Budget impact analysis of revumenib for the treatment of relapsed or refractory acute leukemias with a KMT2A translocation in the United States

Abstract

BACKGROUND:

Acute leukemias (ALs), including acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL), are heterogeneous diseases characterized by different phenotypic, genetic, and molecular alterations that can guide treatment decisions. ALs harboring lysine methyltransferase 2A gene translocation (KMT2t), previously known as mixed-lineage leukemia, are associated with high rates of relapsed or refractory (R/R) disease. Revumenib, a first-in-class oral menin inhibitor, has shown improved clinical outcomes in patients with R/R KMT2At ALs.

OBJECTIVE:

To estimate, using a budget impact model (BIM), the financial impact of introducing revumenib for the treatment of adult patients with R/R KMT2At ALs on the formulary of a hypothetical US 1-million-member commercial health plan.

METHODS:

The BIM compared scenarios with or without revumenib and the resulting impact on commercial US third-party payers over a 3-year time horizon. Although no other therapies specifically targeted for R/R KMT2At ALs were approved during BIM development, 11 additional pharmacotherapies for R/R ALs (5 for AML and 6 for ALL, not including revumenib) were included as treatment options in the model. Clinical data included adverse event (AE) rates, duration of treatment, time to subsequent treatment, and survival outcomes. Cost inputs (USD 2024) included in the model comprised drug acquisition and administration, grade 3 or greater AEs, treatment-related supportive care and monitoring, subsequent treatment, and end-of-life costs. The differential cost per member per month (PMPM) was estimated. One-way sensitivity analyses varying the costs of drug acquisition and toxicity by ±20% and scenario analyses varying uptake of revumenib and epidemiology inputs, as well as excluding costs related to supportive care and posttreatment discontinuation, were performed.

RESULTS:

An estimated 1.7 adult patients (AML, 1.1; ALL, 0.6) were treatment eligible annually. Estimated 3-year total plan costs without and with revumenib were $2,146,564 and $2,126,919, respectively, for savings of −$19,646. Including revumenib was estimated to yield a differential PMPM cost of −$0.0005 over 3 years. The total number of grade 3 or greater AEs was lower over 3 years (10.82 vs 10.99, respectively) in the plan with revumenib vs without. Sensitivity and scenario analyses validated the robustness of the model.

CONCLUSIONS:

The BIM demonstrated that including revumenib in a formulary for adult patients with R/R KMT2At ALs was approximately cost neutral, offering patients access to a targeted treatment with potential for improved clinical outcomes.

Plain language summary

Revumenib, US Food and Drug Administration approved in 2024, has shown improvements for patients with certain types (KMT2At) of relapsed/refractory acute leukemias. A budget impact model (BIM) was developed to estimate the financial impact of adding revumenib as a treatment option for adults with relapsed/refractory acute leukemias on a US-based health plan. The results showed revumenib was approximately cost neutral, potentially offering patients a new treatment option with improved outcomes.

Implications for managed care pharmacy

The BIM assessed a hypothetical US 1-million-member commercial health plan with/without revumenib over 3 years. Annually, an estimated 1.7 adults were treatment eligible. The estimated 3-year total plan costs without ($2,146,564)/with ($2,126,919) revumenib resulted in savings of −$19,646. Overall, the BIM demonstrated that including revumenib in a formulary for adult patients with relapsed/refractory KMT2At acute leukemias was approximately cost neutral, offering patients access to a targeted treatment with potential for improved clinical outcomes.

Acute leukemias (ALs), including acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL), are heterogeneous diseases characterized by different phenotypic, genetic, and molecular alterations that can guide prognostication and treatment decisions. ^{1 – 5} The estimated incidence of AML and ALL in the United States for 2024 was 20,800 and 6,550, respectively. ^{4 , 5} Relapsed (ie, return of leukemia cells in the marrow in patients who were previously in remission) and refractory (ie, residual leukemic cells persistent in the marrow even after treatment) disease commonly occur among patients with ALs. ^{6 , 7} Within 12 months of diagnosis, up to 57% of patients with AML experience primary refractory or relapsed (R/R) disease, or death. ^{6 , 7} Similarly, more than 50% of adult patients with ALL eventually relapse, despite initial complete response rates greater than 90%. ⁸ Ultimately, R/R ALs are associated with poor survival outcomes, ^{7 , 8} substantial health care resource utilization (HCRU), and negative effects on health-related quality of life. ^{9 – 14}

AL harboring gene lysine methyltransferase 2A translocation (KMT2At), previously known as mixed-lineage leukemia, occur in approximately 4% to 11% of adult AML cases ^{15 – 18} and 5% to 20% of adult ALL cases. ^{19 – 22} Patients with KMT2At ALs experience high rates of R/R disease and poor survival outcomes. ^{19 , 21} Limited survival outcomes have been reported for patients with KMT2At AML, with a median overall survival (OS) of 2.4 months in patients who have received 3 or more prior lines of therapy. In KMT2At ALL, the 5-year OS rate is reported to be 18%. ²¹

Until recently, and recognizing advancements in disease biology understanding, treatment options recommended by the National Comprehensive Cancer Network (NCCN) to reinduce remission in R/R AML included intensive combinations of chemotherapy or less intensive venetoclax- or hypomethylating agent (HMA)–based regimens, with or without targeted agents. ^{23 , 24} In R/R ALL, immunotherapy agents, such as blinatumomab and inotuzumab, are often used alone or in combination with chemotherapy. Chimeric antigen receptor T-cell (CAR-T) therapies have also been used as salvage treatment. For high-risk leukemias, including KMT2At ALs, allogeneic hematopoietic cell transplantation remains the only potentially curative intent strategy. However, not all patients are eligible to receive allogeneic hematopoietic cell transplantation depending on disease-related (eg, prior therapies, disease status) or patient-related (eg, performance status, comorbidities) factors, and the procedure is often associated with significant short- and long-term morbidity and mortality risks. ²⁵ There were no approved therapies targeting KMT2At ALs until November 15, 2024, when the US Food and Drug Administration (FDA) approved revumenib (Revuforj) for the treatment of R/R ALs with KMT2At in adult and pediatric patients 1 year and older.

Revumenib, a first-in-class oral menin inhibitor of the menin-KMT2A interaction, was evaluated in the phase 1/2 dose-escalation study and phase 2 registration-enabling study AUGMENT-101 (NCT04065399). Revumenib was shown to improve clinical outcomes among patients with R/R KMT2At ALs while maintaining an overall well-tolerated safety profile. ^{26 – 28} Revumenib is now recommended in the most recent AML and ALL NCCN Guidelines as the sole targeted therapy for patients with KMT2At-rearranged AML and ALL. ^{23 , 24} Hence, this study aimed to provide a representative national projection of the cost implications of adding revumenib to the formulary of a hypothetical 1-million-member commercial health plan in the United States for the treatment of adult patients with R/R KMT2At AL, using a budget impact model (BIM).

Methods

MODEL OVERVIEW

A de novo BIM was developed using Microsoft Excel from the perspective of commercial US third-party payers. Considering the reported median OS of less than 12 months (Table 1) for most of the included treatment options, as well as the time it takes for a new drug to increase uptake and market share, a 1-year to 3-year time horizon was chosen for the BIM. The BIM was developed following the guidelines and recommendations from the Academy of Managed Care Pharmacy and the Professional Society for Health Economics and Outcomes Research. ^{29 , 30} The quality of reporting was assessed using the Consolidated Health Economic Evaluation Reporting Standards checklist. ³¹

TABLE 1.

Model Inputs

Population
% of commercial plan population adults (age 20 to <65)	70.2%	US Census Bureau 44
Annual incidence rate of AML per 100,000 (age 20 to <65) a	2.76	National Cancer Institute 45
Proportion of adult AML that is KMT2A	8.0%	Point estimate derived from the literature reported rates of 3.9%-11.3% 15 – 18
Proportion of adult KMT2A AML that is R/R	70.0%	Point estimate assumed from the literature reported CIR of 55%-85%
Annual incidence rate of ALL per 100,000 adults (age 20 to <65) a	0.96	National Cancer Institute 45
Proportion of adult ALL that is KMT2A	10.0%	Point estimate derived from the literature reported rates of 4.6%-14% 19 – 22
Proportion of adult KMT2A ALL that is R/R	85.0%	Based on the literature, KMT2A-specific RFS rates of 15% 19 , 21
Percent receiving subsequent systemic treatment	100.0%	Analysis assumption b
Market shares	Before revumenib 35	Year 1 35	Year 2 35	Year 3 35
AML
Revumenib (AML)	0.0%	10.0%	15.0%	20.0%
AML aggressive chemotherapy	40.0%	35.0%	33.0%	32.0%
HMA/LDAC	15.0%	12.0%	11.0%	10.0%
Venetoclax + azacitidine	25.0%	23.0%	23.0%	22.0%
Gilteritinib	10.0%	10.0%	9.0%	8.0%
Gilteritinib + azacitidine	10.0%	10.0%	9.0%	8.0%
Total	100.0%	100.0%	100.0%	100.0%
ALL
Revumenib (ALL)	0.0%	5.0%	10.0%	15.0%
ALL aggressive chemotherapy	30.0%	30.0%	29.0%	28.0%
Blinatumomab	30.0%	28.0%	26.0%	24.0%
Inotuzumab ozogamicin	25.0%	23.0%	22.0%	21.0%
Inotuzumab ozogamicin + mini-CVD + blinatumomab	10.0%	9.0%	8.0%	7.0%
Brexucabtagene autoleucel (CAR-T cell)	2.5%	2.5%	2.5%	2.5%
Tisagenlecleucel (CAR-T cell)	2.5%	2.5%	2.5%	2.5%
Total	100.0%	100.0%	100.0%	100.0%
Clinical outcomes	Median duration of treatment (months)	Median time to subsequent treatment (months) c	Median overall survival (months)	Source of grade ≥3 AE rates in Supplementary Table 2 (209.9KB, pdf)
AML therapy
Revumenib (AL)	2.39 46 , 47	2.80 47	8.00 46	Literature 28
AML aggressive chemotherapy	0.92 48 – 50	3.70 48 , 50	7.30 48 , 49 , 51 – 54	USPI 55
HMA/LDAC	0.92 56	0.92 56 , 57	6.02 56 , 57	USPI 55
Venetoclax + azacitidine	1.84 58	2.30 58	5.90 58	USPI 59
Gilteritinib	4.60 56	4.60 56	9.30 56	USPI 55
Gilteritinib + azacitidine	4.60 56	4.60 56	9.30 56	Literature 60
ALL therapy
Revumenib (AL)	2.39 46 , 47	2.80 47	8.00 46	Literature 28
ALL aggressive chemotherapy	1.00 a	1.80 61	6.70 61	USPI 55 , b
Blinatumomab	2.76 62	7.30 62	7.70 62	USPI 63
Inotuzumab ozogamicin	2.53 61	5.00 61	7.70 61	USPI 64
Inotuzumab ozogamicin + mini-CVD + blinatumomab	2.76 65	13.00 65	17.00 65	Literature 66
Brexucabtagene autoleucel (CAR-T cell)	0.03 c	11.60 67	25.40 67	USPI 68
Tisagenlecleucel (CAR-T cell)	0.03 c	24.00 69	54.01 69	USPI 70

AML aggressive chemotherapy included the following: (cladribine + cytarabine) ± (mitoxantrone or idarubicin) (20% of basket); (high-dose cytarabine [if not previously received]) ± (idarubicin or daunorubicin) (20% of basket); fludarabine + cytarabine ± idarubicin (20% of basket); etoposide + cytarabine ± mitoxantrone (20% of basket); clofarabine + cytarabine ± idarubicin (10% of basket); clofarabine ± idarubicin (10% of basket).

HMA/LDAC included the following: azacitidine (34% of basket), decitabine (33% of basket), cytarabine (33% of basket).

ALL aggressive chemotherapy included the following: (cladribine + cytarabine) ± (mitoxantrone or idarubicin) (20% of basket); (high-dose cytarabine [if not previously received]) ± (idarubicin or daunorubicin) (20% of basket); fludarabine + cytarabine ± idarubicin (20% of basket); etoposide + cytarabine ± mitoxantrone (20% of basket); clofarabine + cytarabine ± idarubicin (10% of basket); clofarabine ± idarubicin (10% of basket).

Median time to subsequent treatment had to equal or exceed the median duration treatment used in the model; thus, if event-free survival or duration of response was less than the reported duration of treatment, median time to subsequent treatment was set equal to median duration of treatment.

^a The analysis for AML was run on November 13, 2023, and the analysis for ALL was run on October 22, 2023.

^b A scenario analysis was run excluding subsequent treatment for all included treatments.

^c The median time to subsequent treatment is the median time from the start of the model to the time that patients start subsequent treatment.

AE = adverse event; AL = acute leukemia; ALL = acute lymphoblastic leukemia; AML = acute myeloid leukemia; BIM = budget impact model; CAR-T = chimeric antigen receptor T cell; CIR = cumulative incidence rate; CVD = cyclophosphamide + vincristine + dacarbazine; HMA/LDAC = hypomethylating agents and low-dose cytarabine; KMT2A = histone-lysine N-methyltransferase 2A; RFS = relapse-free survival; R/R = relapsed or refractory; USPI = US prescribing information.

The model estimated the budget impact of revumenib under 2 scenarios: current formulary (scenario “without” revumenib) and future formulary (scenario “with” revumenib) (Figure 1). ³⁰ Patients entered the model on an annualized basis, and associated patient costs were carried over into subsequent years as appropriate through the selected time horizon. Patient flow was tracked separately for the R/R KMT2At AML and ALL adult populations to accommodate the differential treatments for each population, while the budget impact was calculated for the combined populations.

FIGURE 1.

Budget Impact Model Design

AE = adverse event; AL = acute leukemia; ALL = acute lymphoblastic leukemia; AML = acute myeloid leukemia; KMT2A = lysine methyltransferase 2A; R/R = relapsed/refractory.

MODEL INPUTS

Eligible Population.

The eligible population for the BIM included adult patients with R/R KMT2At AL, where population inputs were based on published epidemiology data for the target population (Table 1).

Treatment Options and Market Shares.

Prior to revumenib, there was no approved therapy specifically targeting R/R KMT2At; hence, a landscape analysis of real-world practice was performed to identify appropriate comparators. A total of 11 pharmacotherapies, including 5 drugs for AML and 6 for ALL, were selected based on the real-world evidence and clinical input. These 11 regimens were included in the “without” scenario of the base-case analysis, to which revumenib was added to constitute the “with scenario” (Table 1). ^{10 , 32 – 35} The market shares for the included treatment options were allocated based on the likelihood of achieving a subsequent complete response and/or the presence of a targetable genetic alteration. Given the treatment assumption of first salvage, and the overall treatment strategy using more aggressive options for most patients to achieve an adequate response, revumenib monotherapy uptake was assumed to increase gradually over 3 years from 10% to 20% in AML and from 5% to 15% in ALL, as a subset of these patients may not be able to tolerate or elect to undergo more aggressive treatment options.

Clinical Inputs.

Clinical inputs included duration of treatment, time to subsequent treatment, survival outcomes (including OS and progression-free survival/event-free survival), and frequency of grade 3 or greater AEs. Clinical inputs for revumenib were informed by the AUGMENT-101 phase 2 KMT2At ALs subgroup analysis. ²⁸ For the other treatment options, data were sourced from US Prescribing Information (USPI) and publications associated with pivotal clinical studies (Table 1). Similarly, the dosing regimens and route of administration for the included treatments in the BIM were based on USPI and published trial data (Tables 1 and 2). The median duration of a given treatment was used to calculate its cost, including drug acquisition and administration costs. Time to subsequent treatment and median OS were used to calculate postdiscontinuation costs (described in Cost Inputs). For treatments where time to subsequent treatment was not reported, event-free survival, duration of response, or duration of treatment served as a proxy.

TABLE 2.

Cost Inputs

AML treatments (RED BOOK 37 ) a			Acquisition cost per 28-day cycle	Administration cost per cycle 38 , 39
Revumenib 110-mg tablet			$36,867	$0
Revumenib 160-mg tablet			$36,867	$0
AML aggressive chemotherapy			$4,801	$234,439
HMA/LDAC			$1,593	$1,163
Venetoclax + azacitidine			$17,050	$407
Gilteritinib			$26,259	$0
Gilteritinib + azacitidine			$29,696	$1,492
ALL treatments (RED BOOK 37 ) a			Acquisition cost per 28-day cycle (except where noted otherwise)	Administration cost per cycle 38 , 39
Revumenib 110-mg tablet			$36,867	$0
Revumenib 160-mg tablet			$36,867	$0
ALL aggressive chemotherapy			$4,801	$215,822
Blinatumomab			$137,204 (per 42-day cycle)	$22,813 (per 42-day cycle)
Inotuzumab ozogamicin			$135,414	$639
Inotuzumab ozogamicin + mini-CVD + blinatumomab			$417,200 (per 42-day cycle)	$6,072 (per 42-day cycle)
Brexucabtagene autoleucel (CAR-T cell)			$462,000 (single administration)	$109,765
Tisagenlecleucel (CAR-T cell)			$519,418 (single administration)	$109,765
Postdiscontinuation
Subsequent systemic treatment costs per month			$42,889	Market share–weighted average WAC cost of AL treatment regimens
End-of-life costs			$17,459	US Bureau of Labor Statistics 36 ; Hoverman et al 41
Drug administration
CPT code	Infusion type	Description	Cost per unit commercial	Source
96402	Intramuscular injection	Chemotherapy administration, subcutaneous or intramuscular	$58.21	CMS; Avalere Health
96413	IV infusion cost (initial hour)	Chemotherapy administration, intravenous infusion technique; up to 1 hour, single or initial substance/drug	$213.11
96417	Each additional IV infusion (initial hour)	Chemotherapy administration, intravenous infusion technique; each additional sequential infusion (different substance/drug), up to 1 hour	$104.91
Health care resource utilization		Description	Cost per unit	Source
Inpatient admissions		Cost per inpatient hospital day	$6,738.00	Shah et al 71 ; US Bureau of Labor Statistics 36
Outpatient hospital visits		Cost per outpatient visit	$149.94	CMS 39 ; Avalere Health 38
Outpatient prescription visits		Outpatient prescription drugs	$0	Assumption
Comparator treatment		Description	Total cost
Brexucabtagene autoleucel (CAR-T cell)		Based on a hospitalization with an average of 4 ICU days and 12.67 non-ICU days	$109,765
Tisagenlecleucel (CAR-T cell)			$109,765
AML aggressive chemotherapy		FLAG-IDA: length of hospital stay for salvage induction was 43 days (range, 10-305) MEC: median duration of hospitalization was 30 days (range, 14-78) Assume median of 36.5 inpatient days Subtracted AE costs to remove double counting	$232,246
ALL aggressive chemotherapy			$232,246
Blinatumomab		Hospitalization is recommended for the first 3 days of the first cycle and the first 2 days of the second cycle	$33,692

Cost inputs based on monthly costs and expected durations of treatment.

Medicare costs were adjusted to commercial values using an American Hospital Association Private to Medicare ratio (1.65) based on the average values from 2014 to 2018 reported in Avalere Health’s TrendWatch Chartbook. ³⁸

^a Generic pricing was used where available; for drugs with multiple manufacturers, the prices were averaged. Hospitalization costs during administration, sourced from published literature, were included explicitly for 5 treatments: 2 CAR-T cell therapies, ⁷¹ AML and ALL aggressive chemotherapy baskets, ^{72 , 73} and blinatumomab (based on USPI). ^{63 , 71} To avoid double counting, the AML and ALL aggressive chemotherapy-related AE costs were subtracted from the hospitalization costs during administration.

AE = adverse event; AL = acute leukemia; ALL = acute lymphoblastic leukemia; AML = acute myeloid leukemia; CAR-T = chimeric antigen receptor T-cell; CMS = Centers for Medicare and Medicaid Services; CPT = Current Procedural Terminology; CVD = cyclophosphamide + vincristine + dacarbazine; FLAG-IDA = fludarabine + cytarabine + G-CSF + idarubicin; HCUP = Healthcare Cost and Utilization Project; HMA/LDAC = hypomethylating agents and low-dose cytarabine; ICU = intensive care unit; IV = intravenous infusion; KMT2A = histone-lysine N-methyltransferase 2A; R/R = relapsed or refractory; WAC = wholesale acquisition cost.

Cost Inputs.

Costs associated with treatment, management of grade 3 or greater AEs, treatment-related supportive care and monitoring, and postdiscontinuation were included in the BIM. All cost values were adjusted to 2024 USD using the medical care component of the CPI. ³⁶ Key cost inputs are included in Table 2 and further described in Supplementary Tables 1 and 2 ^{(209.9KB, pdf)} (available in online article).

Treatment Costs.

Treatment costs included drug acquisition and administration, which were calculated using duration of treatment to estimate number of cycles required for the treatments. The model included the use of 2 tablet strengths of revumenib (160 mg and 110 mg). The wholesale acquisition cost (WAC) was used for the drug costs of treatments as well as supportive care, which were sourced from the RED BOOK online database (Truven Health Analytics). ³⁷ Treatment administration costs included subcutaneous injections, intravenous infusion, and hospitalization. The costs of administration-related hospitalization were based on the published literature. The reported hospitalization length of stay for administrating aggressive chemotherapies ranged from 10 to 305 days (Table 2); hence, a median value of 36.5 days was used in the base-case analysis and different scenario analyses were performed to assess the parameter uncertainty. In addition, the costs of AEs were subtracted to avoid double counting for AEs that would be included in those hospitalizations. Component costs for intravenous treatments were retrieved from the Centers for Medicare & Medicaid Services physician fee schedule and adjusted to commercial values. ^{38 , 39}

AE Costs.

Costs associated with the management of grade 3 or greater AEs were included. The Clinical Classifications Software Refined category associated with grade 3 or greater AEs reported in the pivotal clinical trials was used to retrieve the cost from the Healthcare Cost and Utilization Project website. ⁴⁰

Treatment-Related Supportive Care and Postdis-continuation Costs.

Per-regimen treatment-related costs included monitoring-related costs and comedication supportive care (Supplementary Table 3 ^{(209.9KB, pdf)} ). Two additional types of costs, subsequent treatment and end-of-life costs, were included in the postdiscontinuation costs. Subsequent treatment costs were determined by multiplying the monthly cost of subsequent treatment by the difference between the time to subsequent treatment and the median OS. Of note, a single cost was used for all subsequent treatments in the BIM as there were limited data on the distribution of treatments used in subsequent lines, particularly across the large array of included comparators. By using a single consistent cost for subsequent treatment, the differentiation between regimens was simplified to be based only on differences in the time on subsequent treatment, which could be estimated from published data on the time to next treatment and OS. Thus, the differentiation between total costs of subsequent lines of treatment was not confounded by assumptions on differences in treatment patterns. End-of-life costs were assigned to each treatment regimen and were applied at the time of death. The monthly cost was based on a market share–weighted average of BIM treatment regimens; end-of-life costs were taken from the literature. ⁴¹

MODEL ASSUMPTIONS

To address data limitations and challenges in the modeling exercise, several assumptions were applied. Focusing on the commercial plan population, the BIM assumed that all plan members were adults younger than 65 years for both AML and ALL. Commercially insured Medicare Advantage members were not included. In addition, because limited treatments were approved in pediatric populations, the BIM did not include pediatric patients.

Based on the AUGMENT-101 KMT2At ALs subgroup analysis, ²⁸ clinical inputs for revumenib, for both the AML and ALL populations, were assumed to be the same. Approximately 20% of the AUGMENT-101 study population were younger than 18 years; this was assumed not to impact clinical outcomes for an adult-only study population. Pivotal studies of other treatment therapies did not specifically report KMT2At subgroup–specific data; therefore, clinical inputs were based on data from the overall R/R study populations. It was assumed that intensive chemotherapy consisted of the same combination of aggressive chemotherapies for both AML and ALL.

Although revumenib and other treatments reported different grades of AEs, as commonly done in hematologic oncology analyses, only grade 3 or greater AEs were included in the model, assuming that high-grade AEs would incur significant HCRU and associated costs (Supplementary Table 2 ^{(209.9KB, pdf)} ). The AE rates were assumed to be 0% if not reported for any treatments in the USPI or literature. The same AE rates for revumenib were used for both the AML and ALL inputs. In addition, all included AEs were assumed to require hospitalization and to be independent events.

Postdiscontinuation costs were applied to all patients during the period from time to subsequent treatment to the end of life. The same monthly cost was applied to all patients, regardless of prior therapy, and was assumed to be the subsequent salvage treatment.

ANALYSIS

Base-Case Analysis.

The BIM results were reported for current formulary (without revumenib) and future formulary (with revumenib), with the differential impact reflecting the budgetary impact of adding revumenib to the formulary. The per-member-per-month (PMPM) cost was calculated as the difference in total plan costs in each scenario divided by the total member-months of exposure. The per-treated-member-per-month (PTMPM) cost was defined as the difference in average total costs per patient in each scenario divided by the total treated patient-months of exposure.

Sensitivity and Scenario Analyses.

One-way sensitivity analyses (OWSAs) and different scenario analyses were conducted to assess the uncertainties in model parameters and assumptions. The OWSAs were conducted by varying 20% change in drug acquisition and toxicity costs (sum of AE rates multiplied by corresponding AE costs across all the AEs) for revumenib and other treatment options.

There was some variation in the literature regarding incidence rates of KMT2At R/R AL and median OS for various treatment options. In addition, as hospital length of stay ranged from 10 days to 305 days, a wide variation in associated costs for administrating aggressive chemotherapies was also seen in the literature. Therefore, more than 10 different scenarios were performed to examine parameter uncertainty in the BIM, such as varying the uptake of revumenib, epidemiology rates, time on treatment and time to next treatment, OS, and aggressive chemotherapy administration costs, and excluding cost components related to monitoring, supportive care, and postdiscontinuation costs (Supplementary Table 1 ^{(209.9KB, pdf)} ).

Results

BASE-CASE ANALYSIS

In a hypothetical 1-million-member commercial health plan, an estimated 1.7 adults with R/R ALs (AML,1.1; ALL, 0.6) per year would be eligible for treatment. Based on the estimated market share of revumenib (Table 1), 0.14, 0.22, and 0.30 patients would initiate revumenib in year 1, year 2, and year 3, respectively. In the base-case analysis (Table 3), the estimated 3-year total plan costs before the introduction of revumenib are $2,146,564. Total plan costs decline to $2,126,919 over the 3-year period following introduction owing to lower costs for drug acquisition (−$10,205), treatment administration (−$50,538), and AE management (−$2,498). This yields a net budget impact of savings totaling −$19,646 over the 3-year period, which translates into estimated PMPM and PTMPM cost savings of −$0.0005 and −$329, respectively. Additional results from the BIM are included in Supplementary Table 3 ^{(209.9KB, pdf)} .

TABLE 3.

Aggregated Costs ($) and Budget Impact (1-3 Years) With and Without Revumenib

	1-Year			2-Year			3-Year
	With revumenib	Without revumenib	Differential	With revumenib	Without revumenib	Differential	With revumenib	Without revumenib	Differential
Aggregated costs
Drug acquisition	205,345	204,180	1,165	406,210	408,359	–2,150	602,334	612,539	–10,205
Treatment administration	121,897	133,321	–11,425	237,658	266,643	–28,985	349,426	399,964	–50,538
Total adverse event cost	65,470	66,102	–632	130,998	132,204	–1,206	195,808	198,306	–2,498
Supportive care	11,104	10,828	276	22,288	21,656	632	33,546	32,484	1,063
Postdiscontinuationa	209,139	204,644	4,495	548,601	529,690	18,911	945,804	903,272	42,532
Total	612,954	619,075		1,345,753	1,358,551		2,126,919	2,146,564
Budget impact of revumenib market entry
Net budget impact			–6,121			–12,798			–19,646
PMPM cost	0.0511	0.0516	–0.0005	0.0561	0.0566	–0.0005	0.0591	0.0596	–0.0005
PTMPM cost	30,820	31,127	–308	33,833	34,154	–322	35,647	35,977	–329

^a Includes systemic treatment and end-of-life care.

PMPM = per member per month; PTMPM = per treated member per month.

Introducing revumenib to the formulary was estimated to increase patient life-years during the 3-year time horizon from 3.80 to 3.83 cumulatively for the cohort of patients. The total number of grade 3 or greater AEs was lower over 3 years (10.82 vs 10.99 for the cohort of patients) in the scenario with revumenib vs without.

The 1-year and 2-year results were generally consistent with the 3-year results, as presented in Table 3.

SENSITIVITY ANALYSIS

The OWSA results demonstrated that decremental total costs associated with introduction of revumenib varied from −$5,740 to −$33,551 (−$0.0009 to −$0.0002 PMPM) over 3 years (Figure 2). Results for the 10 factors the model was most sensitive to are shown in Figure 2. The model was most sensitive to the drug acquisition cost of revumenib, followed by acquisition costs for ALL-specific treatments with inotuzumab and blinatumomab.

FIGURE 2.

One-Way Sensitivity Results Relative to Base Case Showing the Top 10 Factors the Budget Impact Model Was Most Sensitive to

AE = adverse event; AML = acute myeloid leukemia.

SCENARIO ANALYSES

In general, different scenario analyses demonstrated a decremental impact on the budget (−$0.0018 to −$0.0003 PMPM) and clinical consequences, given the small treatment-eligible population (Supplementary Table 1 ^{(209.9KB, pdf)} ). Specifically, including only the treatment-related costs (acquisition and administration) and AE management costs, while excluding costs related to monitoring, supportive care, and postdiscontinuation care in the BIM, resulted in cost savings of −$63,240 over the 3-year period (PMPM, −$0.0018; PTMPM, −$1,060). Variation in uptake of revumenib or epidemiology rates had minimal impact on the model. Changing clinical outcomes such as time on treatment and time to next treatment, median OS, and hospital administration time for aggressive chemotherapy induction by 20% from base case had minimal impact on the model.

Discussion

This budget impact analysis indicates that introducing revumenib into a hypothetical 1-million-member commercial health plan would result in nominally lower costs. Estimated base-case cost savings total −$19,646 over a 3-year time horizon. Sensitivity analyses show that these savings could vary from −$5,740 to −$33,551, and scenario analyses suggest potential savings between −$10,324 and −$63,240 depending on variations in treatment costs included and excluded, upward or downward variations in revumenib, and lower and higher population estimates for KMT2At R/R AML and R/R ALL.

The decremental budget impact was mainly driven by the lower drug administration costs of revumenib—an oral menin inhibitor, which eliminates the need for visits to the cancer clinic or hospitalizations for parenteral drug administration, which current treatment options may require. Notably, the base-case and sensitivity analyses underscored the benefit of oral agent treatment, which lowered treatment administration costs (−$50,538).

To our knowledge, this is the first budget impact analysis targeting a population with R/R KMT2At ALs, so there is no other budget impact evidence against which to compare our findings. We attempted to counter this by comprehensively including alternate, nontargeted and R/R KMT2At-agnostic treatment options for both the AML and ALL populations, with various cost components ranging from drug acquisition and AE management to subsequent care management after discontinuation. In addition, the one-way sensitivity and the different scenario analyses to account for uncertainties in the model assumptions and data limitations underscored the robustness of the base-case results.

Revumenib is the first FDA-approved therapy for this rare indication, with the potential to improve outcomes for patients without increasing financial implications. Although the number of eligible patients may be small, this limited population size is consistent with other orphan drug and rare indications. ^{42 , 43} As revumenib is the only targeted therapy approved for R/R KMT2At ALs, other therapies included in the BIM are indicated for other subgroups or overall R/R AML and ALL populations. As such, the clinical efficacy data for the other treatment options were not KMT2At-specific. Additionally, clinical inputs (eg, OS, duration of treatment) were sourced from clinical trials of each treatment without performing meta-analyses or indirect treatment comparisons. Given that R/R KMT2At ALs have a worse prognosis than overall R/R ALs, it is likely that the improvement in clinical outcomes in the scenario with revumenib was underestimated.

LIMITATIONS

Our analysis has some limitations while also identifying areas for future research. The patient population was limited to adult patients younger than 65 years and hence excluded patients in commercially insured Medicare Advantage Plans (Part D) as well as pediatric patients. Therefore, our study may have underestimated the number of treatment-eligible patients. However, we do not expect this to materially impact our estimates, given that R/R KMT2At AL is a rare and difficult-to-treat disease and scenario analyses varying population estimates were consistent with base-case results; yet future updates may be recommended. Additionally, owing to the absence of previously approved treatments for R/R KMT2At AL and clinical data on heavily pretreated patient populations, coupled with factors such as co-mutation rates, availability of other targeted agents, and lack of specific treatment sequencing data in the targeted rare population, it was challenging to estimate market shares for treatment options and uptake patterns for revumenib. As such, the proposed market mix and uptake of revumenib in the BIM may differ from the real-world setting. Updating the analyses using real-world market mix data is recommended.

Publicly available data on WAC were used in the BIM instead of assuming net payment of insurance considering rebates or discounts, in order to maintain transparency and simplicity of the data. However, different scenario analyses and sensitivity analyses were performed, incorporating acquisition costs for revumenib and other pharmacotherapy options to account for parameter uncertainty. Additionally, in the BIM, 10%, 15%, and 20% of patients with AML and 5%, 10%, and 15% of patients with ALL were assumed to initiate revumenib in years 1, 2, and 3, respectively. Therefore, only those patients had a chance to gain life-years in this analysis. Thus, the incremental life-years in this analysis may have been underestimated.

Conclusions

Revumenib is a first-in-class menin inhibitor and the only FDA-approved targeted therapy for patients with R/R KMT2At ALs as of November 2024. While aiding in fulfilling an unmet need among these patients with the potential for improving clinical outcomes, the net budget impact of revumenib was approximately cost neutral.