessed in a specific questionnaire, a higher incidence of ADRs was reported (42–63%) [31, 34].

In 12 studies of ioversol clinical development (658 patients), the median incidence of ADRs was 1.65% (range: 0–33.3%), with 6 studies reporting no ADRs (Table 2). The highest incidence was reported in a pharmacokinetic study [21], where 6 of 18 patients reported ADRs, none of which was severe. Overall, most of the reported ADRs were minor and consisted of nausea, vomiting, and headache.

Three other studies reported incidences between 0.13 and 0.23% [17, 42, 43]. Vogl et al [42] reported ADRs in 0.28% of 10,836 patients, mainly urticaria (0.12%), nausea (0.10%), and erythema (0.06%). Four serious ADRs (0.037%) were reported, including 3 anaphylactoid reactions requiring hospitalization (0.028%).

An et al [17] reported an incidence of ADRs with ioversol of 0.23%, with urticaria (47.3%) and itching (43.9%) being the most frequent acute ADRs, and maculopapular rash (88%) the most frequent delayed ADR. The incidence of serious ADRs with ioversol was 0.01% (no deaths reported) (Table 2).

Chen et al [43] showed that ADRs were mainly evocative of HSRs, with an incidence of 0.13% for ioversol. Only one anaphylactic shock reaction (0.019%) and no case of laryngeal edema was reported with ioversol for 5261 patients exposed. The incidence of moderate and severe ADRs with ioversol was 0.02%, no deaths induced by ICM were reported, and all ADRs resolved.

The incidence of HSRs with ioversol was explicitly reported in two studies (0.2–0.66%) [40, 44] (Table 2). Morales et al [40] included patients with a previous history of HSRs to ICM. The incidence of HSRs was 0.2% with ioversol (mostly cutaneous symptoms [88.7%]), and severe HSRs represented 6.4% of all cases (no specific data with ioversol). In the study by Cha et al [44], HSR incidence was 0.66% and no severe HSRs were reported among 24,220 patients who received ioversol.

The incidence of acute ADRs was explicitly reported in two studies [18, 37], and in a third study, acute ADRs represented the majority of the reported ADRs (88.6%) [17]. The incidence was 0.23–1.8% [17, 18, 37]. In the study by Gomi et al [37], the acute ADR incidence was significantly lower with ioversol (1.8%) compared to iomeprol (3.9%) and iopromide (3.5%). Overall, 0.7% of the reported reactions required treatment and resolved, with no association with the type of ICM. No patient experienced life-threatening severe complications requiring immediate transfer to the emergency department.

In the study by Juchem et al [18], acute ADRs corresponding to two cases of vomiting (1%) were reported with ioversol, while the incidence of acute ADRs with meglumine diatrizoate was 12.5% (85% were anaphylactoid reactions). All acute ADRs were mild and patients recovered spontaneously.

Furthermore, in the study by Motosugi et al [36], acute allergic-like reaction incidence with ioversol was 1.8% and that of acute physiologic reactions was 1.1%, and none were severe.

Anaphylactoid reaction incidence in patients exposed to ioversol was reported in two studies ranging from 0.18% [42] to 2.5% [45]. Federle et al [45] reported more than a threefold higher incidence of anaphylactoid reactions with iothalamate compared to ioversol at both slow (8.3% vs. 2.0%, respectively) and fast (9.1% vs. 2.5%, respectively) injection rates.

The incidence of ADRs in pediatric patients exposed to ioversol for CT or urography was reported by Callahan et al [38], with a total of 12,494 pediatric patients and a mean (SD) age of 9.5 (5.9) years. Mild symptoms such as nausea, warm sensation, altered taste, and anxiety were not recorded as ADRs in this study. No ADRs were reported among 941 patients who underwent excretory urography. Only mild (0.38%) and moderate ADRs (0.08%) were reported. In patients aged ≤ 6 years old, only ADRs of mild intensity were reported. Two other pediatric studies from ioversol clinical development (mean age ≈ 5 years) reported ADRs in 3 of 65 patients (4.6%): metallic taste, nausea, and vomiting in two patients and not defined in the third patient [27, 30].

Studies with a comparison with other ICM

Ioversol was compared to a non-ionic, monomeric LOCM in 5 studies [20, 24, 25, 29, 31] during its clinical development, and no difference was shown regarding ADR incidence (Table 2). In 6 other studies [17, 36, 37, 40, 43, 44], the incidence of all ADRs and HSRs and severe/serious events (when reported) with ioversol was among the lowest (Table 2). In 3 studies [17, 43, 44], also including data with the IOCM iodixanol, the incidences of ADRs and HSRs with ioversol were 0.13–0.66% vs. 0.27–0.99% with iodixanol, and severe/serious events were 0.00–0.02% vs. 0.03–0.48%, respectively.

Five studies reported that the incidence of ADRs or HSRs was significantly different between ICM, with the highest incidences reported with iomeprol and/or iopromide [17, 37, 40, 43, 44]. Two studies compared the nature of ADRs between ICM. In Chen et al, rash was the predominant ADR reported with all ICM, but was more frequent with iodixanol. Facial swelling was more often reported with iodixanol compared with iopamidol and iopromide and was not reported with ioversol [43]. An et al analyzed the prevalence of ADRs by system organ class (SOC) and reported that “skin and appendages disorders” were more frequent with iodixanol, and “gastrointestinal system disorders” and “respiratory system disorders” more frequent with iomeprol [17].

Post-contrast acute kidney injury

PC-AKI prophylactic measures were described in two studies, and consisted of oral or IV hydration [16, 19]. A large heterogeneity in PC-AKI incidence was observed among the 5 studies (1–42%), due to heterogenous patient populations and differences in used PC-AKI definitions (Table 4).

Table 4 Incidence of PC-AKI after intravenous administration of ioversol
Full size table

In Louvel et al [46], one patient (1.1%) aged 82 years had a 25% increase in sCr (87 to 109 mmol/L) which rapidly improved. An increase > 10% in sCr was observed in 8 patients aged > 69 years and 4 patients aged < 60 years, with no significant difference between the two age groups. In Gomez et al [19] (98 diabetic patients using metformin), PC-AKI was observed for only one patient (1%) with an eGFR < 60 mL/min/1.73 m2 (incidence of 4.7% in this subpopulation), without clinical repercussion. During a 1-month follow-up period, no patient had alteration of renal function requiring medical care.

Ng et al [47] included two matched groups of patients who underwent CT with or without ioversol, and showed no difference in PC-AKI incidence (17%), sCr increase (0.25 and 0.11 mg/dL, respectively), need for hemodialysis (2% and 1%, respectively), and in-hospital mortality (17% and 21%, respectively). Moura et al [16] included a high-risk population of patients admitted to intensive care unit (ICU) with a length of stay > 3 days. The broader PC-AKI definition used in this study resulted in an incidence of 42%. Hemodialysis was needed for seven patients (12%) and deaths reported for 9 patients (6.5%).

Gilligan et al [39] included two matched groups of pediatric patients exposed to ioversol (aged 8 [6] years), and those who underwent abdominal US, and showed no difference in PC-AKI incidence (2.4% and 2.6%, respectively). In patients with an eGFR < 60 mL/min/1.73 m2, PC-AKI incidence was lower with ioversol (5.6% vs. 11.1%, respectively), although not statistically significant.

Discussion

This systematic literature review showed a large heterogeneity between studies regarding the way ADRs were collected and the type of ADRs reported. The median (range) incidence of ADRs with IV ioversol was 0.23% (0–33.3%). This variability is mainly emanating from ioversol clinical development studies, which included a low number of patients, and where heat and pain were specifically assessed in some studies. In the other studies, the incidence of ADRs in adults was low, independent of the type of ADR reported: 0.13–0.28% for all ADRs [17, 42, 43], 0.23–1.8% for acute ADRs [17, 18, 36, 37], and 0.2–0.66% for HSRs [40, 44]. In two studies, the relatively high incidence of events could be due to the systematic interview of patients [36] and a higher incidence of mild events (> 90% [36], 83% [44]). These incidences are comparable to those reported with other ICM. Indeed, two large retrospective studies with more than 246,000 patients who received IV non-ionic LOCM, reported an ADR incidence of 0.3% [49, 50].

The incidence of severe reactions to IV ioversol was low (0–0.02%) [18, 36, 38, 43, 44] and similar (if not lower) to what has been reported with other ICM (0.01–0.08%) [49,50,51,52]. Anaphylactic shock was reported in only one study, with a low incidence (0.019%) [43], consistent with a previous study using other non-ionic ICM (0.016%) [53]. Thus, the occurrence of severe events can be considered as rare with non-ionic ICM.

The risk of ADRs after using ICM in pediatric patients, and particularly life-threatening reactions, is low [54, 55]. Callahan et al reported a low incidence of ADRs (0.46%) and absence of severe events [38]. In one study, where non-ionic ICM were administered in 13,461 pediatric patients, the overall incidence of ADRs was 3.4%, and that of severe ADRs was 0.07% [55]. Another study reported an incidence of allergic-like reactions of 0.18% overall and 0.027% for severe reactions on 11,306 IV administrations [56]. This variability could be due to the different reporting (all ADRs or specific types, some mild symptoms not recorded as ADRs) [38]. ADR incidence was previously associated with the age of the patients with lower incidences observed in patients aged ≤ 10 years (0.22%) [50]. This could be linked to weak immune responses in pediatric patients compared to adults. Overall, it can be concluded that ioversol has a similar safety profile as other non-ionic ICM when IV administered to pediatric patients.

Several large retrospective studies investigated the safety profile of different ICM. Two studies using different non-ionic ICM reported that cutaneous and gastrointestinal disorders were the most frequent for mild events (51–69% and 12–14%, respectively) [49, 50]. In contrast, in a comparison of the safety profile of seven ICM, it was reported that skin (69.4%) and respiratory system disorders (8.9%) were the most frequent, followed by gastrointestinal disorders (5.7%). For ioversol, the proportion of gastrointestinal disorders and cardiovascular disorders was significantly higher than the general profile of LOCM (8% vs. 6% and 2% vs. 1%, respectively) and skin disorders significantly lower (65% vs. 70%) [57]. Despite some differences between LOCM, cutaneous and gastrointestinal manifestations are the most frequent and it could be concluded that ioversol has a similar safety profile to other LOCM.

PC-AKI incidence was highly variable, with the highest incidence reported in a critical care population with strong competing risk factors for AKI [16]. It is advised to use the lowest dose of ICM as possible in patients with diabetes and other co-morbidities and/or in patients with impaired renal function [7, 58, 59]. Consistent with what has been reported by Gomez et al [19], others reported a PC-AKI incidence of 1% in patients with normal renal function, which increased to 14% in those with severe renal impairment [60].

In the two studies comparing CT with ioversol to unenhanced CT or abdominal US, IV administration of ioversol per se did not increase the incidence of PC-AKI in adult and pediatric patients [39, 47]. Others reported that IV ICM administration for CT was not associated with an increased risk of PC-AKI [60], and large retrospective studies using propensity score matching suggested a lower incidence of PC-AKI than previously estimated [61]. In studies comparing the safety profile of iodixanol to that of other non-ionic LOCM, urinary system disorders were more frequently reported than with non-ionic LOCM [57]. However, this could be due to iodixanol being used more frequently in high-risk patients with underlying renal diseases [17]. The proportion of urinary system disorders with ioversol was comparable to the general profile of LOCM, suggesting a similar safety profile with regard to PC-AKI [17, 57]. In procedures involving IV administration of ICM, several meta-analyses showed that iodixanol was not associated with a reduction in PC-AKI compared to non-ionic LOCM [62,63,64].

In conclusion, the safety profile of ioversol, by IV route, is good and comparable to that of other non-ionic LOCM, with a low incidence of ADRs overall and particularly severe/serious ADRs, in adult and pediatric patients. PC-AKI incidence following IV administration of ioversol was not higher than in patients unexposed to ICM. Further well-designed studies are warranted in order to confirm these results.