Economic evaluation of therapeutic cancer vaccines and immunotherapy: A systematic review

Abstract

Cancer immunotherapy is a rapidly growing field in oncology. One attractive feature of cancer immunotherapy is the purported combination of minimal toxicity and durable responses. However such treatments are often very expensive. Given the wide-spread concern over rising health care costs, it is important for all stakeholders to be well-informed on the cost and cost-effectiveness of cancer immunotherapies. We performed a comprehensive literature review of cost and cost-effectiveness research on therapeutic cancer vaccines and monoclonal antibodies, to better understand the economic impacts of these treatments. We summarized our literature searches into three tables by types of papers: systematic review of economic studies of a specific agent, cost and cost-effectiveness analysis. Our review showed that out of the sixteen immunotherapy agents approved, nine had relevant published economic studies. Five out of the nine studied immunotherapy agents had been covered in systematic reviews. Among those, only one (rituximab for non-Hodgkin lymphoma) was found to be cost-effective. Of the four immunotherapy drugs not covered in systematic reviews (alemtuzumab, ipilimumab, sipuleucel-T, ofatumumab), high incremental cost-effectiveness ratio (ICER) was reported for each. Many immunotherapies have not had economic evaluations, and those that have been studied show high ICERs or frank lack of cost-effectiveness. One major hurdle in improving the cost-effectiveness of cancer immunotherapies is to identify predictive biomarkers for selecting appropriate patients as recipients of these expensive therapies. We discuss the implications surrounding the economic factors involved in cancer immunotherapies and suggest that further research on cost and cost-effectiveness of newer cancer vaccines and immunotherapies are warranted as this is a rapidly growing field with many new drugs on the horizon.

Introduction

Cancer immunotherapy or immune-oncology was first noted by William B. Coley in the 19th century when bacterial toxins injected into cancer patients led to occasional cures. The field remained dormant until the late 20th century when tumor-associated antigens and tumor-infiltrating lymphocytes were discovered and non-specific pro-inflammatory cytokines like interleukin-2 (IL-2) were noted to generate robust immune responses against some neoplasms.¹ The development of advanced biotechnology techniques over the last decade has accelerated our understanding of the immune system. It has recently been proposed that one of the eight hallmarks of malignant growth of tumors is the evasion of cancer cells from attack and elimination by the immune system.^2,3

Although there is no one standard definition of cancer immunotherapy, over the years three very broad categories have evolved. The first category consists of non-specific cytokines such as IL-2 and interferons (IFN) that have been shown to have modest success in melanoma and renal cell carcinoma, but drugs in this category carry significant toxicities and are infrequently used today.^4,5 The second category comprises of cancer vaccines, a form of active immunotherapy, designed to actively kill neoplasms via the body's own induction of tumor antibodies or T cells.⁶

The third and largest class of cancer immunotherapeutics includes monoclonal antibodies (mAbs) which are a form of passive immunotherapy and are the most commonly used form of immunotherapy in oncology.⁷ Monoclonal antibodies can be further subdivided into five categories: (1) mAbs directed against a tumor-specific antigen such as rituximab (CD-20) or alemtuzumab (CD-52); (2) mAbs directed against cellular receptors such as cetuximab (anti-EGFR) or trastuzumab (anti-Her2 receptor); (3) mAbs that target soluble growth factors such as bevacizumab that interacts with vascular endothelial growth factor (VEGF); (4) mAbs that are immunoconjugates, combining a chemotherapy or radioisotope agent with an anti-immune antibody such as T-DMI (trastuzumab with emtansine) or ibritumomab tiuxetan; and (5) mAbs that are directed against negative regulators of the immune system such as ipilimumab (checkpoint inhibitors).

The field of cancer immunotherapy is witnessing a renaissance in the past 5 years, with rapid development of novel monoclonal antibodies and therapeutic cancer vaccines for multiple tumor types. For example, the first two immunotherapeutics to be FDA approved based on improvements in overall survival in large phase III trials have both been approved since 2010: sipuleucel-T (Provenge®), a therapeutic cancer vaccine for castrate-resistant prostate cancer and ipilimumab (Yervoy), a T cell potentiating monoclonal antibody for metastatic melanoma.^8,9 With further phase III development of therapeutic vaccines (e.g., clinicaltrials.gov NCT01582672, NCT01322490) and exciting early clinical results using the anti-programmed cell death (anti-PD-1) and anti-programmed cell death ligand (anti-PD-L1) checkpoint inhibitor mAbs, cancer immunotherapy is bound to become a more prominent pillar of oncologic care.

As the oncology community celebrates therapeutic innovations made possible from modern cancer immunotherapy, the high price tag associated with these drugs raises concern over the affordability of cancer care.^10,11 mAbs such as rituximab (the first FDA approved mAb in oncology in 1997), trastuzaumab and bevacizumab have through the years been top-selling cancer drugs, generating billions of dollars in revenues for drug manufacturers.¹² The high cost associated with new oncologics, many of which are immunotherapy agents, has cast doubts on the ‘value’ of these novel agents and triggered widespread debates over their cost and cost-effectiveness throughout the world.¹³ A prime example is sipuleucel-T, approved by the FDA in 2010, with a cost of $93,000 for three injections and a median overall survival benefit of 4.1 months.^9,14 The cost of sipuleucel-T sparked a media outcry.¹⁵ The Centers for Medicare and Medicaid Services (CMS) conducted its own evaluation of the drug, a very unusual move for a cancer drug already approved by the FDA, and issued a ruling in March of 2011 that they will allow national coverage.¹⁶ The drug, although approved in the European Union as of 2013, is currently only beginning to be marketed in Europe, and the National Institute for Health and Care Excellence (NICE) is currently working on an analysis of sipuleucel-T to be released in 2015. Whether it will be paid for by EU countries remains unclear. A common tactic of pharmaceutical companies is to benchmark prices of new oncologics against existing products and to set the price at a level that is at least as high as the highest priced competing product.^17,18 Thus, unless therapeutic gains of future immunotherapies are significantly better than sipuleucel-T, similar questions and concerns will likely be raised again and again.

In an era of clearly limited resources and widespread concern over rising health care costs, the debate over the value of immunotherapeutics in oncology care will continue among all stakeholders – clinicians, payers, patients and their family members. The objective of our paper was to review the current status of knowledge on the cost and cost-effectiveness of cancer immunotherapeutics. Due to their high toxicity and limited use in current practice, we excluded non-specific cytokines from our review and focused on therapeutic cancer vaccines and mAbs. Also excluded were prophylactic vaccines because of our focus on immunotherapy for treatment purposes.

Results

Figures 1 and 2 depict the flowchart of the literature search process as suggested by Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).¹⁹ In brief, our phase I search (Figure 1) identified five systematic reviews of economic evaluations for five drugs: trastuzumab, rituximab, cetuximab, bevacizumab, and panitumumab.^20-24 The second-phase search excluded the above five agents and focused on the following eleven drugs: alemtuzumab, brentuximab vedotin, catumaxomab, ibritumomab tiuxetan, ipilimumab, obinutuzumab, ofatumumab, pertuzumab, sipuleucel-T, tositumomab, and trastuzumab emtansine. As shown in Figure 2, we identified 91 studies in our initial screening from the phase II search. After reviewing title/abstract, we removed duplicated, non-English studies, and those irrelevant to economic analysis. Full-text review for the remaining studies led to the exclusion of studies incompatible with our inclusion criteria. Eight studies examining the costs or cost-effectiveness of alemtuzumab, ipilimumab, sipuleucel-T, or ofatumumab, were included in the final analyses.^25-32

Figure 1. Phase 1 search flow chart to identify systematic review(s) regarding the cost or cost-effectiveness of therapeutic cancer vaccines and immunotherapy published after January 1, 2012.

Figure 2. Phase 2 search flow chart to identify studies regarding the cost or cost-effectiveness of therapeutic cancer vaccines and immunotherapy based on our modified drug list.

Table 2 summarizes findings of review articles identified from the phase I search regarding the cost-effectiveness of therapeutic cancer vaccines and immunotherapy in the current literature. As shown, rituximab combined with chemotherapy was considered to be cost-effective for non-Hodgkin's lymphoma (NHL) when compared to chemotherapy alone.^20,21 Trastuzumab was largely considered to not be cost-effective for HER2 positive metastatic breast cancer when compared to conventional chemotherapy, hormonal therapy, or best supportive care (BSC).²² Cetuximab (+/- chemotherapy) was mainly considered not cost-effective for metastatic colorectal cancer when compared to conventional chemotherapy or BSC.^23,24 Bevacizumab (plus chemotherapy) was considered not cost-effective for mCRC when compared to chemotherapy alone.²⁴ Panitumumab was also considered not cost-effective for mCRC when compared to BSC alone.^23,24 However, Kirsten ras oncogene (KRAS) mutation testing before treatment with cetuximab or panitumumab was considered to be cost-effective compared to no testing.²⁴

Table 1. List of all immunotherapy agents approved (bolded therapies were reviewed; italicized therapies were summarized; rest have no published data)

Drug Name	Drug Type	Original Year of FDA Approval	FDA Labeled Oncologic Indication (s)/Target	Cost per administration in the U.S. (AWP)*
Rituximab	mAb	1997	NHL,CLL/CD20	$5,740
Trastuzumab	mAb	1998	Breast cancer, gastric cancer, esophageal cancer/Her-2	$4,022
Alemtuzumab^1	mAb	2001	CLL/CD52	N/A^1
Ibritumomab tiuxetan	mAb (conjugated to Y 90 or In 111)	2002	NHL/CD20	$36,000^2
Tositumomab^3	mAb (conjugated to I 131)	2003	NHL/CD20	N/A^3
Bevacizumab	mAb	2004	Colorectal cancer, lung cancer, GBM, RCC/VEGF	$5,451
Cetuximab	mAb	2004	Colorectal cancer, head and neck cancer/EGFR	$2,871
Panitumumab	mAb	2006	Colorectal/EGFR	$4,669
Catumaxomab^4	mAb	N/A	Malignant ascites/EpCAM	N/A^4
Ofatumumab	mAb	2009	CLL/CD20	$11,150
Sipuleucel-T	Vaccine	2010	Prostate Cancer	$37,200
Brentuximab vedotin	mAb	2011	Hodgkin lymphoma, ALCL/CD30	$15,800
Ipilimumab	mAb	2011	Melanoma/CTLA-4	$31,617
Obinutuzumab	mAb	2013	CLL/CD20	$6,192
Pertuzumab	mAb	2013	Breast Cancer/Her2	$4,890
Trastuzumab emtansine	mAb (conjugated to DM1)	2013	Breast Cancer/Her2	$8,386

¹Alemtuzumab is no longer approved in the United States, but may be provided free of charge for appropriate patients.

² http://interactive.snm.org/docs/HOPPS%202011F%20vs%202012F-Nov-4-2011.pdf

³Tositumomab has been discontinued as of 2/20/2014.

⁴Approved in the EU in 2009.  Not approved by the FDA.

*Based on a 70 kg, 178 cm, 1.86 M² individual; doses based on the following common indications (Rituximab = 375 mg/M²; Trastuzumab = 6 mg/kg; Bevacizumab = 10 mg/kg; Cetuximab = 250 mg/M²; Panitumumab = 6 mg/kg; Ofatumumab = 2000 mg; Ibritumomab tiuxetan = 28 mCi; Sipuleucel-T = per dose price; Brentuximab vedotin = 1.8 mg/kg; Ipilimumab = 3mg/kg; Obinutuzumab = 1000 mg; Pertuzumab = 420 mg; Trastuzumab emtansine = 3.6 mg/kg); source UpToDate® except where indicated.

Table 2. Summary of systemic review of economic evaluation for cancer immunotherapy/vaccine published since 2012

Study & year of publication	]Drug(s), disease, & treatment setting	Numbers of studies included	Results	Authors’ Conclusion	Comments
Parkinson et al (2014)	Trastuzumab Her2 + MBC treatment containing trastuzumab versus a comparator (C/T, hormonal therapy, or BSC)	15	1st line: ICUR: $47,332 - 186,000/QALY; 2^nd line: ICUR: $77,476/QALY	“none of the evaluations reported an ICER in the range that trastuzumab would be considered cost-effective” “There were numerous drivers of the different conclusions regarding the cost-effectiveness of trastuzumab, many of which are due to judgments made by the authors when translating data from RCTs”	Reference year for cost: 2008-2009
Papaioannou et al (2012)	Rituximab FL (a subtype of NHL) Rituximab + C/T vs C/T alone	4	ICUR (based on the author's own estimates of HTA reports) $11,677 – 16,388/ QALY	“This assessment provides an indication of the cost-effectiveness of the addition of rituximab to CVP, CHOP and MCP in a UK setting”	This study is a UK review of technology appraisal (No. 110)
Auweiler et al (2012)	Rituximab NHL Rituximab + standard C/T vs standard C/T alone	14	All ICER below country-specific threshold: $9,836 –39,328/ QALY	“Adding rituximab to standard chemotherapy is considered a cost-effective treatment option for NHL”	Follicular lymphoma was included but not separated from other subtypes of NHL in this study.
Hoyle et al (2013)	Cet, Bev, Pan mCRC (after 1st line); Cet (mono- or combination C/T), Bev (combination with non-oxaliplatin C/T) and Pan (monotherapy) vs irinotecan- or oxaliplatin-based C/T or BSC	5	ICUR (based on the authors’ own estimates of HTA reports) Cet vs BSC: $144,543/QALY; Cet + irinotecan vs BSC: $129,794/QALY; Pan vs BSC: $221,239/QALY; Bev: not modeled	“used for third- and subsequent-line treatment relative to best supportive care, cetuximab plus best supportive care, cetuximab plus irinotecan plus best supportive care and panitumumab plus best supportive care are effective but not cost-effective if a decision threshold of £20,000 per QALY or £30,000 per QALY is used.”	This study is a UK review of technology appraisal (No. 150 and No. 118) All included studies were also included in Lange et al (2014)
Lange et al (2014)	Cet, Bev, Pan mCRC various combinations	15	ICER/ICUR Bev plus C/T vs C/T (1^st line): $1,047 - $149,614/LY; Bev + C/T vs Cet + C/T (1^st line): $19,893/LY; Cet (+/- C/T or KRAS testing) vs other treatment (conventional or BSC or Cet without KRAS testing): $21,033 – 2,932,767/LY; $37,363 – 416,648 /QALY; Pan vs BSC: $41,812 – 269,649 /QALY; Different treatment sequences: $170,896 –243,096/LY	“The treatment with bevacizumab, cetuximab and panitumumab is mainly considered to be not cost-effective in patients with mCRC. However, testing for Kirsten ras oncogene (KRAS) mutation prior to the treatment with cetuximab or panitumumab is found to be clearly cost-effective compared to no testing “	Reference year for cost: converted to US dollars at the exchange rate prevalent in the year of publication.

Bev: bevacizumab; BSC: best supportive care; Cet: cetuximab; C/T: chemotherapy; FL: follicular lymphoma; ICER: incremental cost-effectiveness ratio; ICUR: incremental cost-utility ratio; KRAS: Kirsten ras oncogene; LY: life-year; MBC: metastatic breast cancer; mCRC: metastatic colorectal cancer; NHL: non-Hodgkin's lymphoma; Pan: panitumumab; QALY: quality-adjusted life years.

There were four cost papers identified from the phase II search, covering alemtuzumab (for chronic lymphocytic leukaemia, CLL), ipilimumab (for metastatic melanoma) and sipuleucel-T (for castration-resistant prostate cancer, CRPC) (table 3).^25-28 Scott et al reported that alemtuzumab was cost saving ($11,181/patient) in New Zealand when compared to chemotherapy.²⁷ Jarkowski et al reported that ipilimumab with dose rounding was cost saving ($7,503/patient) in the United States when compared to ipilimumab without dose rounding.²⁸ Two commentary-type papers calculated the incremental drug cost ($98,780 for a full course) of sipuleucel-T and commented on the high cost of this immunotherapy agent.^25,26

Table 3. Summary of cost studies

Study & year of publication, country	Target population & intervention	Study details	Results	Conclusion
Scott et al (2007) New Zealand	3^rd line treatment of CLL Intervention: Alemtuzumab vs Chemotherapy (FCR)	Perspective: Payer (New Zealand Pharmaceutical Management Agency (PHARMAC)/District Health Board) Cost type: Direct medical cost Reference year for cost: 2006 Time horizon: Lifetime Parameter sources: The literature Discount rate: NS Approach: Modeling	Cost: $36,758 vs 47,938 One-way sensitivity analysis: elasticity: -1.58 ∼ 2.72	Incremental cost: Cost saving $11,181/patient Main conclusion: “Third-line treatment of eligible patients with alemtuzumab was found to be …less costly than FCR.”
Jarkowski et al (2014) US	Metastatic melanoma Intervention: Ipilimumab with dose rounding vs ipilimumab without dose rounding (hypothetically)	Perspective: Payer Cost type: Direct medical cost (drug) Reference year for cost: 2011 Time horizon: NS Parameter sources: The literature Discount rate: NS Approach: Hybrid of clinical database and modeling, with straightforward calculation	Cost: Individual cost saving ranged from $0 ∼ $16,824; One-way sensitivity analysis: NS	Incremental cost: Overall cost saving: $7,503 /patient Main conclusion from the authors: “Dose rounding of ipilimumab … has the potential to result in a significant cost savings.”
Vasani et al (2011) US	CRPC Intervention: Sipuleucel-T vs placebo (design of IMPACT trial)	Perspective: Payer Cost type: Direct medical cost (drug) Reference year for cost: 2011 Time horizon: During treatment Parameter sources: The literature Discount rate: NS Approach: Modeling, with straightforward calculation	Cost: NS; One-way sensitivity analysis: NS	Incremental cost: $ 98,780; One-way sensitivity analysis: NS; Main conclusion from the authors: “With these costly therapies all having potential application in the patient with mCRPC, multiple entities (industry, government, and the general public) must strategize to determine how the cost burden of these agents can be balanced” Note: This paper is a commentary rather than a primary study. The data extracted was assumed based on Table 1 which cites The New York Times as a source.
Klemm et al (2012) US	CRPC Intervention: Sipuleucel-T vs placebo (design of IMPACT trial)	NA	Drug cost $ 98,780 for a full course (assumed the same reference year as Vasani et al 2011; no detail or reference provided)	Note: This paper is a primer rather than a primary study.

CLL: chronic lymphocytic leukaemia; CRPC: castration-resistant prostate cancer; FCR: fludarabine, cyclophosphamide and rituximab; NA: not applicable; NS: not specified; NZ: New Zealand Dollar.

Four cost-effectiveness studies were identified from the Phase II search, covering alemtuzumab (for T-cell prolymphocytic leukemia, T-PLL), ipilimumab (for previously treated advanced melanoma), sipuleucel-T (CRPC), and ofatumumab (for CLL) (table 4).^29-32 Lu et al reported that while alemtuzumab (vs. chemotherapy) was likely to be cost-effective for T-PLL, especially if used earlier in the course of treatment, the cost-effectiveness of alemtuzumab is highly uncertain (ICER $42,710 ∼ $119,701/QALY among four scenarios).²⁹ Barzey et al reported that the probability that ipilimumab was cost-effective relative to BSC at willingness-to-pay (WTP) of $146,000 USD/QALY was 95% for patients with previously treated advanced melanoma. It should be noted that while the authors concluded that ipilimumab was cost-effective, the ICER estimated from their study was $140,811/QALY, which was higher than commonly accepted cost-effectiveness thresholds (i.e., $50,000 ∼ $100,000/QALY).³⁰ Holko et al reported that sipuleucel-T plus standard treatment was not cost-effective (ICER $289,964/QALY) for CRPC when compared to standard treatment alone.³¹ In an appraisal of a manufacturer's submission, Hoyle et al reported that ofatumumab for CLL was associated with a high ICER ($126,756/QALY) when compared to BSC.³²

Table 4. Summary of cost-effectiveness studies

Study & year of publication, country	Target population & intervention	Study details	Results	Conclusion
Hoyle et al (2011) UK	CLL refractory to fludarabine and alemtuzumab Intervention: Ofatumumab vs BSC	Perspective: Payer (NHS in manufacture's submission, see table 1) Cost type: NS Reference year for cost: 2011 (published in 2011, NS in the text) Time horizon: 10 years Parameter sources: Clinical trial & the literature; Discount rate:3.5% (according to manufacturer's submission) Approach: Modeling, Area-Under The-Curve	Effect: Incremental survival more than 5 months (according to manufacturer's submission); Cost: Incremental cost $21,098 (according to manufacturer's submission); One-way sensitivity analysis: NS	ICER/ICUR $59,723/QALY (according to manufacturer's submission) $126,756/QALY (calculated by the authors); PSA: NS by the author; Main conclusion: “Ofatumumab is not recommended for the treatment of chronic lymphocytic leukaemia that is refractory to fludarabine and alemtuzumab.” Note: This paper is an appraisal of a manufacturer's submission rather than a primary study.
Lu et al (2012) UK	T-PLL Intervention: Alemtuzumab vs Chemotherapy	Perspective: Payer (NHS) Cost type: Direct medical cost (drugs, drug administration and management of opportunistic infections) Reference year for cost: 2012 Time horizon: Lifetime Parameter sources: Case series & the literature Discount rate: 3.5% Approach: Decision-analytic modeling	Effect: Incremental QALY: 0.1596 ∼ 0.2449 (among four scenarios); Cost: Incremental cost: $8,031 ∼ $ 19,889 (among four scenarios); One-way sensitivity analysis: ICER > = $28,129/QALY;The results were most sensitive to changes in survival curves and utility values.	ICER/ICUR: $42,710 ∼ $119,701/QALY (among four scenarios); PSA: Probability cost-effective at WTP €36,300 ($39,729)/QALY: 0 ∼ 53% (among four scenarios); Main conclusion: “Cost-effectiveness is highly uncertain and future research is clearly justified.”
Barzey et al (2013) US	Previously treated advanced melanoma Intervention: Ipilimumab vs BSC	Perspective: Payer Cost type: Direct and indirect cost (drug, disease, toxicity) Reference year for cost: 2010 Time horizon: Lifetime Parameter sources: Clinical trial & the literature Discount rate: 3% Approach: Markov model	Effect: 2.88 vs 1 LY; 1.76 vs 0.62 QALY; Cost: $184,531 vs $23,954; One-way sensitivity analysis: ICER > = around $100,000/QALY The results were most sensitive to changes in time horizon and utility values.	ICER/ICUR: $85,608/LY; $140,811/QALY; PSA: Probability cost-effective at WTP $113,000 ∼ 146,000/QALY: 6% ∼ >95%; Main conclusion: “The estimated cost-effectiveness of ipilimumab … to be acceptable.”
Holko et al (2013) US	Asymptomatic or minimally symptomatic CRPC Intervention: Sipuleucel-T + standard treatment vs standard treatment	Perspective: National health system Cost type: Direct medical cost Reference year for cost: 2012 Time horizon: Lifetime Parameter sources: Clinical trial & the literature Discount rate: 3% Approach: Markov model	Effect: 2.64 vs 2.16 LY; 2.01 vs 1.64 QALY; Cost: $159,818 vs $52,710; One-way sensitivity analysis: ICER > = $154,716/QALY; The results were most sensitive to changes in hazard ratio for overall survival and cost of sipuleucel-T.	ICER/ICUR: $222,340/LY; $289,964/QALY; PSA: Probability cost-effective at WTP $149,000/QALY: 3.5%; Main conclusion: “Sipuleucel-T immunotherapy treatment at the current price … is not cost-effective.”

BSC: best supportive care; CLL: chronic lymphocytic leukaemia; CRPC: castration-resistant prostate cancer; ICER: incremental cost-effectiveness ratio; ICUR: incremental cost-utility ratio; NS: not specified; PSA: probabilistic sensitivity analysis; QALY: quality-adjusted life years; T-PLL: T-cell prolymphocytic leukemia; WTP: willingness-to-pay.

Discussion

To our knowledge this is the most up-to-date comprehensive systematic review of economic evaluations of immunotherapies in cancer in the recent era. We a priori excluded studies examining non-specific cytokines (IL-2 and IFN) and non-therapeutic vaccines, focusing our review on therapeutic cancer vaccines and mAbs. This left sixteen agents, all but one of which is a form of a monoclonal antibody. Among those, we identified economic studies for nine immunotherapy agents, leaving both costs and cost-effectiveness of seven agents unexplored. We reached the following conclusions from our review of economic studies of cancer immunotherapeutics.

First, we found that for around one-third (5/16) of agents (rituximab, trastuzumab, bevacizumab, cetuximab and panitumumab) systematic reviews regarding their cost-effectiveness have already been performed since 2012. Most (4/5) of these agents were considered as not cost-effective in the systematic reviews (Table 2) with the noted exception of rituximab which has revolutionized the field of lymphoma.³³ Nevertheless, all are widely used in oncology practice throughout the world (and in particular in the United States) and clearly have provided clinical benefit for some (but not all) patients who are treated with them. The degree of that clinical benefit varies, and the lack of robust predictive biomarkers (for example, patients who are HER2 positive or KRAS wildtype do not all respond similarly to trastuzumab or cetuximab) makes it impossible to predict a response for an individual patient with any certainty, thus leading to widespread use of the drugs for all eligible patients. Inappropriate use of these targeted agents could be detrimental to patients, as exemplified by an increased risk of heart disease among cancer patients who received traztuzumab without documented evidence of HER2-testing.³⁴

Second, among the remaining 11 immunotherapy drugs, only four (alemtuzumab, ipilimumab, sipuleucel-T, ofatumumab) had been studied for either their cost or cost-effectiveness. Four of the 7 drugs not studied have been approved after 2011 (Table 1), and therefore their exclusion is not surprising. Of the remaining three, one (catumaxomab) has never been approved by the FDA while the other two (tositumomab and ibritumomab tiuxetan) are rarely used in practice and in fact as of this writing tositumomab has been withdrawn from the market by GlaxoSmithKline due to its current limited use and continual projected decline given the other treatment options available for patients with non-Hodgkin lymphoma. We expect an increasing number of economic evaluation studies of immunotherapy agents as time goes by. It should be noted that due to the paucity of literature in this area, our review also included several commentaries that touched on the economics of immunotherapy agents, which is a major limitation of this review.

Third, in our review of the cost-effectiveness analyses mentioned above, all drugs were associated with a high ICER, with the highest ICER found in sipuleucel-T vs. standard treatment for castration-resistant prostate cancer. The ICERs were in line with many of the modern targeted oncologic drugs,³⁵ whether IV or oral, and support the thesis that modern day oncologic care is expensive with often times mild gains for significant prices.^10,36 In part this is due to the simple premise that cancer can be a deadly diagnosis for which for decades physicians had limited tools. Thus, when new therapies become available in the market both patients and doctors want to use them even though many of them may at best improve overall survival by only a few months, with some notable exceptions such as imatinib. As such, an increasing number of cancer patients are reported to experience financial hardship.^37,38 Concerns of financial and emotional distress from the use of costly new cancer drugs that offer marginal benefits, especially among patients with incurable cancers, have prompted the IOM Committee on the Quality of Cancer Care to recommend providers to communicate information on both total and out-of-pocket costs with patients and to promote early communications of supportive care.³⁹

Over the last several years, immunotherapy has experienced a revival. This has been driven by the approval of agents such as ipilimumab and sipuleucel-T, and also by a host of early phase trials with extremely promising results using both vaccines and mAbs.^40-42 In particular, the elucidation of the importance of immune checkpoint pathways in the progression of cancer has led to the development of novel mAbs aimed and disinhibiting the immune system in the presence of cancer. A prime example is the development of PD-1 agents targeting T-cell activation, which has shown exciting results in many solid tumors.^42-44 Several other “checkpoint” pathways are now being targeted in the clinical pipeline, as well as the development of a host of other antibody-drug conjugates, adoptive cell therapies and therapeutic vaccines.^45,46 Recent attempts to develop therapeutic HBV (e.g., clinicaltrials.gov NCT01374308) or HPV (e.g., clinicaltrials.gov NCT01957878) vaccines for use in already infected individuals have also showed early promise and may pave the way to the interruption of the infection-to-cancer progression.^47,48

Part of the allure of immunotherapy is the purported combination of minimal toxicity with durable responses due to immunologic memory, the specificity of the binding mAbs to a single target antigen, and the absence of mutation-induced resistance that has made it attractive both to scientists and the lay press.^{49, 50} However, as with any anti-cancer therapy, immunotherapy carries possible side-effects, and these occasionally can be quite significant, eliciting for example a robust auto-immune reaction. Thus, their capacity to be used in practice outside carefully monitored clinical trials remains to be carefully defined.

Despite the excitement, there are many concerns about the cost-effectiveness and budget impact of the coming wave of checkpoint inhibitors.¹³ Given their apparent efficacy, pharmaceutical companies are likely to seek prices as high or higher even than those for recent novel drugs. With innovator firms seeking regulatory approvals for combinations of checkpoint inhibitors with other immunotherapies and with targeted oral therapies, and for indications in multiple high prevalence cancers including melanoma and lung cancer and for extended or even chronic care use,⁵¹ overall expenditures on these inhibitors threaten to be unsustainable.

Competition among the new therapies offers some measure of hope for payers attempting to restrain prices. Drug development costs may well be dropping as new clinical trial protocols such as the multi-agent, multi-sponsor lung cancer Master Protocol, combined with designation of many new drugs under the Breakthrough Therapy FDA approval pathway, lead to shorter development times.⁵² As one example, Merck's MK-3475 checkpoint inhibitor moved from initial human testing to regulatory filing in under three years.⁵¹ Lower development costs would be expected to attract more entrants and competition in this drug class. In addition, the historical willingness by oncologists to prescribe off-label combinations of drugs may offer opportunities to avoid very high pricing of regulator-approved drug combinations. Public payers in a number of countries frequently do not cover many high-priced therapies on grounds of cost-effectiveness.⁵³ However, private payers have shown only tentative signs of being willing to engage in tough management of novel oncologics, even for cancers with multiple therapeutic options offering room for price negotiations.⁵⁴ Finally, other potential ways to decrease the high cost of immunotherapy may be by reducing the number of doses administered in a particular treatment regimen or even by reducing the actual dose amount per administration of drug. For example, it is currently unknown how much or for how long one needs to receive ‘checkpoint inhibition therapy’ in order to develop and sustain the appropriate immunologic response and future trials will be necessary to address this question. Simply continuing therapy indefinitely once a response is attained may not be necessary when being treated with novel immunotherapy agents as opposed to the more traditional chemotherapy drugs.

Finding robust predictive biomarkers that would help select appropriate drugs for specific patients could be of major help in reducing costs & improving treatment effectiveness, but faces major obstacles. This challenge can be clearly illustrated by the example of sipuleucel-T. With a cost of close to $100,000 per patient and a median overall survival advantage of 4.1 months in those with metastatic castration resistant prostate cancer, the drug at first glance appears not to be cost-effective. The challenge then is whether a subset of patients can be identified prior to therapy initiation who may be predicted to derive significant benefit from the vaccine, while not subjecting those predicted to obtain minimal to no benefit to the treatment. Restricting the drug to the appropriately targeted subset of patients would lower the ICER and theoretically make the intervention cost-effective. Carefully matching patients with therapies based on their biomarkers will continue to be a focus of future trials and should hopefully help to improve the cost-effectiveness of immunotherapy in the future. The success of using biomarkers to enhance the cost-effectiveness of immunotherapy will hinge upon the accuracy as well as costs of tests for biomarkers.

In summary, our systematic review has demonstrated that with the exception of rituximab, most immunotherapies in oncology today either have not had an economic evaluation or have a relatively high ICER. In several instances, mAbs have been found not to be cost-effective. Given that oncology is just beginning to explore and evaluate modern day immunotherapies such as checkpoint inhibitors and therapeutic vaccines, their role may indeed prove to be vital and cost-effective in the coming years, but further evaluation is required.

Methods

We performed a comprehensive literature search for articles examining the cost or cost-effectiveness of therapeutic cancer vaccines and immunotherapy. Given the lack of standard definition of cancer immunotherapy, we determined the drugs to be included in our review as those that had been approved by the Food and Drug Administration (FDA)⁵⁵ or the European Medicines Agency (EMA)⁵⁶ as therapeutic cancer vaccines or mAbs before 2014. The master list of drugs included: trastuzumab, pertuzumab, cetuximab, panitumumab, bevacizumab, rituximab, ofatumumab, alemtuzumab, obinutuzumab, ipilimumab, tositumomab, ibritumomab tiuxetan, trastuzumab emtansine, brentuximab vedotin, sipuleucel-T, and catumaxomab (Table 1). Because some drugs on the list have been on the market for several years, there are recently published systematic reviews of economic evaluations of these drugs. Our approach was to avoid replicating information already available in the literature and to focus on economic studies of immunotherapy agents that have not been covered in review articles. However, we did provide a brief summary of findings from the systematic reviews so as to inform readers of the current state of knowledge on the economics of selected cancer immunotherapeutics. Thus, we conducted our searches in two phases. Phase I identified immunotherapy agents that have already been discussed in systematic reviews published after January 1, 2012, so as to restrict to recently published systematic reviews. We used balanced search filters for costs or economics in the PubMed ((“costs and cost analysis”[MeSH] OR costs[Title/Abstract] OR cost effective*[Title/Abstract]) OR (cost*[Title/Abstract] OR “costs and cost analysis”[MeSH:noexp] OR cost benefit analysis*[Title/Abstract] OR cost-benefit analysis[MeSH] OR health care costs MeSH : noexp)) as used in the literature^57-59 and “cancer” and the above drug list in Pubmed⁶⁰ and limited to systematic reviews published after January 1, 2012. After excluding the subset of the drugs identified from phase I (decided by consensus between two authors (Chien & Shih)), we then proceeded to phase II in which we performed a comprehensive literature search using the remaining drugs on the master list (the modified list) AND the above search filters for costs or economics AND cancer in Pubmed on Feb 6th, 2014. We revised the search terms (any drugs on the modified list) AND cancer AND cost* in title/abstract/keywords for Cochrane database⁶¹ and also searched for drugs on the modified drug list in ‘Title’ in the advanced search of Health Technology Assessment in UK.⁶² After removing duplicated and non-English studies, potentially relevant studies were selected by two independent reviewers (Chien & Shih) who reviewed titles and/or abstracts. Disagreement was resolved by discussion between the reviewers to reach a final consensus. Further review of the full paper of the above searches led to the decision on studies to be included in the final analyses, again by two independent reviewers (Chien & Shih). Disagreement was resolved by consensus after discussion.

We summarized studies identified from phases I and II searches in tables, categorized by type of papers: systematic review, cost, and cost-effectiveness. For systematic reviews, we synthesized information retrieved from the papers as follows: authors and year of publication, drug(s) and disease and treatment setting, numbers of studies included, results, and authors’ conclusion. For cost or cost-effectiveness papers, we synthesized information as follows: authors and year of publication, country, target population & intervention, study details (including perspective, cost type, reference year for cost, time horizon, parameters sources, discount rate and approach), results (including cost and/or effectiveness, and findings from one-way sensitivity analysis if available), and conclusions (including incremental cost or incremental cost-effectiveness ratio (ICER), probabilistic sensitivity analysis (PSA) if available, and the main conclusion from the authors). When multiple outcomes (such as life-year (LY) or quality-adjusted life (QALY)) were reported, we mainly extracted the one(s) that allowed for comparison with other studies. For reference year for cost, we used the number as reported in the paper or based on the most relevant year of data used or the year of publication if not clearly specified in the studies. Cost was expressed in 2013 USD by applying the local consumer price index and purchasing power parity index (PPP)^63,64 For the threshold value reported in probabilistic sensitivity analyses based on currency other than $US, we cited the original value as well as $US converted using PPP in 2013.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

Drs. Shih and Smieliauskas are supported by a grant from the Agency for Healthcare Research and Quality (R01 HS018535), and Dr. Shih as well by The University of Chicago Cancer Research Foundation Women's Board. Dr. Chien is supported by a grant from the China Medical University Hospital (DMR-103-043), and the Ministry of Health and Welfare (MOHW 103-TD-B-111-03). Dr. Geynisman is in part supported by an independent grant from Pfizer, Inc. # 11703561.