Modern nonionic iodine-based contrast media are among the most widely used diagnostic drugs in medicine, with market projections estimating that nearly 250 million doses were administered worldwide in 2025 for angiography and computed tomography. A new narrative review published in European Radiology Abdomen by Francisco Vega and Aart J. van der Molen examines whether the individual agents in this class differ measurably in their rates of adverse drug reactions and hypersensitivity reactions, and whether those differences should guide the choice of contrast agent in clinical practice.
The review, published as an open-access article on 1 November 2026, synthesised original studies, systematic reviews and meta-analyses identified through a comprehensive literature search of PubMed, Embase, Emcare, Web of Science and the Cochrane Library covering publications from January 2010 onwards. The authors included only studies reporting agent-specific adverse drug reaction data, hypersensitivity data, cross-reactivity information or outcomes after switching contrast media. General safety guidelines lacking granular per-agent data, studies based solely on spontaneous reporting systems, case reports, early-phase clinical trials and studies of agents no longer commercially available in Europe were excluded.
The authors are explicit that their work does not meet the rigorous criteria of a systematic review. It lacks a formal framework such as PICO, omits a structured risk-of-bias and quality assessment of the included studies, and does not perform quantitative synthesis or meta-analysis of its own. Screening was performed independently by all authors, first by title and abstract and then by full text, with discrepancies resolved through consensus, but standardised dual-investigator procedures aimed at minimising selection bias and ensuring strict reproducibility were not employed. Because of these limitations, the article is classified as a narrative review intended to provide a structured overview of current knowledge and emerging research trends.
The headline numbers are reassuring. Overall acute adverse drug reaction rates remained below 1% for most nonionic contrast agents, with severe events occurring in fewer than 0.02% of administrations. The median hypersensitivity reaction incidence across aggregated data was 0.71%, with severe reactions comprising approximately 3% of all hypersensitivity events. Three meta-analyses reported overall adverse drug reaction incidences ranging between 0.77% and 1.38%, though most demonstrated high statistical heterogeneity. The first pooled 30 studies published between 1990 and 2017, encompassing 1,360,488 administrations, and found an overall acute adverse drug reaction incidence of 1.03% with severe reactions in 0.014% of cases. A second, covering seven studies of 2,570,986 contrast-enhanced CT examinations in South Korea, found an overall incidence of 0.82%, categorised as mild in 0.72%, moderate in 0.11% and severe in 0.013%. The most recent meta-analysis pooled 32 studies representing 13,398,078 injections administered between 1989 and 2023, reporting a pooled incidence of 0.91% overall, 0.098% moderate and 0.016% severe.
When the authors aggregated all available data, the overall median adverse drug reaction rate was 0.74% in pooled individual studies and 0.85% in meta-analyses. Moderate reactions accounted for 17.6% of total adverse drug reactions in the pooled studies and 14.85% in the meta-analyses, while severe reactions represented 1.5% to 1.7% of the total across all aggregated data. For hypersensitivity reactions specifically, the data came from pooled individual studies rather than meta-analyses, yielding an overall median rate of 0.71%, with moderate and severe reactions accounting for 12.4% and 3.0% of these cases respectively.
The agent-specific analysis produced a nuanced picture. Iobitridol displayed the lowest adverse drug reaction rates across both individual studies, at 0.45%, and meta-analyses, at 0.77%, yielding the lowest relative risk of all evaluated agents, and it maintained low rates for both severe and moderate reactions. Iohexol exhibited a low overall adverse drug reaction frequency with a median rate of 0.69% across individual studies and 0.79% in meta-analyses, and in several studies it showed significantly lower rates than iopromide, iomeprol, iopamidol and ioversol. Ioversol recorded a median overall rate of 0.67% in pooled studies and 0.88% in meta-analyses, placing it consistently in the low relative risk tier. Conversely, iomeprol consistently presented the highest overall adverse drug reaction relative risk, with a rate of 2.59% in pooled data and 1.38% in meta-analyses, followed by iopromide in pooled studies and iopamidol in meta-analyses.
The hypersensitivity data told a partly different story, which the authors interpret as evidence that the two reaction types involve distinct pathophysiological mechanisms. The highest median hypersensitivity frequencies were observed with iomeprol, iobitridol and iopamidol, whereas iohexol and above all ioversol maintained the lowest baseline risk, with ioversol at a median of 0.40%, nearly half the incidence seen with alternative agents. However, when a hypersensitivity reaction did occur with ioversol, it exhibited the second-highest proportion of non-mild events at 21.3%, with moderate and severe fractions of 15.6% and 4.9% respectively. Iodixanol also exhibited high hypersensitivity frequencies at 0.79% and carried the greatest risk for both moderate and severe reactions, each at 36.4%, though the moderate figure derives from a single large study. Iomeprol showed the highest hypersensitivity frequency overall at a median of 0.95%.
The authors caution that these trends cannot support definitive claims of superiority for any single agent. Geographical bias is substantial, since most published data originate from East Asian populations, particularly South Korea and Japan, where contrast-related reactions are tracked more systematically than in Europe or North America. Reported rates vary geographically, with specific cohorts showing figures such as 0.75% in the USA and 0.76% in South Korea, compared with lower rates of 0.41% in China and 0.52% in Europe. Additional confounders include differences in contrast media utilisation, market share variations, with iohexol representing roughly one-third of global volume, methodological heterogeneity in patient selection, reaction definitions, severity grading and reporting, insufficiently reported premedication protocols, poorly characterised genetic and ethnic factors, and the scarcity of head-to-head prospective comparative studies and randomised controlled trials.
Despite these limitations, the review supports incorporating contrast agent selection into individualised risk assessment for high-risk patients, such as those with a documented history of prior hypersensitivity reactions or an active allergic predisposition. Because the baseline incidence of severe hypersensitivity is extremely low, a two- to three-fold relative increase between agents represents a minimal absolute risk increase in unselected low-risk populations, but in high-volume centres or high-risk patients these relative differences translate into meaningful opportunities for prevention. The authors emphasise that switching to a different contrast agent significantly outperforms premedication alone for preventing recurrent reactions, a finding supported by robust retrospective and prospective studies as well as meta-analyses, and that effective switching should be culprit-based, either through allergy study or by empirically choosing an alternative agent from a different structural group based on side chain characteristics. Recent large cohort studies show that breakthrough reactions still occur in up to 10 to 12% of premedicated patients.
The review also highlights a mechanistic insight: hypersensitivity reactions to iodine-based contrast media are not triggered by the elemental iodine core, but by immunological recognition of an allergenic epitope presumably located on the side chains of the molecule. Variations in carbamoyl side chains appear to influence reactivity capacity and hypersensitivity incidence, and agents featuring an N-(2,3-dihydroxypropyl)-carbamoyl side chain, which includes iohexol and ioversol, appeared to carry an elevated relative risk for severe hypersensitivity reactions in the review’s severity distribution, despite those two agents showing the lowest overall hypersensitivity rates. The authors call for large prospective multicentre studies with standardised reaction definitions, systematic allergy work-up protocols and detailed characterisation of both culprit and alternative agents, integrated with molecular structure analyses and pharmacogenetic data, to determine whether the reported variations reflect intrinsic agent properties or confounding factors. In the meantime, they recommend maintaining detailed records of the exact agent administered at each examination, since such documentation is critical for subsequent allergy evaluation and for identifying safer alternatives in future imaging studies.
Subject of Research: Comparative adverse and hypersensitivity reaction rates of nonionic iodine-based contrast media
Article Title: Adverse and hypersensitivity reactions after intravascular administration of nonionic iodine-based contrast media: Do agent-specific differences inform contrast agent selection?
Article References: Vega, F., & van der Molen, A. J. (2026). Adverse and hypersensitivity reactions after intravascular administration of nonionic iodine-based contrast media: Do agent-specific differences inform contrast agent selection?. European Radiology Abdomen, 1(1), Article 9. https://doi.org/10.1007/s44501-026-00012-6
Image Credits: AI Generated
DOI: 10.1007/s44501-026-00012-6
Keywords: iodine-based contrast media, adverse drug reactions, hypersensitivity, contrast agent switching, radiology, computed tomography, angiography, drug safety, narrative review, cross-reactivity, premedication, patient risk stratification
News Source: Ophelia Keating. (November 4, 2026). Which iodine contrast agent is safest? New review finds trends but no clear winner. Scienmag.
