FormalPara Key Summary Points

Why carry out this study?

Cutaneous T-cell lymphoma (CTCL) is associated with substantial economic burden due to high healthcare resource utilization and costs. Brentuximab vedotin has been approved as a treatment option for CTCL in Canada, specifically for adult patients with primary cutaneous anaplastic large cell lymphoma (pcALCL) or CD30-expressing mycosis fungoides (MF), two types of CTCLs, treated previously with systemic therapy.

It is important to analyze the cost-effectiveness of brentuximab vedotin as results from the ALCANZA study show improved, clinically meaningful, durable responses.

We hypothesized that brentuximab vedotin is a cost-effective treatment option for patients with pcALCL or CD30-expressing MF.

What has been learned from the study?

Based on clinical data from the ALCANZA trial and Canadian clinical expert opinion on model inputs, this analysis found that brentuximab vedotin is a cost-effective treatment for CTCL; treatment with brentuximab vedotin compared with MTX or BEX incurred higher drug costs, but most of those costs were offset through reduced costs in the post-progression and end-stage management health states.

Brentuximab vedotin provides a cost-effective treatment option with superior efficacy and a comparable safety profile to physician’s choice of MTX or BEX from the perspective of the Canadian publicly funded healthcare system.

Introduction

Cutaneous T-cell lymphoma (CTCL) is a relatively rare, heterogeneous group of non-Hodgkin lymphomas that present in the skin without evidence of extracutaneous disease at the time of diagnosis [1]. The annual incidence of CTCL is estimated at 11.32 cases per million people in Canada [2]. The most common CTCL subtypes are mycosis fungoides (MF; 60% of CTCLs) and primary cutaneous anaplastic large cell lymphoma (pcALCL); pcALCL is characterized by the expression of cell-surface CD30 antigen [3]. By definition, pcALCL exhibits a high level of CD30 expression (≥ 75% of tumor cells) [4], whereas CD30 expression in MF is more variable [5].

CTCLs often have a chronic disease course and a devastating impact on quality of life, arising from persistent and frequently disabling symptoms, including disfiguring lesions, debilitating pruritus, and frequent skin infections [3, 6]. Advanced stages of CTCL (stages IIB+) can be associated with severe skin tumors, erythroderma, and inferior HRQoL compared with earlier stages of the disease [6]. CTCL is also associated with substantial economic burden due to high healthcare resource utilization and costs [7].

The goals of therapy for CTCL, as indicated by CTCL treatment guidelines, are to improve symptoms, reduce the burden of disease, maintain remission, delay disease progression, and improve or preserve HRQoL [8,9,10,11,12]. Currently, there is no standard therapy for CTCL due to its diverse clinical presentation. Generally, patients with advanced disease are treated with systemic therapies [e.g., oral methotrexate (MTX), bexarotene (BEX), interferon-alpha, targeted therapies, chemotherapy], with the goal of achieving a response, maintaining adequate organ function, and becoming eligible to receive a potentially curative allogeneic stem cell transplant. For those who are transplant ineligible, the goals are to increase survival, control disease, reduce symptoms, and improve HRQoL [13].

Given the chronic and recurrent nature of the disease, patients may survive for many years with treatment, although they frequently require several types of treatment, repeat treatment courses, and maintenance regimens for ongoing disease and symptom control. These data demonstrate the need for effective treatments that extend duration of response, time in remission, and improve quality of life.

Brentuximab vedotin (Adcetris®; Seagen; Bothell, WA, USA) is a CD30-directed antibody–drug conjugate approved in Canada for six indications, including the treatment of adult patients with pcALCL or CD30-expressing MF treated previously with systemic therapy. The CADTH pan-Canadian Oncology Drug Review Expert Review Committee has recommended the reimbursement of brentuximab vedotin for adult patients with CD30-positive MF or pcALCL previously treated with systemic therapy based on the ALCANZA trial results [14].

ALCANZA was a phase 3 multicenter, open-label, randomized, active controlled trial designed to evaluate the efficacy and safety of brentuximab vedotin (n = 66) compared with physician’s choice (n = 65) of MTX or BEX, in patients with CD30-expressing MF and pcALCL [15]. A clinically meaningful and statistically significant benefit in favor of brentuximab vedotin compared with MTX or BEX was demonstrated in patients achieving an objective response that lasted at least 4 months (ORR4, 56.3% vs. 12.5%, P < 0.0001). Brentuximab vedotin compared with MTX or BEX also demonstrated statistically significant improvements in complete response (CR; 16% vs. 2%, P = 0.0046), progression-free survival (PFS; 16.7 vs. 3.5 months, P < 0.0001), and the Skindex-29 symptom domain [maximum reduction (SD): − 27.96 (26.877) vs. − 8.62 (17.013), P < 0.0001].

The objective of this economic evaluation was to estimate the cost-effectiveness of brentuximab vedotin compared with MTX or BEX for the treatment of patients with pcALCL or CD30-expressing MF who have had prior systemic therapy over a lifetime horizon from the perspective of the Canadian publicly funded healthcare system. This analysis is the first cost-effectiveness analysis of brentuximab vedotin in the treatment of CTCL in the Canadian healthcare setting.

Methods

Patient Population

The target population of this economic evaluation reflects the treatment of adult patients with pcALCL or CD30-expressing MF who have had prior systemic therapy and aligns with the Health Canada-approved indication. The modeled population is consistent with the intent-to-treat (ITT) patient population enrolled in the ALCANZA trial [15]. The control arm in the ALCANZA trial was physician’s choice of MTX or BEX, which serves as the base-case model comparator. Although MTX is more commonly used in Canada than BEX, BEX is a relevant comparator as evidenced by its inclusion in the Alberta Health Guidelines and the British Columbia Cancer Benefit Drug List, as well as by its validation by Canadian clinical expert opinion [9, 16, 17].

Model Overview

The model satisfies the standards of health economics required by core health technology assessment authorities, and is aligned with ISPOR Task Force on Good Modeling Practices and CADTH economic guidelines [18, 19]. This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

The model was developed in Microsoft Excel® using a partitioned survival analysis in combination with a payoff approach to estimate the proportion of patients in each health state (Fig. 1). Partitioned survival models are accepted approaches to cost-effectiveness models used to inform health technology assessment submissions of oncology medicines [19, 20]. A payoff approach gives a lump sum of costs and quality-adjusted life years (QALYs) upon reaching a milestone; in this analysis, the milestone is progression. The payoff is continuously discounted, such that it is equivalent to incurring those costs and QALYs over time. The health states modeled were pre-progression (on or off treatment), post-progression (receiving active subsequent therapy, no active subsequent therapy, or receiving end-stage disease management), and death.

Fig. 1
figure 1

Model structure diagram. SCT stem-cell transplant

Full size image

Additional health states for patients who receive allogeneic SCT include SCT (disease-free), SCT relapse (receiving active subsequent therapy, no active subsequent therapy, receiving end-stage disease management), or death. The treatment pathways for brentuximab vedotin and either MTX or BEX were the same, although the time spent pre-progression differed between brentuximab vedotin and either MTX or BEX.

All patients enter the model in the pre-progression state, from which most patients will progress to the non-SCT post-progression state, with a small percentage of transplant-eligible patients progressing to SCT, as described in the model inputs. A proportion of patients who respond during pre-progression are eligible to receive allogeneic SCT; these patients may remain disease-free following the procedure or relapse. After progression, patients who did not receive allogeneic SCT undergo multiple lines of active subsequent therapy. Near the end of their lives, patients begin end-stage disease management. Patients who relapse from allogeneic SCT receive the same active subsequent therapy and end-stage care as non-SCT patients, although they do not receive brentuximab vedotin as a subsequent therapy.

The model utilizes a payoff approach when patients reach the post-progression health state, where total post-progression costs and QALYs are determined based on time remaining alive. In each cycle, the proportion of patients leaving the pre-progression state is calculated, and the payoff is applied. Post-progression states include a fixed period spent on active therapy, a period without active therapy, and then the remaining time is spent in end-stage disease management. The difference in payoff between brentuximab vedotin and MTX or BEX in the base case was derived by calculating the area between the overall survival (OS) and PFS curves for non-SCT outcomes, and by calculating the area between the OS and disease-free survival curves for SCT outcomes.

The duration of end-stage management in the base case is 3 months for brentuximab vedotin and 6 months for MTX or BEX, based on Canadian clinician opinion [17]. The duration of active subsequent treatment is the same for the brentuximab vedotin and the MTX or BEX arms, as the treatment mix is identical. Time spent off active treatment in the post-progression health state is calculated as the total time in post-progression, less the active subsequent treatment and the end-stage durations for both arms.

The model used a 1-week cycle length, which aligns with the trial dosing schedule and allows for sufficient granularity in transitions between health states. In the base case, a half-cycle correction was applied to all costs other than those for frontline drugs, which are incurred at the start of each cycle.

A time horizon of 45 years was sufficient to capture all lifetime costs and outcomes across the intervention and the comparators [19]. Uncertainty in the time horizon was explored using a range of scenario analyses examining time horizons of 10, 20, and 30 years.

A discount rate of 1.5% annually was applied to both costs and outcomes in the base case, and discount rates of 0% and 3% were used in scenario analyses, as recommended in the most recent CADTH Guidelines for Economic Evaluation of Health Technologies [19]. In addition, a continuous discounting approach was employed to accurately discount costs and QALYs accrued as a payoff.

The base-case analysis takes the perspective of the Canadian publicly funded healthcare system. Model parameters for the base-case analysis are summarized in Table 1, while inputs and assumptions are described in the following subsections and the Supplemental Appendix. The analysis considered all direct medical costs but not indirect costs.

Table 1 Model parameters for the base-case analysis
Full size table

Model Inputs

Model inputs included efficacy, adverse events, active subsequent therapy, costs, and utilities. Information on these model inputs is provided in the Supplementary Appendix.

In the base case, OS for MTX or BEX was assumed to follow a log-normal distribution. PFS for the ALCANZA ITT population was relatively mature and demonstrated a significant benefit for brentuximab vedotin compared with MTX or BEX. In the base case, an exponential model was chosen for brentuximab vedotin and a log-logistic model for MTX or BEX. The Kaplan–Meier curves and fitted distributions for OS and PFS in the base-case analysis are shown in Fig. 2.

Fig. 2
figure 2

Kaplan–Meier curves and fitted distributions for OS and PFS in the base-case analysis, ITT population. BV brentuximab vedotin, ITT intent to treat, KM Kaplan-Meier, OS overall survival, PC physician's choice, PFS progression-free survival

Full size image

Statistical Analyses

Base-case analyses were conducted using the model specifications noted in Table 1. Probabilistic sensitivity analyses (PSAs) were undertaken to vary inputs simultaneously based upon their assigned distribution to capture the total parameter uncertainty; 10,000 PSA iterations were run, with results stabilizing at approximately 2000 iterations. Scenario analyses were performed primarily to test structural assumptions made throughout the model. The scenarios tested included changes to discount rates, time horizon, parametric survival model choice, survival assumptions, SCT rate, and end-stage management duration.

Results

Base-case Probabilistic Analysis

For the base-case probabilistic analysis, brentuximab vedotin yielded greater average LYs and QALYs gained relative to the MTX or BEX treatment arm, mainly driven by improved PFS of brentuximab vedotin compared with MTX or BEX (Table 2). Correspondingly, the total average cost associated with brentuximab vedotin was slightly higher than MTX or BEX, mainly driven by the drug acquisition cost of brentuximab vedotin compared with MTX or BEX. The estimated incremental costs and QALYs accrued by brentuximab vedotin relative to MTX or BEX resulted in an incremental cost-effectiveness ratio (ICER) of $43,790 per QALY gained.

Table 2 Discounted base-case results: probabilistic analysis
Full size table

Probabilistic disaggregated costs showed a large upfront cost associated with brentuximab vedotin due to the higher drug costs compared to MTX, in addition to the additional time spent pre-progression and receiving treatment. Brentuximab vedotin SCT costs were also higher due to a higher proportion of brentuximab vedotin patients achieving a CR or PR, and therefore becoming eligible to receive allogeneic SCT. Increased costs due to allogeneic SCT were offset by increased costs in medical resource use in the MTX or BEX arm, again due to the PFS benefit provided by brentuximab vedotin. The savings from the reduced medical resource use also arises from less time spent in the resource-intensive end-stage state, as shown in the disaggregated QALY and LY results (Table 3).

Table 3 Disaggregated costs, QALY, and LY breakdown (discounted)
Full size table

The cost-effectiveness acceptability plane representing 10,000 PSA iterations shows predominately higher incremental QALYs with either lower or higher incremental costs (Fig. 3). The cost-effectiveness acceptability curve shows that, at a willingness-to-pay threshold of $50,000, brentuximab vedotin has a 50.6% chance of being cost-effective and, at a willingness-to-pay threshold of $100,000, a 69.9% chance of being cost-effective (Fig. 4). As the willingness to pay increases, the probability of brentuximab vedotin being cost-effective increases monotonically.

Fig. 3
figure 3

Cost-effectiveness plane. QALY quality-adjusted life year

Full size image
Fig. 4
figure 4

Cost-effectiveness acceptability curve. BEX bexarotene, MTX methotrexate

Full size image

Scenario Analyses

Results from the scenario analyses are shown in Supplemental Table 5. Brentuximab vedotin was more cost-effective in scenarios with lower discount rates, a 10-year time horizon, increased SCT use, and alternative post-progression utility inputs [21].

Scenarios that changed or discounted the end-stage duration led to the largest increase in the ICERs. In two scenarios that assumed the same end-stage duration of 6 and 3 months for brentuximab vedotin and MTX or BEX, ICERs increased to about $286,200 at 6 months and to $283,500 at 3 months, as these scenarios equalize the highly resource-intensive health states.

Substantially increasing the discount rate or shortening the time horizon also reduced the impact of costly post-progression and end-stage care, leading to increased ICERs. In 3 scenarios, the discount rate increased from 0 to 5% and brentuximab vedotin went from dominant to an ICER of nearly $275,000. Similarly, in scenarios where the time horizon increased from 10 to 30 years, brentuximab vedotin went from being dominant to an ICER close to the base-case PSA result. Both of these scenarios reduced, or eliminated, the impact of outcomes further in the future.

Discussion

This cost-effectiveness analysis demonstrated the economic value of brentuximab vedotin compared with MTX or BEX for the treatment of adult patients with pcALCL or CD30-expressing MF who received prior systemic therapy, which aligns with the Health Canada-approved indication. The analysis was based on results for the ITT population from the international, open-label, randomized, phase 3 ALZANCA trial [15]. In the ALCANZA trial, brentuximab vedotin significantly delayed progression, with a median PFS of 16.7 months compared with 3.5 months in the MTX or BEX arm. Delaying progression was important within the economic model, both for maintaining better patient HRQoL for a longer period (pre-progression utility, 0.77; post-progression utility, 0.61–0.38) and for reducing the amount of time spent in more severe post-progression health states.

A greater proportion of patients in the brentuximab vedotin arm than in the MTX or BEX arm also achieved either a CR or PR, and were therefore potentially eligible to receive allogeneic SCT (brentuximab vedotin, 9.76%; MTX or BEX, 3.10%). Although outcomes for the new Stanford protocol of allogeneic SCT were informed by a single center trial of 53 patients [22], the curative potential of this approach could translate to improved outcomes for eligible patients. Using brentuximab vedotin as the initial therapy for CTCL offers significant value, as an increased response rate means more patients could become eligible to receive potentially curative allogeneic SCT.

The results of this cost-effectiveness analysis were particularly driven by the modeling assumption regarding the OS of brentuximab vedotin compared with that of MTX or BEX. In the base case, OS for brentuximab vedotin and MTX or BEX was assumed to be equivalent. This assumption was supported by clinical expert opinion that an OS benefit was not likely to be observed, as patients often receive several courses of therapy with various agents to provide disease control over the prolonged disease course. In addition, Kaplan–Meier data, as well as results from a parametric survival analysis, showed brentuximab vedotin and MTX or BEX curves crossing, consistent with the assumption that there is no observable difference in OS.

Because a key element of brentuximab vedotin’s value is the shorter duration spent in the end-stage management state, it was important to accurately capture the resource use and care required in this state. An end-of-life questionnaire was developed specifically for the model and completed by Canadian providers to collect data to inform resource use in end-stage disease management [23].

Due to the chronic nature of CTCL and the duration of time spent in post-progression health states, the total costs accrued were extremely high. In the case of brentuximab vedotin, cost savings in these health states offset the initial drug costs of brentuximab vedotin and the costs of additional allogeneic SCT procedures. Base-case results were $43,790 per QALY for treatment with brentuximab vedotin compared with MTX or BEX, which falls below the ICERs of other approved therapies for oncology indications in Canada [24]. The model investigated how results were affected by varying key parameters and structural assumptions. Scenario analyses supported the finding that resource use costs during post-progression and end-stage management were a key model driver. Additionally, results from scenario analyses showed that assumptions around end-stage duration and assumptions governing survival extrapolations were the most influential on the cost-effectiveness of brentuximab vedotin. Scenarios assuming equal end-stage duration or substantially discounting post-progression resource use led to the largest ICER increases.

Limitations

This model relied on treatment data obtained from the PROCLIPI Registry and confirmed by expert opinion. Differences in the treatment of CTCL may exist between those used in the real world and those included in this model. For example, although romidepsin can be used for CTCL in Canada, it was not included in the PROCLIPI Registry data and therefore not included in the model. However, analysis of our model results shows that subsequent treatments other than brentuximab vedotin have very little effect on our modeled results. The impact of newer therapies that were not approved at the time of this analysis for CTCL, such as mogamulizumab, is unknown.

While ALCANZA’s clinical trial results cannot be generalized to external populations, the use of physician’s choice as a comparator was employed to control for this. Furthermore, trial data were supplemented with expert input and external data to inform post-SCT survival and subsequent treatments. However, the level of care and clinical monitoring in clinical trials is often higher than in real-world settings, which likely increased the overall effectiveness of treatment and adverse event recognition, and may impact the distribution of subsequent treatments.

Patient quality of life is always difficult to measure, and the EQ-5D-3L instrument used to determine utilities may not be fully sensitive to changes in CTCL HRQoL. Comparing the observed utilities from the EQ-5D with those predicted by the Skindex-29 instrument indicates that the EQ-5D lacks sensitivity to this severe condition, and that, by using Skindex-29, the impact on patient HRQoL may be captured to a greater extent [25]. Although the Skindex-29 instrument was shown to be a good predictor of patient HRQoL, the full impact of CTCL on patient HRQoL may not have been captured, and further work is required to determine how best to capture the impact of CTCL on patient quality of life.

Lastly, the model base case was from the perspective of the Canadian publicly funded healthcare system. Although the model offers a comprehensive look at the total cost of treating CTCL, results may not be applicable to other health systems.

Conclusions

Based on clinical data from the ALCANZA trial and Canadian clinical expert opinion, this analysis found that brentuximab vedotin is a cost-effective treatment for CTCL that significantly improved response rate, delayed disease progression, and allowed a higher proportion of patients to achieve symptom control by reducing disease burden compared with MTX or BEX for the treatment of CTCL. Treatment with brentuximab vedotin compared with MTX or BEX incurred higher drug costs; however, most of those costs were offset through reduced costs in the post-progression and end-stage management health states. Results from sensitivity analyses showed that brentuximab vedotin remained cost-effective when key model parameters were varied under many different structural assumptions. In summary, brentuximab vedotin provides a cost-effective treatment option with superior efficacy and a comparable safety profile to physician’s choice of MTX or BEX, from the perspective of the Canadian publicly funded healthcare system.