Quantitative evaluation of the efficacy and safety profiles of two types of targeted inhibitors combined with endocrine therapy in ER+/HER2- metastatic breast cancer

doi:10.21203/rs.3.rs-4163267/v1

Download PDF

Research Article

Quantitative evaluation of the efficacy and safety profiles of two types of targeted inhibitors combined with endocrine therapy in ER+/HER2- metastatic breast cancer

https://doi.org/10.21203/rs.3.rs-4163267/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 20 Jun, 2024

Read the published version in European Journal of Clinical Pharmacology →

You are reading this latest preprint version

Purpose

The aim of this study was to quantitatively compare the efficacy and safety of CDK4/6 inhibitors and PI3K/AKT/mTOR inhibitors for ER+/HER2- metastatic breast cancer.

Methods

A parametric survival function was used to analyze the time course of overall survival (OS) and progression-free survival (PFS). The objective response rate (ORR) and the incidence of any grade and grade 3–4 adverse events were summarized using the random-effects model of a single-arm meta-analysis.

Results

This study included 44 arms from 48 publications, with a total sample size of 7,881 patients. Our study revealed that CDK4/6 inhibitors had a median OS of 40.7 months, a median PFS of 14.8 months, and ORR of 40%. Whereas, PI3K/AKT/mTOR inhibitors had a median OS of 29.8 months, a median PFS of 8.3 months, and ORR of 20%. Additionally, this study also found that the proportion of patients with visceral metastases and specific endocrine therapy used in combination significantly impact OS and PFS. In terms of adverse events, CDK4/6 inhibitors exhibited a relatively high incidence of hematological adverse events.

Conclusion

Our study provides solid quantitative evidence for the first-line recommendation of CDK4/6 inhibitors combined with endocrine therapy for ER+/HER2- metastatic breast cancer in clinical guidelines.

ER+/HER2- metastatic breast cancer

model-based meta-analysis

efficacy

CDK4/6 inhibitors

PI3K/AKT/mTOR inhibitors

PFS

influencing factors

Breast cancer has emerged as the most frequently diagnosed cancer type globally. In 2020, approximately 2.3 million women were diagnosed with breast cancer, marking the first time that the incidence of breast cancer surpassed that of lung cancer[1]. Even for patients with early diagnosis and intervention, about 30% will eventually progress to recurrent or metastatic breast cancer. For the predominant subtype of ER+/HER2- metastatic breast cancer, initial treatment selection is crucial because disease will progress aggressively once drug resistance happened[2]. Historically, endocrine therapy which includes aromatase inhibitors (AIs), selective estrogen receptor modulators (SERMs), and selective estrogen receptor degraders (SERDs), constitutes the cornerstone of treatment for ER+/HER2- metastatic breast cancer[3]. However, resistance to endocrine therapy is unavoidable, numerous novel drugs are developed to add on to overcome drug resistance. The emergence of cyclin-dependent kinases 4 and 6 (CDK4/6) inhibitors and phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) inhibitors bring hope for these patient population and have been explored as add-on therapies because of synergistic effects[4–6]. But till now there is no head-to-head clinical trial to compare the efficacy between these two types of inhibitors, furthermore, controversies and inconsistent conclusions persist regarding their relative efficacy and safety based on current available meta-analysis. For example, the investigation conducted by Wang et.al[7] suggested that CDK4/6 inhibitors remarkably surpass PI3K inhibitors in terms of PFS, both as primary and subsequent lines of treatment. However, this advantage does not extend to OS and ORR. In contrast, research by Leung et.al[8] revealed a significant superiority in OS for CDK4/6 inhibitors over PI3K/AKT/mTOR inhibitors when given as second-line therapy. Meanwhile, another study by Xu et.al[9] indicated that after adjusting therapy lines and metastasis site, CDK4/6 and PI3K/AKT/mTOR inhibitors had no significant difference in PFS and OS whether as first- or second-line treatment. The above situation emphasizes the necessity for systematic, quantitative and direct comparisons between these two types of treatment to determine the optimal choice as first -line treatment.

Model-Based Meta-Analysis (MBMA) effectively integrates pharmacological knowledge, meta-analysis, and mathematical modeling to consolidate data from various sources across multiple dimensions[10, 11]. By establishing time-course and covariate models, MBMA can diminish the influence of heterogeneity among various studies and predict OS and PFS at any time rather than median survival time, which takes full advantage of data and thus enhancing credibility of results. Our study employs MBMA to aggregate and compare the efficacy and safety of combination treatment of CDK4/6 or PI3K/AKT/mTOR inhibitors with endocrine therapy in patients with ER+/HER2- metastatic breast cancer, aiming to provide more credible and scientific evidence.

Searching strategies and inclusion criteria

The Cochrane Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for meta-analyses were used to collect data and report results accordingly. A comprehensive literature search was performed using the public database of PubMed, Embase, and the Cochrane Library for clinical trials of ER+/HER2- metastatic breast cancer by two independent authors between January 1, 2010, and August 23, 2023. The searching key words included “ER+/HER2- breast cancer”, “CDK4/6 inhibitors” and “PI3K/AKT/mTOR inhibitors”. The details of searching strategies are presented in Table S1.

The inclusion criteria were as follows: (1) single-arm, placebo-controlled, or positive drug-controlled clinical trials; (2) involving adults with ER+/HER2- metastatic breast cancer; (3) participants were treated with regimens containing CDK4/6 or PI3K/AKT/mTOR inhibitors; (4) at least one of the efficacy indicators (OS, PFS, or ORR) was used to evaluate efficacy.

Data extraction and risk-of-bias assessment

Microsoft Excel (version 2021) was used for data collection. The collected information included literature characteristics (authors, publication year, and clinical trial registration number), trial design (groups, administration timing, and sample size), participant characteristics (including age, race, ECOG score, proportion of patients with visceral metastasis, proportion of patients with bone metastases only, and proportion of patients who had previously received treatment), efficacy outcomes (OS, PFS, ORR), and safety outcomes (incidence of grade 3–5 adverse events).

If outcome measures are presented graphically in the literature, data extraction will be conducted using Engauge Digitizer (version 12.1). Two independent researchers will extract data from the studies, and any discrepancies will be resolved by a third researcher. During digitization, the data extraction error should not exceed 2%. If the error surpasses this threshold, digitization will be repeated, and the final value is the average of the two extractions.

Model development

Parametric survival models were used to analyze OS and PFS data. The survival model is associated with a hazard function h(t), which can be interpreted as an instantaneous hazard function determined as follows^{[12, 13]}:

$$S(t)=\exp ( - \int\limits_{0}^{t} {h(t)dt} )$$

We evaluated four different hazard functions, namely, the constant, Weibull, Gompertz, and log-normal models (Eqs. 2, 3, 4 and 5). In Eqs. (2–4), parameter λ and β represent the hazard rate at time 0 and the regression coefficient of the hazard rate changing over time, respectively. While in Eq. (5), µ and σ are the mean and standard deviation of the lognormal distribution, respectively, which will determine the shape of the function curve.

Constant $h(t)=\lambda$ (2)

Gompertz $h(t)=\lambda \cdot \exp (\beta \cdot \ln (t))$ (3)

Weibull $h(t)=\lambda \cdot \exp (\beta \cdot t)$ (4)

Lognormal $h(t)={\frac{{(\sigma t\sqrt {2\pi } )}}{{1 - \phi (z)}}^{ - 1}}{e^{( - \frac{1}{2}{z^2})}}$ (5)

The selection of the four hazard functions was based on the minimum value of the objective function (OFV), parameter plausibility, and diagnosis plots.

The inter-trial variation (Eq. 6) and residuals (Eq. 7) were added to the model parameters in the following form:

$${P_i}={P_{pop}} \times {e^{{\eta _i}}}$$

$$Ob{s_{j,i}}=Pre{d_{j,i}}+S{E_{j,i}} \times {\varepsilon _{j,i}}$$

$$S{E_{j,i}}=\sqrt {\frac{{Ob{s_{j,i}} \times (1 - Ob{s_{j,i}})}}{{{N_{j,i}}}}}$$

In Eq. 6, P_i is the model parameter of the i^th study, P_pop is the typical value of the model parameter, η_i is the interstudy variation and η_i follows a normal distribution with a mean of 0 and variance of ω². In Eq. (7), Obs_j,i is the observed survival data at j^th time in i^th study; Pred_j,i is the predicted survival data at j^th time of i^th study; ε_j,i is the residual error of j^th time of i^th study and ε_j,i conforms to a normal distribution with a mean of 0 and variance of σ². ε_j,i is weighted by the standard error of the survival data at j^th time of i^th study (SE_j,i), which can be acquired using Eq. (8), where N_j,i is the sample size of j^th time in i^th study.

Covariate models were used to examine potential factors affecting OS and PFS, including age, race, ECOG score, proportion of patients with visceral metastases, proportion of patients with bone metastases only, proportion of patients who had previously received treatment and the types of endocrine therapy. Missing covariate information was imputed using the median value of the entire study population. Missing covariate proportions > 30% were not assessed.

Continuous variables were introduced using Eq. (9) and Eq. (10), and categorical variables were introduced using Eq. (11):

$${P_{pop}}={P_{typical}}+(COV - CO{V_{_{{median}}}}) \times {\theta _{\operatorname{cov} }}$$

$${P_{pop}}={P_{typical}} \times {(\frac{{COV}}{{CO{V_{median}}}})^{{\theta _{\operatorname{cov} }}}}$$

$${P_{pop}}={P_{typical}}+COV \times {\theta _{\operatorname{cov} }}$$

In Eqs. (9–11), P_pop is the model parameter at different covariable levels, and P_typical is the typical population value of the model parameter when the categorical covariate is equal to 0 or the continuous covariate is equal to COV_median. COV is the covariate value, COV_median is the median value of this covariable in the analysis dataset, and θ_COV is the correction coefficient of the covariable for the model parameters.

The stepwise method is commonly used to develop a covariate model. Forward inclusion and backward elimination methods were used for covariate screening. The bound of OFV decrease in the forward method was set at 3.84 (P < 0.05), whereas in the backward method, the bound was set at 6.63 (P < 0.01).

Model evaluation

The goodness-of-fit plots were used to describe the fitness of models. A visual predictive check was performed to evaluate the predictive performance of the model by comparing the 95% confidence interval (CI) of the model predictions using the observed values. The stability of the model was assessed using a bootstrap with 1,000 new datasets acquired from the original dataset to obtain the median of the model parameter distribution and its 95% CI, which were compared with the estimated values of the model parameters obtained from the original dataset.

Model prediction

Having established the final model, Bayesian feedback was used to obtain the individual model parameters and their standard errors. A random-effects model in a single-arm meta-analysis was used to summarize the model parameters for different targeted drugs. Finally, Monte Carlo simulations (1,0000 times) were used to obtain the typical OS and PFS time curves and their 95% CIs for two types of targeted drugs[14].

ORR and grade 3–5 adverse events

To determine the characteristics of ORR and safety, ORR and incidence of common grade 3–5 adverse events were pooled using a random-effect model in a single-arm meta-analysis.

Correlations among ORR, PFS and OS

Investigating the correlation between OS and potential short-term efficacy endpoints (ORR and PFS), can accelerate drug development. Weighted linear regression analysis was used to evaluate correlations among the following efficacy indicators: (1) 0.5-year and 1-year PFS, (2) 2-year OS, (3) ORR. The Pearson correlation coefficient (R²) was used to measure the strength of correlation among them. We considered a prespecified R² values of ≥ 0.8 to indicate a high correlation, 0.6–0.8 to indicate a moderate correlation, and < 0.6 to indicate a low correlation[12, 15].

Software

Model development and simulation were performed using NONMEM 7.5 (ICON Development Solutions, USA). Statistical analysis、meta-analysis and plotting were performed using R4.3.2 (The R Foundation of Statistical Computing). Quality assessment of literatures was performed using Review Manager (Version 5.4. The Cochrane Collaboration, 2020).

Characteristics of included studies

Forty-eight studies with 7,881 subjects were included in the analysis. The literature screening process is shown in Fig. 1. Among the 48 included studies, 26 studies focusing on CDK4/6 inhibitors (N = 4750) and 22 studies focusing on PI3K/AKT/mTOR inhibitors (N = 3131). Detailed information about the included studies is provided in Table S2 .

As it shown in Table 1, included patients had an average age of 62 years (ranging from 43 to 64 years), Among them, 59% (ranging from 40–84%) had visceral metastases and 21% (ranging from 5–30%) only had bone metastases. Additionally, the proportion of patients receiving first-line treatment was found to be 43% (ranging from 0 to 100%).

The risk of bias assessment of the included studies is summarized in Fig. S1.

Table 1

Baseline characteristics of included patients, median (min-max)
	CDK4/6 inhibitors	PI3K/AKT/mTOR inhibitors	Overall
Number of trials(arms)	21	23	44
Total sample size	4750	3131	7881
Median age(year)	60(43–64)	62(43.9–64)	62(43–64)
Race-Asian(%)	26(0-100)	11(0–23)	15(0-100)
Visceral disease(%)	59(40–67)	60.9(49–84)	59(40–84)
ECOG(%)
0	58(36–73)	60(51–100)	59(36–100)
1	41(26–53)	37.5(0-45.7)	40(0–53)
Primary treatment population(%)	100(0-100)	32.5(0-100)	43(0-100)
Bone only metastasis(%)	21(8–28)	18.8(5–30)	21(5–30)

Model development and evaluation

The results showed that the log-normal model had lower OFV values and a smaller relative standard error (RSE%) of the model parameters. Therefore, the log-normal model was selected to fit OS and PFS data, and its model parameters are presented in Table 2. The detailed OS and PFS model exploration processes are described in Table S2-S3.

The 1,000 times bootstrap-replaceable repeated sampling method successfully converged 995 times and 959 times, respectively, with 95% CIs that were extremely close to the parameter estimates of the final models of OS and PFS, thus indicating that these models are robust and relatively less affected by individual study data (Table 2). The goodness-of-fit and individual fit plots of the final model revealed that the model provided a good fit of the measured data (Fig. S2-S5) and no systematic bias was observed. In addition, most of the measured data generally fell within the 95% CI predicted values and accordingly indicates that the model had good predictability (Fig. S6).

The results of covariate analysis revealed that the proportion of patients with visceral metastases and the type of endocrine therapy drug used in combination have a significant impact on both OS and PFS models. The covariate model can be expressed as follows:

OS model: $h{(t)_{final}}=h{(t)_{base}} \times {e^{({\theta _{ET\_OS}}+(VISCERAL - 0.59) \times 0.557)}}$ (12)

where: ${\theta _{ET\_OS}}=\left\{ \begin{gathered} 0 \hfill \\ 0.442 \hfill \\ 0.297 \hfill \\ \end{gathered} \right.$ $\begin{gathered} AI \hfill \\ Fulvestrant \hfill \\ Tamoxifen \hfill \\ \end{gathered}$

PFS model: $h{(t)_{final}}=h{(t)_{base}} \times {e^{({\theta _{ET\_PFS}}+(VISCERAL - 0.59) \times 2.23)}}$ (13)

where: ${\theta _{ET\_PFS}}=\left\{ \begin{gathered} 0 \hfill \\ 0.421 \hfill \\ 0.377 \hfill \\ \end{gathered} \right.$ $\begin{gathered} AI \hfill \\ Fulvestrant \hfill \\ Tamoxifen \hfill \\ \end{gathered}$

In Eqs. (12) and (13), VISCERAL represents the proportion of patients with visceral metastases, with a median percentage of 0.59. θ_ET is the correction of coefficient for the initial risk of death when different endocrine therapy is used in combination.

Table 2

Parameter estimates of the final model for OS and PFS
Parameters	PFS				OS
	Final model		Bootstrap(n = 1000)		Final model		Bootstrap (n = 1000)
	Value (RSE%)	Shrinkage (%)	Median	95% CI	Value (RSE%)	Shrinkage (%)	Median	95% CI
Survival parameters
SIGM-ET	1.24 (6)		1.22	1.07–1.40	0.951 (4)		0.953	0.889–1.056
SIGM-PAM	1.17 (4)		1.17	1.07–1.35	1.03 (7)		1.064	0.844–1.260
SIGM-CDK	1.21 (4)		1.20	1.11–1.29	0.973 (6)		0.976	0.889–1.112
Mu-ET	2.44 (9)		2.38	1.96–2.82	3.77 (3)		3.807	3.598–4.115
Mu-PAM	2.57 (6)		2.53	2.17–2.87	3.76 (4)		3.766	3.389–4.203
Mu-CDK	3.07 (6)		3.04	2.71–3.40	4.06 (3)		4.051	3.783–4.358
TV-fulvestrant	0.421 (41)		0.41	0.08–0.74	0.442 (36)		0.432	0.160–0.712
TV-tamoxifen	0.377 (56)		0.35	0.02–0.71	0.297 (71)		0.370	0.076–0.690
TV-visceral	2.23 (33)		2.15	0.57–3.51	0.557 (330)		1.424	-0.002-4.012
Variability parameters
η-SIGM	0.0286 (19)	5	0.0278	0.0152–0.0410	0.0318 (41)	3	0.0363	0.0117–0.0595
η-Mu	0.0441 (31)	0	0.0504	0.0231–0.0860	0.00661 (18)	1	0.0056	0.00132–0.00793
ε	0.00104 (14)	5	0.00117	0.00085–0.00174	0.0006 (18)	5	0.000612	0.000456–0.000844

SIGM standard deviation of the lognormal distribution, MU mean of the lognormal distribution, η interstudy variability, ε residual error, CI confidence interval

Typical overall and progression-free survival simulation

Based on the final model, typical OS and PFS and their 95% CIs for CDK4/6 and PI3K/AKT/mTOR inhibitors were simulated. The results (Table 3, Fig. 2) showed that when the proportion of patients with visceral metastases was adjusted to 59% (median), the median OS for CDK4/6 and PI3K/AKT/mTOR inhibitors was 40.7 months (95% CI: 33.3–48.2) and 29.8 months (95% CI:23.8–36.6), respectively. The 1-year OS rates for CDK4/6 and PI3K/AKT/mTOR inhibitors were 91.8% (95% CI: 87.5–95.2) and 83.8% (95% CI: 76.3–90.4), respectively.

The median PFS for CDK4/6 and PI3K/AKT/mTOR inhibitors (Table 3, Fig. 2) was 14.8 months (95% CI: 11.3–18.7) and 8.3 months (95% CI: 7.1–9.9), respectively. The 1-year PFS rates for CDK4/6 inhibitors and PI3K/AKT/mTOR inhibitors were 57.6% (95% CI: 47.8–66.2) and 36.0% (95% CI: 30.2–42.4), respectively.

Besides, among all the covariates, our study found that the two covariates greatly impacted OS and PFS were the proportion of patients with visceral metastases and different endocrine therapy used in combination. When the proportion of patients with visceral metastases was adjusted from 40–84%, the median OS of CDK4/6 inhibitors was decreased from 63.5 to 51.6 months, respectively, and its PFS was decreased from 34.6 to 12.5 months, respectively. For PI3K/AKT/mTOR inhibitors under same adjustment, the median OS changed from 47.3 to 38.0 months, respectively, and its median PFS decreased from 20.6 to 7.7 months, respectively. (Fig. 2)

In terms of the endocrine therapy used in combination, we simulated OS and PFS for CDK4/6 and PI3K/AKT/mTOR inhibitors combined with different endocrine therapy. The results showed that the median OS increased in the following order: CDK4/6 inhibitors/AI (57.8 months) > CDK4/6 inhibitors/tamoxifen (45.6 months) > PI3K/AKT/mTOR inhibitors/AI (42.8 months) > CDK4/6 inhibitors/fulvestrant (40.7 months) > PI3K/AKT/mTOR inhibitors/tamoxifen (33.2 months) > PI3K/AKT/mTOR inhibitors/fulvestrant (29.6 months); the median PFS increased in the following order: CDK4/6 inhibitors/AI (21.7 months) > CDK4/6 inhibitors/tamoxifen (14.9 months) > CDK4/6 inhibitors/fulvestrant (14.2 months) > PI3K/AKT/mTOR inhibitors/AI (13.2 months) > PI3K/AKT/mTOR inhibitors/tamoxifen (9.2 months) > PI3K/AKT/mTOR inhibitors/fulvestrant (8.8 months). However, due to small number of included trials for CDK4/6 inhibitors/tamoxifen (N = 3) and PI3K/AKT/mTOR inhibitors/tamoxifen (N = 1), their typical efficacy estimates may have sampling errors and should be interpreted with caution. (Fig. 3)

Table 3

Predicted typical values and 95% CIs of OS and PFS in two types of drugs
	CDK4/6 inhibitors	PI3K/AKT/mTOR inhibitors
Overall survival
Median OS, months	40.7 (33.3–48.2)	29.8 (23.8–36.6)
1-year OS rate, %	91.8 (87.5–95.2)	83.8 (76.3–90.4)
2-year OS rate, %	73.0 (64.8–80.1)	59.4 (49.5–68.6)
5-year OS rate, %	32.2 (24.1–40.0)	21.4 (14.9–28.9)
Progression-free survival
Median PFS, months	14.8 (11.3–18.7)	8.3 (7.1–9.9)
1-year PFS rate, %	57.6 (47.8–66.2)	36.0 (30.2–42.4)
2-year PFS rate, %	32.3 (23.8–40.7)	14.6 (11.3–19.0)
5-year PFS rate, %	9.1 (5.7–13.7)	2.4 (1.5–3.6)

Meta-Analysis for Secondary and Safety Outcomes

The ORR for CDK4/6 inhibitors combined with endocrine therapy was 40% (95% CI, 34%-46%), which was significantly higher than PI3K/AKT/mTOR inhibitors and endocrine therapy combination whose ORR was 20% (95% CI, 14%-27%).

Figure 4 depicts the adverse effects associated with CDK4/6 inhibitors and PI3K/AKT/mTOR inhibitors, respectively. The results of single-arm meta-analysis showed that the hematologic toxicity of CDK4/6 inhibitors was very high. Among them, neutropenia was the most common hematological adverse reaction, with the incidence of grade 3–5 neutropenia exceeding 50%. Common adverse events associated with PI3K/AKT/mTOR inhibitors included hyperglycemia, nausea, diarrhea, fatigue, stomatitis, and rash, with incidence rates of all grade adverse events exceeding 30%.

Correlation among ORR, PFS and OS

As it is shown in Fig. 5, 1-year PFS was moderately correlated with 2-year OS (R² = 0.624). The corresponding equation (y = 0.4716x + 0.4353) indicates that for every 10% increase in 1-year PFS, the 2-year OS would increase by 4.7%. Subgroup analysis revealed a stronger correlation between 1-year PFS and 2-year OS in CDK4/6 inhibitors group (R² = 0.784). Moderate correlation was also detected between 0.5-year PFS and 2-year OS in CDK4/6 inhibitor group (R² = 0.688). The correlation between ORR and PFS or OS is fairly weak compared with the correlation between PFS and OS.

As far as we know, our study, was the first time to employ MBMA to conduct a comprehensive analysis of two types of targeted inhibitors combined with different endocrine therapy in the context of three efficacy indicators (OS, PFS, ORR), aiming to determine the most effective treatment strategies and factors affecting its efficacy, thereby providing crucial quantitative information for clinical guidelines.

Our study found that CDK4/6 inhibitors had significant advantages in PFS and OS compared with PI3K/AKT/mTOR inhibitors, whether as first-line or later-line therapy. Considering that the timing of treatment with PI3K/AKT/mTOR inhibitors is generally later than CDK4/6 inhibitors, previous studies were usually stratified by treatment lines. We collected information on patients’ previous endocrine therapy in clinical trials and included it as a covariate in the model to eliminate the potential bias. However, by analyzing the data in this way, we did not find a significant impact of the treatment line on PFS and OS. Another point to note was that there are only data on OS for PI3K inhibitors as first-line treatment. Therefore, we could not conclude that CDK4/6 inhibitors were superior to all PAM pathway inhibitors. In terms of safety, CDK4/6 inhibitors have a fairly high incidence of neutropenia, and it is considered to be target related, because CDK6 plays a key role in hematologic precursor proliferation[4]. PI3K/AKT/mTOR inhibitors generally cause more diarrhea, stomatitis, vomiting, cough, rash, asthenia, and hepatotoxicity. In addition, hyperglycemia due to glucose metabolism dysregulation caused by PI3K/AKT/mTOR inhibitors is the most frequently reported adverse effect, which is considered an on-target side effect of these drugs that warrants prevention, monitoring, and treatment[16, 17]. As it showed that the efficacy and safety profile of these two types of drugs varies, the treatment selection depends on the individual situation. If the patient can tolerate severe hematological adverse events, CDK4/6 inhibitors is recommended as first-line treatment for ER+/HER2- metastatic breast cancer, and specific agent (Palbociclib, Ribociclib, Abemaciclib, and Dalpiciclib) should be carefully selected based on the benefit risk ratio of the drug.

Our study found that the proportion of patients with visceral metastases and the type of endocrine drugs in combination were the two most important factors affecting efficacy. Visceral metastases are common in postmenopausal women with ER+/HER2- metastatic breast cancer, and its presence not only indicates increased disease aggressiveness, but also significantly impacts treatment selection, prognosis, and patients' quality of life[18–20]. Our findings suggested that as the proportion of patients with visceral metastases increases, OS and PFS decrease significantly, and confirmed results from previous research and suggested that heterogeneity in the proportion of patients with visceral metastases should be considered in cross-study comparisons and sample size estimation in clinical trial design.

There is an ongoing clinical debate about the best endocrine therapy backbone[21]. Although current clinical guidelines recommend the use of AIs as the first-line and fulvestrant as the second-line, results from some previous studies favor the fulvestrant over AIs which warrants more evidence [22–25]. Our findings suggested that, whether with CDK4/6 or PI3K/AKT/mTOR inhibitors, combining with AIs demonstrates a better efficacy over fulvestrant and tamoxifen, which provides additional evidence for clinical guideline recommendations. Nevertheless, it is important to note that studies have suggested greater benefits from fulvestrant therapy in endocrine therapy-eligible patients with secondary AI resistance due to ESR1 mutations[26]. The presence of ESR1 mutations may have predictive value in determining whether CDK4/6 inhibitors combined with AIs or fulvestrant[27].

OS have always been considered as gold standard for efficacy endpoint, however, it usually requires long-time follow-up which cannot be obtained easily, therefore surrogate endpoints have been explored. A main surrogate endpoint to OS is PFS, which is captured more quickly and is not influenced by subsequent treatments[28]. However, PFS is often criticized because its improvement does not always presage improvement in OS[15]. Given this, we analyzed the correlations between early efficacy indicators (e.g., ORR and 0.5-year PFS) and clinical endpoints (e.g., 1-year PFS and 2-year OS). The results showed a moderate correlation between 1-year PFS and 2-year OS for ER+/HER2- metastatic breast cancer and this correlation was even more significant in CDK4/6 inhibitors group. Based on this, a 1-year PFS could be used as a potential surrogate endpoint for 2-year OS in ER+/HER2- metastatic breast cancer treatment. Conversely, a poor correlation was found between ORR and OS, which is consistent with previous studies as well as the current clinical consensus[29].

Despite our important findings, this study has certain limitations. The data used for analysis were obtained from literatures, which narrows the scope of covariate distribution, and diminishes the ability to effectively explore potential factors influencing efficacy. Critical variables such as race, prior treatments, and the proportion of patients with solely bone metastases—absent in over 30% of the included studies—could not be examined and, therefore could potentially bias the findings. Additionally, the heterogeneity of dosage was not considered due to data constraints. Finally, the exclusive inclusion of English-language studies may introduce publication bias.

In this study, we comprehensively compared the efficacy and safety of two types of targeted inhibitors combined with different endocrine therapy. In terms of PFS, OS and ORR, CDK4/6 inhibitors showed superior efficacy compared with PI3K/AKT/mTOR inhibitors, specifically CDK4/6 inhibitors combined with AIs showed the best efficacy among all the combination regimens. Therefore, for ER+/HER2- metastatic breast cancer patients, CDK4/6 inhibitors combined with AIs could be considered as first-line treatment, with the caution of hematological side effects.

Acknowledgements

We thank all staff of center of Drug Metabolism and Pharmacokinetics, China Pharmaceutical University.

Author contributions

Meiyu Pan and Yan Lin: conceptualization, data curation, formal analysis, investigation, writing—original draft, and writing—review and editing. Meiyu Pan and Yan Lin contributed equally to this work. Yinhui Liu: Software. Ruijuan Xu: conceptualization. Jin Yang: conceptualization and methodology. Jin Yang is the corresponding author of this article and Ruijuan Xu is the co-corresponding author.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflict of interest

The authors declare no competing interests.

Arnold M, Morgan E, Rumgay H et al (2022) Current and future burden of breast cancer: Global statistics for 2020 and 2040. Breast (Edinburgh, Scotland);66:15-23
AlFakeeh A, Brezden-Masley C (2018) Overcoming endocrine resistance in hormone receptor-positive breast cancer. Curr Oncol;25(Suppl 1):S18-S27
Burstein HJ (2020) Systemic Therapy for Estrogen Receptor-Positive, HER2-Negative Breast Cancer. N Engl J Med;383(26):2557-70
Asghar U, Witkiewicz AK, Turner NC et al (2015) The history and future of targeting cyclin-dependent kinases in cancer therapy. Nature Reviews Drug Discovery;14(2):130-46
Campbell RA, Bhat-Nakshatri P, Patel NM et al (2001) Phosphatidylinositol 3-kinase/AKT-mediated activation of estrogen receptor alpha: a new model for anti-estrogen resistance. The Journal of biological chemistry;276(13):9817-24
Presti D, Quaquarini E (2019) The PI3K/AKT/mTOR and CDK4/6 Pathways in Endocrine Resistant HR+/HER2- Metastatic Breast Cancer: Biological Mechanisms and New Treatments. Cancers (Basel);11(9)
Wang L, Da H, Li X et al (2020) Comparison of CD4/6 inhibitors, PI3K inhibitors, and mTOR inhibitors for patients with advanced HR+/HER2‐ Breast cancer: A network meta‐analysis of 19 RCTs. The breast journal;26(9):1862-6
Leung JH, Leung HWC, Wang SY et al (2021) Efficacy and safety of CDK4/6 and PI3K/AKT/mTOR inhibitors as second-line treatment in postmenopausal patients with hormone receptor-positive, HER-2-negative metastatic breast cancer: a network meta-analysis. Expert opinion on drug safety;20(8):949-57
Xu H, Wang Y, Han Y et al (2022) CDK4/6 inhibitors versus PI3K/AKT/mTOR inhibitors in women with hormone receptor-positive, HER2-negative metastatic breast cancer: An updated systematic review and network meta-analysis of 28 randomized controlled trials. Frontiers in oncology;12:956464
Marshall SF, Burghaus R, Cosson V et al (2016) Good Practices in Model‐Informed Drug Discovery and Development: Practice, Application, and Documentation. CPT: Pharmacometrics & Systems Pharmacology;5(3):93-122
Mould DR (2012) Model-Based Meta-Analysis: An Important Tool for Making Quantitative Decisions During Drug Development. Clinical Pharmacology & Therapeutics;92(3):283-6
Li T, Yu J, Hou M et al (2023) Quantitative evaluation of therapy options for relapsed/refractory diffuse large B-cell lymphoma: A model-based meta-analysis. Pharmacol Res;187:106592
Wu L, Chen J, Cai R et al (2023) Difference in Efficacy and Safety of Anti-CD19 Chimeric Antigen Receptor T-Cell Therapy Containing 4-1BB and CD28 Co-Stimulatory Domains for B-Cell Acute Lymphoblastic Leukemia. Cancers (Basel);15(10)
Gong Y, Sui Z, Lv Y et al (2023) LABA/LAMA versus LABA/ICS fixed-dose combinations in the prevention of COPD exacerbations: a modeling analysis of literature aggregate data. Eur J Clin Pharmacol;79(10):1321-32
Seidman AD, Maues J, Tomlin T et al (2020) The Evolution of Clinical Trials in Metastatic Breast Cancer: Design Features and Endpoints That Matter. American Society of Clinical Oncology educational book American Society of Clinical Oncology Annual Meeting;40:1-11
Tankova T, Senkus E, Beloyartseva M et al (2022) Management Strategies for Hyperglycemia Associated with the α-Selective PI3K Inhibitor Alpelisib for the Treatment of Breast Cancer. Cancers (Basel);14(7)
Nunnery SE, Mayer IA (2019) Management of toxicity to isoform α-specific PI3K inhibitors. Ann Oncol;30(Suppl_10):x21-x6
Harb WA (2015) Management of patients with hormone receptor-positive breast cancer with visceral disease: challenges and treatment options. Cancer Manag Res;7:37-46
Robertson JFR, Di Leo A, Johnston S et al (2021) Meta-analyses of visceral versus non-visceral metastatic hormone receptor-positive breast cancer treated by endocrine monotherapies. NPJ Breast Cancer;7(1):11
Rosa Mendoza ES, Moreno E, Caguioa PB (2013) Predictors of early distant metastasis in women with breast cancer. Journal of cancer research and clinical oncology;139(4):645-52
Spring LM, Wander SA, Andre F et al (2020) Cyclin-dependent kinase 4 and 6 inhibitors for hormone receptor-positive breast cancer: past, present, and future. Lancet (London, England);395(10226):817-27
Llombart-Cussac A, Pérez-García JM, Bellet M et al (2021) Fulvestrant-Palbociclib vs Letrozole-Palbociclib as Initial Therapy for Endocrine-Sensitive, Hormone Receptor-Positive, ERBB2-Negative Advanced Breast Cancer: a Randomized Clinical Trial. JAMA oncology;7(12):1791‐9
Slamon DJ, Neven P, Chia S et al (2018) Phase III Randomized Study of Ribociclib and Fulvestrant in Hormone Receptor-Positive, Human Epidermal Growth Factor Receptor 2-Negative Advanced Breast Cancer: MONALEESA-3. Journal of clinical oncology;36(24):2465‐72
Slamon DJ, Neven P, Chia S et al (2020) Overall Survival with Ribociclib plus Fulvestrant in Advanced Breast Cancer. New England journal of medicine;382(6):514‐24
Slamon DJ, Neven P, Chia S et al (2021) Ribociclib plus fulvestrant for postmenopausal women with hormone receptor-positive, human epidermal growth factor receptor 2-negative advanced breast cancer in the phase III randomized MONALEESA-3 trial: updated overall survival. Annals of oncology : official journal of the european society for medical oncology;32(8):1015‐24
Fribbens C, O'Leary B, Kilburn L et al (2016) Plasma ESR1 Mutations and the Treatment of Estrogen Receptor-Positive Advanced Breast Cancer. J Clin Oncol;34(25):2961-8
Brett JO, Spring LM, Bardia A et al (2021) ESR1 mutation as an emerging clinical biomarker in metastatic hormone receptor-positive breast cancer. Breast Cancer Res;23(1):85
Saad ED, Katz A, Hoff PM et al (2010) Progression-free survival as surrogate and as true end point: insights from the breast and colorectal cancer literature. Ann Oncol;21(1):7-12
Ritchie G, Gasper H, Man J et al (2018) Defining the Most Appropriate Primary End Point in Phase 2 Trials of Immune Checkpoint Inhibitors for Advanced Solid Cancers: A Systematic Review and Meta-analysis. JAMA oncology;4(4):522-8

No competing interests reported.

supplementalinformation.docx

Download PDF

Journal Publication

published 20 Jun, 2024

Read the published version in European Journal of Clinical Pharmacology →

Editorial decision: Revision requested
20 May, 2024
Reviews received at journal
20 May, 2024
Reviews received at journal
21 Apr, 2024
Reviewers agreed at journal
14 Apr, 2024
Reviewers agreed at journal
07 Apr, 2024
Reviewers invited by journal
07 Apr, 2024
Submission checks completed at journal
26 Mar, 2024
Editor assigned by journal
26 Mar, 2024
First submitted to journal
25 Mar, 2024

You are reading this latest preprint version

Quantitative evaluation of the efficacy and safety profiles of two types of targeted inhibitors combined with endocrine therapy in ER+/HER2- metastatic breast cancer

Status:

Journal Publication

Version 1

Abstract

Purpose

Methods

Results

Conclusion

Figures

Introduction

Methods

Results

Typical overall and progression-free survival simulation

Meta-Analysis for Secondary and Safety Outcomes

Correlation among ORR, PFS and OS

Discussion

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1