Prediction of dose reductions and impact of reduced starting doses of CDK4/6i on effectiveness in HR‐positive/HER2‐negative metastatic breast cancer: Real‐world data from the OPAL registry

Patrick Marschner; Kai Ringwald; Marc Thill; Mark‐Oliver Zahn; Anja Welt; Arnd Nusch; Gabriele Kaltenecker; Yolanda Rodemer; Volker Hagen; Caroline Schock; Lisa Kruggel; Anja Kaiser‐Osterhues; Dunja Klein; Nina Haug; Elmar Stickeler; Nadia Harbeck; Achim Wöckel; Norbert Marschner; Thomas Decker; the OPAL Registry Group

doi:10.1002/ijc.70433

. 2026 Mar 18;159(1):244–256. doi: 10.1002/ijc.70433

Prediction of dose reductions and impact of reduced starting doses of CDK4/6i on effectiveness in HR‐positive/HER2‐negative metastatic breast cancer: Real‐world data from the OPAL registry

Patrick Marschner ^1,^✉, Kai Ringwald ², Marc Thill ³, Mark‐Oliver Zahn ⁴, Anja Welt ⁵, Arnd Nusch ⁶, Gabriele Kaltenecker ⁷, Yolanda Rodemer ⁸, Volker Hagen ⁹, Caroline Schock ¹⁰, Lisa Kruggel ¹¹, Anja Kaiser‐Osterhues ¹², Dunja Klein ¹², Nina Haug ², Elmar Stickeler ^13,¹⁴, Nadia Harbeck ¹⁵, Achim Wöckel ¹⁶, Norbert Marschner ^17,¹⁸, Thomas Decker ¹⁹; the OPAL Registry Group

PMCID: PMC13139999 PMID: 41851017

Abstract

Cancer patients prone to toxicities might benefit from dose reduction over fixed‐dose recommendations. We develop a predictive index to identify patients with increased probability of dose reduction, intolerable toxicities, or therapy discontinuation (hereafter: dose reduction) in metastatic breast cancer (MBC) and compare real‐world effectiveness of reduced (RSD) versus full starting dose (FSD) using this index. This analysis included 618 patients with HR‐positive, HER2‐negative MBC from the prospective, observational, multicenter registry OPAL (NCT03417115), receiving first‐line palbociclib (n = 386) or ribociclib (n = 232) plus endocrine therapy. A logistic regression model was employed to derive the predictive index. Inverse probability of treatment weighting was used to emulate a head‐to‐head comparison of RSD and FSD by analyzing progression‐free (PFS) and overall survival (OS). Within 6 months, 215 patients (35%) underwent dose reduction, including 109 (51%) with RSD. Predictors for dose reduction were age ≥65 years and Charlson comorbidity index (CCI) ≥1. Among patients with increased probability of dose reduction (index ≥1: ≥65 years and/or CCI ≥ 1), median PFS and OS were 30.1 [21.7, 54.0] and 57.6 [40.0, NA] months with RSD vs. 29.3 [24.9, 32.0] and 43.1 [38.8, 50.3] months with FSD. For low‐probability patients (index = 0: <65 years and CCI = 0), median PFS and OS were 17.6 [9.1, 29.8] and 32.9 [23.1, 40.9] months with RSD vs. 24.5 [19.8, 32.0] and 54.2 [48.8, NA] months with FSD. In this real‐world MBC setting, patients ≥65 years and/or with CCI ≥ 1 had an increased probability of dose reduction and may benefit from RSD, as this yielded outcomes comparable to FSD. Younger, fitter patients may require full dosing. Future studies, ideally randomized controlled trials, should aim to confirm these findings.

Keywords: breast neoplasms, dose reduction, palbociclib, predictive index, registries, ribociclib

What's New?

Balancing efficacy and tolerability is challenging in oncology, as standard dosing often prioritizes maximum tolerated doses over other patient needs. Here, the authors developed a predictive index to identify patients with increased probability of dose reduction in metastatic breast cancer and compared the real‐world effectiveness of reduced versus full starting doses. The results show that patients aged ≥65 and/or with Charlson comorbidity index ≥1 have an increased probability of dose reduction/toxicity and discontinuation. In these patients, reduced and full starting doses of palbociclib or ribociclib achieved comparable progression‐free and overall survival, while the younger, fitter patients benefit from full starting doses.

graphic file with name IJC-159-244-g001.jpg

Abbreviations

ABC: advanced breast cancer
BMI: body mass index
CCI: Charlson comorbidity index
CDKI: cyclin‐dependent kinase 4/6 inhibitor
CI: confidence interval
CTA: chemotherapy agents
DFI: disease‐free interval
ECOG: Eastern Cooperative Oncology Group
ET: endocrine therapy
FDA: US Food and Drug Administration
FSD: full starting dose
HER2: human epidermal growth factor receptor 2
HR: hazard ratio
HR‐positive: hormone receptor‐positive
IPCW: inverse probability of censoring weighting
IPTW: inverse probability of treatment weighting
MBC: metastatic breast cancer
MTD: maximum tolerated dose
OR: odds ratio
OS: overall survival
PALBO: palbociclib
PFS: progression‐free survival
PPV: positive predictive value
QoL: quality of life
RCT: randomized controlled trial
RIBO: ribociclib
RSD: reduced starting dose
SMD: standardized mean difference
SmPC: summary of product characteristics

1. INTRODUCTION

Despite the ongoing advancement of oncology drug development, the design of phase I clinical trials and dose selection remain predominantly based on the maximum tolerated dose (MTD) approach, which was historically developed for cytotoxic chemotherapy agents (CTAs). ¹ In contrast to CTAs, molecularly targeted drugs, such as kinase inhibitors and antibodies, are more selective, exhibit a broader therapeutic window, and may achieve maximal modulation of tumor growth at doses lower than the MTD. ² Consequently, the dose‐finding paradigm with its “more is better” approach may result in the recommendation of an unnecessarily high dose which has led to the initiation of several post‐marketing requirement studies for dose optimization. ² In 2021, the US Food and Drug Administration (FDA) launched the Project Optimus with the objective of reforming dose optimization in oncology drug development. ³ A guidance for industry has recently been released. ⁴ In routine clinical practice, physicians face the challenge of identifying the optimal treatment and appropriate (starting) dose, especially for elderly patients and those with comorbidities underrepresented in clinical trials. Minimizing therapy‐limiting toxicities is crucial to ensure that effective therapies are not discontinued prematurely, which may negatively impact patients' outcomes.

The current standard of care in the first‐line treatment of HR‐positive, HER2‐negative advanced breast cancer (ABC) is the combination of endocrine therapy (ET) with cyclin‐dependent kinase 4/6 inhibitors (CDKIs), that is, palbociclib, ribociclib, and abemaciclib. Although head‐to‐head phase III randomized controlled trials (RCTs) comparing the three approved CDKIs are lacking, real‐world data suggest that these drugs have comparable effectiveness. ⁵ , ⁶

According to the Summary of Product Characteristics (SmPC), palbociclib and ribociclib were approved as recommended full starting doses (FSD) of 125 and 600 mg per day, respectively, for this indication. ⁷ , ⁸ Dose reductions generally occur due to adverse events, especially hematological or gastrointestinal toxicities. ⁹ Reduced starting doses (RSD) for patients at risk of toxicities may improve quality of life (QoL) ¹⁰ while maintaining efficacy, for example by decreasing treatment discontinuations due to toxicity. Although research has aimed to determine predictive factors for dose reduction in patients treated with CDKIs and ET, ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ there are currently no prospectively validated risk stratification models to identify patients who may benefit from an RSD.

Data from clinical and retrospective studies suggest that patients undergoing dose reductions of palbociclib and ribociclib may demonstrate equal or even superior efficacy compared to those without dose reductions. ¹¹ , ¹² , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ However, evaluating the impact of dose reduction can be prone to immortal time bias, ²⁶ which most studies fail to account for. Patients with a FSD who decease early in the study are less likely to receive a reduced dose, which hinders the assessment of whether efficacy truly differs between full and reduced doses. Moreover, patient and disease characteristics affecting dose reductions post‐baseline may also causally affect outcomes, thereby leading to time‐varying confounding if these variables are not adjusted for. Additionally, as previously mentioned, dose reductions are usually triggered by adverse events. Therefore, the hypothetical outcome of patients, had they maintained full dosage irrespective of these adverse events, may not be a meaningful estimate—it would be neither practical nor ethical to enforce such treatment strategy. We therefore resort to the comparison of the effects of a FSD with an RSD, where both compared treatment strategies do not preclude post‐baseline dose adjustments.

This real‐world analysis emulated a hypothetical randomized trial ²⁷ on data from the prospective clinical cohort study OPAL, with palbociclib and ribociclib as the most frequently used CDKIs. We identified patient‐related factors significantly associated with dose reduction or toxicity discontinuation within the first 6 months of CDKI treatment and subsequently developed a predictive index. Stratified by this index, we compared progression‐free survival (PFS) and overall survival (OS) between patients with RSD versus FSD. We hypothesized that patients with an increased probability of dose reduction/toxicity discontinuation would have comparable OS and PFS when treated with RSD versus full dosing.

2. PATIENTS AND METHODS

2.1. Study design and cohort definition

The clinical registry OPAL is an ongoing, open, prospective observational, longitudinal multicenter cohort study on patients with early and advanced (locally advanced and/or metastatic) breast cancer in Germany. Eligible patients are ≥18 years of age with documented diagnosis of early or advanced breast cancer at start of first systemic treatment. Inclusion is possible until up to 6 weeks after start of treatment. There are no restrictions on the type of systemic treatment. Sites participating in OPAL include outpatient practices and clinics as well as hospitals. Details on all (sequential) treatments, patient and tumor characteristics, biomarker testing, clinical and patient‐reported outcomes are collected. Patients are followed up until death or for a maximum of 5 years. Further details on the methodology of OPAL have been published previously. ⁶ , ²⁸

In the present work, we analyzed data on all female patients with HR‐positive, HER2‐negative metastatic breast cancer (MBC) who received ET and either palbociclib or ribociclib as palliative first‐line treatment. Patients with switch of CDKI during first‐line treatment, or those who received another CDKI or chemotherapy agents in this line, were excluded from the present analysis. Switch of ET was allowed. Data base cut was on August 31, 2024.

2.2. Derivation of a predictive index for dose reduction and toxicity discontinuation

Dose reductions and treatment discontinuation due to toxicity in the real‐world setting indicate that the full dose may be too high for the respective patient. Reduced dose applications were defined as <125 mg for palbociclib (SmPC dose reduction levels of 100 and 75 mg) and <600 mg for ribociclib (SmPC dose reduction levels of 400 and 200 mg). As the probability of receiving a reduction depends on the treatment duration (i.e., therapies that last only a few cycles can only be reduced during the course of these few cycles, while treatments that take years can also be reduced at a later point), a comparison between patients with and without reduction may potentially reveal factors that contribute to prolonged therapies rather than those that lead to a reduction (i.e., immortal time bias). ²⁶ This bias can be minimized by focusing solely on the early phase of treatment, such as the first 6 months, when most primary dose reductions occurred in our data. Therefore, the objective of this analysis was to compare patients who (1) started at RSD or (2) discontinued treatment due to toxicity and/or underwent an on‐treatment dose reduction within the first 6 months with patients without a dose reduction to identify predictive factors for dose reduction and toxicity discontinuation. For a better readability, the term “RSD/on‐treatment dose reduction/toxicity discontinuation” is hereafter referred to as “dose reduction.”

However, patients who did not remain under observation or discontinued their treatment prior to 6 months (e.g., due to progression or death) without an (observed) prior dose reduction cannot be unambiguously assigned to one of the two groups, as they could have experienced a reduction or toxicity within the first 6 months if they had remained under observation and had not discontinued their treatment. To deal with such intercurrent events, ²⁹ we used a “hypothetical strategy” and estimate the probability of dose reduction had patients remained on treatment for at least 6 months, in other words, had all intercurrent events been eliminated. To this end, we censored patients who experienced intercurrent events and performed our analysis on the uncensored patients, weighted according to their inverse probability of being censored (i.e., we use inverse probability of censoring weighting, IPCW). This analysis thus assigns greater weight to those unambiguously assignable patients (non‐censored patients) whose characteristics are more similar to those patients who cannot be assigned to a group (censored patients). More specifically, non‐censored patients were assigned, stabilized propensity score weights, calculated using a generalized linear model ³⁰ , ³¹ containing the following baseline variables as potentially predictive of outcome: age, metastases, Charlson comorbidity index (CCI), Eastern Cooperative Oncology Group (ECOG) performance status, disease‐free interval (DFI), prior ET, number of metastases at start of treatment, estrogen receptor expression.

Additionally, a sensitivity analysis using the “while on treatment strategy” (i.e., treating non‐assignable patients as those without dose reduction) was performed to address the intercurrent event (for more information on the rationale behind this decision and the description of the methodology, please refer to Methods section of Data S1, Supporting Information).

Following the calculation of the IPCW weights, the non‐censored patients were divided into a derivation set (comprising approximately 70% of the weights) and a validation set (comprising approximately 30% of the weights) to later validate the identified predictive factors in an independent subsample of the OPAL cohort. This splitting was stratified by “reduction/toxicity within the first 6 months” (yes/no) in order to achieve a more accurate prediction assessment in the validation set. For the sensitivity analysis, no splitting was conducted.

The predictive index for dose reduction was derived based on the derivation set using backward elimination from a weighted logistic regression model. In this model, age, weight, BSA, BMI, metastases, chronic pulmonary disease, liver disease, diabetes, CCI, and ECOG served as predictor variables (Figure S2). Variable selection was based on the Akaike information criterion. ³² All variables in the final model with confidence intervals (CI) of their reduction odds ratio (OR) that do not include the value of “one” were identified as factors for an increased probability of dose reduction and incorporated into the predictive index. For the sensitivity analysis, backward elimination logistic regression was also used, though unweighted and without data splitting (and therefore without independent sample validation).

To validate our predictive index in an independent sample, the weighted positive predictive value (i.e., the probability of being reduced if our predictive index is above a certain threshold) was calculated in both the derivation and the validation sets (see Methods section of Data S1 for details).

2.3. Progression‐free and overall survival of patients with a full versus a reduced starting dose

The objective of this primary analysis was to compare PFS and OS between patients with FSD and RSD according to the SmPC (i.e., <125 and <600 mg for palbociclib and ribociclib, respectively). We were particularly interested in patients with an increased probability of dose reduction, since we hypothesized that RSD would not compromise outcome in these patients.

Inverse probability of treatment weighting (IPTW) using stabilized propensity score weights was used to adjust for confounding between patients with RSD and FSD. ³⁰ , ³¹ For the calculation of the weights, the same variables were considered as in the IPCW analysis (see Methods section of Data S1). SMDs greater than 0.1 were used as indicators for imbalance. ³³ The sample stratified by patients' predictive index for dose reduction (i.e., index = 0 vs. index ≥1) were analyzed. The stabilized propensity score weights were calculated for each subgroup separately to allow for a better balance in covariates.

PFS and OS were estimated using the Kaplan–Meier method. ³⁴ PFS was defined as the interval between start of first‐line treatment and date of progression or death. Patients without tumor progression or death were censored at their date of last contact. OS was defined as the interval between start of first‐line treatment and date of death from any cause. Patients without documented death were censored at their date of last contact.

The main components of the hypothetical target trial emulation (i.e., direct head‐to‐head comparison or RCT) ²⁷ were listed and are summarized in Table S1.

2.4. Statistical analysis

All statistical analyses and data visualization were performed using R version 4.3.2 (2023‐10‐31) “Eye Holes” (Platform: x86_64‐pc‐linux‐gnu [64‐bit]). Propensity‐score weighting was performed using the WeightIt package (0.14.2), survival analyses were performed using the survival (3.4‐0) and survminer (0.4.9) packages.

3. RESULTS

3.1. Cohort description and patient characteristics

At data base cut on August 31, 2024, a total of 2086 patients with advanced breast cancer had been recruited into OPAL by 189 sites all over Germany (Figure 1). Starting from 955 patients with HR‐positive, HER2‐negative metastatic breast cancer, the present analysis is based on 618 patients who received either palbociclib plus ET (n = 386, 62%) or ribociclib plus ET (n = 232, 38%) as palliative first‐line treatment. A total of 215 patients (35%) underwent a dose reduction within the first 6 months of treatment including 109 patients (51%, i.e., 18% of the total population) who received an RSD. In contrast, 403 patients (65%) received the full dose without any documented reduction (Figure 1). Of those patients with a documented dose reduction, 14 patients underwent a subsequent increase in dosage within the initial 6 months of treatment. However, all following analyses revealed similar results when patients were treated as full‐dose patients.

Flowchart. Patient flowchart of all patients included in this analysis, starting from the total number of patients with advanced breast cancer recruited into the OPAL registry from December 2017 until August 2024. The analysis cohort consists of those patients with HR‐positive, HER2‐negative metastatic breast cancer who received either palbociclib or ribociclib as first‐line treatment (n = 618). CDKI, cyclin‐dependent kinase 4/6 inhibitor; ET, endocrine therapy; HER2, human epidermal growth factor receptor 2; HR, hormone receptor; in/ex criteria, inclusion/exclusion criteria.

Unadjusted baseline characteristics of patients are presented in Table 1. Compared to patients without dose reduction, patients who received a dose reduction were older (70 vs. 66 years in median), less often had a good to moderate performance status (ECOG 0/1; 69%, n = 148 vs. 76%, n = 308), and presented with comorbidities more often (CCI ≥1; 30%, n = 64 vs. 19%, n = 77).

TABLE 1.

Baseline characteristics (unadjusted).

Characteristic at start of first‐line treatment	Dose reduction within 6 months			No documented dose reduction within 6 months
Characteristic at start of first‐line treatment	Reduced starting dose n = 109	Full starting dose (reduced later) n = 106	Total reduced n = 215	n = 403
Age in years, median (min–max)	70.0 (36.1–88.6)	70.3 (43.3–86.4)	70.2 (36.1–88.6)	65.7 (34.8–89.7)
<65 years	43 (39.4%)	36 (34.0%)	79 (36.7%)	197 (48.9%)
≥65 years	66 (60.6%)	70 (66.0%)	136 (63.3%)	206 (51.1%)
Weight in kg, median (min–max)	70.0 (38.0–150.0)	70.0 (47.0–127.0)	70.0 (38.0–150.0)	71.0 (40.0–170.0)
<60 kg	15 (13.8%)	22 (20.8%)	37 (17.2%)	61 (15.1%)
≥60 kg	78 (71.6%)	75 (70.8%)	153 (71.2%)	298 (73.9%)
Missing	16 (14.7%)	9 (8.5%)	25 (11.6%)	44 (10.9%)
Body surface area in m²
<1.6	11 (10.1%)	15 (14.2%)	26 (12.1%)	41 (10.2%)
≥1.6	77 (70.6%)	78 (73.6%)	155 (72.1%)	300 (74.4%)
Missing	21 (19.3%)	13 (12.3%)	34 (15.8%)	62 (15.4%)
Body mass index in kg/m²
<25	35 (32.1%)	39 (36.8%)	74 (34.4%)	136 (33.7%)
≥25	53 (48.6%)	54 (50.9%)	107 (49.8%)	205 (50.9%)
Missing	21 (19.3%)	13 (12.3%)	34 (15.8%)	62 (15.4%)
ECOG performance status
0/1	70 (64.2%)	78 (73.6%)	148 (68.8%)	308 (76.4%)
≥2	18 (16.5%)	13 (12.3%)	31 (14.4%)	37 (9.2%)
Missing	21 (19.3%)	15 (14.2%)	36 (16.7%)	58 (14.4%)
Patients with comorbidity
CCI 0 ^a	75 (68.8%)	76 (71.7%)	151 (70.2%)	326 (80.9%)
CCI ≥ 1 ^a	34 (31.2%)	30 (28.3%)	64 (29.8%)	77 (19.1%)
Diabetes	14 (12.8%)	18 (17.0%)	32 (14.9%)	50 (12.4%)
Chronic pulmonary disease	7 (6.4%)	6 (5.7%)	13 (6.0%)	21 (5.2%)
Liver disease	3 (2.8%)	1 (0.9%)	4 (1.9%)	6 (1.5%)
Type of metastases
Bone only	27 (24.8%)	37 (34.9%)	64 (29.8%)	112 (27.8%)
Non‐visceral ^b and non‐bone only	12 (11.0%)	13 (12.3%)	25 (11.6%)	42 (10.4%)
Visceral ± non‐visceral	68 (62.4%)	55 (51.9%)	123 (57.2%)	237 (58.8%)
Unknown	2 (1.8%)	1 (0.9%)	3 (1.4%)	12 (3.0%)

Open in a new tab

Note: Some percentages might not add up to 100% due to rounding.

Abbreviations: CCI, Charlson comorbidity index; ECOG, Eastern Cooperative Oncology Group.

^{^a}

Charlson comorbidity index (CCI) according to Quan et al. ⁴³

^{^b}

Non‐visceral disease includes bone, soft tissue, skin, and lymph node metastases; all other metastases are visceral.

3.2. Predictors for dose reduction: Predictive index

3.2.1. Propensity score weighting

Of the 618 patients included, 70 were censored prior to 6 months without prior dose reduction. Consequently, IPCW was used to reduce selection bias (see section 2 and Methods section of Data S1). SMDs before and after weighting demonstrate an acceptable balance for the total population (Figure S1).

3.2.2. Identified predictive factors and indexing

In the derivation sample (n = 385), the variables ≥65 years and CCI ≥1 were associated with an increased probability of dose reduction with an OR of 1.82 [1.19, 2.79] for age ≥65 years and an OR of 2.12 [1.29, 3.49] for CCI ≥1 (Figure 2). Square brackets denote 95% CIs. See Figure S2 for the full model including all potentially predictive factors.

Forest plot of predictive factors (derivation sample). Predictors that lead to a dose reduction within the first 6 months of treatment, calculated in the weighted derivation sample (n = 385). CCI, Charlson comorbidity index; CI, confidence interval.

Based on the identified factors for dose reduction (age ≥65 years, CCI ≥1), a predictive index was developed (Figure 3). Patients <65 years with a CCI of 0 were assigned an index of 0 (low probability). Patients ≥65 years or with a CCI ≥1 were classified as having an index of 1 (intermediate probability), while those aged ≥65 years and with a CCI ≥1 were assigned an index of 2 (high probability). Of the 618 patients included in this analysis, 387 (63%) had an index ≥1, indicating an increased probability of dose reduction, while 231 patients (37%) had an index of 0, indicating a low probability of dose reduction. Of those patients with an index ≥1, 20% (n = 76) received an RSD, compared to 14% (n = 33) with an index = 0.

Predictive index. Decision‐making process for determining the predictive index for dose reduction/RSD/toxicity discontinuation. CCI, Charlson comorbidity index; RSD, reduced starting dose.

In the validation sample, the probability of experiencing a dose reduction within the first 6 months was 39.6% in the total population and 52.4% in patients with an index = 2 (Table S2A). A higher proportion of patients with dose reduction was also observed in the subgroup of patients receiving either palbociclib or ribociclib (Table S2B,C), suggesting that our index performs similarly for both drugs. The corresponding confusion matrices are provided in Table S2D–F.

The sensitivity analysis, including patients who had discontinued their treatment prior to 6 months as patients without dose reduction, also revealed the factors CCI and age (Figure S3).

3.3. Progression‐free and overall survival according to probability of dose reduction

3.3.1. Propensity score weighting

Unadjusted baseline characteristics of patients stratified by predictive index (0 or ≥1) and starting dose (RSD or FSD) are presented in Table S3. To reduce selection bias between patients with FSD compared to patients with RSD, IPTW was used (see section 2). The covariate balance of IPTW for the total population and stratified by predictive index (0 or ≥1) is shown in Figure S4A–C. The SMDs before and after propensity score weighting revealed a good balance for the subgroup with predictive index ≥1 (Figure S4B) and an acceptable balance for the subgroup with predictive index = 0 (Figure S4C).

3.3.2. Progression‐free and overall survival of the weighted cohorts

IPTW‐weighted outcome data are presented in Figures 4 and 5. Independent of patients' probability of dose reduction, median PFS was 25.1 [18.8, 30.2] months with RSD and 27.1 [24.1, 30.9] months with FSD (Figure 4A). For patients with an increased probability of dose reduction (index ≥1), median PFS was 30.1 [21.7, 54.0] months with RSD versus 29.3 [24.9, 32.0] months with FSD (Figure 4B). In contrast, patients with low probability of dose reduction (index = 0) had a median PFS of 17.6 [9.1, 29.8] months with RSD compared to 24.5 [19.8, 32.0] months with full dosing (Figure 4C). Median OS for patients independent of the predictive index for dose reduction was 40.9 [32.9, 60.3] months with RSD versus 49.1 [42.4, 53.6] months with FSD (Figure 5A). For patients with an elevated probability of dose reduction (index ≥1), median OS was 57.6 [40.0, NA] months with RSD compared to 43.1 [38.8, 50.3] months with FSD (Figure 5B). For patients at low probability (index = 0), median OS was 32.9 [23.1, 40.9] months with RSD versus 54.2 [48.8, NA] months with full dosing (Figure 5C). Square brackets denote 95% CIs.

Progression‐free survival (adjusted, after IPTW). Progression‐free survival after IPTW in patients with full or reduced starting dose, (A) for the total population independent of the predictive index for dose reduction (including RSD and toxicity discontinuation) and stratified by (B) increased (index ≥1) and (C) low probability of dose reduction (index = 0) according to our newly developed predictive index. Numbers at risk refer to the sum of the weights of the respective patients at risk for a given time point. CI, confidence interval; FSD, full starting dose; IPTW, inverse probability of treatment weighting; RSD, reduced starting dose.

Overall survival (adjusted, after IPTW). Overall survival after IPTW in patients with full or reduced starting dose, (A) for the total population independent of the predictive index for dose reduction (including RSD and toxicity discontinuation) and stratified by (B) increased (index ≥1) and (C) low probability of dose reduction (index = 0) according to our newly developed predictive index. Numbers at risk refer to the sum of the weights of the respective patients at risk for a given time point. CI, confidence interval; FSD, full starting dose; IPTW, inverse probability of treatment weighting; RSD, reduced starting dose.

While the OS and PFS curves were found to be quite similar between RSD and FSD for patients at index ≥1, a difference between RSD and FSD was observed for patients with an index = 0, particularly in the OS curve.

4. DISCUSSION

The MTD paradigm for oncology drugs has been criticized for its “more is better” approach, which can result in recommended doses being unnecessarily high and poorly tolerated, potentially leading to treatment discontinuation and comprising patients' QoL due to side effects. In this real‐world setting of HR‐positive, HER2‐negative MBC, the development of a predictive index to identify patients at increased probability for dose reduction and toxicity discontinuation represents a significant advancement in optimizing treatment. Age ≥65 and presence of comorbidities were identified as independent predictors. For these patients, reducing the starting dose of palbociclib or ribociclib achieved comparable PFS and OS to starting therapy with a full dose and reducing it as necessary during the course of treatment.

Clinical trials indicate no significant loss of efficacy following CDKI dose reduction in ABC, with dose reduction rates and survival times largely comparable with those observed in this real‐world analysis. A pooled analysis of the PALOMA trials revealed that 36.9% of patients required a dose reduction of palbociclib, predominantly during the first 6 months of treatment. ¹³ Two publications concluded that dose adjustments of palbociclib in PALOMA‐2 reduced the severity of hematological adverse events while maintaining efficacy. ¹⁸ , ²¹ According to pooled data from the MONALEESA trials, dose reductions of ribociclib occurred in 41.8% of patients. ²² Remarkably, patients appeared to continue to benefit from treatment, with median PFS remaining similar between 24.8 and 29.6 months, depending on the extent of dose reduction. ¹⁹ , ²² By comparison, in the overall populations, median PFS and OS were 25.3 and 63.9 months for ribociclib in MONALEESA‐2 ³⁵ , ³⁶ and 27.6 and 53.9 months for palbociclib in PALOMA‐2. ³⁷ , ³⁸ In this analysis, 35% of the total patient population underwent a dose reduction of palbociclib or ribociclib within the first 6 months of treatment. Notably, half of these patients initiated treatment on a reduced dose which, per protocol, does not occur in the clinical trial setting.

Compared to patients selected for clinical trials, those treated in routine practice often present with less favorable baseline characteristics, which can lead to differences in clinical outcomes between real‐world and clinical trial settings. Most retrospective analyses on patients treated in routine practice suggest equal or even superior effectiveness with any CDKI dose reduction, ¹¹ , ¹² , ²⁰ , ³⁹ which have been reported to range from 30% to 58% for palbociclib or ribociclib. ¹¹ , ¹² , ²⁰ , ²⁴ , ²⁵ , ³⁹ In a recent large retrospective analysis involving ~1300 patients from the US Flatiron Health Analytic Database, median real‐world PFS was maintained with any palbociclib dose modification (20.1 vs. 19.7 months), while median OS was significantly prolonged in the dose‐reduced group compared to that treated with the FSD (59.0 vs. 49.1 months), potentially due to longer treatment. ²⁵ In the same study, 14.5% of patients initiated palbociclib at a reduced dose, a similar proportion to our findings of 18%. ²⁵

Importantly, interpreting the efficacy results from these clinical trials and observational studies, however, is challenging because they either did not account for post‐baseline confounding or their study design induced immortal time bias.

The phase II AMALEE study compared RSD and FSD of ribociclib (400 vs. 600 mg), effectively eliminating the risk of these sources of bias. ²³ Despite a lower overall response rate as the primary endpoint with the 400 mg dose, median PFS was similar (24.9 months for 400 mg vs. 25.1 months for 600 mg) and thus comparable to our findings (overall median PFS: 25.1 months for RSD vs. 27.1 months for FSD). The authors concluded that the use of dose‐reduced ribociclib is a viable option of alleviating adverse events. ²³ , ⁴⁰ Immortal time bias was explicitly addressed in a Danish real‐world retrospective study involving over 500 patients with HR‐positive, HER2‐negative MBC. ⁹ Results showed that dose reduction of a CDKI (palbociclib, ribociclib, and abemaciclib) within the first 12 weeks of treatment was associated with reduced OS (39.9 vs. 54.3 months). However, the authors only analyzed patients who were treated for more than the 12‐week landmark (81%). ⁹

It is also important to note that most of the aforementioned studies either did not adjust for confounders or only included them as covariates in a multivariate survival analysis. Furthermore, none of these studies examined patients at increased risk of toxicities.

Some studies aimed to identify predictive factors for dose reduction and/or progression in patients with ABC or MBC treated with CDKIs and ET. ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ Despite these efforts, routine clinical practice lacks definitive predictors and validated predictive indices that could assist in determining the optimal starting dosage for the individual patient. Based on the existing literature and medical expertise, the present analysis considered a range of baseline variables as potentially predictive of dose reduction and progression. Notably, age ≥65 years and CCI ≥1 were found to be associated with an increased probability of dose reduction. This is a crucial consideration, since these factors are often overlooked in clinical trials, which typically focus on younger, healthier populations. Our findings are consistent with a recent retrospective multicenter study on HR‐positive, HER2‐negative MBC, which demonstrated that age ≥65 years and a CCI ≥2 were associated with greater dosage reductions of either palbociclib or ribociclib. ⁴¹ In another real‐world setting, age (≥70 years) was also a significant predictor for dose reduction of palbociclib. ¹⁶ In the same study, comorbidities were not identified as predictors, but were the second most common reason (28.6%) for an initial dose reduction after older age (42.8%). ¹⁶ While age (≥60 years) was not correlated with palbociclib dose reductions in a single‐center, retrospective study, older patients with ABC were more likely to start palbociclib at lower doses, ¹⁴ which was also shown in the PALOMAGE study. ⁴² In another retrospective study, patients who underwent dose reduction were significantly older compared to those in the full‐dose group (73 vs. 67 years in median), ⁹ similar to the baseline age of our unadjusted cohorts (70 years with dose reduction vs. 66 years without).

By combining routinely available factors—age ≥65 years and CCI ≥1—our predictive index offers an easy‐to‐use tool for identifying patients not being suitable for treatment with full‐dose palbociclib or ribociclib. In addition to predicting dose reductions, a more accurate prediction of outcomes could also help to avoid unnecessary side effects in a considerable proportion of patients.

In the present analysis, most patients (63%) were classified with an index ≥1 (age ≥65 years and/or CCI ≥1), indicating an increased probability for dose reduction. However, only 20% of these patients started treatment with an RSD. This finding is of particular importance, as increased‐probability patients receiving an RSD showed similar outcomes to those treated with the full dose (median PFS and OS of 30.1 and 57.6 months vs. 29.3 and 43.1 months). Conversely, patients under 65, without comorbidities and a low probability of dose reduction (index = 0) showed inferior outcomes with RSD (median PFS and OS of 17.6 and 32.9 months vs. 24.5 and 54.2 months with FSD). These patients closely resemble the populations included in the dose‐finding RCTs and may be less susceptible to toxicity. Therefore, initiating treatment with the FSD appears to be the most appropriate choice for younger and fitter patients.

5. LIMITATIONS

Strengths of this work include its prospective and longitudinal design, as well as the underlying, large real‐world dataset. A major advantage is the combination of stratification for dose reduction probability with the emulation of a hypothetical target trial, which addresses imbalances between treatment groups and considers immortal time bias. Although the adjustment covariates included were chosen based on the existing literature as well as medical expertise and considered sufficient, there is still a risk that unadjusted sources of confounding could lead to biased estimates. Crucially, while RSD in the group of patients with predictive index ≥1 can be explained by observed patient covariates, initial dose reductions in the group of patients with index = 0 may be caused by unobserved variables, which therefore could not be adjusted for. Hence, the observed superiority of FSD in patients with index = 0 may be due to unobserved confounding. The IPCW balance in developing the predictive index was only acceptable. However, sensitivity analysis using an alternative method for patients discontinuing the study before 6 months without dose reduction or treatment cessation identified the same variables, suggesting the results were not due to poor balance. Additionally, the balance in the effectiveness analysis was suboptimal for the subgroup with an index = 0, but the imbalanced variables mainly involved categories with few patients (e.g., ECOG unknown), making it unlikely that these influenced the significant differences in survival curves. Given the observational design of OPAL, there are no specifications concerning the timing, frequency, or criteria for tumor assessment. Therefore, PFS data should be considered as the best clinical approximation and might not be identical to the PFS determined in clinical trials. Despite the large real‐world cohort analyzed and the robust statistical design, the findings of this study may not be generalizable to other patient populations and other treatments.

6. CONCLUSION

To our knowledge, this is the first real‐world analysis to combine stratification for dose reduction probability with a target trial emulation in patients with HR‐positive, HER2‐negative MBC. Our findings indicate that patients aged ≥65 years and/or with comorbidities are at an increased probability of dose reduction and discontinuation due to toxicity. These patients may benefit from initial CDKI dose reduction, as their clinical outcomes were comparable to those receiving the full dose. Younger, fitter patients under 65 without comorbidities, resembling those included in RCTs, may require the full dose to avoid poorer outcomes. Thus, a more nuanced, individualized approach to dosing warrants consideration in routine clinical practice. The predictive index developed here could serve as a valuable tool to help guide dosing, not only to improve clinical outcomes, but also to enhance patients' QoL by minimizing treatment‐limiting toxicities. Future research should aim to confirm the observed effects of an initial CDKI dose reduction and establish causal inference with greater certainty through an RCT.

AUTHOR CONTRIBUTIONS

Patrick Marschner: Conceptualization; investigation; resources; supervision; writing – review and editing. Kai Ringwald: Conceptualization; data curation; formal analysis; methodology; validation; visualization; writing – review and editing. Marc Thill: Resources; investigation; writing – review and editing. Mark‐Oliver Zahn: Resources; writing – review and editing; investigation. Anja Welt: Resources; investigation; writing – review and editing. Arnd Nusch: Investigation; writing – review and editing; resources. Gabriele Kaltenecker: Investigation; writing – review and editing; resources. Yolanda Rodemer: Investigation; writing – review and editing; resources. Volker Hagen: Investigation; writing – review and editing; resources. Caroline Schock: Resources; investigation; writing – review and editing. Lisa Kruggel: Conceptualization; data curation; project administration; writing – review and editing. Anja Kaiser‐Osterhues: Conceptualization; writing – original draft; visualization. Dunja Klein: Conceptualization; writing – review and editing. Nina Haug: Conceptualization; methodology; writing – review and editing. Elmar Stickeler: Investigation; writing – review and editing; resources. Nadia Harbeck: Investigation; writing – review and editing; resources. Achim Wöckel: Investigation; writing – review and editing; resources. Norbert Marschner: Conceptualization; funding acquisition; investigation; resources; writing – review and editing. Thomas Decker: Investigation; resources; writing – review and editing.

FUNDING INFORMATION

The OPAL‐ABC Registry has received continuous financial support from Roche Pharma AG and Onkovis GmbH and temporary financial support from Daiichi Sankyo Deutschland GmbH, Daiichi Sankyo Europe GmbH, Eisai GmbH, Gilead Sciences GmbH, Lilly Deutschland GmbH, Mundipharma GmbH, Mylan Germany GmbH, and Pfizer GmbH. None of the funders had any role in study design, data collection and analysis, interpretation of results, decision to publish, or preparation of the manuscript.

CONFLICT OF INTEREST STATEMENT

KR, GK, LK, AKO, DK, NiH, and YR declare no conflict of interest concerning the topic of this publication. Outside of the published work, the institutions of MT, MOZ, AnW, AN, PM, GK, YR, VH, CS, ES, NaH, AcW, and TD received remuneration for the documentation of patient data. Furthermore, TD has received remuneration for advisory boards from Novartis. PM: employment at and stocks from iOMEDICO AG; consulting or advisory role with Lilly, Novartis, Stemline‐Menarini, Beigene, AstraZeneca and Otsuka; honoraria from Lilly, abbVie, Beigene, Novartis, BMS, Otsuka, Roche, AstraZeneca and FOMF; travel expenses from Lilly, AOP Health, Beigene and J&J. MT: advisory board: Agendia, Amgen, AstraZeneca, Aurikamed, Becton/Dickinson, ClearCut, Daiichi Sankyo, Eisai, Exact Sciences, Gilead Science, Grünenthal, GSK, Lilly, MSD, Neodynamics, Novartis, Onkowissen, Organon, Pfizer, pfm Medical, Pierre‐Fabre, Roche, Saman Tree, Seagen, Sirius Medical, Sysmex, Titanium Textiles; manuscript support: Amgen, Cairn Surgical ClearCut, Clovis, Lilly, Organon, pfm medical, Roche, Servier Travel expenses: Amgen, Art Tempi, AstraZeneca, Clearcut, Daiichi Sankyo, Eisai, Exact Sciences, Gilead, Hexal, I‐Med‐Institute, Lilly, Menarini Stemline, MSD, Neodynamics, Novartis, Pfizer, pfm Medical, Roche, RTI Surgical, Seagen, ZP Therapeutics; congress support: Amgen, AstraZeneca, Daiichi Sanyko, Gilead, Hexal, Lilly, Menarini Stemline, Neodynamics, Novartis, Pfizer, pfm medical, Pierre Fabre, Roche, Sirius Medical; lecture honoraria: Agendia, Amgen, Art Tempi, AstraZeneca, Eisai, Endomag, Exact Sciences, Gilead Science, GSK, Hexal, I‐Med‐Institute, Jörg Eickeler, Johnson and Johnson, Laborarztpraxis Walther et al., Lilly, Medscape, Menarini Stemline, MSD, Novartis, Onkowissen, Pfizer, pfm medical, Roche, Seagen, Sirius Medical, StreamedUp, Sysmex, Vifor, Viatris, ZP Therapeutics; trial funding: Endomag, Exact Sciences; trial honoraria (institutional): AstraZeneca, Biom'Up, CairnSurgical, Clearcut, Neodynamics, Novartis, pfm medical, Roche, RTI Surgical; memberships in academic institutions: Berufsverband der Frauenärzte e.V, Deutsche Gesellschaft für Ultraschall in der Medizin, Deutsche Gesellschaft für Senologie, Deutsche Krebsgesellschaft, Deutsche Gesellschaft für Gynäkologie und Geburtshilfe (DGGG), Arbeitsgemeinschaft Gynäkologische Onkologie (AGO), Breast Competence Club der European Academy of Senology, Arbeitsgemeinschaft Gynäkologische Endoskopie (AGE), European Society of Gynecological Endoscopy (ESGE), Arbeitsgemeinschaft für ästhetische, plastische und wiederherstellende Operationsverfahren in der Gynäkologie (AWOgyn), Bundesarbeitsgemeinschaft Leitender Ärztinnen und Ärzte in der Frauenheilkunde und Geburtshilfe e.V., European Breast Cancer Research Association of Surgical Trialists (EUBREAST), OncoNet Rhein‐Main e.V., Hessische Krebsgesellschaft, European Society of Medical Oncology (ESMO). MOZ: honoraria: Astra Zeneca; holds shares of: Bayer und Gilead. AnW: travel support: Pfizer and Lilly; member of the steering committee in the Breast cancer and Gynaecological cancer; track of the AIO (Arbeitsgemeinschaft Internistische Onkologie), of the DKG (Deutsche Krebsgesellschaft); member of the Consensus Conference Update S3‐Guideline Breast Cancer of the DKG; participated in advisory boards: MSD, Roche, Novartis, Pfizer, Seagen, Lilly, Menarini Stemline, Pierre Fabre; participated on a data safety monitoring board for iOMEDICO; received honoraria for lectures/presentations: Roche, Eisai, Novartis, Pfizer, Lilly, Iomedico, Interplan, MSD, MCI, RCC. AN: honoraria for studies: Roche, Novartis, AstraZeneca, BeiGene. VH: consulting fees from Novartis, Roche, Pfizer, BMS and Boehringer Ingelheim; honoraria for lectures/presentations from Roche, Amgen, Pfizer, Celgene, BMS, Boehringer Ingelheim, Novartis, AstraZeneca, Lilly, mteacdemy and Gilead; travel expenses from Roche, Pfizer, Janssen Cilag, iOMEDICO, Gilead, Celgene, Daiichi Sankyo and Novartis; stocks from JNJ. CS: Advisory boards: Roche, Novartis, Janssen, Stemline. Speaker engagement fees: BMS, Roche, Pfizer, Beigene, Takeda. ES: consulting fees: AstraZeneca, Novartis, Roche; honoraria for lectures/presentations: AstraZeneca, Novartis, Roche, Onkowissen, Lilly, MSD; travel expenses: Pfizer, Novartis, MSD; participated on a data safety monitoring board or advisory board: iOMEDICO, AstraZeneca, MSD; holds leadership or fiduciary role in other board, society, committee, or advocacy group, paid or unpaid: GBG, AGO, IQTIQ. NaH: consulting fees: Gilead, Novartis, Roche, Seagen; honoraria for lectures/presentations: AstraZeneca, Daiichi‐Sankyo, Gilead, Lilly, MSD, Novartis, Pierre‐Fabre, Pfizer, Roche, Seagen, Viatris, Zuelligpharma; participation on a data safety monitoring board or advisory board: Gilead, Roche, Seagen; leadership or fiduciary role in other board, society, committee, or advocacy group, paid or unpaid: West German Study Group (WSG). AcW: consulting fees: Amgen, AstraZeneca, Celgene, Lilly, Novartis, Pfizer, Roche, MSD, Genomic Health, Organon, Seagen, Exact Sciences, Gilead, Dajiichi Sanko, Stemline; honoraria for lectures/presentations: AstraZeneca, Aurikamed, Celgene, Lilly, Novartis, Pfizer, Roche, MSD, Genomic Health, Seagen, Exact Sciences, Gilead, Dajiichi Sanko, Onkowissen, Stemline; travel expenses: Gilead, Dajiichi Sanko, Pfizer, Stemline. NM: Chief Scientific Officer and shareholder of iOMEDICO AG; consulting or advisory role with Amgen, AstraZeneca, Beigene, BMS, Celgene, Daiichi Sankyo, Deciphera, Eisai, Eusapharm, Gilead, GSK, Janssen, Lilly, Mundipharma, Merck, Mylan, MSD, Novartis, Oncopeptides, Oncovis, Pfizer, Pierre Fabre, Roche, Seagen, Servier and Stemline; honoraria from Amgen, Beigene, Eusapharm, GSK, Lilly, Pierre Fabre, Roche, Seagen, Servier and Stemline; research funding from Amgen, AstraZeneca, Beigene, BMS, Celgene, Daiichi Sankyo, Deciphera, Eisai, Eusapharm, Gilead, GSK, Janssen, Lilly, Mundipharma, Merck, Mylan, MSD, Novartis, Oncopeptides, Oncovis, Pfizer, Pierre Fabre, Roche, Seagen, Servier and Stemline. TD: honoraria for lectures/presentations from Lilly, Novartis and AstraZeneca; payment for expert testimony from Novartis, Pfizer and Lilly; holds leadership or fiduciary role in other board, society, committee, or advocacy group, paid or unpaid for iOMEDICO, AIO.

ETHICS STATEMENT

The clinical registry OPAL which started in 2017 was approved by the responsible ethics committee (Baden‐Württemberg, F‐2017‐086) and is registered at ClinicalTrials.gov (NCT03417115). Written informed consent was obtained from all patients. All experiments comply with the current laws in Germany, where they were performed. All procedures performed in studies involving human participants were in accordance with the ethical standards of the national research committee and with the 1964 Helsinki declaration and its later amendments.

Supporting information

Data S1. Supporting Information.

IJC-159-244-s001.pdf^{(36MB, pdf)}

ACKNOWLEDGMENTS

OPAL‐ABC is designed, managed, and analyzed by iOMEDICO and has received continuous financial support from Roche Pharma AG and Onkovis GmbH and temporary financial support from Daiichi Sankyo Deutschland GmbH, Daiichi Sankyo Europe GmbH, Eisai GmbH, Gilead Sciences GmbH, Lilly Deutschland GmbH, Mundipharma GmbH, Mylan Germany GmbH, and Pfizer GmbH. The OPAL Study Group collaborates with the Arbeitsgemeinschaft Internistische Onkologie in der deutschen Krebsgesellschaft e.V. We thank Dr. Katja Gratzke (iOMEDICO) for her critical comments on the manuscript. During the preparation of this work, the authors used the following AI tools: Scite (a smart citation index, Scite Inc., Henderson, USA) for literature search and Langdock (a model‐agnostic AI platform, Langdock GmbH, Berlin, Germany) in order to improve the style of some text passages due to the authors' non‐native English proficiency. The tool was not used to produce content or analyze data. After using these tools/services, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Marschner P, Ringwald K, Thill M, et al. Prediction of dose reductions and impact of reduced starting doses of CDK4/6i on effectiveness in HR‐positive/HER2‐negative metastatic breast cancer: Real‐world data from the OPAL registry. Int J Cancer. 2026;159(1):244‐256. doi: 10.1002/ijc.70433

The authors confirm that this work has not been published previously. Parts of it have been presented at the annual meeting of the German, Austrian and Swiss Societies for Hematology and Medical Oncology organized by the DGHO German Society for Hematology and Medical Oncology eV, 2024 (oral presentation).

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.

REFERENCES

1. Papachristos A, Patel J, Vasileiou M, Patrinos GP. Dose optimization in oncology drug development: the emerging role of pharmacogenomics, pharmacokinetics, and pharmacodynamics. Cancer. 2023;15:3233. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Soltantabar P, Lon H‐K, Parivar K, Wang DD, Elmeliegy M. Optimizing benefit/risk in oncology: review of post‐marketing dose optimization and reflections on the road ahead. Crit Rev Oncol Hematol. 2023;182:103913. [DOI] [PubMed] [Google Scholar]
3. Liao JJZ, Asatiani E, Liu Q, Hou K. Three steps toward dose optimization for oncology dose finding. Contemp Clin Trials Commun. 2024;40:101329. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. FDA F and DA . Optimizing the dosage of human prescription drugs and biological products for the treatment of oncologic diseases; 2024 [cited 2024 Nov 13]. Available from: https://www.fda.gov/regulatory‐information/search‐fda‐guidance‐documents/optimizing‐dosage‐human‐prescription‐drugs‐and‐biological‐products‐treatment‐oncologic‐diseases
5. Gennari A, André F, Barrios CH, et al. ESMO clinical practice guideline for the diagnosis, staging and treatment of patients with metastatic breast cancer. Ann Oncol. 2021;32:1475‐1495. [DOI] [PubMed] [Google Scholar]
6. Thill M, Zahn M‐O, Welt A, et al. Head to head comparison of palbociclib and ribociclib in first‐line treatment of HR‐positive/HER2‐negative metastatic breast cancer with real‐world data from the OPAL registry. Int J Cancer. 2024;156:1770‐1782. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. European Medicines Agency . Summary of product characteristics‐IBRANCE®, Pfizer Europe MA EEIG, INN‐palbociclib; 2021 [cited 2025 Jan 15]. Available from: https://www.ema.europa.eu/en/documents/product-information/ibrance-epar-product-information_en.pdf
8. European Medicines Agency . Summary of product characteristics‐Kisqali®, Novartis Europharm Limited, INN‐ribociclib; 2022 [cited 2025 Jan 15]. Available from: https://www.ema.europa.eu/en/documents/product-information/kisqali-epar-product-information_en.pdf
9. Bjerrum A, Henriksen AF, Knoop AS, et al. Overall survival after CDK4/6 inhibitor dose reduction in women with metastatic breast cancer. BJC Rep. 2024;2:1‐8. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Loeser A, Kim JS, Peppercorn J, et al. The right dose: results of a patient advocate–led survey of individuals with metastatic breast cancer regarding treatment‐related side effects and views about dosage assessment to optimize quality of life. JCO Oncol Pract. 2024;20:972‐983. [DOI] [PubMed] [Google Scholar]
11. Kristensen KB, Thomsen IMN, Berg T, Kodahl AR, Jensen AB. Dose modifications of ribociclib and endocrine therapy for treatment of ER+ HER2− metastatic breast cancer. Breast Cancer Res Treat. 2021;188:799‐809. [DOI] [PubMed] [Google Scholar]
12. Shikdar S, Hall S, Taylor B, et al. Evaluating the response of palbociclib on progression‐free survival in hormone receptor‐positive metastatic breast cancer. Oncology. 2022;100:337‐343. [DOI] [PubMed] [Google Scholar]
13. Ettl J, Im S‐A, Ro J, et al. Hematologic adverse events following palbociclib dose reduction in patients with hormone receptor‐positive/human epidermal growth factor receptor 2‐negative advanced breast cancer: pooled analysis from randomized phase 2 and 3 studies. Breast Cancer Res. 2020;22:27. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Dennison T, Heiling H, Deal A, et al. Tolerability of palbociclib in younger and older patients with advanced breast cancer. J Oncol Pharm Pract. 2023;29:96‐104. [DOI] [PubMed] [Google Scholar]
15. Ogliari FR, Caspani F, Rossi L, Nerone M, Vallini I, Gueli R. Predictive variables of palbociclib dose reduction and its impact on survival outcomes: a retrospective bicentric study. Breast J. 2021;27:415‐416. [DOI] [PubMed] [Google Scholar]
16. Savignon S, Boyer M, Messier C, Monfette ML, Letarte N. Predictive factors of dose reduction among patients treated with palbociclib for advanced or metastatic breast cancer in a real‐world setting. Poster presented at the Hematology/Oncology Pharmacy Association Annual Conference 2024; April 3–6, 2024; Tampa, Florida. 2024 [cited 2024 May 8]. Available from: https://jhoponline.com/issue‐archive/latest‐issue/19651‐predictive‐factors‐of‐dose‐reduction‐among‐patients‐treated‐with‐palbociclib‐for‐advanced‐or‐metastatic‐breast‐cancer‐in‐a‐real‐world‐setting
17. Vazquez L, Coussirou J, Arvers P, et al. Palbociclib associated with endocrine therapy in patients with metastatic breast cancer: predictive factors of severe early hematological toxicity. Breast Cancer. 2019;108:544‐552. [Google Scholar]
18. Diéras V, Harbeck N, Joy AA, et al. Palbociclib with letrozole in postmenopausal women with ER+/HER2− advanced breast cancer: hematologic safety analysis of the randomized PALOMA‐2 trial. Oncologist. 2019;24:1514‐1525. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Ji Y, Darstein C, Yang S, et al. Quantitative assessment of ribociclib exposure‐response relationship to justify dose regimen in patients with advanced breast cancer. J Clin Pharmacol. 2023;63:1359‐1370. [DOI] [PubMed] [Google Scholar]
20. Fernández‐Cuerva C, Chinchilla‐Alarcón T, Alcaraz‐Sánchez JJ. Real‐world effectiveness of ribociclib in metastatic breast cancer patients: Does dose affect survival? J Oncol Pharm Pract. 2023;29:1619‐1627. [DOI] [PubMed] [Google Scholar]
21. Zheng J, Yu Y, Durairaj C, Diéras V, Finn RS, Wang DD. Impact of dose reduction on efficacy: implications of exposure‐response analysis of palbociclib. Target Oncol. 2021;16:69‐76. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Burris HA, Chan A, Bardia A, et al. Safety and impact of dose reductions on efficacy in the randomised MONALEESA‐2, ‐3 and ‐7 trials in hormone receptor‐positive, HER2‐negative advanced breast cancer. Br J Cancer. 2021;125:679‐686. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Cardoso F, Jacot W, Küemmel S, et al. Abstract PD17‐12: primary efficacy and safety results from the AMALEE trial evaluating 600 mg vs 400 mg starting doses of first‐line ribociclib in patients with HR+/HER2− advanced breast cancer. Cancer Res. 2023;83:PD17‐12. [Google Scholar]
24. Ismail RK, van Breeschoten J, Wouters MWJM, et al. Palbociclib dose reductions and the effect on clinical outcomes in patients with advanced breast cancer. Breast. 2021;60:263‐271. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Layman RM, Liu X, Li B, McRoy L, Brufsky A. Real‐world palbociclib dose modifications and clinical outcomes in patients with HR+/HER2− metastatic breast cancer: a Flatiron Health Database analysis. Breast. 2025;81:104448. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Suissa S. Immortal time bias in pharmaco‐epidemiology. Am J Epidemiol. 2008;167:492‐499. [DOI] [PubMed] [Google Scholar]
27. Hernán MA, Robins JM. Using big data to emulate a target trial when a randomized trial is not available. Am J Epidemiol. 2016;183:758‐764. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Thill M, Zahn M‐O, Welt A, et al. Treatment and outcome in metastatic lobular breast cancer in the prospective German research platform OPAL. Breast Cancer Res Treat. 2023;198:545‐553. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. European Medicines Agency . ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials; 2020 [cited 2024 Dec 11]. Available from: https://www.ema.europa.eu/en/documents/scientific‐guideline/ich‐e9‐r1‐addendum‐estimands‐and‐sensitivity‐analysis‐clinical‐trials‐guideline‐statistical‐principles‐clinical‐trials‐step‐5_en.pdf
30. Cole SR, Hernan MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168:656‐664. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34:3661‐3679. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Heinze G, Wallisch C, Dunkler D. Variable selection—a review and recommendations for the practicing statistician. Biom J. 2018;60:431‐449. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Stuart EA, Lee BK, Leacy FP. Prognostic score‐based balance measures can be a useful diagnostic for propensity score methods in comparative effectiveness research. J Clin Epidemiol. 2013;66:S84‐S90.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457‐481. [Google Scholar]
35. Hortobagyi GN, Stemmer SM, Burris HA, et al. Updated results from MONALEESA‐2, a phase III trial of first‐line ribociclib plus letrozole versus placebo plus letrozole in hormone receptor‐positive, HER2‐negative advanced breast cancer. Ann Oncol. 2018;29:1541‐1547. [DOI] [PubMed] [Google Scholar]
36. Hortobagyi GN, Stemmer SM, Burris HA, et al. Overall survival with ribociclib plus letrozole in advanced breast cancer. N Engl J Med. 2022;386:942‐950. [DOI] [PubMed] [Google Scholar]
37. Rugo HS, Finn RS, Diéras V, et al. Palbociclib plus letrozole as first‐line therapy in estrogen receptor‐positive/human epidermal growth factor receptor 2‐negative advanced breast cancer with extended follow‐up. Breast Cancer Res Treat. 2019;174:719‐729. [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Finn RS, Rugo HS, Dieras VC, et al. Overall survival (OS) with first‐line palbociclib plus letrozole (PAL+LET) versus placebo plus letrozole (PBO+LET) in women with estrogen receptor–positive/human epidermal growth factor receptor 2–negative advanced breast cancer (ER+/HER2− ABC): analyses from PALOMA‐2. J Clin Oncol. 2022;40:LBA1003. [Google Scholar]
39. Müller C, Kiver V, Solomayer E‐F, et al. CDK4/6 inhibitors in advanced HR+/HER2− breast cancer: a multicenter real‐world data analysis. Breast Care (Basel). 2023;18:31‐41. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Dr. Jacot on ribociclib dose reduction in HR+/HER2− advanced breast cancer; OncLive2022 [cited 2025 Mar 17]. Available from: https://www.onclive.com/view/dr‐jacot‐on‐ribociclib‐dose‐reduction‐in‐hr‐her2‐advanced‐breast‐cancer
41. Güren AK, Guliyev M, Alan Ö, et al. Evaluation of dose reduction factors and impact on progression‐free survival in patients treated with CDK 4/6 inhibitors. J Clin Med. 2025;14:1071. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Brain E, Pulido M, Paillaud E, et al. Abstract P1‐18‐04: feasibility of palbociclib in women aged 70 and older with resistant and/or pretreated advanced breast cancer in the PALOMAGE study. Cancer Res. 2022;82:P1‐1804. [Google Scholar]
43. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173:676‐682. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1. Supporting Information.

IJC-159-244-s001.pdf^{(36MB, pdf)}

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

[ijc70433-bib-0001] 1. Papachristos A, Patel J, Vasileiou M, Patrinos GP. Dose optimization in oncology drug development: the emerging role of pharmacogenomics, pharmacokinetics, and pharmacodynamics. Cancer. 2023;15:3233. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0002] 2. Soltantabar P, Lon H‐K, Parivar K, Wang DD, Elmeliegy M. Optimizing benefit/risk in oncology: review of post‐marketing dose optimization and reflections on the road ahead. Crit Rev Oncol Hematol. 2023;182:103913. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0003] 3. Liao JJZ, Asatiani E, Liu Q, Hou K. Three steps toward dose optimization for oncology dose finding. Contemp Clin Trials Commun. 2024;40:101329. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0004] 4. FDA F and DA . Optimizing the dosage of human prescription drugs and biological products for the treatment of oncologic diseases; 2024 [cited 2024 Nov 13]. Available from: https://www.fda.gov/regulatory‐information/search‐fda‐guidance‐documents/optimizing‐dosage‐human‐prescription‐drugs‐and‐biological‐products‐treatment‐oncologic‐diseases

[ijc70433-bib-0005] 5. Gennari A, André F, Barrios CH, et al. ESMO clinical practice guideline for the diagnosis, staging and treatment of patients with metastatic breast cancer. Ann Oncol. 2021;32:1475‐1495. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0006] 6. Thill M, Zahn M‐O, Welt A, et al. Head to head comparison of palbociclib and ribociclib in first‐line treatment of HR‐positive/HER2‐negative metastatic breast cancer with real‐world data from the OPAL registry. Int J Cancer. 2024;156:1770‐1782. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0007] 7. European Medicines Agency . Summary of product characteristics‐IBRANCE®, Pfizer Europe MA EEIG, INN‐palbociclib; 2021 [cited 2025 Jan 15]. Available from: https://www.ema.europa.eu/en/documents/product-information/ibrance-epar-product-information_en.pdf

[ijc70433-bib-0008] 8. European Medicines Agency . Summary of product characteristics‐Kisqali®, Novartis Europharm Limited, INN‐ribociclib; 2022 [cited 2025 Jan 15]. Available from: https://www.ema.europa.eu/en/documents/product-information/kisqali-epar-product-information_en.pdf

[ijc70433-bib-0009] 9. Bjerrum A, Henriksen AF, Knoop AS, et al. Overall survival after CDK4/6 inhibitor dose reduction in women with metastatic breast cancer. BJC Rep. 2024;2:1‐8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0010] 10. Loeser A, Kim JS, Peppercorn J, et al. The right dose: results of a patient advocate–led survey of individuals with metastatic breast cancer regarding treatment‐related side effects and views about dosage assessment to optimize quality of life. JCO Oncol Pract. 2024;20:972‐983. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0011] 11. Kristensen KB, Thomsen IMN, Berg T, Kodahl AR, Jensen AB. Dose modifications of ribociclib and endocrine therapy for treatment of ER+ HER2− metastatic breast cancer. Breast Cancer Res Treat. 2021;188:799‐809. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0012] 12. Shikdar S, Hall S, Taylor B, et al. Evaluating the response of palbociclib on progression‐free survival in hormone receptor‐positive metastatic breast cancer. Oncology. 2022;100:337‐343. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0013] 13. Ettl J, Im S‐A, Ro J, et al. Hematologic adverse events following palbociclib dose reduction in patients with hormone receptor‐positive/human epidermal growth factor receptor 2‐negative advanced breast cancer: pooled analysis from randomized phase 2 and 3 studies. Breast Cancer Res. 2020;22:27. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0014] 14. Dennison T, Heiling H, Deal A, et al. Tolerability of palbociclib in younger and older patients with advanced breast cancer. J Oncol Pharm Pract. 2023;29:96‐104. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0015] 15. Ogliari FR, Caspani F, Rossi L, Nerone M, Vallini I, Gueli R. Predictive variables of palbociclib dose reduction and its impact on survival outcomes: a retrospective bicentric study. Breast J. 2021;27:415‐416. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0016] 16. Savignon S, Boyer M, Messier C, Monfette ML, Letarte N. Predictive factors of dose reduction among patients treated with palbociclib for advanced or metastatic breast cancer in a real‐world setting. Poster presented at the Hematology/Oncology Pharmacy Association Annual Conference 2024; April 3–6, 2024; Tampa, Florida. 2024 [cited 2024 May 8]. Available from: https://jhoponline.com/issue‐archive/latest‐issue/19651‐predictive‐factors‐of‐dose‐reduction‐among‐patients‐treated‐with‐palbociclib‐for‐advanced‐or‐metastatic‐breast‐cancer‐in‐a‐real‐world‐setting

[ijc70433-bib-0017] 17. Vazquez L, Coussirou J, Arvers P, et al. Palbociclib associated with endocrine therapy in patients with metastatic breast cancer: predictive factors of severe early hematological toxicity. Breast Cancer. 2019;108:544‐552. [Google Scholar]

[ijc70433-bib-0018] 18. Diéras V, Harbeck N, Joy AA, et al. Palbociclib with letrozole in postmenopausal women with ER+/HER2− advanced breast cancer: hematologic safety analysis of the randomized PALOMA‐2 trial. Oncologist. 2019;24:1514‐1525. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0019] 19. Ji Y, Darstein C, Yang S, et al. Quantitative assessment of ribociclib exposure‐response relationship to justify dose regimen in patients with advanced breast cancer. J Clin Pharmacol. 2023;63:1359‐1370. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0020] 20. Fernández‐Cuerva C, Chinchilla‐Alarcón T, Alcaraz‐Sánchez JJ. Real‐world effectiveness of ribociclib in metastatic breast cancer patients: Does dose affect survival? J Oncol Pharm Pract. 2023;29:1619‐1627. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0021] 21. Zheng J, Yu Y, Durairaj C, Diéras V, Finn RS, Wang DD. Impact of dose reduction on efficacy: implications of exposure‐response analysis of palbociclib. Target Oncol. 2021;16:69‐76. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0022] 22. Burris HA, Chan A, Bardia A, et al. Safety and impact of dose reductions on efficacy in the randomised MONALEESA‐2, ‐3 and ‐7 trials in hormone receptor‐positive, HER2‐negative advanced breast cancer. Br J Cancer. 2021;125:679‐686. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0023] 23. Cardoso F, Jacot W, Küemmel S, et al. Abstract PD17‐12: primary efficacy and safety results from the AMALEE trial evaluating 600 mg vs 400 mg starting doses of first‐line ribociclib in patients with HR+/HER2− advanced breast cancer. Cancer Res. 2023;83:PD17‐12. [Google Scholar]

[ijc70433-bib-0024] 24. Ismail RK, van Breeschoten J, Wouters MWJM, et al. Palbociclib dose reductions and the effect on clinical outcomes in patients with advanced breast cancer. Breast. 2021;60:263‐271. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0025] 25. Layman RM, Liu X, Li B, McRoy L, Brufsky A. Real‐world palbociclib dose modifications and clinical outcomes in patients with HR+/HER2− metastatic breast cancer: a Flatiron Health Database analysis. Breast. 2025;81:104448. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0026] 26. Suissa S. Immortal time bias in pharmaco‐epidemiology. Am J Epidemiol. 2008;167:492‐499. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0027] 27. Hernán MA, Robins JM. Using big data to emulate a target trial when a randomized trial is not available. Am J Epidemiol. 2016;183:758‐764. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0028] 28. Thill M, Zahn M‐O, Welt A, et al. Treatment and outcome in metastatic lobular breast cancer in the prospective German research platform OPAL. Breast Cancer Res Treat. 2023;198:545‐553. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0029] 29. European Medicines Agency . ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials to the guideline on statistical principles for clinical trials; 2020 [cited 2024 Dec 11]. Available from: https://www.ema.europa.eu/en/documents/scientific‐guideline/ich‐e9‐r1‐addendum‐estimands‐and‐sensitivity‐analysis‐clinical‐trials‐guideline‐statistical‐principles‐clinical‐trials‐step‐5_en.pdf

[ijc70433-bib-0030] 30. Cole SR, Hernan MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168:656‐664. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0031] 31. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34:3661‐3679. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0032] 32. Heinze G, Wallisch C, Dunkler D. Variable selection—a review and recommendations for the practicing statistician. Biom J. 2018;60:431‐449. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0033] 33. Stuart EA, Lee BK, Leacy FP. Prognostic score‐based balance measures can be a useful diagnostic for propensity score methods in comparative effectiveness research. J Clin Epidemiol. 2013;66:S84‐S90.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0034] 34. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457‐481. [Google Scholar]

[ijc70433-bib-0035] 35. Hortobagyi GN, Stemmer SM, Burris HA, et al. Updated results from MONALEESA‐2, a phase III trial of first‐line ribociclib plus letrozole versus placebo plus letrozole in hormone receptor‐positive, HER2‐negative advanced breast cancer. Ann Oncol. 2018;29:1541‐1547. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0036] 36. Hortobagyi GN, Stemmer SM, Burris HA, et al. Overall survival with ribociclib plus letrozole in advanced breast cancer. N Engl J Med. 2022;386:942‐950. [DOI] [PubMed] [Google Scholar]

[ijc70433-bib-0037] 37. Rugo HS, Finn RS, Diéras V, et al. Palbociclib plus letrozole as first‐line therapy in estrogen receptor‐positive/human epidermal growth factor receptor 2‐negative advanced breast cancer with extended follow‐up. Breast Cancer Res Treat. 2019;174:719‐729. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0038] 38. Finn RS, Rugo HS, Dieras VC, et al. Overall survival (OS) with first‐line palbociclib plus letrozole (PAL+LET) versus placebo plus letrozole (PBO+LET) in women with estrogen receptor–positive/human epidermal growth factor receptor 2–negative advanced breast cancer (ER+/HER2− ABC): analyses from PALOMA‐2. J Clin Oncol. 2022;40:LBA1003. [Google Scholar]

[ijc70433-bib-0039] 39. Müller C, Kiver V, Solomayer E‐F, et al. CDK4/6 inhibitors in advanced HR+/HER2− breast cancer: a multicenter real‐world data analysis. Breast Care (Basel). 2023;18:31‐41. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0040] 40. Dr. Jacot on ribociclib dose reduction in HR+/HER2− advanced breast cancer; OncLive2022 [cited 2025 Mar 17]. Available from: https://www.onclive.com/view/dr‐jacot‐on‐ribociclib‐dose‐reduction‐in‐hr‐her2‐advanced‐breast‐cancer

[ijc70433-bib-0041] 41. Güren AK, Guliyev M, Alan Ö, et al. Evaluation of dose reduction factors and impact on progression‐free survival in patients treated with CDK 4/6 inhibitors. J Clin Med. 2025;14:1071. [DOI] [PMC free article] [PubMed] [Google Scholar]

[ijc70433-bib-0042] 42. Brain E, Pulido M, Paillaud E, et al. Abstract P1‐18‐04: feasibility of palbociclib in women aged 70 and older with resistant and/or pretreated advanced breast cancer in the PALOMAGE study. Cancer Res. 2022;82:P1‐1804. [Google Scholar]

[ijc70433-bib-0043] 43. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173:676‐682. [DOI] [PubMed] [Google Scholar]

PERMALINK

Prediction of dose reductions and impact of reduced starting doses of CDK4/6i on effectiveness in HR‐positive/HER2‐negative metastatic breast cancer: Real‐world data from the OPAL registry

Patrick Marschner

Kai Ringwald

Marc Thill

Mark‐Oliver Zahn

Anja Welt

Arnd Nusch

Gabriele Kaltenecker

Yolanda Rodemer

Volker Hagen

Caroline Schock

Lisa Kruggel

Anja Kaiser‐Osterhues

Dunja Klein

Nina Haug

Elmar Stickeler

Nadia Harbeck

Achim Wöckel

Norbert Marschner

Thomas Decker

Abstract

What's New?

Abbreviations

1. INTRODUCTION

2. PATIENTS AND METHODS

2.1. Study design and cohort definition

2.2. Derivation of a predictive index for dose reduction and toxicity discontinuation

2.3. Progression‐free and overall survival of patients with a full versus a reduced starting dose

2.4. Statistical analysis

3. RESULTS

3.1. Cohort description and patient characteristics

FIGURE 1.

TABLE 1.

3.2. Predictors for dose reduction: Predictive index

3.2.1. Propensity score weighting

3.2.2. Identified predictive factors and indexing

FIGURE 2.

FIGURE 3.

3.3. Progression‐free and overall survival according to probability of dose reduction

3.3.1. Propensity score weighting

3.3.2. Progression‐free and overall survival of the weighted cohorts

FIGURE 4.

FIGURE 5.

4. DISCUSSION

5. LIMITATIONS

6. CONCLUSION

AUTHOR CONTRIBUTIONS

FUNDING INFORMATION

CONFLICT OF INTEREST STATEMENT

ETHICS STATEMENT

Supporting information

ACKNOWLEDGMENTS

DATA AVAILABILITY STATEMENT

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases