Chasing Shadows, Catching Smoke, and Estimating Anaphylaxis to Sugammadex

See Article, p 93

Although the incidence of rare adverse events is difficult to measure, clinicians yearn for accurate information to better guide patient care and inform patients of potential risks. When new drugs or devices are introduced, clinician-scientists are keen to characterize the incidence of adverse events. This is an important component of the drug or device regulatory approval process. Even though the Food and Drug Administration requires a rigorous multistep process for new drug development in the United States, the process cannot evaluate all possible rare adverse events before release for widespread use. To do this would require large sample sizes and be cost prohibitive. This is why “Phase IV” or postmarketing assessment of safety is necessary and may even be required for ongoing approval. A clear picture of a new drug’s safety emerges only after months if not years of clinical use. This process has been enabled in the Information Age by the advent of large collections of granular patient data, and the tools to query them.

In this month’s issue of Anesthesia & Analgesia, Burbridge¹ describes the incidence of anaphylaxis to sugammadex in a single-center cohort of nearly 20 thousand patients, captured from the institution’s electronic medical record. He reports an incidence of 1 in 9910. This is in stark contrast to previously reported rates. For example, the sugammadex package insert reports a rate of 1 in 300 as described by Min et al.² In a prior retrospective large cohort study, Miyazaki et al³ reported a rate of 1 in 2500. Which of these widely different numbers should we believe?

Although estimates from each study were reported in reputable journals with rigorous peer review, at face value, the range in anaphylaxis rates is too large to be of clinical use in estimating the rate of an adverse event. As is common in clinical research, the reason lies in the methodology of the individual studies: in this case, how the diagnosis of anaphylaxis was established. To better understand how investigators arrived at these very different rates, a brief review of each study’s design reveals significant differences and limitations in the methods used.

For the data presented in the sugammadex package insert, Min et al² utilized a prospective placebo-controlled volunteer study. Their design used 3 groups of volunteers. One group (n = 151 volunteers) received sugammadex 4 mg/kg (a dose used to reverse deep neuromuscular blockade). A second group (n = 148 volunteers) received sugammadex 16 mg/kg (a dose used to reverse profound neuromuscular blockade). A third group received a placebo. Only 1 subject, in the 16 mg/kg group, exhibited signs of anaphylaxis according to the Sampson criteria developed by the National Institute of Allergy and Infectious Diseases.⁴ To estimate a rate of anaphylaxis, the authors summed up the total number of volunteers that received either dose of sugammadex to generate the rate of 1 in 299 or 0.3%.

Positive features of this experimental design include that (1) it was prospective and enrolled healthy volunteers; (2) doses were more than what is routinely used for reversal of neuromuscular blockade; (3) each volunteer received repeated doses of sugammadex to explore the influence of prior exposure as a source of sensitization; and (4) the authors used published criteria to perform a rigorous assessment of the timing of anaphylaxis onset, skin and mucosal changes, and cardiopulmonary function and serum tryptase levels as a marker of mast cell degranulation. Finally, (5) because no other medications were administered and no surgical procedure was underway, hypersensitivity symptoms were most likely the result of the sugammadex, rather than any confounding factor.

Selected critiques include the following:

1. Although not definitively explored, experts suggest that anaphylaxis is more likely with higher doses and does not appear to worsen with repeated doses.⁵ If so, a 0.3% incidence in volunteers administered only high doses may overestimate the incidence in a typical clinical practice with lower doses used to reverse neuromuscular blockade. Given cost constraints with sugammadex, practitioners are often advised to use the smallest dose possible for reversal. This also highlights the need for vigilant quantitative monitoring of neuromuscular blockade, if available, to avoid administering more sugammadex than necessary.
2. In a prospective trial design using predetermined criteria for detecting anaphylaxis, it is likely that if it did occur, investigators would capture it. Although accurate in volunteers, the sample size was small for characterizing a rare adverse event, limiting its generalizability to clinical practice. Finding a single case of severe hypersensitivity in 450 patients provides little statistical confidence in defining a rate.
3. Measuring serum tryptase levels as a marker for anaphylaxis may be misleading. A high proportion of patients with elevated tryptase levels also exhibit signs of anaphylaxis (specificity between 89% and 100%). However, selected patients with anaphylaxis may have normal tryptase levels⁶ making interpretation of these levels difficult. It is noteworthy that even the 1 patient with a severe reaction in this study did not have a tryptase level above the normal range (<11 ng/mL).
4. This study did not evaluate hypersensitivity and anaphylaxis to the sugammadex-rocuronium complex.

Additional study by this group that same year sought to characterize hypersensitivity and anaphylaxis from a pool of patients and healthy volunteers exposed to sugammadex during industry-sponsored phase 1–3 clinical trials in their sugammadex development program.⁷ In this retrospective analysis, their pooled group consisted of 3519 and 544 subjects who received sugammadex and placebo, respectively. Most (2816) received doses typically used in clinical practice (2–4 mg/kg) and the remaining subjects received doses ranging from <2 to 32 mg/kg. Using the Sampson criteria for anaphylaxis, they found a low incidence of hypersensitivity and anaphylaxis for all subjects (<1%) and no differences between groups. The authors also evaluated all available postmarket adverse events suggestive of hypersensitivity to sugammadex and reported an approximate incidence of 0.01%. They concluded that the incidence of hypersensitivity to sugammadex is low and no different than rates of hypersensitivity in patients that do not receive sugammadex.

In a brief report, Miyazaki et al³ conducted a rigorous retrospective single-center study over 3 years (2012–2015) investigating the incidence of sugammadex-induced anaphylaxis. During this 3-year period, attending anesthesiologists completed an assessment form of perioperative events that included hypersensitivity reactions. The authors reviewed 23,608 assessment forms. Of those, 15,479 patients received sugammadex (65%). Using clinical criteria for diagnosing anaphylaxis as described in the World Allergy Organization (WAO) Guidelines,⁸ they identified 6 patients that had findings consistent with anaphylaxis 1–4 minutes after they had received sugammadex. Of those 6, 1 had elevated serum tryptase levels. These data were used to estimate the incidence of anaphylaxis as 1 in 2500.

Positive features of this experimental design include the following. (1) In this large retrospective analysis, the authors used a detailed assessment form that was voluntarily completed by anesthesiologists on every patient over the study period. The assessment form included data that would allow them to identify potential anaphylaxis episodes. If the assessment form was suggestive of a potential sugammadex-associated anaphylaxis, the authors performed an in-depth review of the patient record and extracted the timing of sugammadex administration and the onset of anaphylaxis. (2) Only those patients who met WAO criteria were included in their final analysis. (3) They reported that the doses of sugammadex were between 2 and 4 mg/kg in those patients who developed anaphylaxis, consistent with common clinical use.

Selected critiques include the following:

1. This retrospective study took advantage of voluntary completion of an assessment form by attending anesthesiologists, but some patients may have been missed if the form was not completed. While taking advantage of the expert opinion of the clinician at the scene, this methodology cannot remove subjectivity from that opinion.
2. Because this was an observation of real-life practice, there may be confounding effects of the surgical procedure and other administered medications.
3. Sugammadex doses used in routine practice can be a risk for anaphylaxis.
4. Serum tryptase levels, as above, were not useful in identifying all patients who developed anaphylaxis to sugammadex.
5. Although rigorous, it is a single-center study with a homogeneous population somewhat limiting its generalizability to other patient groups.

In this article, Burbridge¹ conducted a retrospective single-center study over 3 years (2016–2019), using methodology similar to Miyazaki et al.³ He searched through an institutional electronic medical record database of nearly 4 million patients. He conducted 2 searches. The first search identified records that reported an allergy to sugammadex (again, dependent on self-reporting by the involved clinicians). Two patients were identified. The second search looked for patients who received both sugammadex and epinephrine over the course of the same case. A total of 19,821 patients received sugammadex (0.5%). Of those, 779 also received epinephrine. Using the same WAO criteria for diagnosing anaphylaxis, all 779 patient records were reviewed by clinical experts. They examined the doses of medications administered, the timing of epinephrine relative to sugammadex, and the course of the patient’s vital signs. For example, a patient who received epinephrine mixed with lidocaine as an epidural infusion 2 hours before the dose of sugammadex was not likely being treated for anaphylaxis, whereas a patient receiving intravenous epinephrine 15 minutes after sugammadex dosing in the setting of fluctuating blood pressure was much more likely to represent treatment for acute hypersensitivity. The author arrived at a consensus diagnosis of sugammadex anaphylaxis in exactly the same 2 patients who were reported by the attending anesthesiologists.

When contrasting his findings with those of Miyazaki et al,³ the author points out that regional differences in exposure to cyclodextrins, widely used in several products, may explain the large difference in rates of anaphylaxis reported. This argument is similar to the observation regarding variations in the use of cosmetics as an explanation for the widely varying risk of anaphylaxis to rocuronium observed between American and European populations.⁹

Features of this experimental design include (1) Burbridge¹ presents a very large retrospective observational analysis (nearly 4 million patients) conducted in the United States. (2) Only those patients who met WAO criteria were included in their analysis. (3) He reports that the doses of sugammadex were 1.8 and 2.3 mg/kg in the 2 patients who developed anaphylaxis, again consistent with normal clinical use.

Selected critiques include the following:

1. The percentage of patients who received sugammadex within the database was small (0.5%). This may limit generalizability given that sugammadex may be more widely used in other regions and practices.
2. Using epinephrine in combination with sugammadex as search terms is intuitive because this is the recommended treatment for severe hypersensitivity; however, this approach will underestimate the overall risk of reactions because many mild or moderate cases would not have required epinephrine administration.
3. Although rigorous, it is curious that the 2 search approaches used identified the same 2 patients out of nearly 4 million to have an anaphylactic response to sugammadex. The observation is either reassuring about the sugammadex + epinephrine search strategy or else represents a variety of confirmation bias contaminating the study methodology.
4. Burbridge¹ presents 1 approach to searching large databases. There are others. Selected others include (1) using multiple independent reporting systems to confirm the presence of a diagnosis or (2) using a Delphi technique with subject matter experts to identify pertinent search criteria, utilize these search criteria in electronic search strategies, and then adjudicate each patient record that meets criteria with a panel of subject matter experts to confirm the diagnosis.¹⁰

When interpreting the difference in estimates of anaphylaxis to sugammadex, it is useful to consider their confidence intervals. In the study presented by Burbridge,¹ the 95% confidence intervals were 0.001–0.036 about a mean of 0.01% suggesting a range of rates from 1 in 100,000 to 1 in 2800. In the study presented by Miyazaki et al,³ the 95% confidence intervals were 0.014–0.084 about a mean of 0.039% suggesting a range of rates from 1 in 7142 to 1 in 1190. Although there are large differences in rate estimates of anaphylaxis, the overlap within the 95% confidence intervals suggests that these estimates may originate from similar populations.

To help put allergic reactions to cyclodextrins in perspective, a brief review of their widespread application is useful to review. They are used in the food, chemical, and pharmaceutical industries. They consist of a series of glucose subunits configured in small rings of 6 (α), 7 (β), or 8 (γ) cyclodextrins. They have a cone shape that is hydrophobic on the inside and a hydrophilic on the outside. They are known for their ability to host guest molecules for a variety of reasons and have low immunogenicity and toxicity. This makes them attractive for a number of purposes. For example, in the food industry, they are used to solubilize food coloring and vitamins, stabilize flavors and fragrances, suppress unwanted odors, and provide sustained release of selected foods.¹¹ Foods that use cyclodextrins include breads, breakfast cereals, dry mixes for beverages, salad dressings, and soups among many others. The average person consumes up to 4 g/d of γ cyclodextrins.¹²

In the pharmaceutical industry, cyclodextrins are primarily used to improve the aqueous solubility of poorly soluble hydrophobic drugs or as a scavenger to collect unwanted molecules such as cholesterol. Cyclodextrins have been studied and used with numerous intravenous and oral drug preparations across a variety of drug classes such as antivirals, anesthetics, anticancer drugs, cardiovascular drugs, etc.^13,14 Not all cyclodextrins are the same. They can be modified to improve desired function. Sugammadex, for instance, is a γ cyclodextrin modified with a set of 8 carboxyl thio (sulfur) ether groups added to each glucose subunit to extend the length of the cone to better encapsulate rocuronium.

It is likely many patients who receive sugammadex have had a prior exposure to a cyclodextrin. Why a select few develop anaphylaxis remains a mystery. With the exception to sugammadex, there is a paucity of literature documenting allergic reactions to naturally occurring or modified cyclodextrins. We are left to deduce that it is unlikely that prior exposure sensitizes patients. Some have suggested that the allergic reaction is in response to the sugammadex-rocuronium complex and not the cyclodextrin.¹⁵ To make our appraisal of the antigenicity of sugammadex even more confusing, case reports describe using sugammadex to treat rocuronium anaphylaxis.¹⁶

In conclusion, given that varied approaches to characterizing the rate of anaphylaxis to sugammadex have led to very different estimates, what is the clinician to believe? Does it matter? Does one rate promote a more cavalier approach to administering sugammadex than another (eg, 1 in 300 vs 1 in 10,000)? Clinical experience would suggest that 1 on 300 is too high. If that were the case, in institutions where sugammadex is routinely used, there would be a case of anaphylaxis every week. The actual incidence remains elusive but is likely to be much lower than 1 in 300. At this point, it is best to acknowledge that our understanding of the true incidence of anaphylaxis with sugammadex is immature. Regional differences and associated exposure to cyclodextrins in various forms may be important. As such, data reporting low rates of anaphylaxis in one region may not generalize to other parts of the world. The wise clinician will be aware that any anesthetic medication can provoke a hypersensitivity response, and will remain correspondingly vigilant.

DISCLOSURES

Name: Ken B. Johnson, MD.

Contribution: This author helped prepare the editorial and approved the final manuscript.

Name: Richard P. Dutton, MD, MBA.

Contribution: This author helped prepare the editorial and approved the final manuscript.

This manuscript was handled by: Richard C. Prielipp, MD, MBA.