Reduction of Rescue Treatment With Dupilumab for Chronic Rhinosinusitis With Nasal Polyps in Japan

ABSTRACT

Objective

Add‐on dupilumab provides significant symptomatic improvement for individuals with chronic rhinosinusitis with nasal polyps (CRSwNP); however, little is known about the impact of dupilumab in reducing rates of rescue treatment for CRSwNP in Japanese real‐world practice settings. This observational study examined the impact of add‐on dupilumab treatment on disease burden, including the use of rescue therapy, in Japanese individuals with CRSwNP.

Methods

The study analyzed individual participant data from a Japanese hospital‐based administrative claims database between January 1, 2019, and March 31, 2023. Data were collected for each individual from 12 months prior to the index date (the start date of dupilumab administration) to 12 months after the index date (follow‐up period). The primary outcome was the change in overall rate of rescue treatment (systemic corticosteroid [SCS] use or endoscopic sinus surgery [ESS]) before and after the initiation of dupilumab.

Results

The study included 438 individuals with CRSwNP (mean ± SD age of 53.9 ± 13.4 years; 56.6% male). The overall rate of rescue treatment before and after initiation of dupilumab treatment was 7.89 events per patient‐year (PY) and 2.14 events per PY, respectively (p < 0.0001), and there was a 73% reduction in the rate of rescue treatment. Compared with the pre‐index period, a lower rate of SCS prescriptions (7.54 vs. 2.05) and ESS procedures (0.35 vs. 0.05) during the follow‐up was observed, respectively, and there was a 72% reduction in the rate of SCS prescriptions. The mean ± SD daily SCS dose decreased from 7.1 ± 9.7 mg in the pre‐index period to 6.1 ± 7.3 mg during follow‐up.

Conclusion

In this real‐world study, add‐on dupilumab treatment significantly reduced the requirement for rescue treatment in Japanese individuals with CRSwNP.

Level of Evidence

3.

This observational study examined the impact of add‐on dupilumab treatment on the use of rescue therapy in Japanese individuals with chronic rhinosinusitis with nasal polyps (CRSwNPs) using data from a large hospital‐based administrative claims database. In the 438 individuals who newly initiated dupilumab during the study period, the use of rescue treatments (systemic corticosteroid [SCS] or endoscopic sinus surgery) decreased over the 12‐month study period. We conclude that add‐on dupilumab treatment provides substantial benefits in Japanese individuals with CRSwNP and is, therefore, a valuable therapeutic option for this condition.

1. Introduction

Chronic rhinosinusitis (CRS) is a heterogeneous condition characterized by persistent symptomatic mucosal inflammation of the nasal and sinus cavities [1, 2], resulting in nasal congestion or discharge, facial pressure or pain, olfactory dysfunction, cough, and fatigue [2, 3].

Two main clinical phenotypes exist: CRS without nasal polyps (CRSsNPs), defined by variable chronic inflammation; and CRS with nasal polyps (CRSwNP), typically driven by type 2 inflammation [2, 4, 5]. Eosinophilic CRS, an intractable form of CRSwNP, has recently become more prevalent in Japan [6, 7]. Individuals with eosinophilic CRS often have other type 2 inflammation‐related comorbidities, such as asthma, non‐steroidal anti‐inflammatory drug‐exacerbated respiratory disease (N‐ERD), and allergic rhinitis, that make their condition particularly difficult to treat [6, 8].

Historically, treatment of CRSwNP has included intranasal corticosteroids, low‐dose macrolides, and short courses of systemic corticosteroid (SCS) [7, 9]. Pharmacotherapy (other than SCS) often has limited effectiveness, and long‐term SCS use is associated with severe adverse effects (e.g., osteoporosis, adrenal suppression, and skin disorders) [10]. Individuals with CRSwNP unresponsive to pharmacotherapy may undergo surgery (usually endoscopic sinus surgery [ESS]) to remove nasal polyps; however, polyp recurrence is common, especially in those with eosinophilic CRS [7].

Dupilumab, a fully human monoclonal antibody that blocks the signaling of interleukin (IL)‐4 and IL‐13 (key type 2 inflammatory cytokines), is approved for the treatment of CRSwNP in many countries, including Japan [11]. Dupilumab has demonstrated efficacy and safety as add‐on therapy for severe CRSwNP in the Phase 3 LIBERTY NP SINUS‐24 and LIBERTY NP SINUS‐52 studies [12], including a Japanese subpopulation [13]. A pooled analysis of these studies showed that the proportion of individuals needing rescue treatment with SCS or ESS was significantly reduced with dupilumab vs. placebo [12].

Although rescue treatment may be used to assess disease control among individuals with CRSwNP [9], the data on the impact of dupilumab treatment in real‐world clinical settings are limited. This real‐world, observational study examined the impact of add‐on dupilumab treatment on disease burden, including the use of rescue therapy, in Japanese individuals with CRSwNP.

2. Methods

2.1. Study Design

The RHINO study was a retrospective, observational cohort study that analyzed data from a hospital‐based administrative claims database, EBM Provider provided by Medical Data Vision (MDV) Co. Ltd. (Tokyo, Japan).

The study used individual participant data from January 1, 2019, to March 31, 2023 (data selection period; Figure S1). The index date was defined as the start date of the first dupilumab administration. For each individual, data were collected from 12 months before the index date (pre‐index period) to 12 months after the index date (follow‐up period).

All procedures were in accordance with the principles of the Helsinki Declaration of 1964 (and later versions), the International Society for Pharmacoepidemiology Guidelines for Good Pharmacoepidemiology Practices, and regulations associated with scientific and medical research involving humans in Japan. The study protocol was approved by the Institutional Review Board of the MINS Clinical Trial Committee (Tokyo, Japan; Approval No. 230227 in 2023). Informed consent was not required as the database information had been de‐identified.

2.2. Database

The MDV database contains de‐identified health insurance claims data from outpatients and hospitalized patients in Japan, collected via the Diagnostic Procedure Combination (DPC) data collection system for acute inpatient care [14]. As of September 2023, the database contained data, including individual characteristics, diagnosis, drug prescriptions, and procedures, from approximately 44 million individuals and covered > 27% of acute care facilities in Japan.

2.3. Study Participants

Individuals were screened for study inclusion during the participant selection period (March 25, 2020 to March 31, 2022; Figure S1). Individuals who met all the following criteria were included: at least two medical claims with International Classification of Disease 10th Revision (ICD‐10) code for CRS (J32) during the participant selection period; at least one prescription for dupilumab by an ear, nose, and throat (ENT) specialist during the participant selection period; age ≥ 15 years at index date; continuous database enrollment for ≥ 12 months before and after the index date; and SCS use and/or history of ESS during the pre‐index period. Individuals who received dupilumab treatment before the index date were excluded.

Subpopulations included individuals with or without asthma/N‐ERD in the 6 months prior to the index date, and with or without a history of SCS and ESS in the pre‐index period.

2.4. Outcomes

The primary outcome was the overall rate of rescue treatment required during the pre‐index period vs. the follow‐up period. Rescue treatment was defined as an SCS prescription or ESS procedure, as per the LIBERTY NP SINUS‐24/LIBERTY NP‐52 definitions [12]. The outcome was evaluated by the total number of SCS prescriptions or ESS procedures divided by the total number of patient‐years (PYs) for all individuals receiving dupilumab treatment.

The secondary outcomes included the use of rescue treatment, SCS‐related outcomes (including SCS use), ESS‐related outcomes, and individual baseline characteristics. The details of these outcomes, and those of the exploratory outcomes, are provided in Table S1.

2.5. Statistical Analysis

No hypothesis testing or formal statistical interpretation was performed; therefore, all data analyses are descriptive, and no formal sample size calculations were conducted.

The analyzed population included all individuals who met all inclusion criteria with available MDV data during the data selection period. Baseline demographics and clinical characteristics were summarized using descriptive statistics (i.e., means, standard deviations [SDs], medians, ranges, participant numbers, and/or proportions as applicable).

For the primary outcome, the overall rate of rescue treatment was calculated using the following equation: $\text{Overall rate}={\sum }_1^n{\mathrm{CNT}}_i/{\sum }_1^n{T}_i$, where CNT _i was the number of rescue treatment events that occurred in an individual and T _i was the duration of the evaluation period (in days). The rate difference (and ratio) between the pre‐index and follow‐up periods was also calculated; 95% confidence intervals (CIs) and p values were estimated using the Wald method. The adjusted overall rate was determined using the generalized linear model with negative binomial data and the following covariates: sex, age at index date, asthma/N‐ERD (yes/no), prior SCS history (yes/no), and prior ESS history (yes/no).

Regarding the secondary outcomes, the overall rates of SCS use and ESS were calculated in the same way as described for the primary outcome. The percentage change in mean SCS daily dose was calculated using the following equation: Perc = [Ave _f⁻Ave _p]/Ave _p × 100, where Ave _p and Ave _f were the mean SCS daily dose during the pre‐index and follow‐up periods, respectively. The remaining outcomes were analyzed using descriptive statistics.

Where applicable, p values were used to detect potential signals only. Analyses were based on observed data, with no imputation for missing data. Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA). Statistical factors considered when determining the analyses performed are outlined in the “Statistical considerations” section of Supporting Information S1.

3. Results

3.1. Study Population

Of 430,925 individuals diagnosed with CRS in the MDV database during the data selection period, 443 met all of the inclusion criteria and 438 individuals were included in the study population; 5 individuals were excluded as they had received dupilumab before the participant selection period (Figure 1).

FIGURE 1.

Flow of individuals in the study. ^aDefined as the date of the first dupilumab treatment. ^bDefined as within 365 days prior to the index date. ^§Defined as the number of individuals meeting inclusion criteria #1, then #1 + 2, then #1 + 2 + 3, and so on. CRS = chronic rhinosinusitis; ENT = ear, nose, and throat; ESS = endoscopic sinus surgery; ICD‐10 = International Classification of Diseases, 10th Revision; MDV = Medical Data Vision; SCS = systemic corticosteroids.

The study population had a mean ± SD age of 53.9 ± 13.4 years and 56.6% of individuals were male (Table 1). The most common comorbidities associated with CRS (in ≥ 30% of individuals) were allergic rhinitis (99.3%), asthma (70.6%), and nasal septal deviation (37.2%). Most individuals (97%) had received SCS at least once, and 18.3% had undergone ESS during the year prior to the index date. The most common SCS‐related claims (in ≥ 10% of individuals) were peptic ulcer (25.3%), hypertension (11.6%), and diabetes (11.2%).

TABLE 1.

Demographics and disease characteristics of the study population in the pre‐index period.

Characteristic	Study population
	N = 438
Age at index date, a years
Mean ± SD	53.9 ± 13.4
Median (range)	54.0 (20.0–87.0)
Sex, n (%)
Male	248 (56.6)
Female	190 (43.4)
Comorbidities associated with CRS, b , c n (%)
Allergic rhinitis	435 (99.3)
Asthma/N‐ERD	320 (73.1)
Asthma	309 (70.6)
N‐ERD	27 (6.2)
Nasal septal deviation	163 (37.2)
Eosinophilic otitis media	51 (11.6)
Sleep disorder	42 (9.6)
Anxiety	14 (3.2)
Eczema (atopic dermatitis)	6 (1.4)
Food allergy	3 (0.7)
Depression	3 (0.7)
Drug allergy	1 (0.2)
Treatment history, n (%)
Prior SCS use	425 (97.0)
Prior ESS	80 (18.3)
SCS‐related claims, b , c n (%)
Peptic ulcer	111 (25.3)
Hypertension	51 (11.6)
Diabetes	49 (11.2)
Osteoporosis	33 (7.5)
Adrenal insufficiency	19 (4.3)
Bone fractures	7 (1.6)
MACE	7 (1.6)
Non‐fatal myocardial infarction	4 (0.9)
Non‐fatal stroke	3 (0.7)
Cardiovascular death	0
Glaucoma	3 (0.7)

Abbreviations: CRS = chronic rhinosinusitis; ESS = endoscopic sinus surgery; MACE = major adverse cardiovascular event; N‐ERD = non‐steroidal anti‐inflammatory drug‐exacerbated respiratory disease; SCS = systemic corticosteroid; SD = standard deviation.

^a Defined as the date of the first dupilumab treatment.

^b Each individual may have had multiple comorbidities or claims.

^c In the 6 months prior to the index date.

3.2. Dupilumab Exposure

Over 1 year of follow‐up, individuals received a mean ± SD of 21.7 ± 7.9 injections of dupilumab over a mean ± SD of 339.2 ± 80.8 days (Table S2). In total, 407 individuals (92.9%) used pen injections, while 213 (48.6%) used syringe injections during follow‐up.

3.3. Rescue Treatment

The overall rate of rescue treatment significantly decreased from 7.89 events per PY in the pre‐index period to 2.14 events per PY in the follow‐up period (difference −5.75 events per PY [95% CI: −6.04, −5.45]; p < 0.0001) (Table 2). The pre‐index/follow‐up period ratio of the overall rate of rescue treatment was 0.27 (95% CI: 0.25, 0.29; p < 0.0001). When adjusted for sex, age at index date, presence/absence of asthma/N‐ERD, and prior history of SCS and ESS, the overall rate of rescue treatment decreased from 3.26 to 0.86 events per PY (difference −2.40 events per PY [95% CI: −2.87, −1.98]). The adjusted pre‐index/follow‐up period ratio of the overall rate of rescue treatment was 0.26 (95% CI: 0.22, 0.32; p < 0.0001).

TABLE 2.

Rescue treatment use before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods).

	Study population
	N = 438
	Pre‐index period	Follow‐up period
Rescue treatment needed, n (%)
Yes	438 (100.0)	174 (39.7)
No	0	264 (60.3)
Total number of rescue treatments needed	3454	937
Number of rescue treatments per individual
Mean ± SD	7.9 ± 5.5	2.1 ± 4.3
Median (range)	7.0 (1.0–30.0)	0.0 (0.0–34.0)
Number of rescue treatments, n (%)
0	0	264 (60.3)
1	32 (7.3)	39 (8.9)
2	46 (10.5)	24 (5.5)
3	30 (6.9)	18 (4.1)
4	38 (8.7)	18 (4.1)
≥ 5	292 (66.7)	75 (17.1)
Overall rate, events per PY (95% CI)	7.89 (7.62, 8.15)	2.14 (2.00, 2.28)
Difference (95% CI)	−5.75 (−6.04, −5.45)
p value a	< 0.0001
Ratio (95% CI)	0.27 (0.25, 0.29)
p value a	< 0.0001
Adjusted overall rate, b events per PY (95% CI)	3.26 (2.88, 3.69)	0.86 (0.70, 1.04)
Difference (95% CI)	−2.40 (−2.87, −1.98)
Ratio (95% CI)	0.26 (0.22, 0.32)
p value a	< 0.0001

Abbreviations: CI = confidence interval; ESS = endoscopic sinus surgery; N‐ERD = non‐steroidal anti‐inflammatory drug‐exacerbated respiratory disease; PY = patient‐year; SCS = systemic corticosteroid; SD = standard deviation.

^a Estimated using the Wald method.

^b Adjusted using the marginal model with negative binomial data and the following covariates: sex, age at index date, presence/absence of asthma/N‐ERD, prior history of SCS, and prior history of ESS.

When assessed in subgroups of individuals with or without prior SCS use or ESS in the pre‐index period, the overall rate of rescue treatment was significantly decreased from the pre‐index period to the follow‐up period regardless of prior rescue treatment (p < 0.0001 for all subgroups; Figure 2). Similarly, dupilumab was associated with significantly decreased overall rates of rescue treatment in individuals with or without a history of asthma/N‐ERD (Figure S2).

FIGURE 2.

Overall rate of rescue treatment before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods) in subgroups defined by prior systemic corticosteroid (SCS) use or endoscopic sinus surgery (ESS). *p < 0.0001 (estimated using the Wald method). CI = confidence interval; PY = patient‐year. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

The overall rate of SCS use significantly decreased from 7.54 to 2.09 prescriptions per PY between the pre‐index and follow‐up periods (difference −5.45 prescriptions per PY [95% CI: −5.74, −5.16]; p < 0.0001). The mean ± SD daily SCS dose decreased from 7.1 ± 9.7 mg in the pre‐index period to 6.1 ± 7.3 mg in the follow‐up period. Compared with the pre‐index period, lower proportions of individuals required prescription of short‐ and long‐term SCS courses during the follow‐up period (70.6% vs. 22.6% and 52.7% vs. 11.4%, respectively) (Table 3). The prescription patterns for short‐term and long‐term SCS courses are summarized in Tables S3 and S4, respectively. Mean long‐term SCS use decreased by −63.9% vs. the pre‐index period after 60 days of dupilumab treatment, with further sustained reductions observed for the remainder of the follow‐up period (Figure S3).

TABLE 3.

Systemic corticosteroid (SCS) prescription before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods).

	Study population
	N = 438
	Pre‐index period	Follow‐up period
SCS, n (%)
Yes	425 (97.0)	169 (38.6)
No	13 (3.0)	269 (61.4)
Total number of SCS prescriptions	3302	914
Number of SCS prescriptions per individual
Mean ± SD	7.5 ± 5.4	2.1 ± 4.3
Median (range)	6.5 (0.0–29.0)	0.0 (0.0–34.0)
Number of SCS prescriptions, n (%)
0	13 (3.0)	269 (61.4)
1	29 (6.6)	37 (8.5)
2	39 (8.9)	25 (5.7)
3	34 (7.8)	19 (4.3)
4	35 (8.0)	15 (3.4)
≥ 5	288 (65.8)	73 (16.7)
Duration of SCS prescription, days/year a
Mean ± SD	135.7 ± 111.5	94.2 ± 113.1
Median (range)	102.0 (3.0–365.0)	38.0 (1.0–365.0)
Cumulative SCS dose, mg/year a
Mean ± SD	614.5 ± 585.9	511.8 ± 681.9
Median (range)	431.4 (16.7–3735.3)	220.0 (6.8–3831.3)
Mean SCS daily dose, mg a
Mean ± SD	7.1 ± 9.7	6.1 ± 7.3
Median (range)	4.6 (0.8–111.3)	4.1 (1.0–48.8)
Short‐term SCS course, n (%)
Yes	309 (70.6)	99 (22.6)
No	129 (29.5)	339 (77.4)
Long‐term SCS course, n (%)
Yes	231 (52.7)	50 (11.4)
No	207 (47.3)	388 (88.6)
Overall rate, prescriptions per PY (95% CI)	7.54 (7.28, 7.80)	2.09 (1.95, 2.22)
Difference (95% CI)	−5.45 (−5.74, −5.16)
p value b	< 0.0001
Ratio (95% CI)	0.28 (0.26, 0.30)
p value b	< 0.0001

Abbreviations: CI = confidence interval; PY = patient‐year; SD = standard deviation.

^a Analysis among patients with SCS prescription in each period.

^b Estimated using the Wald method.

The overall rate of ESS significantly decreased from 0.35 to 0.05 procedures per PY between the pre‐index and follow‐up periods (difference −0.29 procedures per PY [95% CI: −0.35, −0.24]; p < 0.0001) (Table 4).

TABLE 4.

Endoscopic sinus surgeries performed before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods).

	Study population
	N = 438
	Pre‐index period	Follow‐up period
ESS, n (%)
Yes	80 (18.3)	14 (3.2)
No	358 (81.7)	424 (96.8)
Total number of ESS procedures	152	23
Number of ESS procedures per individual
Mean ± SD	0.3 ± 0.8	0.1 ± 0.3
Median (range)	0.0 (0.0–6.0)	0.0 (0.0–3.0)
Number of ESS procedures, n (%)
0	358 (81.7)	424 (96.8)
1	13 (3.0)	7 (1.6)
2	65 (14.8)	5 (1.1)
3	1 (0.2)	2 (0.5)
4	0	0
≥ 5	1 (0.2)	0
Type of ESS performed, n (%)
I	26 (5.9)	8 (1.8)
II	3 (0.7)	0
III	12 (2.7)	2 (0.5)
IV	46 (10.5)	5 (1.1)
V	1 (0.2)	0
Overall rate, procedures per PY (95% CI)	0.35 (0.29, 0.40)	0.05 (0.03, 0.07)
Difference (95% CI)	−0.29 (−0.35, −0.24)
p value a	< 0.0001
Ratio (95% CI)	0.15 (0.10, 0.23)
p value a	< 0.0001

Abbreviations: CI = confidence interval; ESS = endoscopic sinus surgery; PY = patient‐year; SD = standard deviation.

^a Estimated using the Wald method.

3.4. Healthcare Resource Utilization

A summary of HCRU before and after initiation of dupilumab treatment is provided in Table 5. All 438 individuals attended outpatient visits during the pre‐index and follow‐up periods, and the mean ± SD number of outpatient visits per individual was numerically lower in the pre‐index vs. follow‐up period (12.5 ± 6.6 vs. 15.2 ± 8.6). Compared with the pre‐index period, lower proportions of individuals required ER visits (7.5% vs. 4.8%) and inpatient hospitalizations (20.8% vs. 6.2%) during the follow‐up period. However, in individuals who had an ER visit or hospitalization, the mean ± SD number of ER visit or hospitalization days per year remained numerically similar before and after the initiation of dupilumab treatment.

TABLE 5.

Healthcare resource utilization before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods).

	Study population
	N = 438
	Pre‐index period	Follow‐up period
Outpatient visit, n (%)
Yes	438 (100.0)	438 (100.0)
No	0	0
Outpatient visits per individual, days/year
Mean ± SD	12.5 ± 6.6	15.2 ± 8.6
Median (range)	11.0 (1.0–51.0)	13.0 (4.0–82.0)
ER visit, n (%)
Yes	33 (7.5)	21 (4.8)
No	405 (92.5)	417 (95.2)
ER visits per individual, days/year
Mean ± SD	2.6 ± 2.4	2.7 ± 2.9
Median (range)	1.0 (1.0–9.0)	1.0 (1.0–10.0)
Inpatient hospitalization, n (%)
Yes	91 (20.8)	27 (6.2)
No	347 (79.2)	411 (93.8)
Hospitalizations per individual, days/year
Mean ± SD	8.9 ± 6.1	8.6 ± 5.1
Median (range)	7.0 (1.0–32.0)	9.0 (2.0–21.0)
Prescribed CRS‐associated medications, a n (%)
Yes	426 (97.3)	392 (89.5)
Nasal corticosteroids	360 (82.2)	265 (60.5)
Nasal drops	317 (72.4)	250 (57.1)
Inhalation spray	147 (33.6)	66 (15.1)
Leukotriene antagonists	326 (74.4)	288 (65.8)
Anti‐allergy drugs	258 (58.9)	237 (54.1)
Carbocisteine	198 (45.2)	125 (28.5)
Anti‐microbial drugs	175 (40.0)	89 (20.3)
Vasoconstrictors	41 (9.4)	17 (3.9)
No	12 (2.7)	46 (10.5)
Prescriptions for CRS‐associated medications per individual, times/year
Mean ± SD	24.1 ± 17.4	18.8 ± 14.5
Median (range)	20.0 (1.0–111.0)	16.0 (1.0–74.0)

Abbreviations: CRS = chronic rhinosinusitis; ER = emergency room; SD = standard deviation.

^a Each individual may have been prescribed more than one CRS‐associated medication.

The mean ± SD number of prescriptions for CRS‐associated medications per individual decreased from 24.1 ± 17.4 per year during the pre‐index period to 18.8 ± 14.5 per year during the follow‐up period (Table 5). The proportion of individuals who were not prescribed CRS‐associated medications increased from 2.7% (n = 12) before dupilumab initiation to 10.5% (n = 46) after dupilumab initiation. The proportion of individuals who were prescribed each type of CRS‐associated medication also decreased in the 12 months after dupilumab was initiated. The most common CRS‐associated medications (prescribed in ≥ 50% of individuals) were nasal corticosteroids (82.2% in the pre‐index period vs. 60.5% in the follow‐up period), leukotriene antagonists (74.4% vs. 65.8%), and anti‐allergy drugs (58.9% vs. 54.1%).

3.5. Disease Control

The proportion of individuals with a “controlled” status of CRSwNP increased from 41.3% (n = 181) during the pre‐index period to 87.2% (n = 382) during the follow‐up period (Figure 3). Conversely, the proportion of individuals with “uncontrolled” CRSwNP decreased from 42.2% (n = 185) to 8.2% (n = 36) after initiation of dupilumab treatment. At 3 months after dupilumab withdrawal, most individuals (n = 423, 96.6%) had maintained disease control.

FIGURE 3.

Control status of chronic rhinosinusitis with nasal polyps before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods) in the study population (N = 438). Disease control status was defined as follows: “controlled” = no or 1 rescue treatment required (short‐term SCS use or ESS) in the last 12 months; “partly controlled” = 2 courses of rescue treatment required (short‐term SCS use or ESS) in the last 12 months; and “uncontrolled” ≥ 3 courses of rescue treatment required (short‐term SCS use or ESS) or ≥ 1 long‐term SCS use required in the last 12 months. ESS = endoscopic sinus surgery; SCS = systemic corticosteroids. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

4. Discussion

Overall, this analysis provides clinically meaningful insights into the prescribing of add‐on dupilumab treatment for CRSwNP in a Japanese real‐world setting.

The age and sex distributions of Japanese individuals with CRSwNP who initiated dupilumab treatment were comparable with those of previous studies [12, 13, 15, 16, 17, 18]. Regarding complications of CRSwNP with type 2 inflammation, the most common was allergic rhinitis (99.3%), which was higher than that in a previous report (78.2%) [18]. The high prevalence of comorbid allergic rhinitis can be attributed to the prescription of nasal steroids, which are commonly prescribed by otorhinolaryngologists, being used when these clinicians make a diagnosis of allergic rhinitis. The finding that 70.6% of individuals in this study had comorbid asthma aligns with the asthma comorbidity rate of 50.0%–70.6% in randomized Phase 3 studies of individuals with CRSwNP treated with dupilumab [12, 13]. During the year prior to the index date, 97.0% of individuals had received SCS treatment and 18.3% had undergone ESS procedures. Given the small number of individuals with a prior history of ESS during the pre‐index period, it is assumed that the dupilumab treatment strategy involved a delayed initiation of dupilumab post‐surgery and the adoption of a wait‐and‐see approach that followed the individual's postoperative course. The status of dupilumab exposure over 1 year of follow‐up revealed that most individuals remained on the prescriptions that they had started.

In this real‐world study, Japanese individuals with CRSwNP who newly initiated dupilumab treatment had a significant reduction in the overall rate of rescue treatment over 12 months after dupilumab initiation, including decreases in SCS prescriptions and ESS procedures. These findings are in line with those reported in the pooled analysis of the LIBERTY NP SINUS‐24 and LIBERTY NP SINUS‐52 studies in individuals with severe CRSwNP, in which the risk of requiring rescue treatment over 12 months was significantly lower in the dupilumab 300 mg every 2 weeks arm vs. the placebo arm (hazard ratio: 0.243; 95% CI: 0.169, 0.351; p < 0.0001) [12]. In the subanalysis of the LIBERTY NP SINUS studies, the decreased need for rescue treatment with add‐on dupilumab therapy was observed regardless of prior SCS or ESS history [19]. Further, the cumulative event rate for rescue treatment was reduced with dupilumab vs. placebo in the pooled Japanese subpopulation of SINUS‐24 and SINUS‐52 (9.1% vs. 31.3%) [13]. The improvements in the overall rate of rescue treatment seen in this real‐world study were observed regardless of concurrent asthma/N‐ERD or prior SCS/ESS history, both of which are categorized prior to the administration of dupilumab. Our analysis found a difference between the overall rate (per PY) and adjusted overall rate (per PY) of rescue treatment before and after dupilumab treatment initiation. This difference might reflect the inclusion of prior history of SCS, as a comparison of overall rates for each covariate revealed a much lower value among individuals without prior SCS use. However, as the number is small, there is no clearly identifiable effect.

Several real‐world studies previously confirmed significant symptomatic improvement with dupilumab treatment in individuals with CRSwNP [15, 16, 17, 20, 21]. However, of these studies, rescue treatment requirements during dupilumab treatment were only described in one Phase 4 study [16] and two single‐center studies [20, 21]. In the Phase 4 study, which enrolled 648 Italian people with severe, uncontrolled CRSwNP, 19 participants (2.9%) required rescue treatment with SCS or ESS over 12 months of dupilumab treatment [16]. In one Italian single‐center study, all 66 individuals with CRSwNP had a “good” response and continued to receive dupilumab treatment after 12 months (according to European Forum for Research and Education in Allergy and Airway Diseases [EUFOREA] criteria [22]), including a reduced need for SCS [20]. In the second single‐center study, conducted in the Netherlands, 2 out of 131 participants with CRSwNP (1.5%) required rescue treatment (both with SCS) during 48 weeks of dupilumab treatment [21].

However, it should be noted that it is difficult to compare the results of our study with those of studies conducted in other countries because of differences in the healthcare systems of Japan vs. other countries. In Japan, the patient criteria for reimbursement of dupilumab administration are determined by the Optimal Use guidelines of the healthcare authority, with the indications for use of the drug in the guidelines based on the inclusion criteria of the LIBERTY NP SINUS‐24 and SINUS‐52 studies (e.g., treatment history of SCS within 2 years or treatment history of ESS, nasal polyp score ≥ 5 points, and nasal congestion score ≥ 2) [12]; therefore, Japanese individuals receiving dupilumab in real‐world settings are likely to be similar to those who were involved in clinical trials. In addition, the proportion of Caucasian individuals with a type 2 endotype of CRSwNP has been shown to be higher than that in Japanese individuals with CRSwNP [7]; however, a recent report suggested similarities in molecular mechanisms in the type 2 endotype in both White/non‐Asian and Japanese individuals with CRSwNP [23]. These points should be considered when assessing the external validity of our findings.

In this study, after the initial prescription of dupilumab, a 59.4% reduction in participants who were prescribed SCS at least once from the pre‐index period was observed. As a limitation of the database, the reason for SCS prescription could not be identified, and the SCS prescription may not have only been for CRS, which means the rate of individuals being prescribed SCS is potentially higher than the exact rate. However, the following result of the study suggests that long‐term SCS prescriptions for CRS were steadily decreasing; a 63.9% reduction in SCS prescription at 60 days from the pre‐index period was observed, and the numbers of both short‐ and long‐term SCS prescriptions decreased. For safety reasons, the European Position Paper on Rhinosinusitis and Nasal Polyps 2020 (EPOS2020) advises limiting short‐term (prescriptions lasting ≤ 4 weeks) SCS prescriptions to two courses per year, while long‐term (> 4 weeks) SCS prescriptions are not recommended [9]. In this analysis, the proportion of individuals with short‐term SCS prescriptions decreased after dupilumab initiation during the follow‐up period. The percentage change from the index date in mean long‐term SCS prescriptions suggests that SCS tapering occurs early after initiation of dupilumab. While these data indicate the possibility of using dupilumab intermittently to manage disease exacerbations rather than prescribing SCS, dupilumab should be considered a disease‐controlling agent with continuous administration, not a short‐term rescue medication for CRSwNP.

In this study, we defined “control” status for CRSwNP by the requirement of rescue treatment, with reference to EPOS2020 guidance [9]. The proportion of individuals with a “controlled” status increased from 41.3% during the pre‐index period to 87.2% during the follow‐up period. According to the EUFOREA and EPOS expert panel members, the requirement of rescue treatment for CRSwNP can be used to assess the status of disease control and also the achievement of clinical remission [24].

The main strengths of this real‐world study are that it addresses a knowledge gap with regard to the impact of add‐on dupilumab therapy on rescue treatment requirements of individuals with CRSwNP; its large sample size compared with the Japanese subcohort in the LIBERTY NP SINUS‐52 study (n = 33 received dupilumab) allowed a more precise assessment of dupilumab's effectiveness [13]; and that the eligible patients in this study are likely to be representative of Japanese individuals with CRSwNP treated with dupilumab in actual clinical practice. Furthermore, the longitudinal approach made possible the assessment of interventions 12 months pre‐ and post‐dupilumab initiation, reducing potential confounding and ensuring internal validity by using individuals as their own controls.

However, the study was limited by factors associated with the MDV source data, which do not provide an individual's information on medical care prior to the acute hospitalization (e.g., treatment history at a different hospital) or post‐transfer information. To mitigate this limitation, the study only included individuals with ≥ 12 months of pre‐ or post‐index information in the database. Other study limitations include potential differences between disease definitions in the database (according to the ICD‐10 code) and individuals' diagnosis and disease status in the real‐life clinical setting, and that the database does not include information on disease severity. Moreover, it was not possible to confirm drug prescription information in the MDV database, including whether dupilumab was administered (i.e., there is a lack of information on actual medication adherence patterns), and whether dupilumab was prescribed for other indications (i.e., atopic dermatitis or asthma). To mitigate this point, the study included individuals who were prescribed dupilumab by ENT specialists only. However, > 70% of individuals included in our study had an “asthma” disease code and the impact of dupilumab on asthma and the interaction between asthma and CRSwNP should be considered when interpreting our results. In addition, the MDV database does not track information about adverse effects of drugs, despite safety issues being an essential consideration. Other real‐world studies have reported injection site reactions and transient increases in eosinophil counts as common adverse effects during treatment with dupilumab [20, 25], and these points should be considered during its clinical use. Finally, the study was observational and uncontrolled and so at risk of biases caused by unmeasured complex confounders present in healthcare databases, including the MDV; thus, a causal relationship between dupilumab treatment and outcomes cannot be confirmed.

5. Conclusion

This real‐world, observational study demonstrated that add‐on dupilumab treatment provides substantial benefits in Japanese individuals with CRSwNP, including a significantly decreased requirement for rescue treatment. Therefore, dupilumab is a valuable therapeutic option for the treatment of CRSwNP.

Conflicts of Interest

M.Y. has received fees for lectures from Sanofi, and research funding from Kissei Pharmaceutical. M.Ok. has received lecture fees from Mitsubishi Tanabe Pharma, Novartis, Sanofi, and Taiho. M.T., Y.S., M.Or., and M.I. are employees of Sanofi K.K.

Supporting information

Data S1. Statistical considerations.

Table S1. Secondary and exploratory outcomes.

Table S2. Summary of dupilumab exposure during the follow‐up period.

Table S3. Summary of prescription patterns for short‐term systemic corticosteroids before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods).

Table S4. Summary of prescription patterns for long‐term systemic corticosteroids before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods).

Figure S1. Study design.

Figure S2. Overall rate of rescue treatment before and after initiation of dupilumab treatment (during the pre‐index and follow‐up periods) in subgroups defined by prior history of asthma/non‐steroidal anti‐inflammatory drug‐exacerbated respiratory disease.

Figure S3. Percentage change from index date in mean long‐term systemic corticosteroid use during dupilumab treatment (follow‐up period) in the study population (N = 438).

Yoshikawa M., Okano M., Takeuchi M., Sunaga Y., Orimo M., and Ishida M., “Reduction of Rescue Treatment With Dupilumab for Chronic Rhinosinusitis With Nasal Polyps in Japan,” The Laryngoscope 135, no. 9 (2025): 3093–3103, 10.1002/lary.32171.

Data Availability Statement

The data generated during this study are available from the author upon reasonable request. MT confirms that she had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.