How valuable is a cancer therapy? It depends on who you ask.

How valuable is a cancer therapy? It depends on who you ask. Comparing and contrasting the ESMO Magnitude of Clinical Benefit Scale with the ASCO Value Framework in Cancer Ram Subramanian Kevin Schorr December 2015 www.simon-kucher.com

Overview Earlier this year, the European Society for Medical Oncology (ESMO) published a tool to assess the magnitude of clinical benefit for cancer therapies the ESMO Magnitude of Clinical Benefit Scale (ESMO-MCBS). The tool allows one to derive a relative ranking of the magnitude of clinically meaningful benefit that can be expected from a cancer therapy based on published data from randomized clinical trials. Shortly after ESMO-MCBS v1.0 was published, the American Society of Clinical Oncology (ASCO) published its Value Framework in Cancer. The ASCO framework enables comparisons of a new treatment regimen with the prevailing standard of care for a specific clinical cancer indication based on data derived from a prospective randomized clinical trial. With two value assessment tools available, the natural questions are how these tools compare and what the implications for manufacturers of cancer therapies are. This article will answer the following questions: What are the ESMO-MCBS and the ASCO Value framework in cancer? How do the two approaches differ in evaluating a cancer therapy? What are the implications for manufacturers of cancer therapies? The ESMO and ASCO value frameworks for cancer therapies The ESMO Magnitude of Clinical Benefit Scale version 1 (ESMO-MCBS v1.0) is a tool to assess the magnitude of clinical benefit for cancer medicines that uses a structured approach to derive a relative ranking of the magnitude of clinically meaningful benefit that can be expected from a cancer therapy. 1 Figures 1, 2a, 2b, and 2c illustrate the algorithm used to calculate the ESMO Clinical Benefit grade. The ASCO value framework is a physician-guided tool to help the physician and patient with shared decision-making. 2 It enables a comparison of a new treatment regimen with the prevailing standard of care for a specific clinical cancer indication based on data derived from a prospective randomized clinical trial. The framework calculates a net health benefit (NHB) score by awarding (or subtracting) points for clinical benefit and toxicity. The NHB is juxtaposed with the direct cost of treatment, to provide an overall summary assessment. Figures 3 and 4 illustrate the algorithm used to calculate the NHB. 1 Cherny et al. Annals of Oncology, May 2015. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) 2 Schnipper et al. Journal of Clinical Oncology, June 2015. ASCO Statement: A conceptual framework to assess the value of cancer treatment options. 2

Differences in approach and implications for manufacturers 1. ESMO scores are likely to play a larger role than ASCO scores in payer evaluations of solid cancers There appears to be a difference in the intent between the two frameworks. The ESMO framework explicitly states that the tool is a first step to addressing the critical public policy issue of value in cancer care, helping to frame the appropriate use of limited public and personal resources to deliver cost-effective and affordable cancer care. The authors of ESMO-MCBS v1.0 state that drugs or treatment interventions that obtain the highest scores on the scale will be emphasized in the ESMO guidelines, with the hope that they will be rapidly endorsed by health authorities across the European Union. ASCO has characterized its framework as a tool to help physicians and patients with shared decision-making. The emphasis appears to be more on educating the oncologist about the importance of discussing costs associated with recommended treatments and empowering patients to ask questions pertaining to the anticipated costs of their treatment options. While ASCO acknowledges that ultimately their framework may end up influencing policy positions, it does not seem that influencing public policy is the primary intent for now. Breakthrough therapies or therapies on an accelerated approval pathway that are being evaluated based on single-arm studies cannot be compared across both frameworks. While ESMO-MCBS v1.0 has only been developed for solid cancers, the ASCO framework applies to all types of cancers. 2. Breakthrough therapies or therapies on an accelerated approval pathway that are being evaluated based on single-arm studies cannot be compared across both frameworks The ESMO-MCBS v1.0 can only be applied to comparative research outcomes and is, therefore, not applicable when evidence of benefit derives from single-arm studies. ESMO acknowledges that this limits its utility in the uncommon situation in which registration is granted on the basis of outcomes reported from single-arm studies. ASCO, on the other hand, allows for the evaluation of therapies that have been approved based on single-arm studies by using response rate (RR) to evaluate the therapy. 3. Differences in the rigor of clinical trial design and performance of the control arm are likely to be reflected in the ESMO score to a greater extent than in the ASCO framework While ASCO considers a hazard ratio (HR) threshold in the assessment of curative therapies, it does not do so in the palliative setting. In contrast, ESMO considers HR when evaluating therapies in the palliative setting (non-curative therapies). ESMO-MCBS v1.0 relies on HR thresholds in the evaluation instead of percent change in median OS/PFS. 3

ESMO-MCBS v1.0 takes into account the absolute performance of the control arm when evaluating non-curative therapies. Form 2a is stratified by median OS of the control arm 12 and >12 months. Similarly, Form 2b is stratified by median duration of PFS of the control arm 6 and >6 months. The ASCO framework does not account for such absolute performance of the control arm. 4. Curative therapies that lack mature survival data will require re-evaluation in the ESMO framework While both the ASCO value framework and ESMO-MCBS v1.0 consider upgrades in score based on non-efficacy measures, only the EMO-MCBS v1.0 accepts improvements in QoL in their grading criteria. The ASCO framework intentionally excludes QoL measurements, which are commonly derived from patient-reported outcomes, as they consider that these measures can be subjective as well as sometimes inconsistent. While ESMO-MCBS v1.0 allows for evaluation of curative therapies based on early data demonstrating high disease-free survival (DFS) in the absence of mature survival data, the evaluation is considered temporary. ESMO states that the therapy will need to be re-evaluated when mature survival data become available. The ASCO framework does not call for a re-evaluation when the initial assessment is made on DFS. 5. Therapies that show late survival gains will benefit from the ESMO evaluation While the ASCO framework is based on an evaluation of median survival gains, ESMO-MCBS v1.0 also allows for an evaluation of non-curative therapies based on a late survival advantage. It does this by looking at HR which captures the reduction in risk across the entire patient timeline (including late responders) and not just improvements in medians. 6. Therapies that show improvements in quality of life (QoL) will benefit in the ESMO evaluation While both the ASCO value framework and ESMO-MCBS v1.0 consider upgrades in score based on non-efficacy measures, only the ESMO-MCBS v1.0 accepts improvements in QoL in their grading criteria. The ASCO framework intentionally excludes QoL measurements, which are commonly derived from patient-reported outcomes, as they consider these measures to be subjective and sometimes inconsistent. In contrast, if a clinical trial includes a QoL improvement as a secondary endpoint, the ESMO score can be upgraded. Additionally, if QoL is reported as a secondary measure, but does not demonstrate improvement, an ESMO score may be downgraded. 7. Reported treatment toxicities are accounted for differently between the ASCO and ESMO scales Although both the ASCO value framework and ESMO-MCBS v1.0 take into account reductions in treatment toxicity, the two scales utilize significantly different methods for incorporating toxicity data into their respective treatment ratings. The ASCO framework toxicity bonus is calculated based upon the relative number of grade 3-5 toxicity categories reported. In con- 4

trast, the ESMO-MCBS v1.0 will evaluate toxicity based upon the number of grade 3-5 adverse events reported within a category. Therefore, a treatment which reduces the frequency of a grade 3-5 toxicity without eliminating it as a reported category may result in a positive score adjustment on the ESMO-MCBS v1.0, but may not be reflected in the ASCO framework toxicity bonus. Additionally, the ESMO-MCBS v1.0 may downgrade a score if the treatment is associated with a particularly severe toxi-city including toxic death, cardiovascular ischemia, hospitalization, congestive heart failure, neurotoxicity, or other irreversible toxicities. 8. The role of acquisition cost is limited to the ASCO evaluation only Acquisition cost is part of the ASCO Value assessment framework. While cost does not influence the score of the therapy, it is information that is required to be provided as part of the assessment. ESMO-MCBS v1.0 does not call for any cost-related information to be considered as part of the evaluation. Overall impressions of the ASCO and ESMO rating scales 1. The ASCO Value framework appears to apply more stringent evaluation criteria than its ESMO counterpart A comparative evaluation of the two oncology treatment rating scales suggests it is more difficult to obtain a score within the upper range of the ASCO framework when compared to the ESMO-MCBS v1.0 for the advanced disease setting (Figure 5). Among the 74 non-curative oncology treatments tested through this analysis, the highest ASCO NHB score produced was found to be a 64 out of 130. In contrast, many of the same treatments when evaluated by the ESMO-MCBS v1.0 received a clinical benefit grade of 4 out of 5, and in several cases 5 out of 5 ratings. Therefore, it is important to consider when using the ASCO framework that the entirety of the rating scale is rarely applied, and that a rating which appears to be in the mid-range of the 130-point scale may in fact have an excellent treatment rating relative to other oncology products. The implications of this finding are particularly important when considering that the ASCO Value framework is intended as a physician and patient shared decision-making tool. It is important that physicians understand the scaling of the ASCO framework grading system, since impressions of a score in the mid-range of the ASCO rating scale will not be interpreted appropriately without the knowledge that the upper range of the scale is rarely applicable. 2. ESMO score adjustments are more frequently applied than ASCO bonuses Due to the fact that the ASCO framework intentionally excludes QoL metrics or patient-reported data, it is far less likely that a treatment will receive a 5

palliation bonus under the ASCO evaluation criteria than a QoL score adjustment under the ESMO-MCBS v1.0 evaluation criteria. Additionally, given the fact that it is more likely to reduce the number of adverse events within a grade 3-5 toxicity than to eliminate a grade 3-5 toxicity as a category, a treatment is also more likely to receive a toxicity bonus under the ESMO rating scale than under the ASCO framework. Of the 74 oncology assesments in the advanced cancer setting using both frameworks, 22 assessments included some form of QoL or toxicity bonus under the ESMO-MCBS v1.0 rating scale whereas only 4 assessments included a palliation or toxicity bonus under the ASCO framework. 3. The score a treatment receives on the ESMO rating scale is not While the treatment ratings generated under the ASCO framework are generally positively correlated with the corresponding treatment ratings calculated using the ESMO-BCBS v1.0 evaluation system, neither scale is predictive of the other. predictive of the score the same treatment will receive under the ASCO framework, or vice versa While the treatment ratings generated under the ASCO framework are generally positively correlated with the corresponding treatment ratings calculated using the ESMO-BCBS v1.0 evaluation system, neither scale is entirely predictive of the other in an analysis of 45 unique cancer treatments in the advanced cancer setting (74 assessments in total due to multiple indications, sub-indications, and, in some cases, comparisons against multiple controls). Treatments which earned an ESMO clinical benefit grade of 1 were found to have an NHB score between 11 and 33 out of 130 using the ASCO framework. Similarly, treatments with an ESMO grade of 2 were found to have an ASCO NHB score between 6 and 55 while treatments with an ESMO grade of 3 were found to range in ASCO NHB score from 16 to 55. Treatments with ESMO grades of 4 and 5 ranged on the ASCO NHB scale from 0 to 64 and 16 to 48, respectively. This wide range of overlap between treatment scores indicates that the ESMO rating is by no means predictive of that individual treatment s ASCO score (or vice versa) which may vary widely depending upon the individual treatment characteristics. Figure 5 compares the ASCO Net Health Benefit Score with the corresponding ESMO Clinical Benefit Grade for several oncology therapies. One particular example of this potential scoring disparity between the ASCO and ESMO grading criteria comes in the form of the lung cancer treatment Alimta (pemetrexed) produced by Eli Lilly. Under the ASCO framework, this treatment originally received a net health benefit score of a 0 for failing to demonstrate superior overall survival outcomes to the standard of care utilized in the trial, cisplatin and gemcitabine. At the urging of the manufacturer, the treatment was re-evaluated under the ASCO framework using trial data specific to patients with non-squamous cancers, the target patient population for the treatment according to the FDA label (as opposed to overall patient population used in the trial which included squamous and non-squamous cases). Using the overall survival statistics specif- 6

ic to non-squamous patients, the treatment was now found to earn a score of 16 out of 130 on the ASCO framework scale. 3 However, even with this score adjustment, the ASCO NHB score for Alimta still varied significantly from the grade earned under the ESMO-BCBS v1.0; Alimta earned a clinical benefit grade of a 4 out of 5 on the ESMO-MCBS v1.0 evaluation system. According to ESMO-MCBS v1.0, this treatment earned a preliminary grade of 3 as the two-year survival rate of patients on Alimta improved between 5% and 10% compared to the standard of care. Additionally, the treatment received a toxicity score adjustment which increased the final grade to 4, as Alimta was found to cause significantly fewer cases of grade 3 to 4 neutropenia, anemia, and thrombocytopenia. This toxicity bonus was not reflected under the ASCO framework which only provides a toxicity bonus for fully eliminating a toxicity. In the case of Alimta, some grade 3 to 4 cases of these toxicities still occurred in patients using the treatment, even if they occurred at a significantly lower rate. ESMO and ASCO have taken steps to guide the medical community in assessing the value of oncology therapies by releasing their value assessment frameworks. However, the frameworks are not identical and there are several instances where therapies appear to be valued differently. The substantial difference in scores for Alimta, a 16 out of 130 on the ASCO framework as opposed to a 4 out of 5 on the ESMO-MCBS v1.0, is illustrative of the notion that the differences between the evaluation criteria of the two frameworks can result in considerably different assessments of value. Conclusion With increasing budget constraints and the need to make trade-offs in health spending, assessing the value of cancer medications has taken center stage in the discussion of drugs and drug prices. ESMO and ASCO have taken steps to guide the medical community in assessing the value of oncology therapies by releasing their value assessment frameworks. However, the frameworks are not identical, and there are several instances where therapies appear to be valued differently. As manufacturers and the medical community get more familiar with these frameworks and start to implement them, it is important to know the details of each framework and the resulting implications. 3 http://www.fiercepharma.com/story/lillys-alimta-gets-rating-boost-update-ascos-drug-scorecard/2015-07-15 7

Figure 1: ESMO Magnitude of Clinical Benefit Scale v1.0 Form 1: For new approaches to adjuvant therapy or new potentially curative therapies Grade A >5% improvement of survival at 3years follow-up Improvements in DFS alone (primary endpoint) (HR <0.65) in studies without mature survival data Grade B 3% but 5% improvement at 3% years follow-up Improvements in DFS alone (primary endpoint) (HR <0.65-0.8) without mature survival data Non inferior OS or DFS with reduced treatment toxicity or improved Quality of Life (with validated scales) Non inferior OS or DFS with reduced treatment cost as reported study outcome (with equivalent outcomes and risks) Grade C <3% improvement of survival at 3% years follow-up Improvements in DFS alone (primary endpoint) (HR >0.8) in studies without mature survival data Mark with X if relevant Magnitude of clinical benefit grade (highest grade scored) A B C Source: Cherny et al. Annals of Oncology, May 2015. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) Figure 2a: ESMO Magnitude of Clinical Benefit Scale v1.0 Form 2a: For therapies that are not likely to be curative with primary endpoint of OS IF median OS with the standard treatment is 1 year Grade 4 HR 0.65 AND Gain 3 months Increase in 2 year survival alone 10% Grade 3 HR 0.65 AND Gain 2.5-2.9 months Increase in 2 year survival alone 5 - <10% Grade 2 Mark with X if relevant HR >0.65 0.70 OR Gain 1.5-2.4 months Increase in 2 year survival alone 3 - <5% Grade 1 HR >0.70 OR Gain <1.5 months Increase in 2 year survival alone <3% Preliminary magnitude of clinical benefit grade (highest grade scored) 4 3 2 1 Quality of Life (QoL)assessment/grade 3-4 toxicities assessment* Does secondary endpoint quality of life show improvement Are there statistically significantly less grade 3-4 toxicities impacting on daily well-being* *This does not include alopecia, myelosuppression, but rather chronic nausea, diarrhoea, fatigue, etc. 8

Adjustments Upgrade 1 level if improved quality of life and/or less grade 3-4 toxicities impacting daily well-being are shown Final adjusted magnitude of clinical benefit grade 5 4 3 2 1 IF median OS with the standard treatment is >1 year Grade 4 HR 0.70 AND Gain 5 months Increase in 3 year survival alone 10% Grade 3 HR 0.70 AND Gain 3-4.9 months Increase in 3 year survival alone 5 - <10% Mark with X if relevant Grade 2 HR >0.70 0.75 OR Gain 1.5-2.9 months Increase in 3 year survival alone 3 - <5% Grade 1 HR >0.75 OR Gain <1.5 months Increase in 3 year survival alone <3% Preliminary magnitude of clinical benefit grade (highest grade scored) 4 3 2 1 QoL assessment/grade 3-4 toxicities assessment* Does secondary endpoint quality of life show improvement Are there statistically significantly less grade 3-4 toxicities impacting on daily well-being* *This does not include alopecia, myelosuppression, but rather chronic nausea, diarrhoea, fatigue, etc. Adjustments Upgrade 1 level if improved quality of life and/or less grade 3-4 toxicities impacting daily well-being are shown Final adjusted magnitude of clinical benefit grade 5 4 3 2 1 Source: Cherny et al. Annals of Oncology, May 2015. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) 9

Figure 2b: ESMO Magnitude of Clinical Benefit Scale v1.0 Form 2b: For therapies that are not likely to be curative with primary endpoint of PFS IF median PFS with standard treatment 6months Grade 3 HR 0.65 AND Gain 1.5 months Grade 2 HR 0.65 AND Gain <1.5 months Grade 1 HR >0.65 Mark with X if relevant Preliminary magnitude of clinical benefit grade (highest grade scored) 3 2 1 Toxicity assessment Is the new treatment associated with a statistically significant incremental rate of: toxic death >2% cardiovascular Ischemia >2% hospitalization for toxicity >10% excess rate of severe CHF >4% grade 3 neurotoxicity >10% severe other irreversible or long lasting toxicity >2% please specify: Mark with X if relevant (Incremental rate refers to the comparison versus standard therapy in the control arm) Quality of Life (QoL)/grade 3-4 toxicities assessment* Was QoL evaluated as secondary outcome Does secondary endpoint QoL show improvement Are there statistically significantly less grade 3-4 toxicities impacting on daily well-being* *This does not include alopecia, myelosuppression, but rather chronic nausea, diarrhoea, fatigue, etc. Adjustments a) Downgrade 1 level if there is one or more of the above incremental toxicities associated with the new drug b) Upgrade 1 level if improved QoL or if less grade 3-4 toxicities that bother patients are demonstrated c) When OS as secondary endpoint shows improvement, it will prevail and the new scoring will be done according to form 2a d) Downgrade 1 level if the drug ONLY leads to improved PFS and QoL assessment does not demonstrate improved QoL Final, toxicity and QoL adjusted, magnitude of clinical benefit grade 4 3 2 1 Source: Cherny et al. Annals of Oncology, May 2015. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) 10

IF median PFS with standard treatment >6months Grade 3 HR 0.65 AND Gain 3 months Grade 2 HR 0.65 AND Gain <3 months Grade 1 HR >0.65 Mark with X if relevant Preliminary magnitude of clinical benefit grade (highest grade scored) 3 2 1 Toxicity assessment Is the new treatment associated with a statistically significant incremental rate of: toxic death >2% cardiovascular Ischemia >2% hospitalization for toxicity >10% excess rate of severe CHF >4% grade 3 neurotoxicity >10% severe other irreversible or long lasting toxicity >2% please specify: Mark with X if relevant (Incremental rate refers to the comparison versus standard therapy in the control arm) Quality of Life (QoL)/grade 3-4 toxicities assessment* Was QoL evaluated as secondary outcome Does secondary endpoint QoL show improvement Are there statistically significantly less grade 3-4 toxicities impacting on daily well-being* *This does not include alopecia, myelosuppression, but rather chronic nausea, diarrhoea, fatigue, etc. Adjustments a) Downgrade 1 level if there is one or more of the above incremental toxicities associated with the new drug b) Upgrade 1 level if improved QoL or if less grade 3-4 toxicities that bother patients are demonstrated c) When OS as secondary endpoint shows improvement, it will prevail and the new scoring will be done according to form 2a d) Downgrade 1 level if the drug ONLY leads to improved PFS and QoL assessment does not demonstrate improved QoL Final, toxicity and QoL adjusted, magnitude of clinical benefit grade 4 3 2 1 Highest magnitude clinic benefit grade that can be achieved is Grade 4. Source: Cherny et al. Annals of Oncology, May 2015. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) 11

Figure 2c: ESMO Magnitude of Clinical Benefit Scale v1.0 Form 2c: For therapies that are not likely to be curative with primary endpoint other than OS or PFS or equivalence studies Primary outcome is Toxicity or Quality of Life (QoL) AND Non-inferiority Studies Grade 4 Reduced toxicity or improved QoL (using validated scale) with evidence for statistical non inferiority or superiority in PFS/OS Mark with X if relevant Grade 3 Improvement in some symptoms (using validated scale) BUT without evidence of improved overall QoL Primary outcome is Response Rate Grade 2 RR is increased 20% but no improvement in toxicity/qol/pfs/os Grade 1 R is increased <20% but no improvement in toxicity/qol/pfs/os Final magnitude of clinical benefit grade 4 3 2 1 Source: Cherny et al. Annals of Oncology, May 2015. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the European Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) 12

Figure 3: Simplified schematic of ASCO Value Framework for adjuvant setting (early stage disease) 1 Determine the regimen s CLINICAL BENEFIT 2 Determine the regimen s TOXICITY YES 1 2 3 4 5 x 16 = OS Score 1A Assign an HR Score Hazard Ratio (HR) for death reported? NO 1B YES 0 1 2 3 4 x 15 = DFS Score Disease-Free - Survival (DFS) reported? Assign a DFS Score Assign a toxicity score by comparing the sum of grade 3-5 toxicities for each regimen Toxicity Score Improvement -20-10 0 +10 +20 Substantially less well tolerated Less well tolerated Toxicity is the same Better tolerated Substantially better tolerated The toxicity score is based on the difference in toxicity between the two regiments. If there are unresolved symptomatic treatmentrelated toxicities at 1 year after completion of treatment, 5 points are subtracted. Clinical Benefit Score = OS Score or DFS Score or RR Score (max. allowable points are 80) Toxicity Score (max. allowable toxicity points are 20) 3 Calculate the regiment s NET HEALTH BENEFIT = 1 CLINICAL BENEFIT Score + 2 TOXICITY Score 4 Determine the Cost (entire course of regimen): Drug Acquisition Cost (DAC): regimen s COST = Patient Co-Pay: 5 SUMMARY 1 Clinical Benefit 2 Toxicity 3 Net Health Benefit 4 Cost ASSESSMENT /80 /20 /100 DAC: Patient Payment: Source: Schnipper et al. Journal of Clinical Oncology, June 2015. ASCO Statement: A conceptual framework to assess the value of cancer treatment options. Figure 4: Simplified schematic of ASCO Value Framework for advanced disease 1 Determine the 2 Determine the 3 Determine the regimen s regimen s regimen s CLINICAL BENEFIT TOXICITY BONUS POINTS YES Assign an OS Score 1 2 3 4 5 x 16 = OSScore 1A Overall Survival (OS) reported? YES 1 2 3 4 5 x 11 = PFSScore NO Progression - free 1B Survival (PFS) reported? Assign a PFS Score NO Response 1C Rate (RR) reported? YES Assign a RR Score 1 2 3 4 5 x 8 = RRScore Clinical Benefit Score = OS Score or PFS Score or RR Score (max. allowable points are 80) Assign a toxicity score by comparing the sum of grade 3-5 toxicities for each regimen -20-10 0 +10 +20 Substantially Less well less well tolerated tolerated Toxicity is the same Better tolerated Substantially better tolerated The toxicity score is based on the difference in toxicity between the two regiments. Toxicity Score (max. allowable toxicity points are 20) YES Award 10 Palliation Bonus Points Are data related to the palliation 3A of symptoms reported? YES NO Toxicity Score Improvement Assign Treatment-Free Interval Bonus Points Bonus 0 5 10 15 20 Points % >0%- 20%- 36%- 50%- 75% Change 19% 35% 49% 74% Are data related to 3B treatment-free interval reported? Total Bonus Points = Palliation Bonus Points + Treatment Free Interval Bonus Points (max. allowable points are 30) NO 3C Calculate Total Bonus Points 4 Calculate the regiment s NET HEALTH BENEFIT = 1 CLINICAL 2 TOXICITY 3 BONUS BENEFIT Score + Score + POINTS Score 5 Determine the Cost(per month) : regimen s = Drug Acquisition Cost (DAC): COST Patient Co-Pay: 1 6 Clinical Benefit 2 3 4 Net Health 5 Toxicity Bonus Points Cost (per month) SUMMARY Benefit ASSESSMENT /80 /20 /30 /130 DAC: Patient Payment: Source: Schnipper et al. Journal of Clinical Oncology, June 2015. ASCO Statement: A conceptual framework to assess the value of cancer treatment options. 13

Figure 5: Comparison of ASCO Net Health Benefit Score with corresponding ESMO Clinical Benefit Grade for several oncology therapies ASCO Net Health Benefit Score 130 120 110 100 90 80 70 60 50 40 30 20 10 0 1 2 3 4 5 ESMO Clinical Benefit Grade Source: Simon-Kucher & Partners analysis 14

About the authors Ram Subramanian is a Director in the life sciences practice of Simon-Kucher & Partners based in Boston. He advises pharmaceutical and biotechnology companies on commercialization of new products across the globe. : Ram.Subramanian@simon-kucher.com : http://bit.ly/1mt5anu :@2RamSubramanian Kevin Schorr is a Consultant in the life science division of Simon-Kucher & Partners in Boston. He has worked on a wide variety of projects, including pricing and market access, commercialization strategy, and value messaging across numerous product lines within the life sciences sector. : Kevin.Schorr@simon-kucher.com : http://bit.ly/1liekyohttp://bit.ly/1ib8wx3 About Simon-Kucher Simon-Kucher & Partners is a global consulting firm with 800 professionals in 29 offices worldwide focusing on TopLine Power. Founded in 1985, the company has 30 years of experience providing strategy and marketing consulting, and is regarded as the world s leading pricing advisor. 15

Boston One Canal Park Cambridge, MA 02141 Tel. +1 617 231 4500 Fax +1 617 576 2751 E-Mail boston@simon-kucher.com www.simon-kucher.com