Economic value of ribociclib in early breast cancer in the United States

Abstract

BACKGROUND:

Many patients with hormone receptor–positive/human epidermal growth factor receptor 2–negative (HR+/HER2−) early breast cancer (eBC) experience recurrence despite treatment with adjuvant endocrine therapy; recurrences bear a substantial cost and can impact life expectancy and quality of life.

OBJECTIVE:

To estimate the lifetime clinical and economic value of ribociclib for the treatment of patients with HR+/HER2− stage II/III eBC at high risk of recurrence including select node-negative and all node-positive patients.

METHODS:

A 4-state Markov model comparing ribociclib+nonsteroidal aromatase inhibitor (NSAI) with NSAI alone estimated the expected lifetime costs and effectiveness for patients with eBC (mean age = 53 years). High risk of recurrence in HR+/HER2− eBC was based on the NATALEE trial criteria: all node-positive and node-negative patients with tumor size larger than 5 cm or 2-5 cm with Grade 3 or Grade 2 and high genomic risk. Model outcomes included invasive disease–free survival (IDFS), locoregional recurrence (LR), distant recurrence (DR), overall survival, adjuvant treatment costs, adverse event costs, disease management costs by health state, and DR treatment costs. Patient-level IDFS for the first 4.5 years observed in the NATALEE trial was parametrically extrapolated thereafter. DR and LR progression risks and treatment utilization patterns were obtained from the literature. Model outputs included total costs, quality-adjusted life-years (QALYs), life-years (LYs), equal-value LYs (evLYs), and the incremental cost-effectiveness ratio (ICER) of ribociclib+NSAI vs NSAI when using IDFS as the effectiveness outcome.

RESULTS:

Ribociclib+NSAI improved IDFS (13.08 vs 11.09 years) and reduced DR (DR state = 1.28 vs 2.35 years), generating QALY gains of 0.68 (ribociclib+NSAI = 11.45; NSAI = 10.77), LY gains of 0.69 (ribociclib+NSAI = 14.98; NSAI = 14.29), and evLY gains of 0.75 (ribociclib+NSAI = 11.52; NSAI = 10.77). Total lifetime incremental costs were $52,385 (ribociclib+NSAI = $831,433; NSAI = $779,047). The ICER was $77,448/QALY, and the probability of ribociclib+NSAI being cost-effective was 68.5% at a willingness to pay of $150,000/QALY.

CONCLUSIONS:

Based on model estimation, ribociclib+NSAI is projected to be a cost-effective therapy in the United States vs NSAI alone by reducing the rate of high-cost metastatic recurrences and increasing QALYs.

Plain language summary

Patients with hormone receptor–positive/human epidermal growth factor receptor 2–negative early breast cancer experience substantial risk of breast cancer recurrence more than 20 years after initial surgery. Ribociclib, a treatment that targets cyclin-dependent kinases 4 and 6, reduces the risk of cancer returning when given with a nonsteroidal aromatase inhibitor (NSAI). Although adding ribociclib to breast cancer therapy costs more at the time of treatment, these costs are partly balanced by avoiding the downstream costs of future recurrences.

Implications for managed care pharmacy

Formulary decisions based primarily on pharmacy costs may favor NSAI monotherapy over ribociclib plus NSAI. However, this analysis demonstrates that combination therapy reduces recurrence risk, with delayed recurrence generating downstream medical cost offsets. When both costs and outcomes are considered, the additional health benefits of ribociclib plus NSAI are achieved at cost-effectiveness levels consistent with commonly accepted US payer thresholds. These findings support favorable formulary positioning of ribociclib plus NSAI based on a comprehensive assessment of clinical benefit and economic value.

In the United States, breast cancer represents 15.2% of all new cancer diagnoses.¹ Hormone receptor–positive (HR+), human epidermal growth factor receptor 2–negative (HER2−) breast cancer is the largest subtype of breast cancer, accounting for 69% of cases.¹ Breast cancer is also the second leading cause of cancer-related deaths among women.¹ In 2023, it was estimated that 297,790 women were diagnosed with breast cancer and that 43,170 died, accounting for 7.1% of all cancer-related deaths.¹ In recent years, patient outcomes for all stages of breast cancer have improved, with a 5-year survival rate of 91.7%.^2,3

In the United States, approximately 90% of HR+/HER2− breast cancer cases are diagnosed at an early stage. Treatment options for early breast cancer (eBC), comprising stage 0 (noninvasive) and stages I-III (early invasive cancer), include surgery, postoperative radiation therapy, and pre- and/or postoperative systemic therapy.^2,4,5 Both the National Comprehensive Cancer Network (NCCN) and European Society for Medical Oncology (ESMO) recommend definitive surgery for patients with operable eBC followed by adjuvant systemic therapy.^2,4 Expression of biomarkers in HR+/HER2− eBC can aid clinicians in selecting the optimal adjuvant systemic therapy based on the risk of recurrence and treatment goals.⁶ Regarding endocrine therapy, the NCCN Clinical Practice Guidelines in Oncology (NCCN guidelines) for premenopausal women specify 5 years of tamoxifen (±ovarian suppression [OFS]) or an aromatase inhibitor (AI)+OFS, with the option to extend tamoxifen to 10 years if still premenopausal. Patients with lymph node involvement may benefit from extended AI duration of 7.5-10 years total. For postmenopausal women, 5 years of AI or sequential tamoxifen and AI (2-3 years each) is preferred, and therapy can be extended up to 10 years using either agent or both sequentially.

Ribociclib is a cyclin-dependent kinase 4 and 6 inhibitor (CDK4/6i) first approved by the US Food and Drug Administration (FDA) in 2017 for use in combination with an AI for the treatment of postmenopausal women with HR+/HER2− advanced or metastatic breast cancer.⁷ In 2024, ribociclib received FDA approval for stage II and III HR+/HER2- eBC at high risk of recurrence.⁷ At a median follow-up of 55.4 months in the most recent updated analysis with 5-year outcome estimates, the clinical benefit of ribociclib+nonsteroidal AI (NSAI) was larger at 5 years than at 3 years (absolute risk reduction in invasive disease–free survival of 4.5% vs 2.7%).⁸ The efficacy benefit was consistent across subgroups and secondary endpoints, including distant disease–free survival and distant relapse–free survival. Overall survival data were immature but favored ribociclib+NSAI (hazard ratio [HR] = 0.80, 95% CI = 0.637-1.003).

Given that longer-term efficacy data from the NATALEE trial are lacking, the cost-effectiveness of ribociclib+NSAI vs NSAI alone in eBC remains uncertain. This analysis modeled the long-term cost-effectiveness of adjuvant ribociclib+NSAI for patients with HR+/HER2− stage II or stage III eBC from a US payer perspective. Additionally, we also assessed the long-term cost-effectiveness of ribociclib+NSAI using alternate measures of health benefit and a modified societal perspective, as Centers for Medicare & Medicaid Services (CMS) guidance notes the Medicare Drug Price Negotiation program (1) cannot use quality-adjusted life-years (QALYs) and (2) may use broader societal impacts (eg, productivity or caregiver burden) as part of the Medicare Drug Price Negotiation Program.⁹

Methods

MODEL OVERVIEW

This analysis was conducted in Microsoft Excel and employed a 4-state Markov model to evaluate the long-term clinical and economic outcomes associated with reductions in locoregional and distant recurrences caused by ribociclib+NSAI treatment vs NSAI alone. All patients entered the model in the invasive disease–free state (IDFS) and in each cycle could progress to locoregional recurrence (LR) or distant recurrence (DR), remain in their current health state, or die (Figure 1). The IDF state included patients who had no occurrence of invasive recurrence or death after treatment. The LR state included patients who experienced recurrence in the same breast, lymph nodes, or chest wall. The DR state included patients who experienced recurrence in other organs or a second primary cancer. This model structure was selected following previous published eBC models^10,11 as well as following key endpoints observed in the NATALEE trial.^12,13 The model used monthly cycles, and patients (mean age = 53 years) entered the model and were followed over lifetime (or until reaching 100 years) following health technology assessment guidelines.^14,15 A list of additional modeling assumptions are available in Supplementary Table 1 ^{(931.4KB, pdf)} (available in online article).

FIGURE 1.

Model Structure

The target population was adults aged at least 18 years with HR+/HER2− stage II and III eBC at high risk of recurrence. High risk of recurrence in HR+/HER2− eBC was defined as all node-positive and node-negative patients with tumor size greater than 5 cm or 2-5 cm with grade 3 or grade 2 and high genomic risk based on the efficacy and safety of ribociclib in combination with NSAI as the adjuvant treatment in the phase 3 multicenter, randomized, open-label NATALEE trial (NCT03701334).¹³ The target population entered the model with a mean weight of 72.3 kg and was assumed to be 99.6% female following pooled estimates of the NATALEE trial intention-to-treat population in the ribociclib+NSAI and NSAI arms.¹³

Treatments included in the analysis were consistent with the investigational arm of the NATALEE trial.

The intervention was ribociclib (400 mg) administered orally once daily for 3 weeks followed by a 1-week break in a 28-day cycle for up to 36 months, disease recurrence, or unacceptable toxicity, whichever occurred first, in combination with an NSAI (letrozole administered orally 2.5 mg once daily or anastrozole administered orally 1.0 mg once daily for ≥5 years) and goserelin (3.6 mg subcutaneous injection) once every 4 weeks, for premenopausal women and men.⁷ The comparator was NSAI (letrozole administered orally 2.5 mg once daily or anastrozole administered orally 1.0 mg once daily) for at least 5 years and for premenopausal women and men, goserelin (3.6 mg subcutaneous injection) once every 4 weeks for up to 60 months, disease recurrence, or unacceptable toxicity, whichever occurred first.^16–18

Clinical and survival outcomes included total LYs, LY gains by health state, and number of DR events avoided. Patient utility outcomes included total QALYs and equal-value LYs (evLYs), QALY and evLY gain by health state, and disutility because of adverse events (AEs). Cost outcomes included adjuvant treatment costs, AE costs associated with adjuvant ribociclib+NSAI and comparator, LR treatment costs, DR treatment costs, disease management costs, terminal care costs, and total costs. For analysis from a modified societal perspective, costs of patient market productivity losses were also considered. LR and DR treatment costs included surgical costs and systemic therapy costs for recurrent disease, whereas disease management costs included monitoring and health care resource utilization costs. Health economic outcomes included the incremental total costs and the incremental cost-effectiveness ratio (ICER) assessed when measuring patient utility gains as QALYs, LYs, and evLYs.

MODEL INPUTS

The model included clinical, utility, and financial parameters (patient and modified societal). Assumptions relied on in the base-case analysis are available in Supplementary Table 1 ^{(931.4KB, pdf)} . The analysis relied on NATALEE data to inform treatment duration for both ribociclib+NSAI and NSAI (Supplementary Figure 1 ^{(931.4KB, pdf)} ).

CLINICAL INPUTS

The primary efficacy, safety, and utility inputs were derived from the phase 3 NATALEE trial. At the data cutoff (April 29, 2024) 4-year IDF survival rates were 88.5% in the ribociclib+NSAI arm vs 83.6% in the NSAI arm only (HR = 0.715, 95% CI = 0.609-0.840).¹³ Median follow-up IDF survival was 44.2 months.

The hazards of leaving the IDFS were modeled using patient-level data from the NATALEE trial until model year 4.5, after which log-logistic and Weibull parametric models were fit to extrapolate IDF survival (Table 1; Supplementary Figures 2 and 3 ^{(931.4KB, pdf)} ).¹³ Parametric extrapolations were selected following National Institute for Health and Care Excellence guidance on the basis of the minimum Akaike information criterion and were validated using visual alignment between the predicted curve and observed data.¹⁹ For external validation and clinical plausibility, a treatment waning effect was considered for ribociclib+NSAI, which was assumed to be linear and started at 10 years and ended at 33 years. The starting time was based on clinical expert opinion and results of the “Arimidex, Tamoxifen, Alone or in Combination” (ATAC) trial (median follow-up = 10 years) that showed lasting IDFS benefit of anastrozole vs tamoxifen up to 8 years, and the ending time was the time point when the IDFS rate of the NSAI arm equaled the mortality of the general population. Breast cancer recurrence events were modeled as the product of the hazard of leaving IDF and the distribution of the type of breast cancer reoccurrence event (LR vs DR) observed in the NATALEE trial (Table 1).²⁰ The transition probability from LR to DR was estimated based on the literature, where a median time until progression of 7.6 years was reported.²¹ Accordingly, the monthly transition probabilities were derived using the median, assuming an exponential distribution with a constant rate of transition. In the DR state a weighted average of the median overall survival was calculated based on the data in relevant metastatic BC trials for a heterogeneous mix of treatments in the metastatic setting (Table 1).^22–32 Then, the monthly transition probabilities for DR to death were derived using the median, assuming exponential distribution with a constant mortality rate.

TABLE 1.

Model Inputs

Category	Parameter	Value	Source
Clinical inputs
	IDF Weibull parameters
	Slope	310.897	13
	Scale	1.097	13
	IDF log-logistic parameters
	Shape	203.119	13
	Scale	1.139	13
	Distribution of first IDF events, %
	Ribociclib+NSAI
	Locoregional recurrence	18.6	20
	Distant recurrence	74.3	20
	Death	7.2	20
	NSAI
	Locoregional recurrence	20.9	20
	Distant recurrence	75.7	20
	Death	3.4	20
	Median OS after experiencing DR, months
	Ribociclib+NSAI	27.8	22 – 32
	NSAI	44.0	22 – 32
Utility
	Patient HRQoL
	IDF on treatment: ribociclib+NSAI	0.811	13
	IDF on treatment: NSAI	0.806	13
	IDF off treatment	0.764	13
	LR	0.728	13
	DR	0.644	13
	Disutility because of AEs
	Alanine aminotransferase increased	−0.115	34
	Neutropenia	−0.150	34
Financial
	Monthly drug acquisition cost
	Ribociclib+NSAI	$14,329	35 , 59
	NSAI	$438	17 , 35
	Administration cost per administrationa
	Ribociclib+NSAI	$34	45
	NSAI	$33	45
	LR treatment costs
	NSAI	$8	17 , 35
	Surgery	$41,843	34
	Total DR treatment costs
	Ribociclib+NSAI entry cost	$132,349	Calculated
	NSAI	$327,020	Calculated
	Monthly disease management costs
	IDF (years 1-3) for ribociclib+NSAI	$435	45
	IDF (years 1-3) for NSAI	$261	45
	IDF (years 4+)	$261	45
	LR	$5,444	37
	DR	$17,440	38
	Additional DR costs for cyclin-dependent kinase	$435	Assumption
	Terminal care cost	$13,754	39; assumption
	AE costs
	Alanine aminotransferase increased	$9,805	46
	Neutropenia	$17,308	46

^a Administration costs were only applied to patients receiving goserelin for ovarian suppression along with endocrine therapy.

AE = adverse event; DR = distant recurrence; HRQoL = health-related quality of life; IDF = invasive disease–free; LR = locoregional recurrence; NSAI = nonsteroidal aromatase inhibitor; OS = overall survival.

DR Treatment Assumptions After Adjuvant CDK 4/6i Therapy

Currently, no established guidelines exist for treatment pathways following recurrence after adjuvant CDK4/6i therapy, and the literature addressing this clinical scenario is limited. Therefore, we relied on clinical expert opinion and published evidence on post-CDK4/6i treatment strategies in the HR+/HER2− metastatic breast cancer setting to develop a framework for potential treatment options, including considerations for re-treatment with a CDK4/6i, following DR.³³ This framework was based on 2 key factors: the type of adjuvant therapy received and the timing of disease recurrence (Supplementary Figure 4 ^{(931.4KB, pdf)} ). Accordingly, it was assumed that a proportion of patients entering DR state would be re-treated with CDK4/6is using clinical criteria. In the ribociclib+NSAI arm, patients with DR longer than 12 months after completing ribociclib treatment but within 12 months of completing NSAI could receive CDK4/6i+fulvestrant. If recurrence occurred more than 12 months after completion of NSAI, rechallenge could include CDK4/6i+fulvestrant or NSAI. In the NSAI arm, patients could be re-treated with CDK4/6i+fulvestrant for recurrence within 12 months of NSAI completion and CDK4/6i+fulvestrant or NSAI for recurrence after more than 12 months.

HEALTH UTILITY INPUTS

Health state utilities were derived from patient responses to the EuroQol-5 Dimension 5-Level questionnaire collected in the NATALEE Trial (Table 1).¹³ Utility index scores were assessed for each health state (ie, IDF, LR, and DR). For IDF, treatment-specific utilities were considered when patients were receiving treatment. When patients discontinued the treatment or transitioned to LR or DR states, non–treatment-specific utilities were considered. Death was the absorbing state with utility 0. The model applied AE-specific utility decrements to each treatment defined as the product of event-specific disutility, incidence in the NATALEE trial, and an assumed duration of 1 month. The model considered 2 AEs of at least grade 3 with a prevalence rate greater than or equal to 5% observed in either arm of the NATALEE trial: alanine aminotransferase increase and neutropenia and event-specific disutility values were derived from a survey of patients with metastatic breast cancer.³⁴

COST INPUTS

The model included treatment, disease management, health care resource utilization, and societal costs. Monthly adjuvant treatment acquisition and administration costs were derived from RED BOOK and the CMS Physician Fee Schedule, respectively (Table 1).^35,36 LR treatment costs included costs of surgery and NSAI. DR treatment costs were calculated as the weighted average of 6 categories of DR treatment composed of 14 unique treatment regimens observed in clinical trials and routine practice, treatment market share estimates, and duration of each treatment (Table 1; Supplementary Tables 1 and 2 ^{(931.4KB, pdf)} ).^{7,28–32,34,35,37–44} In the IDFS, disease management costs were estimated by applying the CMS Physician Fee Schedule to chemotherapies and endocrine therapies recommended in NCCN guidance (Table 1).^4,45 In the LR and DR states, disease management costs including inpatient hospitalization, outpatient visits, and emergency department visits were obtained from the literature.^37,38 The model included additional monitoring utilization costs of electrocardiogram tests associated with ribociclib treatment in each health state following ribociclib’s FDA label and monitoring cost estimates, which were retrieved from the Physician Fee Schedule (Supplementary Tables 3 and 4 ^{(931.4KB, pdf)} ).^7,45 Costs of terminal care were applied for 1 cycle using estimates in the literature.³⁹ Costs associated with AEs were derived from the Healthcare Cost and Utilization Project.⁴⁶ Costs of patient productivity losses associated with each health state (ie, IDF, LR, and DR) because of breast cancer were calculated using wage and employment data from the Bureau of Labor Statistics, and the number of working days per month lost observed among patients with breast cancer obtained from the literature (Supplementary Table 5 ^{(931.4KB, pdf)} ).^47,48 All costs were inflated to 2024 US dollars with the exception of drug unit acquisition costs for CDK4/6i, which relied on 2025 wholesale acquisition cost data to reflect the latest pricing. All costs and utilities were discounted at a constant rate of 3%.

A deterministic, 1-way sensitivity analysis that varied individual parameters by their reported 95% CI or ±15% was conducted. The deterministic sensitivity analysis included 23 additional clinical scenarios of interest. A probabilistic sensitivity analysis included a Monte Carlo simulation with 1,000 iterations of key parameters. A subgroup analysis of the high-risk node-negative (N0) subgroup in the NATALEE trial (defined as N0 with tumor size ≥5 cm or 2-5 cm with grade 3, or grade 2 disease with either Ki-67 index score ≥20% or high genomic risk) was conducted from payer and modified societal perspectives. An additional subgroup analysis of a cohort of patients aged at least 65 years was conducted from a payer perspective. Productivity losses were not considered for this subgroup because the starting age exceeded the modeled retirement age used to compute productivity losses. Finally, to better approximate real-world scenarios in which drug list prices are reduced following loss of exclusivity and new cohorts of patients with eBC are diagnosed each year, a scenario analysis was conducted incorporating genericization and stacked cohorts.⁴⁹ Genericization, following published estimates of price reductions observed in oral cancer therapies,⁵⁰ was modeled as a 60% reduction in ribociclib’s list price beginning in model year 6, following loss of exclusivity dates published in the FDA Orange Book.⁵¹ A new eBC cohort entered the model each cycle, resulting in 566 total cohorts over the 47.2-year horizon.

Results

Over a typical patient’s lifetime horizon, treatment with ribociclib+NSAI was projected to prolong disease-free survival and reduce eBC morbidity. Patients treated with ribociclib+NSAI would experience, on average, an 8.28-month gain in LYs (0.69 LYs gained, 14.98 vs 14.29) (Figure 2; Supplementary Figure 5 ^{(931.4KB, pdf)} ). In addition to an improvement in LYs, patients spent more time in higher quality of life health states. Specifically, patients spent more time in the IDFS (13.08 LYs vs 11.09 years) (Supplementary Figure 5 ^{(931.4KB, pdf)} ) and consequently less time in the LR (0.62 LYs vs 0.85 years) and DR (1.28 LYs vs 2.35 years) health states.

FIGURE 2.

Utility Benefits of Ribociclib

evLY = equal-value life-year; LY = life-year; NSAI = nonsteroidal aromatase inhibitor; QALY = quality-adjusted life-year.

Clinical improvements because of ribociclib+NSAI would generate lifetime utility gains for patients with eBC relative to NSAI treatment alone. Patients treated with ribociclib would experience 0.68 incremental QALYs (11.45 vs 10.77) compared with NSAI alone (Figure 2). Disutility because of AEs had minimal impact on lifetime utility (−0.0064 QALYs vs −0.0003 QALYs) because of ribociclib demonstrating a favorable safety profile in the NATALEE trial (Supplementary Table 6 ^{(931.4KB, pdf)} ). In the model, QALY gains were primarily driven by additional time spent in the IDFS (1.53 QALYs; 10.17 vs 8.64) and less time spent in the LR (−0.17 QALYs; 0.45 vs 0.62) and DR (−0.69 QALYs; 0.82 vs 1.51) health states. When evaluating utility gains using evLYs, the utility benefit of ribociclib+NSAI was 10.3% higher than standard QALYs (0.75 evLY; 11.52 vs 10.77) (Figure 2).

From a US payer perspective, ribociclib+NSAI treatment incurred higher adjuvant treatment costs, which were substantially offset by reductions in LR treatment costs, DR treatment costs, and disease management costs. Adjuvant treatment costs made up the largest share of total cost for ribociclib+NSAI, increasing lifetime costs by $385,716 ($408,360 vs $22,644) (Table 2). Reductions in the likelihood of LR and DR among ribociclib+NSAI–treated patients generated $1,280 ($3,821 vs $5,101) and $112,143 ($55,043 vs $167,186) in cost offsets, respectively. Reductions in eBC morbidity among those treated with ribociclib+NSAI produced $227,833 ($347,862 vs $575,695) in disease management cost offsets. Disease management offsets were driven by fewer ribociclib+NSAI–treated patients experiencing DR, leading to lower use of doctor visits, emergency department/inpatient visits, and procedures and monitoring tests in the metastatic setting (Table 2; Supplementary Figure 6 ^{(931.4KB, pdf)} ). DR treatment cost offsets in the intervention arm were attributed to lower use of CDK4/6i+NSAI, CDK4/6i+fulvestrant, and other targeted therapies in the metastatic setting. As a result, the lifetime incremental cost of ribociclib+NSAI treatment was projected to be $52,385 ($831,433 vs $779,047). From a modified societal perspective, ribociclib+NSAI would be associated with a reduction in patient market productivity losses of $11,736 ($130,277 vs $142,012) on average. This additional offset reduced lifetime incremental cost of ribociclib+NSAI from $52,385 using a payer perspective to $40,650 ($961,709 vs $921,060) from a modified societal perspective. The incremental cost of ribociclib+NSAI represented a smaller percentage increase when calculated on an annual basis vs over lifetime as increases in lifetime costs were partially driven by improvements in LYs. From a payer perspective, annualized incremental costs represent a 1.8% increase compared with NSAI ($55,494 vs $54,516) vs a 6.7% lifetime increase ($831,433 vs $779,047) (Table 2; Supplementary Table 7 ^{(931.4KB, pdf)} ). From a modified societal perspective, the annualized cost of ribociclib+NSAI offered potential cost savings of $263 per LY ($64,190 vs $64,453) (Supplementary Table 7 ^{(931.4KB, pdf)} ).

TABLE 2.

Lifetime Cost Outcomes per Patient

Costs	RIB+NSAI, 2024 USD	NSAI, 2024 USD	RIB+NSAI vs NSAI, 2024 USD
Adjuvant treatment costs	408,360	22,644	385,716
Locoregional recurrence treatment costs	3,821	5,101	−1,280
Distant recurrence treatment costs	55,043	167,186	−112,143
Disease management costs	347,862	575,695	−227,833
Terminal care costs	7,817	8,099	−281
Adverse event costs	8,528	322	8,205
Indirect costs (societal perspective only)	130,277	142,012	−11,736
Total costs (payer perspective)	831,433	779,047	52,385
Total costs (societal perspective)	961,709	921,060	40,650

NSAI = nonsteroidal aromatase inhibitor, RIB = ribociclib; USD = US dollars.

In the base-case analysis, the use of ribociclib was cost-effective under all patient utility gain metrics and model perspectives using a willingness-to-pay (WTP) threshold of $150,000 per QALY/LY/evLY. From a payer perspective, ribociclib+NSAI had an ICER of $77,448 per QALY gained and $75,710 per LY gained (Figure 3). The cost per additional evLY was $69,934. From a modified societal perspective, ribociclib+NSAI had an ICER of $60,098 per QALY gained and $58,749 per LY gained owing to improvements in patient marketplace productivity. The cost per additional evLY was $54,267.

FIGURE 3.

Incremental Cost-Effectiveness Ratio Results of Ribociclib+Nonsteroidal Aromatase Inhibitor vs Nonsteroidal Aromatase Inhibitor Alone

evLY = equal-value life-year; LY = life-year; QALY = quality-adjusted life-year; USD = US dollars.

The modeled ribociclib+NSAI cost-effectiveness results were sensitive to variations in monthly transition probability from DR to death for NSAI, followed by monthly disease management costs in DR state, and monthly transition probability from DR to death for ribociclib+NSAI (Supplementary Figure 7 ^{(931.4KB, pdf)} ). Across 1-way sensitivity analyses, the incremental cost per QALY gained ranged from $28,107 (monthly disease management costs in DR state) to $127,279 (monthly transition probability from DR to death in the NSAI arm). Among the 23 modeled clinical scenarios, the ICER ranged from $5,087 (no treatment waning for ribociclib) to $248,883 (annual discount rate = 5%). The discount rate was highly impactful because of the long survival duration of most patients with eBC; in fact, ribociclib+NSAI dominated NSAI treatment when the discount rate was 0%. Ribociclib+NSAI also dominated NSAI if one were not allowed to rechallenge with CDK4/6i or when DR treatment utilization rates were derived from a survey of physicians. Using an exponential model to extrapolate IDFS for both arms yielded an ICER of $180,102. However, this model assumes a constant hazard rate over time, which is considered less clinically plausible given non–treatment-related factors influencing recurrence risk. Scenario analyses exploring alternative assumptions for the timing and duration of treatment waning yielded ICERs below the $150,000 threshold (Supplementary Table 8 ^{(931.4KB, pdf)} ). In 1,000 iterations of the probabilistic scenario analysis, the mean incremental cost per QALY was $72,800 for ribociclib+NSAI (Supplementary Figure 8 ^{(931.4KB, pdf)} ). At a cost-effectiveness threshold of $150,000 per QALY, the probability of ribociclib+NSAI being cost-effective vs NSAI alone was 68.5%. In an analysis of the high-risk N0 subgroup of the NATALEE trial, the cost per additional QALY from a payer perspective was $4,886, the cost per additional LY was $4,325, and the cost per additional evLY was $4,396 (Supplementary Figure 9 ^{(931.4KB, pdf)} ). From a modified societal perspective, ribociclib+NSAI dominated NSAI alone across all measures of patient utility in the N0 subgroup. In a subgroup analysis of patients aged at least 65 years, the ICER for ribociclib+NSAI compared with NSAI was $2,294 per QALY and $2,078 per evLY (Supplementary Table 9 ^{(931.4KB, pdf)} ). Finally, when accounting for genericization, treatment with ribociclib+NSAI dominated NSAI alone resulting in 0.36 QALYs gained (6.10 vs 5.74) and an incremental cost of −$38,381 ($325,376 vs $363,757) (Supplementary Table 10 ^{(931.4KB, pdf)} ).

Discussion

Base-case findings demonstrate that ribociclib+NSAI is a cost-effective treatment strategy compared with NSAI alone for the treatment of HR+/HER2− stage II and III eBC at high risk of recurrence at acceptable WTP thresholds. Ribociclib+NSAI produced 0.68 additional QALYs and 8.28 months of additional LYs compared with NSAI alone. The use of ribociclib+NSAI incurred $52,385 in additional lifetime costs from a US payer perspective and $40,650 in additional lifetime costs from a modified societal perspective. Cost outcomes were driven by higher adjuvant treatment costs and reduced by LR, DR, treatment, and disease management cost offsets. The cost per additional QALY of ribociclib+NSAI was $77,448 from a US payer perspective and $60,098 from a modified societal perspective. Ribociclib+NSAI treatment is also cost-effective when using non-QALY measures of health benefit (eg, LY or evLYG). Subgroup analysis of the high-risk N0 subgroup in the NATALEE trial showed that ribociclib+NSAI is a dominant strategy from a modified societal perspective, and the cost per additional QALY was $4,886 from a payer perspective. Subgroup analysis of patients aged at least 65 years estimated ribociclib+NSAI to have an ICER of $2,294 per QALY compared with NSAI alone. When incorporating price reductions because of genericization and stacked cohorts, ribociclib+NSAI dominated NSAI alone. The base-case results were robust to univariate changes in key parameters, with no individual parameter variation resulting in an ICER that exceeded the WTP threshold of $150,000/QALY. The probability of ribociclib+NSAI being cost-effective at a WTP threshold of $150,000 per QALY was 68.5%.

The incremental cost and utility outcomes found in our analysis are more favorable compared with findings of a conference abstract by Potnis et al that reported an ICER of $1.2 million per evLY with the addition of ribociclib to endocrine therapy.⁵² In the Potnis et al modeling analysis, higher evLYs among treated patients and lower lifetime incremental costs were because of relying on a 10-year model horizon, which is 4 years shorter than the mean survival in the comparator arm of this analysis. As a result, that model did not capture the long-term survival benefits of ribociclib as well as the higher lifetime costs of treating recurrent disease in the NSAI arm. This finding yields key implications for managed care decision-making because formulary decisions based primarily on pharmacy costs may favor NSAI, but this analysis demonstrated that combination therapies that reduce recurrence risk produce downstream cost offsets over the patient’s lifetime. Despite uncertainty surrounding real-world treatment duration, adherence, and uptake for ribociclib+NSAI in HR+/HER2− eBC, reducing disease recurrence in early oncology has been shown to produce downstream cost offsets across a wide variety of tumor types.^53–56

LIMITATIONS

This modeling analysis is subject to several limitations. First, the results of this study rely on parametric extrapolations of patient outcomes observed in the NATALEE trial, at median follow-up of 44.2 months at the data cutoff.^8,20 Although the efficacy extrapolation assumptions appear clinically plausible based on ATAC and updated NATALEE 5-year results, and are qualitatively supported by longer-term evidence from other CDK4/6i studies (eg, sustained curve separation observed in monarchE over 7 years), longer-term efficacy data from NATALEE will be required to validate the extrapolated findings.^8,57,58 Sensitivity to parametric models of eBC progression was rigorously assessed through deterministic and probabilistic sensitivity analyses. There is limited real-world evidence in the literature on the use and sequencing of treatments in the metastatic setting after adjuvant ribociclib treatment. The distribution of treatment options in the DR state was based on the available literature, and sequencing was based on clinical expert opinion. Further validation with real-world evidence is warranted. This study assumed that the health-related quality of life estimates for IDF survival were treatment-specific, which also needs to be further evaluated in the real world. Also, this study considered a modified societal perspective that included patient productivity losses because of eBC, but there may be other societal benefits from targeted treatments in the adjuvant setting, including nonmarket productivity losses, caregiver burden, and direct nonmedical costs.⁴⁹ Finally, ribociclib is expected to lose patent exclusivity within the modeled horizon, which will likely lead to reductions in acquisition prices because of generic entry.⁴⁹ We conducted a scenario analysis that incorporated price reductions because of genericization and stacked cohorts.

Conclusions

This health economic analysis assessed the cost-effectiveness of ribociclib+NSAI vs NSAI for the treatment of HR+/HER2− stage II and stage III eBC at high risk of recurrence. The effectiveness of ribociclib+NSAI compared with NSAI led to greater QALYs gained on average. Higher initial adjuvant treatment costs were substantially offset by the reduced treatment costs and disease management costs in the LR and DR states. Model findings demonstrate that ribociclib+NSAI is a cost-effective strategy compared with NSAI alone at a WTP threshold of $150,000/QALY.

Disclosures

Dr Brufsky is a consultant with AstraZeneca, Pfizer, Novartis, Lilly, Genentech/Roche, Daiichi Sankyo, Merck, Agendia, Sanofi, Puma Biotechnology, Myriad, Gilead, and BriaCell and has received research support from Agendia and AstraZeneca. Dr Tolaney reports a consulting or advisory role for Novartis, Pfizer, Merck, Lilly, AstraZeneca, Genentech/Roche, Eisai, Sanofi, Bristol Myers Squibb, Seagen, CytomX Therapeutics, Daiichi Sankyo, Gilead, OncXerna, Zymeworks, Zentalis, Blueprint Medicines, and Reveal Genomics. Dr Gadi is a founder (equity) of SEngine Precision Medicine (acquired by TempusAI), Novilla, XTRemisBio, and 3rdEyeBio; scientific board member for Puma Biotechnology and Phoenix Molecular Designs; member of speakers bureau of or consultant for and Hologic, Pfizer, Novartis, AstraZeneca, Stemline, Gilead, Daiichi Sankyo, and Genentech/Roche; and recipient of research support to his institution from Agendia, Gilead, Novartis, AstraZeneca, Glaxo-Smith-Kline, and Illimuna. Ms Alexander, Dr Bollu, Dr Ganapathy, Dr Patel, and Mr Sopher are employees and stockholders of Novartis Pharmaceuticals Corporation. Dr Chai, Mr Chen, Ms Ma, Dr Yang, and Dr Zhou are employees of Analysis Group, Inc., which has received consulting fees from Novartis for the conduct of this research, as has Dr Garrison Jr. Dr Sparano reports a consulting or advisory role for Novartis, Genentech, Pfizer, AstraZeneca, Delphi Diagnostics, and Precise DX.

This work was supported by Novartis, which also provided financial support for medical editorial assistance.