Gadolinium Contrast Safety: What the ACR Group I Versus Group II Classification Means

The short answer

The American College of Radiology (ACR) sorts gadolinium-based contrast agents into three groups based on how many cases of nephrogenic systemic fibrosis (NSF) each agent has been linked to, rather than on chemistry alone. Group I agents account for the greatest number of NSF cases and have largely left the United States market; Group II agents show few, if any, unconfounded NSF cases and are the workhorses of modern imaging; Group III holds agents for which the data remain thin. Reading the classification well means seeing it as an epidemiologic verdict sitting on top of a chemistry story, where the two are related but not identical.

Why one disease reshaped contrast safety

NSF is a rare and sometimes disabling condition that thickens and hardens the skin and connective tissue and can involve internal organs. It was first described in the late 1990s, and by 2006 investigators had tied it to gadolinium exposure in people with severe kidney impairment. The mechanism most researchers favor is that some gadolinium separates from its carrier molecule and deposits in tissue, where it appears to drive fibrosis. Nearly every confirmed case occurred in patients with advanced kidney disease, acute kidney injury, or dialysis dependence, because impaired clearance leaves the agent circulating far longer than the minutes it lingers in a healthy person.

Faced with a growing register of case reports, the ACR needed a way to turn scattered signals into usable guidance. The Group I, II, and III system in the ACR Manual on Contrast Media is that translation. The groups are defined by observed NSF cases, and specifically by unconfounded cases, meaning cases where a single agent can be blamed without another gadolinium agent muddying the picture.

What each group means

Group I Group I contains the agents associated with the greatest number of NSF cases: gadodiamide (Omniscan), gadopentetate dimeglumine (Magnevist), and gadoversetamide (OptiMARK). These carry FDA contraindications in patients at highest risk, such as those with acute kidney injury or chronic stage 4 to 5 disease, and they have effectively disappeared from routine use in the United States.

Group II Group II contains the agents associated with few, if any, unconfounded NSF cases: gadobenate dimeglumine (MultiHance), gadobutrol (Gadavist), gadoterate meglumine (Dotarem and Clariscan), and gadoteridol (ProHance). The ACR added gadopiclenol (Elucirem, Vueway) to this group in 2023 after judging its kinetic stability comparable to the other members. For standard or lower doses of Group II agents, the ACR states that assessing kidney function by questionnaire or laboratory testing before an injection is optional, because the residual NSF risk is low enough to be possibly nonexistent. A joint ACR and National Kidney Foundation consensus reached the same conclusion: Group II agents may be given when needed even to patients on dialysis or with stage 4 to 5 disease, at a risk the panel described as exceedingly low.

Group III Group III is the smallest and least stable category, reserved for agents with limited NSF data but no clear pattern of unconfounded cases. Gadoxetate disodium (Eovist), a liver-specific agent, sat here for years before accumulating enough safety data to be treated alongside Group II in the most recent manual.

The chemistry the groups sit on top of

Every gadolinium agent is a toxic metal ion wrapped in a carrier molecule, or chelate. Two properties determine how tightly that wrapping holds: thermodynamic stability, or how firmly the ion is bound at equilibrium, and kinetic inertness, or how slowly the ion can escape. Macrocyclic agents cage the ion inside a closed ring and are markedly more inert than linear agents, whose open-chain structure lets the metal dissociate more readily. That difference is the leading mechanistic explanation for why Group I, built from less stable linear agents, produced most of the NSF cases.

Here is the point most summaries miss. Group II is not a synonym for macrocyclic. Gadobenate dimeglumine is a linear ionic agent, yet it sits in Group II because its observed safety record earned it a place there. The lesson for reading the evidence is that the ACR groups report outcomes, while the macrocyclic versus linear distinction explains a mechanism. A sound classification uses both, and it does not assume chemistry alone predicts every clinical result.

Retention is a different question than NSF

A separate concern surfaced around 2017, when the FDA reported that trace gadolinium is retained in tissues, including the brain and bone, for months to years after injection, and that linear agents leave behind more than macrocyclic agents. The agency stated that it had identified no adverse health effects from this retention in patients with normal kidney function and that the benefit of these agents continued to outweigh the risk. It subsequently required a class warning and a patient Medication Guide across all gadolinium agents.

Reading the evidence carefully means keeping two questions apart. Retention is a measurable fact. Harm from retention in people with healthy kidneys is not established. The group classification answers the NSF question, not the retention question, and it is a mistake to treat a Group II label as a verdict on every gadolinium safety debate, including the contested cluster of symptoms some patients attribute to gadolinium deposition.

How to read it like a regulator

The AJR special series update on gadolinium safety pulls these threads together: the classification is outcome-driven, it hinges on unconfounded cases, and it is revised as agents accumulate data, as gadopiclenol and gadoxetate both show. When you see a group label, read it as a summary of what has been observed under real dosing, cross-checked against a plausible chemical mechanism, rather than as a fixed property of the molecule. This article is educational and is not medical advice; decisions about contrast belong to a patient and their own clinicians.