The impact of rarity on oncology health technology assessment (HTA) - PowerPoint PPT Presentation

The impact of rarity on oncology health technology assessment (HTA) and funding in Ontario: A review of pCODR recommendations from 2012 - 2017 CADTH Symposium 2019 Presenters: James Keech & Wei Fang Dai Authors: James Keech, Wei Fang Dai, Jaclyn Beca, Kelvin Chan 1 Advancing Health Economics, Services, Policy and Ethics

Conflicts of Interests • We have no conflicts of interests of declare • The views expressed in this presentation are solely those of the authors and do not reflect the views of Cancer Care Ontario 2

Overview 1. Oncology HTA in Canada (pCODR Expert Review Committee framework) 2. What is rarity in cancer? 3. Study objectives 4. Rarity and Canadian HTA/funding in Ontario 5. Conclusions & future directions 3

Oncology Drug Funding in Canada Regulatory approval Health Canada HTA review & recommendation pCODR/CADTH & INESSS Pricing Negotiations pCPA Public Formulary Listing F/P/T Ministries of Health and Provincial Cancer Agencies 4

The pERC Deliberative Framework pCODR expert review committee (pERC) considers 4 criteria when making funding recommendations: Alignment with Overall Clinical Benefit patient values Funding Recommendation Feasibility of adoption Cost-effectiveness into the health system 5 Source: pCODR Expert Review Committee Deliberative Framework. March 2016

Evaluating Rare Indications • pERC’s framework does not explicitly address rarity – Available evidence must suggest drug under review could substantially ↓ morbidity and/or mortality of disease Significant Unmet Need • Rarity of Condition Affects small # of patients • Often genetically based • Heavy burden on health system • Difficult to study • Population Need identified on a population basis • Absence of Absence of effective alternatives • Alternatives Substantial morbidity/mortality 6 Source: Recommendation Framework for CADTH CDR and pCODR programs. March 2016

Evaluating Rare Indications • Challenges of evaluating drugs for rare indications include: 1. High unmet need 2. Uncertainty in clinical evidence – Lack of RCTs and comparative data – No appropriate comparator/standard of care – Uncertainty in burden of illness 3. High costs/challenges with pharmacoeconomic evaluations 4. No universal definition of rarity 7 Source: Nestler-Parr et al. Challenges in Research and Health Technology Assessment of Rare Disease Technologies: Report of the ISPOR Rare Disease Special Interest Group

What is considered rare in cancer? 11,547 • New Cases of Lung Cancer in Ontario • Non-small cell lung cancer (NSCLC) • 85% of cases 9815 • Stage III/IV at diagnosis (advanced) • 70% of cases 6870 • ROS1 mutation • 2% of cases 137 8 Source: 2017 statistics from the Canadian Cancer Society

Research Questions 1. How often are positive funding recommendations made for rare indications by pCODR? 2. How many rare indications are submitted to pCODR with an RCT? In how many instances is conducting an RCT feasible? 3. Are drugs for rare indications less cost-effective (higher ICER)? 4. What is the time to funding for rare indications in Ontario? 9

Methods • No universal definition of rarity • Incidence is one approach to operationalize rarity – Definition 1 (less stringent): Incidence < 5/10,000 diagnoses per year – Definition 2 (more stringent): Incidence < 1/100,000 diagnoses per year 10 Source: Recommendation Framework for CADTH CDR and pCODR programs. March 2016

Methods • pCODR submissions with final recommendations (2012-2017) × 2018/19 recommendations × Submissions that were withdrawn, suspended or requests for advice • Incidence rates for each indication were extracted from: 1. pCODR Final Recommendations or Final Clinical Guidance Reports 2. Canadian Cancer Society Statistics 3. Pivotal Trials 11

Methods Total pCODR Reviews 2012 - 2017 Excluded Reviews N =104 (100%) (Withdrawn, suspended, RFA) N=8 (7.7%) Included Reviews N=96 (92.3%) Pos. Recommendation Neg. Recommendation N=75 (78%) N=21 (22%) With Conditions Without Conditions N=66 (88%) N=9 (12%) Definition 1: Incidence < 5/10,000 All recommendations (positive and negative) were rare 12

How often are positive recommendations made for rare and non-rare indications (definition 2)? Rare (<1/100,000) Non Rare (>1/100,000) Positive Recommendation (n = 75) 12 (75%) 63 (78.8%) Negative Recommendation (n = 21) 4 (25%) 17 (21.2%) Odds Ratio = 0.67 (95% CI: 0.19, 2.4); P-value = 0.5* 60% % of Recommendations 50% 40% 30% 20% 10% 16% 19% 31% 33% 44% 24% 9% 24% 0% <1 1 to 4 5 to 9 10 to 45 Incidence rate (diagnoses/100,000 in Ontario) The frequency of positive recommendations for rare and non-rare indications were similar (75% vs 78.8%; p-value = 0.5). *Fischer’s Exact Test 13 Source: pCODR Website - Find a Review; https://www.cadth.ca/pcodr/find-a-review

Does Biomarker Status affect the definition of rarity? Incidence - Biomarker Incidence - not accounting for biomarker Definition 1 Definition 2 60 Incidence Rate (diagnoses/100,000 in Ontario) 50 40 30 20 10 0 Indication with Biomarker Status 2 indications changed from rare → non rare using definition 1 (<5/10,000). There were no changes in rarity status for definition 2 (<1/100,000). 14 Source: pCODR Website - Find a Review; https://www.cadth.ca/pcodr/find-a-review

How many rare indications are submitted to pCODR with a RCT? Rare (<1/100,000) Non Rare (>1/100,000) No RCT Conducted (n = 16) 8 (50%) 8 (50%) RCT Conducted (n = 80) 8 (10%) 72 (90%) Relative Risk = 5 (95% CI: 2.2, 11.35); Odds Ratio = 9 (95% CI: 2.7, 30.6); P-value <0.01 % of Submitted Requests with RCT 100% 72/80 80/96 90% 80% 70% 60% 8/16 50% 90.0% 83.3% 40% 30% 50.0% 20% 10% 0% Total Non-Rare Rare Rare indications appear to have fewer submissions with RCTs conducted 15 Source: pCODR Website - Find a Review; https://www.cadth.ca/pcodr/find-a-review

% of Positive Recommendations with RCTs RCT No RCT 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Frequency of Positive Recommendations 2 feasible Total (N=75) 9 66 1 feasible 3 Non-Rare (N=63) 60 1 feasible 6 Rare (N=12) 6 For positive recommendations without an RCT, pERC did not consider future RCT trials to be feasible for majority of the indications. 16 Source: pCODR Website - Find a Review; https://www.cadth.ca/pcodr/find-a-review

% of Negative Recommendations with RCTs RCT No RCT 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Frequency of Negative Recommendations All feasible Total (N=21) 14 7 All feasible 12 5 Non-Rare (N=17) All feasible 2 2 Rare (N=4) For negative recommendations without an RCT, pERC deemed it feasible to conduct an RCT for all indications. 17 Source: pCODR Website - Find a Review; https://www.cadth.ca/pcodr/find-a-review

Are drugs for rare indications less cost-effective? 45 Positive - rare (n = 12) Positive - non rare (n = 62) Negative - rare (n = 3) Negative - non rare (n = 16) 40 Incidence rate (diagnoses/100,000 in Ontario) Drugs for rare indications do not appear to be less 35 cost-effective than drugs for non-rare indications. 30 Mean (SD) P-value 25 Negative – Rare (n = 3) 370,001 (188,507) 0.81 Negative – Non-Rare (n = 16) 312,096 (395,001) 20 Positive – Rare (n = 12) 324,493 (292,220) 0.49 15 Positive – Non-Rare (n = 62) 269,055 (248,478) 10 5 0 -200,000 200,000 600,000 1,000,000 1,400,000 1,800,000 Incremental Cost-effectiveness Ratio ($/QALY)* *The upper range of the ICER was plotted for submissions where the range of ICER was provided; 18 Note: Two submissions without ICERs reported were excluded; One submission with the ICER upper range of ~4 million was excluded Source: pCODR Economic Guidance Panel Report;

Time to Funding in Ontario Time to Funding = Date of Notice of Implement to Funding Date 900 800 P-value = 0.45* 700 Time to Funding (Days) 600 Mean (days) N = 10 500 N = 62 400 322 282.85 300 200 100 0 Non-rare Rare There appears to be no significant differences in time to funding between rare and non-rare indications. 19 *Calculated from Wilcoxon-Mann-Whitney test **3 reviews were excluded from the analysis as they were not funded at the time of data extraction Source: pCODR Website - Find a Review; https://www.cadth.ca/pcodr/find-a-review

Limitations 1. Line of therapy or previous treatment(s) not factored in to incidence calculation 2. Limited number of rare indications submitted to pCODR 3. No universal definition of rarity 4. Incidence calculated from multiple data sources and often using best estimates 5. Did not control for multiple factors contributing to funding recommendations and decisions 20

Conclusions 1. Rarity does not significantly impact positive or negative funding recommendations 2. Rare indications appear to have fewer submissions with RCTs conducted 3. Drugs for rare indications do not appear to be less cost- effective than drugs for non-rare indications 4. There appears to be no significant differences in time to funding between rare and non-rare indications. 21

Future Directions  Evaluate real-world utilization/public expenditures on drugs for rare oncology indications  Conduct a sensitivity analysis for the limitations described in our incidence calculations  Further examine how rarity should be incorporated into oncology HTA in Canada 22