
Editor’s Note: At the 2026 European Society for Medical Oncology (ESMO) Congress, Professor Young-Joon Kang from Incheon St. Mary’s Hospital, the Catholic University of Korea, presented the results of the multicenter, randomized AIUR study conducted across seven hospitals in South Korea. The study evaluated the efficacy of compression using regular-sized surgical gloves for preventing paclitaxel-induced peripheral neuropathy (PIPN) in patients with breast cancer. Although the study failed to meet its primary endpoint of improvement in the overall score, exploratory analyses showed that the intervention’s efficacy was limited to the gloved hands, with no benefit observed in the ungloved feet. Against this background, Oncology Frontier invited Professor Young-Joon Kang to discuss the rationale behind the study design, its key findings, and the implications for clinical practice and future research.
Expert Interview
01
Oncology Frontier: Based on the AIUR trial data you presented at this year’s ESMO Congress, could you briefly introduce the background and rationale of the study design? Why was compression therapy chosen as the intervention, and what advantages does using normal-sized surgical gloves as the compression method offer over previous compression devices? How were the primary endpoint and assessment methods defined?
Professor Young-Joon Kang: Hello, nice to meet you. I am Young-Joon Kang, a breast surgical oncologist at Incheon St. Mary’s Hospital, the Catholic University of Korea. Today, I will introduce the AIUR trial.
Currently, no medication is available to prevent paclitaxel-induced peripheral neuropathy. Cooling and compression therapies remain available options, although the evidence supporting these approaches is of low quality. The ESMO guidelines state that compression using surgical gloves may be considered, whereas ASCO has made no recommendation on this approach.
We chose compression with surgical gloves because it does not require any additional equipment. Frozen gloves require a freezer and need to be replaced every 30 minutes, and some patients cannot tolerate the cold. Previous compression approaches used tight-fitting gloves one size smaller than the patient’s hand, which could cause pain. In contrast, normal-sized gloves match the patient’s actual hand size, require no equipment to set up, and cost approximately 60 cents per pair.
We also wanted to address some gaps in the existing evidence. Most previous studies involving gloves had patients wear a glove on one hand while using the other hand as a control, with outcomes assessed by investigators.
We randomized the patients, and the primary endpoint was a patient-reported outcome: the proportion of patients whose FACT-NTX score worsened by 5 points or more from baseline to the end of paclitaxel treatment. Secondary endpoints included changes in the score over time, clinician-assessed peripheral neuropathy according to CTCAE, and comparisons between hand-related and foot-related items.
02
Oncology Frontier: As a multicenter randomized controlled trial, what were the results of the AIUR trial? What differences were observed between the compression therapy group and the control group in terms of the incidence, severity, and duration of paclitaxel-induced peripheral neuropathy (PIPN)? What were the findings regarding safety, patient tolerability, and adherence?
Professor Young-Joon Kang: We enrolled 104 women with stage II or III breast cancer across seven university-affiliated hospitals in South Korea and randomized them 1:1. In one group, patients wore two layers of normal-sized surgical gloves on both hands, starting 30 minutes before each infusion and continuing until 30 minutes after the infusion. The other group served as the control and did not wear gloves.
The study did not meet its primary endpoint. In the glove group, 74.5% of patients experienced a worsening of 5 points or more in their scores, compared with 82.6% in the no-glove group (P=0.45). However, when we examined the scores as continuous variables, the degree of worsening was smaller in the glove group: the FACT-NTX scores worsened by 9.7 points versus 13.7 points, respectively (P=0.03).
The clearest difference was observed according to the affected body region. In both groups, scores for the feet, which were not covered by gloves, worsened by approximately 5 points. By contrast, scores for the gloved hands worsened by 2.9 points, compared with 6.1 points in the control group—approximately half the degree of worsening. In addition, the interaction between treatment group and body region was statistically significant.
According to clinician assessments, approximately 59% of patients in the glove group experienced grade 2 or higher peripheral neuropathy, compared with 82.2% in the no-glove group (P=0.02).
Regarding the duration of neuropathy, our assessments only continued until the end of paclitaxel treatment. Therefore, it remains unclear whether this difference persists over time.
In terms of safety, no glove-related adverse events were recorded, and 92% of patients rated the gloves as satisfactory or very satisfactory.
03
Oncology Frontier: What do you think is the clinical significance of the AIUR trial results for preventing PIPN during paclitaxel chemotherapy in patients with breast cancer? Given that this intervention is simple, low-cost, and easy to implement, are there plans for larger-scale validation studies or for exploring its combination with other neuroprotective strategies?
Professor Young-Joon Kang: I would not consider this study sufficient to change clinical practice because it did not meet its primary endpoint. What the data support is a localized protective effect on the hands. Patients who wore gloves had better outcomes in their hands, whereas no improvement was observed in the ungloved feet. Because the overall score included results from both the hands and the feet, this was also why the trial failed to meet its primary endpoint.
From a clinical practice perspective, glove use is safe, the cost is almost negligible, and patients can tolerate it. Furthermore, healthcare institutions do not need any additional equipment beyond the gloves they already have. Therefore, I believe gloves can be considered an option, but they are not a definitive solution.
They could be used in combination with other measures, such as cooling therapy when patients can tolerate it, as well as symptomatic treatment, including medication when symptoms develop. This combination was not evaluated in the present study, but I believe it could help patients manage these adverse effects.
The direction for the next study is quite clear. We need to target different body regions separately, apply compression to the feet as well, and use prespecified patient-reported outcomes as study endpoints, while setting a realistic effect size. Whether and when we will conduct such a trial has not yet been decided.
Study Overview
Background
Both cooling therapy and compression gloves rely on equipment and can be uncomfortable, limiting patient adherence. Normal-sized surgical gloves do not have these disadvantages and can be put on within seconds.
Previous studies of compression therapy, including the POLAR study, used within-patient control designs and investigator-assessed outcomes, potentially underestimating symptoms as perceived by patients. A double-blind randomized controlled trial also reported negative results. AIUR is the first parallel-group randomized controlled trial to use patient-reported outcomes (PROs) as its primary endpoint to evaluate the efficacy of this approach.
Methods
Across seven study centers in South Korea (NCT05771974), patients with stage II–III breast cancer who were scheduled to receive at least 12 weeks of paclitaxel treatment were randomized 1:1 to either wear two pairs of normal-sized surgical gloves (from 30 minutes before to 30 minutes after each infusion) or receive no intervention.
The primary endpoint was the proportion of patients whose score on the 11-item FACT-NTX scale increased by ≥5 points at the end of paclitaxel treatment. Secondary endpoints included changes in the total FACT-NTX score and NCI-CTCAE grading.
Hand- and foot-specific item scores were compared, with the ungloved feet serving as an internal control.
Results
A total of 104 patients were randomized, with 52 patients in each group.
The primary endpoint did not reach statistical significance. The primary endpoint was based on the total score across 11 items: four hand-related items (which showed less worsening in the glove group), three foot-related items, and four systemic items (which showed no difference between the groups).
However, the efficacy was limited to the upper extremities, with a statistically significant treatment group-by-region interaction (P<0.001). All four upper-extremity items and fingertip pain were statistically significant (BH q<0.05), whereas no statistically significant differences were observed in the lower-extremity, systemic, or non-neuropathic hand-related items.
The incidence of CTCAE grade ≥2 sensory neuropathy was 58.7% versus 82.2% (P=0.02).
No serious adverse events were reported. Among the 48 evaluable patients, 44 (92%) rated glove use as satisfactory or better.
Conclusions
The AIUR trial did not meet its prespecified primary endpoint. Exploratory analyses showed that the efficacy was limited to the gloved hands, with no benefit observed in the ungloved feet. Therefore, combining outcomes from both regions into a single overall score may have diluted the localized benefit observed in the hands.
This approach can be readily extended to the feet and warrants further investigation in trials that assess each body region separately and use prespecified PRO endpoints.
Professor Young-Joon Kang
