Indirect Treatment Comparison between Ribociclib Combined with Non-Steroidal Aromatase Inhibitors and Ovarian Function Suppression vs. Tamoxifen in Premenopausal Women with Early Breast Cancer

Keywords

ribociclib - indirect treatment comparison - HR-positive - HER2-negative - early breast cancer

Schlüsselwörter

Ribociclib - indirekter Behandlungsvergleich - HR-positiv - HER2-negativ - Brustkrebs im Frühstadium

Introduction

Breast cancer (BC) is one of the most common types of cancer in women worldwide. The majority of newly diagnosed BC patients suffer from early-stage hormone receptor (HR)-positive and human epidermal growth factor receptor 2 (HER2)-negative BC [1] [2]. Locoregional treatment may initially lead to a disease-free status, but a metastatic recurrence is common and is a primary cause of death in patients with early BC [1] [3].

Adjuvant endocrine therapies (ET) such as tamoxifen as well as steroidal and non-steroidal aromatase inhibitors (NSAI) lowering estrogen levels have been shown to decrease locoregional and distant recurrences as well as BC-specific mortality in HR-positive early BC [4]. As NSAI are contraindicated in women with intact ovarian function due to their mechanism of action, ovarian function suppression (OFS) is mandatory for adjuvant ET with NSAI in premenopausal women. Tamoxifen ± OFS and NSAI + OFS are commonly applied treatment options for premenopausal women with HR-positive early BC in German routine care [5] and recommended by guidelines [2] [6] [7] [8]. However, recurrence remains an important issue especially for patients who are at intermediate or high risk as assessed by anatomic staging or supported by genomic testing. Recurrence, mostly as distant metastases, is usually incurable leading to BC-related death in most of the patients [3].

For HR-positive, HER2-negative advanced BC, the drug class of cyclin dependent kinase (CDK) 4/6 inhibitors has been introduced as a novel treatment option. The CDK4/6 inhibitors palbociclib, abemaciclib, and ribociclib have been shown to act synergistically with ET [9] and were approved in combination with NSAI or fulvestrant as initial therapy for advanced BC or after disease progression following prior ET in pre- and postmenopausal patients with HR-positive, HER2-negative advanced BC [10] [11] [12] [13] [14] [15] [16] [17]. Additionally, abemaciclib was approved for adjuvant treatment of HR-positive, HER2-negative, node-positive early BC with high risk of recurrence based on the monarchE clinical trial [17].

Similar efficacy has been demonstrated with ribociclib, as evidenced by the MONALEESA and NATALEE trials. In the MONALEESA clinical trial program, ribociclib in combination with estrogen inhibiting therapies significantly improved overall survival (OS) and progression-free survival (PFS, MONALEESA-2, -3, -7) in patients with HR-positive, HER2-negative advanced BC [18]. These benefits of ribociclib in advanced BC extend to early BC, as shown in the NATALEE trial, in which adjuvant ribociclib + NSAI significantly improved invasive disease-free survival (iDFS) in patients with HR-positive, HER2-negative early BC when compared to NSAI alone in patients with Stage II and Stage III disease [19]. This data led to the approval of ribociclib + NSAI for the adjuvant treatment of patients with HR-positive, HER2-negative early BC at high risk of recurrence [19].

As a selective small-molecule inhibitor with a highly specific nanomolar inhibitory activity against CDK4/cyclin-D1 and CDK6/cyclin-D3 complexes, ribociclib is generally well tolerated, with neutropenia and nausea being most prevalent adverse effects [20].

There is currently no clinical evidence on the effectiveness and safety of ribociclib + NSAI + OFS in premenopausal women with HR-positive, HER2-negative early BC when compared to a frequently used treatment option in German clinical routine (tamoxifen). We therefore conducted the IRINA study, an indirect treatment comparison (ITC) based on secondary use of data derived from the NATALEE trial [21] and a retrospective data collection (CLEAR-B) as external control [22]. Using the best practices approach for adjusted ITC [23], the IRINA study therefore provides an effectiveness and safety comparison of ribociclib + NSAI + OFS vs. tamoxifen ± OFS.

Material and Methods

Study design

IRINA (Indirect treatment comparison of the effectiveness of RIbociclib combined with Non-steroidal Aromatase inhibitors vs. tamoxifen for the adjuvant treatment of premenopausal women with HR-positive, HER2-negative early BC) is a retrospective, non-interventional study to assess effectiveness and safety of ribociclib + NSAI + OFS vs. tamoxifen ± OFS in premenopausal women with HR-positive, HER2-negative early-stage BC. This assessment is based on an ITC of clinical trial data (ribociclib arm) and external control data collected in German routine clinical care (tamoxifen ± OFS arm). Data on ribociclib + NSAI + OFS were derived from the subset of premenopausal patients of the NATALEE trial, a randomized phase III multicenter trial on efficacy and safety of ribociclib in HR-positive, HER2-negative adjuvant breast cancer [19]. In the NATALEE trial, 5101 patients were recruited at 393 sites globally. Data cut-off was on 21 July 2023.

External control data were retrieved from the CLEAR-B data subset of tamoxifen-treated patients [22]. The CLEAR-B study (NCT05870813; AGO-B-059) collected data in a retrospective setting from medical records of patients treated in 75 BC centers in Germany. Only anonymized data were captured in electronic case report forms, and automated plausibility checks ensured data quality. The CLEAR-B study inclusion period (January 2016 – June 2019) overlapped with the NATALEE trial recruitment period to ensure a minimum follow-up of 3 years.

Eligibility criteria

The following eligibility criteria were applied to both data sources: Premenopausal women (18–60 years) with HR-positive, HER2-negative early-stage BC without distant metastases were included if they met the criteria of anatomic stage III, IIB, or IIA (N1 or N0 grade 3 or N0 grade 2 with any of the following criteria: Ki-67 ≥ 20%, Oncotype DX Breast Recurrence Score ≥ 26 [24], high risk according to Prosigna/PAM50 [25], MammaPrint [26], or EndoPredict EPclin Risk Score [27]). Patients undergoing therapy with a CDK4/6 inhibitor other than ribociclib as well as patients with limited life expectancy (< 5 years) or poor general condition (Eastern Cooperative Oncology Group [ECOG] status > 1 [28]) were excluded.

Study endpoints

The effectiveness endpoints iDFS, distant disease-free survival (dDFS), recurrence-free survival (RFS), and OS were time-to-event endpoints, calculated from the reference date to the date of the first contributing event according to the Standardized Definitions for Efficacy End Points in Adjuvant Breast Cancer Trials (STEEP [29]). For the NATALEE arm, the reference date was the date of randomization, which, according to protocol, had to be performed within 12 months after the beginning of adjuvant ET. For the control arm, the reference date was the start of adjuvant ET. Due to the retrospective nature of this study, the definition of endpoints as primary or secondary was omitted.

Safety-related endpoints included early treatment termination (TT) and toxicity-related treatment termination (TTtox), excluding the regular cessation of ribociclib after 3 years. In the ribociclib arm, patients could either discontinue both ribociclib + NSAI (TT-a) or stop ribociclib while continuing NSAI treatment (TT-b). Patients who discontinued ribociclib, but continued NSAI (TT-b) were further divided into two groups: those who stopped NSAI before the data cut-off (b1) and those who continued NSAI at least until the cut-off date (b2). In the tamoxifen ± OFS arm, treatment discontinuation was defined as the permanent discontinuation of tamoxifen. For each category, the time to TT or TTtox was calculated as the time (in months) from the reference date to permanent discontinuation of treatment. For safety analyses, the time to complete treatment discontinuation (TT-a, TTtox-a) was compared between the two treatment arms. Additionally, the discontinuation of ribociclib or NSAI in the ribociclib + NSAI + OFS arm (TT-b, TTtox-b) was compared to the discontinuation of tamoxifen in the tamoxifen ± OFS arm.

Reasons for early treatment discontinuation were summarized using descriptive statistics. For both effectiveness and safety-related endpoints, hazard ratios (HR) were calculated using Cox proportional hazards regression, median time-to-event and follow-up, as well as Kaplan-Meier estimates at yearly intervals, all with 95% confidence intervals (CI). All HRs were calculated based on a 4-year observation period for both treatment arms.

Statistics

All statistical analyses were pre-specified in the study protocol. To address potential imbalances, the methodology of Desai and Franklin [23] was applied. Propensity scores (PS) were calculated using logistic regression, with treatment as the dependent variable and confounders as independent variables (neoadjuvant chemotherapy, pT stage, pN stage, age, and grading; pre-specified based on a systematic literature review and expert discussion). Standardized mean differences (SMD) were used to assess balance between treatment arms before and after weighting (target: SMD < 0.25). Patients in non-overlapping PS density regions were trimmed from the dataset.

Two weighting approaches were applied to estimate the average treatment effect among the treated population (ATT): fine stratification weights and standardized mortality ratio weights (SMRW) [23]. The choice of weighting method was based on its performance in achieving balance in confounders, as assessed by SMD distributions and density plots, following the decision scheme outlined by Desai and Franklin (2019) [23] (Supplementary Fig. S1) (appendix, online). The comparability of the cohorts, including baseline characteristics such as systemic therapies/chemotherapy, was evaluated both before and after adjustment, as shown in the baseline tables ([Table 1] and Supplementary Table S2) (appendix, online).

Table 1 Unweighted patient characteristics along with analysis of propensity score variations (standardized mean differences) when appropriate.

Characteristic	Ribociclib + NSAI + OFS N = 1115	Tamoxifen ± OFS N = 822	SMD unweighted	SMD SMRW
NSAI = nonsteroidal aromatase inhibitors; OFS = ovarian function suppression; SD = standard deviation; SMD unweighted = standardized mean differences before imputation; SMD SMRW = SMD after propensity score adjustment by standardized mortality ratio weights. Missing values are noted where present.
Age
Mean (SD), years	44.1 ± 6.10	45.1 ± 6.14
Median (Range), years	45.0 (24.0–60.0)	46.0 (23.0–64.0)
> 40 years, n (%)	805 (72.2%)	645 (78.5%)	0.143	−0.023
≤ 40 years, n (%)	310 (27.8%)	177 (21.5%)	−0.142	0.023
Anatomic stage group, n (%)
I	3 (0.3%)	0 (0.0%)
II	419 (37.6%)	654 (79.6%)
III	693 (62.2%)	168 (20.4%)
Clinical tumor category at diagnosis, n (%)
cT0	2 (0.2%)	2 (0.2%)
cT1	176 (15.8%)	289 (35.2%)
cT > 1	856 (76.8%)	505 (61.4%)
cTis	1 (0.1%)	0 (0.0%)
cTx	62 (5.6%)	0 (0.0%)
Missing	18 (1.6%)	26 (3.2%)
Pathological tumor category at surgery, n (%)
pT0	29 (2.6%)	44 (5.4%)	0.145	0.009
pT1	365 (32.7%)	322 (39.2%)	0.126	−0.010
pT > 1	706 (63.3%)	447 (54.4%)	−0.182	0.007
pTis	7 (0.6%)	0 (0.0%)
pTx	8 (0.7%)	0 (0.0%)
(Missing)	0 (0.0%)	9 (1.1%)
Pathological nodal status at surgery, n (%)
pN0	144 (12.9%)	238 (29.0%)	0.458	0.018
pN1	491 (44.0%)	402 (48.9%)	0.155	0.004
pN > 1	479 (43.0%)	133 (16.2%)	−0.591	−0.020
pNx	1 (0.1%)	0 (0.0%)
(Missing)	0 (0.0%)	49 (6.0%)
Grading at diagnosis, n (%)
G1	95 (8.5%)	87 (10.6%)	0.023	−0.031
G2	637 (57.1%)	489 (59.5%)	−0.143	0.003
G3	210 (18.8%)	231 (28.1%)	0.140	0.018
(Missing)	173 (15.5%)	15 (1.8%)
Prior neo-/adjuvant chemotherapy, n (%)
Yes	1032 (92.6%)	645 (78.5%)
No	83 (7.4%)	177 (21.5%)
Neoadjuvant chemotherapy received, n (%)
Yes	548 (49.1%)	301 (36.6%)	0.250	0.007
No	567 (50.9%)	521 (63.4%)	−0.250	−0.007
Adjuvant chemotherapy received, n (%)
Yes	514 (46.1%)	374 (45.5%)
No	601 (53.9%)	448 (54.5%)
Prior radiation therapy, n (%)
Yes	1017 (91.2%)	698 (84.9%)
No	0 (0.0%)	121 (14.7%)
(Missing)	98 (8.8%)	3 (0.4%)
Ovarian suppression therapy, n (%)
Yes	1088 (97.6%)	130 (15.8%)
No	0 (0.0%)	692 (84.2%)
(Missing)	27 (2.4%)	0 (0.0%)
KI67 score at diagnosis, n (%)
Ki-67 > 20	438 (39.3%)	351 (42.7%)
Ki-67 ≤ 20	393 (35.2%)	417 (50.7%)
(Missing)	284 (25.5%)	54 (6.6%)
KI67 score at surgery, n (%)
Ki-67 > 20	210 (18.8%)	200 (24.3%)
Ki-67 ≤ 20	369 (33.1%)	284 (34.5%)
(Missing)	536 (48.1%)	338 (41.1%)

Cox proportional hazards regressions and Kaplan Meier estimates were conducted using the PS weights. Missing values were handled by multiple imputations (MI) according to Leyrat et al. [30], incorporating PS values across imputed datasets (MI.ps) in both effectiveness and safety analyses. Potential immortal time bias – arising from the absence of recurrence events between adjuvant ET initiation and randomization in the NATALEE arm – was assessed using a landmark approach [31], which defined follow-up start times at points when increasing proportions (10% steps) of ribociclib + NSAI + OFS patients initiated therapy. Untreated patients at each landmark were reassigned to the control group and effects on all endpoint results at each landmark were interpreted.

Descriptive analyses were performed for categorical variables (means of absolute and relative frequencies) and continuous variables (mean, standard deviation [SD], minimum, median, maximum, and number of non-missing values). All statistical analyses were conducted using R (version 4.3.3 and higher) and R studio.

Results

Patients

A total of 1115 patients treated with ribociclib + NSAI + OFS (NATALEE) and 822 out of 1562 patients treated with tamoxifen ± OFS in 56 centers (CLEAR-B) met the eligibility criteria and were included ([Fig. 1]). Primary analyses revealed an imbalance between study populations in most confounding variables (SMD > 0.25), thus requiring confounder adjustment. In propensity score variation analysis, SMRW achieved the best adjustment with regards to the sum of absolute SMDs, with all confounders being below the pre-specified threshold of SMD 0.25. Therefore, SMRW was applied in confounder adjustment ([Table 1] and Supplementary Table S1) (appendix, online). As a result of SMRW confounder adjustment, the weighted patient number in the comparator arm (tamoxifen ± OFS) was 1103 patients. Additionally, subgroup analyses were conducted to evaluate the consistency of treatment effects for both effectiveness and safety outcomes. The results of these analyses are provided in Supplementary Table S2 (appendix, online).

Fig. 1 Study design of IRINA, a retrospective non-interventional study to assess effectiveness and safety of ribociclib + NSAI + OFS vs. tamoxifen ± OFS in premenopausal women with early-stage breast cancer. For indirect treatment comparisons, the methodology of Desai and Franklin was applied [20] to compensate for the imbalance within the study populations (CDK = cyclin dependent kinase; ECOG = Eastern Cooperative Oncology Group; HER2 = human epidermal growth factor receptor 2; HR = hormone receptor; NSAI = nonsteroidal aromatase inhibitors; OFS = ovarian function suppression).

Effectiveness outcomes

The analyses ([Table 2]) showed that patients in the ribociclib + NSAI + OFS arm had a significantly better 4-year iDFS probability (90%) compared to patients in the tamoxifen ± OFS arm (78%), resulting in an absolute difference of 12% at year 4 (HR = 0.5 [95% CI 0.35; 0.71]; p < 0.01). Similar benefits favoring ribociclib + NSAI + OFS over tamoxifen ± OFS were observed across additional efficacy endpoints. For dDFS, the absolute difference at 4 years was 10%, with ribociclib + NSAI + OFS showing a 4-year survival probability of 91% compared to 81% for tamoxifen ± OFS (HR = 0.52 [95% CI 0.35; 0.77]; p = 0.01). RFS reflected a 13% absolute difference at 4 years, with ribociclib + NSAI + OFS at 91% and tamoxifen ± OFS at 78% (HR = 0.42 [95% CI 0.29; 0.62]; p < 0.01). Notably, substantial benefit was also observed in OS, with ribociclib + NSAI + OFS at 97% and tamoxifen ± OFS at 91%, resulting in a 6% absolute increase in survival probability at 4 years (HR = 0.34 [95% CI 0.18; 0.63]; p = 0.01). Landmark sensitivity analyses confirmed the robustness of all effectiveness outcomes (Supplementary Fig. S2–S5) (appendix, online). Kaplan-Meier plots for analyses of effectiveness outcomes are depicted in [Fig. 2].

Table 2 Indirect treatment comparison (ITC) of effectiveness and safety endpoints.

	Ribociclib + NSAI + OFS					Tamoxifen ± OFS					ITC
	Patients (under risk)	Event n (%)	Q1 time to event (95% CI)	Median time to event (95% CI)	Median follow-up (95% CI)	Patients (under risk)	Event n (%)	Q1 time to event (95% CI)	Median time to event (95% CI)	Median follow-up (95% CI)	HR (95% CI)	Log-rank test
Indirect treatment comparison (ITC) between ribociclib + NSAI + OFS vs. tamoxifen ± OFS (4 years). Endpoints are time-to-event endpoints (months). Patient data from propensity score overlaps were included for analysis after adjustment by standardized mortality ratio weights. Effectiveness endpoints comprise invasive disease-free survival (iDFS), distant disease-free survival (dDFS), recurrence-free survival (RFS), and overall survival (OS). Safety endpoints are treatment discontinuation of ribociclib + NSAI + OFS and tamoxifen ± OFS for any reason (TTa) and due to toxicity of treatment (TTa_Tox). Partial discontinuation of ribociclib (TTb, TTb_Tox) was compared with total treatment discontinuation of tamoxifen. Partial discontinuation of ribociclib was further differentiated according to subsequent NSAI continuation (discontinuation before data cut-off date, b1; continuation at least until cut-off date, b2). ITC = indirect treatment comparison; NSAI = nonsteroidal aromatase inhibitors; OFS = ovarian function suppression (HR = hazard ratio; CI = confidence interval; NE = not estimable). The number of patients in the tamoxifen ± OFS arm represents the weighted sample size in the comparator arm, derived from statistical adjustments and not reflecting the naive sample size.
Effectiveness endpoints
iDFS	1115 (1115)	83 (7.4%)	NE [NE; NE]	NE [NE; NE]	33.71 [33.28; 35.55]	1103 (1103)	243 (22%)	59.5 [45.14; 79.84]	NE [NE; NE]	53.45 [51.06; 56.51]	0.5 [0.35; 0.71]	p < 0.01
dDFS	1115 (1115)	73 (6.5%)	NE [NE; NE]	NE [NE; NE]	33.68 [33.28; 35.42]	1103 (1103)	201 (18.2%)	NE [48.33; NE]	NE [NE; NE]	53.59 [51.22; 56.38]	0.52 [0.35; 0.77]	p = 0.01
RFS	1115 (1115)	70 (6.3%)	NE [NE; NE]	NE [NE; NE]	33.68 [33.28; 35.42]	1103 (1103)	237 (21.5%)	60.71 [45.54; NE]	NE [NE; NE]	53.55 [51.09; 56.51]	0.42 [0.29; 0.62]	p < 0.01
OS	1115 (1115)	24 (2.2%)	NE [NE; NE]	NE [NE; NE]	35.94 [35.91; 36.83]	1103 (1103)	105 (9.5%)	NE [79.15; NE]	NE [NE; NE]	54.18 [51.32; 56.8]	0.34 [0.18; 0.63]	p = 0.01
Safety endpoints
TTa	1106 (1106)	45 (4.1%)	NE [NE; NE]	NE [NE; NE]	36.01 [35.91; 38.11]	1103 (998)	210 (21%)	67.06 [62.78; 70.67]	80.62 [74.74; 84.24]	57.69 [54.6; 60.06]	1.2 [0.71; 2.01]	p = 0.48
TTa_Tox	1106 (1106)	11 (1%)	NE [NE; NE]	NE [NE; NE]	35.91 [35.88; 36.8]	1103 (998)	72 (7.2%)	NE [82.17; NE]	NE [NE; NE]	55.2 [51.84; 57.76]	0.54 [0.22; 1.3]	p = 0.23
TTb1	1106 (1106)	459 (41.5%)	27.37 [25.59; 27.4]	NE [NE; NE]	38.87 [38.7; 39.26]	1103 (998)	210 (21%)	67.06 [62.78; 70.67]	80.62 [74.74; 84.24]	57.69 [54.6; 60.06]	11.48 [7.58; 17.4]	p < 0.01
TTb1_Tox	1106 (1106)	162 (14.6%)	NE [NE; NE]	NE [NE; NE]	35.98 [35.91; 37.95]	1103 (998)	72 (7.2%)	NE [82.17; NE]	NE [NE; NE]	55.2 [51.84; 57.76]	6.81 [3.49; 13.29]	p < 0.01
TTb2	1106 (1106)	314 (28.4%)	32.89 [30.16; 32.95]	NE [NE; NE]	38.67 [38.67; 38.74]	1103 (998)	210 (21%)	67.06 [62.78; 70.67]	80.62 [74.74; 84.24]	57.69 [54.6; 60.06]	8.03 [5.23; 12.33]	p < 0.01
TTb2_Tox	1106 (1106)	105 (9.5%)	NE [NE; NE]	NE [NE; NE]	35.91 [35.88; 36.11]	1103 (998)	72 (7.2%)	NE [82.17; NE]	NE [NE; NE]	55.2 [51.84; 57.76]	4.54 [2.3; 8.94]	p < 0.01

Fig. 2 Kaplan-Meier plots for time-to-event effectiveness endpoints in patients treated with ribociclib + NSAI + OFS (red line) and tamoxifen ± OFS (blue line). Patient data from propensity score overlaps were included for analysis after adjustment by standardized mortality ratio weights (iDFS: invasive disease-free survival, dDFS: distant disease-free survival, RFS: recurrence-free survival, OS: overall survival).

Safety-related Outcomes

Analyses ([Table 2]) showed no statistically significant differences in the risk of simultaneous early treatment discontinuation of ribociclib + NSAI vs. tamoxifen ± OFS. This was consistent for both overall early treatment terminations (TT-a; HR = 1.2 [95% CI 0.71; 2.01], p = 0.48) and toxicity-related early treatment terminations (TTtox-a; HR = 0.54 [95% CI 0.22; 1.30], p = 0.23).

As expected, the risk of treatment discontinuation of ribociclib only was significantly higher compared to tamoxifen ± OFS, where partial discontinuation was not possible. Results are consistent between patients, who discontinued ribociclib and also discontinued NSAI before the data cut off (TT-b1; HR = 11.48 [95% CI 7.58; 17.40], p < 0.01; TTtox-b1; HR = 6.81 [95% CI 3.49; 13.29], p < 0.01) and patients who discontinued ribociclib but continued NSAI at least until the cut-off date (TT-b2; HR = 8.03 [95% CI 5.23; 12.33], p < 0.01; TTtox-b2; HR = 4.54 [95% CI 2.30; 8.94], p < 0.01).

Discussion

The results of the IRINA study demonstrated the superior efficacy of ribociclib + NSAI + OFS compared to tamoxifen ± OFS, a commonly used adjuvant treatment regimen in Germany for premenopausal women with HR-positive, HER2-negative early BC. Ribociclib + NSAI + OFS showed statistically significant benefits across all effectiveness outcomes, including overall survival while therapy discontinuation rates were comparable between ribociclib + NSAI + OFS and tamoxifen ± OFS.

In Germany, NSAI + GnRH agonists have become a standard adjuvant endocrine therapy for high-risk premenopausal women, reflecting evolving treatment practices and guideline recommendations [7] [32] [33] [34] [35]. The SOFT and TEXT trials were pivotal in demonstrating the benefits of adding ovarian suppression to NSAI or tamoxifen, particularly for high-risk patients [36]. Furthermore, a meta-analysis from EBCTCG confirmed the superiority of NSAI + OFS over tamoxifen ± OFS in improving iDFS for premenopausal women with estrogen receptor positive BC [37]. The ADAPTCycle study further supported these findings in the neoadjuvant setting [38]. Recent real-world data from the CLEAR-B study highlight the increasing adoption of NSAI + OFS in clinical practice, with recommendations rising from 8.4% in 2016–2019 to 42.1% in 2022–2023 [35]. These findings contextualize the IRINA results and support the ongoing shift toward individualized, risk-adapted endocrine therapies.

Since no head-to-head randomized clinical trials directly compare ribociclib + NSAI + OFS and tamoxifen ± OFS, we applied an established ITC approach [23]. General comparability between treatment arms was ensured through consistent inclusion/exclusion criteria across both data sources. To balance populations and minimize confounding, a propensity score approach using SMRW was applied [23]. Unlike previous ITC studies on CDK4/6 inhibitors [39] [40], our analysis benefited from access to individual patient-level data (IPD), allowing for more precise adjustments for potential confounders. Before adjustment, notable differences existed in anatomic stage distribution: in the ribociclib + NSAI + OFS arm, 37.6% of patients were in stage II and 62.2% in stage III, while in the tamoxifen ± OFS arm, 79.6% were in stage II and only 20.4% in stage III. After adjustment, the cohorts show improved comparability across most characteristics, though minor differences persist. For instance, the comparability of anatomic stage groups improved, with stage II patients accounting for 55.2% and stage III for 44.8% in the tamoxifen ± OFS arm, aligning more closely with the ribociclib + NSAI + OFS arm. Additionally, differences in ovarian suppression therapy remained due to varying inclusion criteria across the trials ([Table 1] and Supplementary Table S2) (appendix, online).

To address the risk of immortal time bias – arising from the design of the NATALEE trial, where patients were randomized within 12 months of starting adjuvant ET – we conducted a sensitivity analysis using the landmark method [41]. This method synchronized follow-up times by defining specific landmarks (in 10% increments) at which increasing proportions of ribociclib + NSAI + OFS patients had initiated therapy. Patients who had not started ribociclib by a given landmark were considered part of the control group, to allow for an assessment of potential confounding effects of immortal time bias.

The landmark analysis revealed significant benefits for ribociclib + NSAI + OFS in iDFS, as early as the landmark, at which 20% of patients had initiated ribociclib + NSAI + OFS therapy. For OS, significant results were observed at the 50% landmark, likely reflecting the low number of OS events during the IRINA study observation period. The robustness of effects at relatively early landmarks supports the primary results and indicates that immortal time bias had a low impact on the observed treatment effects.

Regarding safety, no statistically significant differences were observed in the risk of simultaneous discontinuation of ribociclib + NSAI compared to the complete discontinuation of tamoxifen ± OFS, regardless of whether discontinuation occurred for any reason or due to toxicity. However, the discontinuation rate of ribociclib alone was significantly higher. This difference likely reflects the asymmetry in treatment structures: in the ribociclib + NSAI arm, patients could partially discontinue therapy by stopping ribociclib or NSAI, whereas in the tamoxifen ± OFS arm, therapy discontinuation meant stopping all endocrine treatment. Moreover, variations in real-world practices, such as differences in drug discontinuation protocols or AE management within the CLEAR-B arm, may have influenced the higher discontinuation rates observed for ribociclib.

We recognized several limitations in our IRINA study:

1. Differences between clinical trial data and real-world data – such as variations in patient monitoring, drug handling, and missing data in health records – may introduce bias. Despite using multiple imputation and advanced statistical methods to address missing data, some residual bias remains possible.

2. As no data on adverse effects were available in the real-world data, a detailed safety comparison was not possible. Consequently, we could not assess the frequency of clinically relevant adverse effects such as neutropenia and liver enzyme elevations.

3. Furthermore, the NATALEE arm was globally recruited, while the CLEAR-B arm included only patients treated in Germany. This geographic discrepancy may have introduced unmeasured confounders, including differences in healthcare systems, access to care, and patient management.

4. Additionally, subgroup analyses in indirect comparisons with an external control arm have inherent limitations, such as potential heterogeneity between data sources, risk of bias due to unequal group compositions, and limited statistical power. As a result, the results of subgroup analyses in this study were heterogeneous and did not show clear signs of effect modification.

Conclusion

In this ITC, ribociclib + NSAI + OFS demonstrated consistent advantages over tamoxifen ± OFS across all effectiveness endpoints, including overall survival, in premenopausal women with HR-positive, HER2-negative early BC with high risk of recurrence. Safety-related analyses confirmed generally comparable tolerability for both regimens. These findings support ribociclib + NSAI + OFS as a promising and well-tolerated treatment option for this patient population.

Ethics

The manuscript describes analyses using patient-level data from the NATALEE trial (NCT03701334), which has been previously published. This manuscript also uses data from the CLEAR-B study (NCT05870813). This IRINA study was approved by the institutional review board of each participating institution.

Informed Consent

Written informed consent was obtained from all individual participants included in the NATALEE trial and the CLEAR-B study. The data has been shared with the involved institute conducting the analyses for this study in a fully anonymized form.

Supplementary Material

• Supplementary Table S1. Weighted patient characteristics after propensity score weighting.

• Supplementary Table S2. Subgroup analysis by endpoints.

• Supplementary Fig. S1. Flow chart for the adjustment of confounders.

• Supplementary Fig. S2. Whisker plots for landmark sensitivity analysis of iDFS (weighed by SMRW).

• Supplementary Fig. S3. Whisker plots for landmark sensitivity analysis of dDFS (weighed by SMRW).

• Supplementary Fig. S4. Whisker plots for landmark sensitivity analysis of RFS (weighed by SMRW).

• Supplementary Fig. S5. Whisker plots for landmark sensitivity analysis of OS (weighed by SMRW).

• References

• 1 Ye H, Lin G, Wang X. A narrative review: research progress of adjuvant intensive endocrine therapy for early breast cancer. Transl Breast Cancer Res 2024; 5: 20
• 2 Loibl S, André F, Bachelot T. et al. Early breast cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol 2024; 35: 159-182
• 3 Pan H, Gray R, Braybrooke J. et al. 20-Year Risks of Breast-Cancer Recurrence after Stopping Endocrine Therapy at 5 Years. N Engl J Med 2017; 377: 1836-1846
• 4 Pistilli B, Lohrisch C, Sheade J. et al. Personalizing Adjuvant Endocrine Therapy for Early-Stage Hormone Receptor-Positive Breast Cancer. Am Soc Clin Oncol Educ Book 2022; 42: 1-13
• 5 Fasching P, Aktats B, Hübner H. CLEAR-B – Adjuvant therapy effectiveness in patients with primary breast cancer: A retrospective registry study (P4–11–12). San Antonio, TX, USA: San Antonio Breast Cancer Symposium; 2024
• 6 Hammond MEH, Hayes DF, Dowsett M. et al. American Society of Clinical Oncology/College Of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Clin Oncol 2010; 28: 2784-2795
• 7 AWMF. Interdisziplinäre S3-Leitlinie für die Früherkennung, Diagnostik, Therapie und Nachsorge des Mammakarzinoms. 2021 Accessed February 17, 2025 at: https://www.leitlinienprogramm-onkologie.de/fileadmin/user_upload/Downloads/Leitlinien/Mammakarzinom_4_0/Version_4.4/LL_Mammakarzinom_Langversion_4.4.pdf
• 8 Schneeweiss A, Bauerfeind I, Fehm T. et al. Therapy Algorithms for the Diagnosis and Treatment of Patients with Early and Advanced Breast Cancer. Breast Care (Basel) 2020; 15: 608-618
• 9 Hunter RJ, Park J, Asprer KJ. et al. Updated Review Article: Cyclin-Dependent Kinase 4/6 Inhibitor Impact, FDA Approval, and Resistance Pathways. J Pharm Technol 2023; 39: 298-308
• 10 Turner NC, Ro J, André F. et al. Palbociclib in Hormone-Receptor-Positive Advanced Breast Cancer. N Engl J Med 2015; 373: 209-219
• 11 Loibl S, Turner NC, Ro J. et al. Palbociclib Combined with Fulvestrant in Premenopausal Women with Advanced Breast Cancer and Prior Progression on Endocrine Therapy: PALOMA-3 Results. Oncologist 2017; 22: 1028-1038
• 12 Goetz MP, Toi M, Campone M. et al. MONARCH 3: Abemaciclib As Initial Therapy for Advanced Breast Cancer. J Clin Oncol 2017; 35: 3638-3646
• 13 Hortobagyi GN, Stemmer SM, Burris HA. et al. Ribociclib as First-Line Therapy for HR-Positive, Advanced Breast Cancer. N Engl J Med 2016; 375: 1738-1748
• 14 Slamon DJ, Neven P, Chia S. et al. Phase III Randomized Study of Ribociclib and Fulvestrant in Hormone Receptor-Positive, Human Epidermal Growth Factor Receptor 2-Negative Advanced Breast Cancer: MONALEESA-3. J Clin Oncol 2018; 36: 2465-2472
• 15 Finn RS, Martin M, Rugo HS. et al. Palbociclib and Letrozole in Advanced Breast Cancer. N Engl J Med 2016; 375: 1925-1936
• 16 O’Keefe K, Desai NV, Tan AR. Practical Guidance on Abemaciclib in Combination with Adjuvant Endocrine Therapy for Treating Hormone Receptor-Positive, Human Epidermal Growth Factor Receptor 2-Negative High-Risk Early Breast Cancer. Breast Cancer (Dove Med Press) 2024; 16: 517-527
• 17 Johnston SRD, Toi M, O’Shaughnessy J. et al. Abemaciclib plus endocrine therapy for hormone receptor-positive, HER2-negative, node-positive, high-risk early breast cancer (monarchE): results from a preplanned interim analysis of a randomised, open-label, phase 3 trial. Lancet Oncol 2023; 24: 77-90
• 18 Parati MC, Pedersini R, Perego G. et al. Ribociclib in the Treatment of Hormone-Receptor Positive/HER2-Negative Advanced and Early Breast Cancer: Overview of Clinical Data and Patients Selection. Breast Cancer (Dove Med Press) 2022; 14: 101-111
• 19 Slamon D, Lipatov O, Nowecki Z. et al. Ribociclib plus Endocrine Therapy in Early Breast Cancer. N Engl J Med 2024; 390: 1080-1091
• 20 Stanciu I-M, Parosanu AI, Nitipir C. An Overview of the Safety Profile and Clinical Impact of CDK4/6 Inhibitors in Breast Cancer-A Systematic Review of Randomized Phase II and III Clinical Trials. Biomolecules 2023; 13: 1422
• 21 ClinicalTrials.gov. A Trial to Evaluate Efficacy and Safety of Ribociclib With Endocrine Therapy as Adjuvant Treatment in Patients With HR+/HER2− Early Breast Cancer (NATALEE). Accessed November 06, 2024 at: https://clinicaltrials.gov/study/NCT03701334%23more-information
• 22 Institut für Frauengesundheit. Cancer Landscape – Early Adjuvant Retrospective Registry – Breast Cancer (CLEAR-B). Accessed November 28, 2024 at: https://clear-b.de/%23about
• 23 Desai RJ, Franklin JM. Alternative approaches for confounding adjustment in observational studies using weighting based on the propensity score: a primer for practitioners. BMJ 2019; 367: l5657
• 24 Schaafsma E, Zhang B, Schaafsma M. et al. Impact of Oncotype DX testing on ER+ breast cancer treatment and survival in the first decade of use. Breast Cancer Res 2021; 23: 74
• 25 Ohnstad HO, Borgen E, Falk RS. et al. Prognostic value of PAM50 and risk of recurrence score in patients with early-stage breast cancer with long-term follow-up. Breast Cancer Res 2017; 19: 120
• 26 Xu Y, Qi Y, Lu Z. et al. Navigating precision: the crucial role of next-generation sequencing recurrence risk assessment in tailoring adjuvant therapy for hormone receptor-positive, human epidermal growth factor Receptor2-negative early breast cancer. Cancer Biol Ther 2024; 25: 2405060
• 27 Jung W, Kim K, Moon B-I. Treatment Outcomes according to the EndoPredict Score in ER-Positive, HER2-Negative Early Breast Cancer. Breast Care (Basel) 2022; 17: 561-566
• 28 ECOG Cancer Research Group. ECOG Performance Status Scale. Accessed November 13, 2024 at: https://ecog-acrin.org/resources/ecog-performance-status/%23
• 29 Hudis CA, Barlow WE, Costantino JP. et al. Proposal for standardized definitions for efficacy end points in adjuvant breast cancer trials: the STEEP system. J Clin Oncol 2007; 25: 2127-2132
• 30 Leyrat C, Seaman SR, White IR. et al. Propensity score analysis with partially observed covariates: How should multiple imputation be used?. Stat Methods Med Res 2019; 28: 3-19
• 31 Morgan CJ. Landmark analysis: A primer. J Nucl Cardiol 2019; 26: 391-393
• 32 Park-Simon T-W, Müller V, Jackisch C. et al. Arbeitsgemeinschaft Gynäkologische Onkologie Recommendations for the Diagnosis and Treatment of Patients with Early Breast Cancer: Update 2023. Breast Care (Basel) 2023; 18: 289-305
• 33 Nabieva N, Altmann F, Apel K. et al. The Endocrine Treatment Landscape for Patients with HR+ HER2− Early-stage Breast Cancer in Germany Before the Introduction of CDK4/6 Inhibitor Therapy – A Real-World Analysis. Geburtshilfe Frauenheilkd 2023; 83: 1127-1137
• 34 Ma L, Yang B, Wu J. Revisiting ovarian function suppression with GnRH agonists for premenopausal women with breast cancer: Who should use and the impact on survival outcomes. Cancer Treat Rev 2024; 129: 102770
• 35 Mueller V, Aktas B, Huebner H. et al. Adjuvant therapy effectiveness in patients with primary breast cancer: A retrospective registry study to evaluate endocrine therapy strategies and outcomes for premenopausal primary breast cancer patients at intermediate and high risk of recurrence. 2024 Accessed April 02, 2025 at: https://clear-b.de/downloads/sabcs-poster.pdf
• 36 Francis PA, Pagani O, Fleming GF. et al. Tailoring Adjuvant Endocrine Therapy for Premenopausal Breast Cancer. N Engl J Med 2018; 379: 122-137
• 37 EBCTCG. Aromatase inhibitors versus tamoxifen in premenopausal women with oestrogen receptor-positive early-stage breast cancer treated with ovarian suppression: a patient-level meta-analysis of 7030 women from four randomised trials. Lancet Oncol 2022; 23: 382-392
• 38 Gluz O, Christgen M, Nitz U. et al. Impact of age and ovarian function suppression (OFS) on endocrine response to short preoperative endocrine therapy (ET): Results from the multicenter ADAPTcycle trial (n = 4,334). Cancer Research 2024; 84: LBO1–05
• 39 Law E, Gavanji R, Walsh S. et al. Palbociclib versus abemaciclib in HR+/HER2- advanced breast cancer: an indirect comparison of patient-reported end points. J Comp Eff Res 2022; 11: 109-120
• 40 Zhao JJ, Fong KY, Chan YH. et al. Indirect Treatment Comparison of First-Line CDK4/6-Inhibitors in Post-Menopausal Patients with HR+/HER2− Metastatic Breast Cancer. Cancers (Basel) 2023; 15: 4558
• 41 Chang M, Balser H, Roach JM. et al. Innovative strategies, statistical solutions and simulations for modern clinical trials. (Chapman & Hall/CRC Biostatistics Series; ). Boca Raton, London, New York: CRC Press Taylor & Francis Group; 2019