Estimation of the Percentage of US Patients With Cancer Who Are Eligible for and Respond to Checkpoint Inhibitor Immunotherapy Drugs

Key Points Question What is the estimated percentage of US patients with cancer who are eligible for and respond to checkpoint inhibitor drugs approved for oncology indications by the US Food and Drug Administration? Findings This cross-sectional study found that the estimated percentage of US patients with cancer who are eligible for checkpoint inhibitor drugs increased from 1.54% in 2011 to 43.63% in 2018. The percentage of patients estimated to respond to checkpoint inhibitor drugs was 0.14% in 2011 and increased to 12.46% in 2018. Meaning The estimated percentages of patients who are eligible for and who respond to checkpoint inhibitor drugs are higher than reported estimates for drugs approved for genome-driven oncology but remain modest.


Introduction
Cancer checkpoint inhibitors have received considerable and broad interest because of their ability to generate durable responses in many hitherto intractable malignant tumors 1,2 and for improvements in overall survival in several randomized trials.These promising drugs, and their underlying preclinical science, formed the basis of the 2018 Nobel Prize in Medicine. 3eaning The estimated percentages of patients who are eligible for and who respond to checkpoint inhibitor drugs are higher than reported estimates for drugs approved for genome-driven oncology but remain modest.

+ Invited Commentary + Supplemental content
Author affiliations and article information are listed at the end of this article.The first approved agent, ipilimumab, received FDA marketing authorization in 2011 for metastatic melanoma.Since then, 5 more checkpoint inhibitor drugs have been approved for a total of 14 different indications (eTable in the Supplement).

Open
Between 2015 and 2017, the number of clinical trials using PD-1 and PD-L1 inhibitors has increased nearly 600%, from 215 trials to more than 1500. 4The market is expected to grow similarly, from $1 billion dollars in 2013 to $7 billion dollars in 2020. 5However, despite growing interest in checkpoint inhibitors, [6][7][8][9][10] empirical analyses have quantified the use of these drugs only in certain tumor types. 11To our knowledge, there has been no empirical analysis of the potential use or benefit among all US patients with cancer.For this reason, we sought to estimate what percentage of US patients with cancer are eligible for checkpoint inhibitors and what percentage might respond to them.Our analysis of checkpoint inhibitors is similar to a prior analysis 12 of the estimation of genomedriven cancer therapies in US patients with cancer, which estimated a benefit of less than 5%.We hypothesized that the benefit of response from checkpoint inhibitors will be modest.

Methods
We sought to estimate the percentage of US patients with cancer who are (1) eligible for and (2) may respond to FDA-approved checkpoint inhibitor drugs in a cross-sectional analysis.We defined persons as eligible for checkpoint inhibitors if they had the tumor type and notable inclusion criteria of the drug approval (eg, PD-L1 level).We defined persons as responding to therapy based on the best available response rate (FDA drug label) for that indication.We report annual findings from 2011 to present.Our methods are comparable to a prior analysis of genome-driven cancer therapies. 12We

Data Set
We selected all checkpoint inhibitor oncology drugs that were approved by the FDA through August 17, 2018. 13We included all indications, even those that had conditional approval under the accelerated pathway.For each drug and year for which the drug was approved, the approved indication and the overall response rate (complete plus partial) were extracted from the FDA label.
We used the overall response rate in the experimental group if the drug was compared with a placebo control, a different drug control, or was assessed in a single-group trial.The difference in overall response was used if a trial was run with a treatment backbone given with or without the checkpoint inhibitor.Overall response rates that were from exploratory analysis were also not used (eg, nivolumab for hepatocellular carcinoma with a Modified Response Evaluation Criteria in Solid Tumors).

Statistical Analysis
To determine the percentage of patients who could potentially be candidates for, and therefore benefit from, a checkpoint inhibitor drug, we used annual cancer deaths as a stand-in for annual incidence of advanced or metastatic disease.This was similar to our prior work. 12Notably, stage IV disease incidence could not be used, as this does not include relapsed, metastatic disease that initially presented as early stage. 14ta on cancer deaths were obtained from American Cancer Society publications on cancer statistics, published between 2011 and 2018, to correspond with years that checkpoint inhibitor

JAMA Network Open | Oncology
Eligibility for and Response to Checkpoint Inhibitor Immunotherapy Drugs for Cancer drugs had approval(s).Death statistics on cancers were aggregated by overall type, and the drugs for which there were approvals were sometimes approved only for a subgroup of that cancer.For example, lung cancer was subdivided into non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC).The category of NSCLC could be further subdivided by PD-L1 expression.In this situation, the number of deaths from lung cancer was multiplied by 85% to get an estimate of eligible patients with NSCLC and by 15% to get an estimate of eligible patients with SCLC.To estimate the number of patients with PD-L1 expression greater than 50%, the NSCLC estimate was then multiplied by 25%. 15,16The estimate for patients with PD-L1 expression of 0% to 50% was then calculated by taking the difference between the NSCLC estimate and the estimate for PD-L1 greater than 50%.In the case of urothelial carcinoma, pembrolizumab was approved in 2017 but reported different objective response rates for people who had been previously treated and people who had urothelial cancer that was cisplatin ineligible.In this case, for the years 2017 and 2018, the estimated benefit was calculated separately for cisplatin-ineligible and cisplatin-independent objective response rates based on previously reported estimates of individuals with cisplatin-ineligible disease and then the estimated benefits were combined into a single urothelial cancer measure. 17r each cancer type, the number of people estimated to be eligible for therapy was multiplied by the overall response rate reported in the FDA drug label for each year the drug was approved for that indication.This provided an estimate of the cancer-specific benefit or, in other words, the number of people who could potentially benefit.In selecting the overall response rate, the highest response rate reported by each drug approved for the specific cancer was used, thus giving the most generous estimate of benefit.Similarly, if multiple drugs were approved for a certain indication, the response rate from the drug showing the highest benefit was used.For each year, the estimated cancer-specific benefit for all indications that had an FDA-approved drug were summed.The sum of the estimated cancer-specific benefits for each year was then divided by the total number of people who died during that year from cancers for which there was an FDA-approved checkpoint inhibitor drug to derive an overall estimate of responders.
To calculate an estimate of the percentage of people who were eligible for checkpoint inhibitors, the number of cancer-specific deaths for which there were FDA-approved checkpoint inhibitors was divided by the total number of cancer deaths.To provide context, a ratio was also calculated dividing the percentage benefit from immunotherapy checkpoint inhibitors by the percentage of cancers affected by immunotherapy drugs (percentage eligible) (eFigure 1 in the Supplement).This descriptive analysis was done in Excel (Microsoft Corp).Calculation of 95% confidence intervals for the percentage of patients who were eligible and who could benefit were done in the MASS package of R statistical software version 3.5.0(R Project for Statistical Computing).

Results
Six checkpoint inhibitor drugs were approved for 14 indications between March 25, 2011, and August

Eligible for Immunotherapy
In  2).
The Table and eFigure 1 in the Supplement show the percentage benefit of checkpoint inhibitors for the specific cancer.Melanoma, the first condition for which a checkpoint inhibitor was approved, potentially resulted in responses in 0.14% of all patients with cancer during 2011, the year that it was first approved.Estimated responders increased to 0.53% when a PD-1 drug was approved for this indication.Responders increased again in 2015 to 1.01%, when higher response rates were reported in patients with BRAF V600 mutations, and the benefit has leveled off since then.
Urothelial carcinomas have also seen a noticeable increase in the estimated number of patients who respond, with the percentage increasing from 0.43% in 2016 to 0.75% in 2018 after a checkpoint inhibitor drug was approved in 2014 for use in patients ineligible for cisplatin treatment.Response for renal cell carcinoma was somewhat stable from 2015 to 2017 (0.51% for all years) but increased to 1.02% in 2018 with the approval of a different checkpoint inhibitor drug combination that reported a higher response rate.In 2015, with the approval of drugs for NSCLC, it was estimated that 4.33% of cancers would benefit from a checkpoint inhibitor drug.Most of the benefit for checkpoint inhibitor drugs after 2015 was due to the responders with NSCLC: 5.92% in 2016, 6.78% in 2017, and 7.09% in 2018, although the benefit of checkpoint inhibitors for NSCLC was a lower proportion in later years than in earlier years, with the approval of multiple checkpoint inhibitors for more indications.
The ratio of percentage benefit to percentage of cancers with FDA-approved checkpoint inhibitor drugs was 0.09 in 2011, peaked in 2014 at 0.32, and in 2018 was estimated to be 0.28 (eFigure 2 in the Supplement).

Discussion
Checkpoint inhibitor drugs have generated deserved excitement in the field of oncology and are enjoying rapid uptake. 11Here, we present upper bound estimations of the percentage of US patients with cancer eligible for and responding to these drugs based on publicly collected and available data, assuming universal access to these medications.
The results of our analysis suggest that checkpoint inhibitors may at best lead to responses among less than 13% of patients with cancer in the United States.While the estimated percentage of people who respond to checkpoint inhibitor immunotherapy drugs is small, the benefit may be greater than some other drug classes in oncology owing to reports of durability.Recently, genometargeted therapy, which has also generated considerable excitement, was estimated to benefit only 4.9% of patients with cancer. 12r the first few years after checkpoint inhibitors were initially approved for any oncology indication, the percentage of eligible patients was small and curves remained flat.In 2015, checkpoint inhibitors gained FDA approval for NSCLC, leading to a noticeable increase in benefit.As of 2015, only 3 checkpoint inhibitors were approved for 3 cancers, but since then, 3 other drugs have been approved, and the number of cancers for which these drugs have been approved has grown to 14, including the most recent approval of nivolumab for SCLC.However, even though the number of indications has increased substantially in recent years, the increase in benefit from these drugs in terms of the percentage of patients responding has slowed.Moreover, the percentage of individuals who are eligible to receive these drugs has grown since 2014 at a faster rate than the estimated percentage of individuals who actually benefit from these drugs.
One striking observation from the figures presented here is that the overall estimated benefit is driven mainly from NSCLC checkpoint inhibitor drugs, for which the benefit is somewhere around 7% of cancer deaths.This can be seen in Figure 2B, where the percentage of benefit is much larger for NSCLC than for other indications.This finding suggests that large improvements in US cancer statistics may be driven by drugs that are active in the most common tumor types.
The higher cancer-specific benefit in the group of patients with PD-L1 from 0% to 50% may seem counterintuitive at first glance, given that the response rate for NSCLC with PD-L1 greater than 50% is much higher than for NSCLC with PD-L1 between 0% and 50%.However, we estimated that there were approximately 3 times as many people who died from NSCLC with PD-L1 between 0% and 50% than NSCLC with PD-L1 greater than 50% during 2018, and the absolute numbers of individuals who develop and die from NSCLC with PD-L1 between 0% and 50% is what is driving the benefit.
For these analyses, we did include drugs that had conditional approval, such as nivolumab for hepatocellular carcinoma and other types of cancers.There is the possibility that these drugs fail to show benefit in confirmatory studies; in fact, pembrolizumab was conditionally approved for hepatocellular carcinoma but recently failed to improve more salient outcomes, such as overall What is the estimated percentage of US patients with cancer who are eligible for and respond to checkpoint inhibitor drugs approved for oncology indications by the US Food and Drug Administration?Findings This cross-sectional study found that the estimated percentage of US patients with cancer who are eligible for checkpoint inhibitor drugs increased from 1.54% in 2011 to 43.63% in 2018.The percentage of patients estimated to respond to checkpoint inhibitor drugs was 0.14% in 2011 and increased to 12.46% in 2018.

Figure 1 .
Figure 1.Percentage of US Patients With Cancer Who May Benefit From and Respond to Checkpoint Inhibitor Immunology Drugs (2011-2018) 100

Figure 2 .
Figure 2. Percentage of US Patients With Cancer Eligible to Receive Checkpoint Inhibitor Drugs and Percentage Who Respond, by Cancer Type, in 2018

Table .
Eligibility and Benefit for All Cancers and Benefit for Specific Cancers From Checkpoint Inhibitor Drugs From 2011 to 2018 Abbreviations: NA, not applicable; PD-L1, programmed cell death ligand 1.