Cost and Duration of Clinical Trials in Drug Development by Japanese Pharmaceutical Companies

Abstract

Background

Understanding parameters such as cost and duration, which reflect the scale of clinical trials, is essential for government agencies to formulate pharmaceutical policies and for pharmaceutical companies to strategically allocate resources. As most studies on clinical trial costs and durations have primarily focused on pharmaceutical companies developing drugs in the USA, understanding actual trial costs and durations for Japanese pharmaceutical companies engaged in global development or development within their own countries remains challenging.

Objectives

This study aimed to clarify the actual phase-specific costs and durations of clinical trials conducted by leading Japanese pharmaceutical companies and to identify the factors influencing these parameters.

Methods

We conducted a questionnaire-based survey among 22 pharmaceutical companies headquartered in Japan. Respondents from each company retrospectively identified clinical trials for new drug applications or for expanding approved disease indications, conducted between 2012 and 2022, providing cost and duration data of each clinical trial. We calculated summary statistics for costs and durations by phase and conducted an exploratory analysis of the variables influencing these parameters.

Results

In total, 631 projects with clinical trials conducted by these companies were analyzed. The median costs in the analysis population were 859 million Japanese yen (JPY) for Phase 1, 1,739 million JPY for Phase 2, and 4,183 million JPY for Phase 3. For reference, the median annual average exchange rates during the study period were 109 JPY per US dollar (USD) and 130 JPY per euro (EUR). The median durations were 20 months for Phase 1, 29 months for Phase 2, and 31 months for Phase 3. For projects targeting approval only in Japan, the median Phase 3 cost was 3,050 million JPY, while for projects targeting approval involving the USA or Europe, it was 8,083 million JPY.

Conclusions

This study quantifies the costs and durations incurred by Japanese pharmaceutical companies in conducting clinical trials, revealing the high costs associated with projects targeting approval involving the USA or Europe. The findings highlight the substantial investments required by the Japanese pharmaceutical industry to bring new drugs to the global market.

Key Points

Although Japanese pharmaceutical companies contribute to the global market, limited research exists on the actual costs and durations of clinical trials conducted by these companies.

Based on a large-scale survey of clinical trial data in cooperation with 22 pharmaceutical companies headquartered in Japan, this study provides accurate information on the costs and durations of clinical trials conducted between 2012 and 2022.

High costs for clinical trials in projects by Japanese companies targeting approval involving the USA or Europe were quantified, highlighting the substantial investments required by the Japanese pharmaceutical industry to bring new drugs to the global market.

Introduction

Bringing a new drug to market requires evaluation through clinical trials, which are among the most expensive and time-intensive processes in pharmaceutical research and development [1, 2]. Furthermore, the low success rate of these trials increases the risk for pharmaceutical companies investing in drug development [3, 4]. Key parameters, such as cost and duration of clinical trials, are essential for making informed investment decisions [5, 6]. Pharmaceutical companies base their investment decisions on expected sales and profits relative to research and development costs. Therefore, balancing development costs and drug prices is a critical factor in motivating companies to pursue new drugs. Understanding the financial and time burdens of drug discovery and development, and accurately reflecting these investments in drug pricing, will improve patient access to essential drugs.

A study using funding databases reported that average clinical trial costs in the USA, stratified by therapeutic area, ranged from US $1.4 million to US $6.6 million for Phase 1, US $7.0 million to US $19.6 million for Phase 2, and US $11.5 million to US $52.9 million for Phase 3 [7]. Another study using clinical financial data from major global pharmaceutical companies reported median costs of US $3.4 million for Phase 1, US $8.6 million for Phase 2, and US $21.4 million for Phase 3 [8]. Regarding durations, a study utilizing databases derived from institutional press releases and financial reports showed that the median duration of clinical trials was 1.6 years for Phase 1, 2.9 years for Phase 2, and 3.8 years for Phase 3 [9]. Although several studies have analyzed the cost and duration of clinical trials, most relied on estimates [10–12]. Few studies have collected large-scale actual data from individual clinical trials in cooperation with multiple companies [1, 8]. Additionally, most studies using confidential company data have largely focused on major pharmaceutical companies involved in US-based drug development. Consequently, understanding actual trial costs and durations for companies outside the USA or Europe engaged in global development or development within their own countries remains challenging. As drug development becomes increasingly global, companies worldwide, including those in Japan, are actively participating in this process. Among them, it has been suggested that the development strategies of Japanese pharmaceutical companies may differ from those of foreign pharmaceutical companies, primarily based in the USA or Europe [13]. Studies focusing specifically on clinical trials conducted by Japanese pharmaceutical companies are limited, indicating a need for closer examination.

This study aimed to clarify the actual phase-specific costs and durations of clinical trials conducted by leading Japanese pharmaceutical companies and to identify the factors influencing these parameters. By deriving these results quantitatively, the study will clarify the burdens of conducting clinical trials faced by companies in Japan and other regions outside the USA or Europe, highlighting inherent challenges in the pharmaceutical industry business model. Preliminary results estimating the total research and development costs to bring a new drug to market, based on the survey used in this study, have been released in Japanese on the Office of Pharmaceutical Industry Research website; however, these results have not been peer-reviewed [14]. The study focused on individual phases, providing additional detailed analyses of costs and durations. Some results presented herein (medians and interquartile ranges in Figs. 2 and 4) are based on a study first reported on the website.

Fig. 2.

Distribution and median of trial costs and durations; (a) trial costs for each phase; (b) trial durations for each phase

Fig. 4.

Trial costs by development region

Methods

Data Collection

To clarify the current status of clinical trials in Japanese pharmaceutical companies, the Office of Pharmaceutical Industry Research, a research institute established by the Japan Pharmaceutical Manufacturers Association (JPMA), conducted a questionnaire-based survey among 28 pharmaceutical companies headquartered in Japan. These companies are members of the Research and Development Committee of JPMA. In total, 22 of these companies agreed to participate in the survey. The participating companies included nine of the top ten Japanese pharmaceutical companies based on 2022 sales. The survey focused on projects conducting clinical trials for new drug applications or for expanding approved disease indications. Projects were categorized by therapeutic area for each product. If a single product addressed multiple therapeutic areas, it was classified into separate projects according to the number of therapeutic areas. For each project, data on clinical trials started and completed between 2012 and 2022 were collected. Respondents from each company retrospectively identified clinical trials that they sponsored for new drug applications or for expanding approved disease indications. They provided information, including the costs and durations of each trial, by reviewing their company’s clinical trial databases. Respondents submitted their answers to an external organization, which checked the data for consistency. Project names were pseudonymized by the organization before analysis to protect trade secrets, ensuring that the actual drug names and companies associated with each project could not be identified. The pseudonymized data were provided to us for analysis.

The project-specific survey questions included the costs and start dates of each clinical trial, the date of the decision to discontinue the project, drug modality, product origin, geographic region of development, therapeutic area, rare disease applicability, and annual sales forecasts for the Japanese market. The costs of clinical trials included both internal and external expenses, covering outsourced trial costs, personnel costs, and all other identifiable costs. The start date of the clinical trial and the discontinuation decision date of the project were provided in years and months. For projects where development region and rare disease applicability were not determined at the time of the survey, respondents were asked to provide the expected classification. Development regions were categorized into three groups: projects targeting approval in Japan only, projects targeting global approval including the USA or Europe (whether or not Japan was included), and projects targeting approval in other regions, such as Asia. Rare diseases were defined as conditions affecting fewer than 50,000 patients in Japan, consistent with the definition used by the Japanese Ministry of Health, Labour, and Welfare. The detailed survey items are available on the website [14].

Statistical Analyses

Analyses were conducted on a phase-by-phase basis. Therefore, even if all phases of a project were not completed within the study period, individual phases completed within the study period were included in the analysis population. The analysis population comprised projects that completed at least one phase of clinical development between 2012 and 2022 and provided sufficient information for the analysis. Clinical trials with multiple phase roles, such as Phase 1/2, were treated as the later phase (e.g., Phase 1/2 as Phase 2 and Phase 2/3 as Phase 3). Missing values were not imputed. Consequently, projects with missing information on phases or trial costs were excluded from the analysis population for trial costs. Projects with missing information on phases or trial start dates were excluded from the analysis population for trial duration. If project characteristics had missing values, the project was excluded from the analysis of the relevant category based solely on that missing value.

The costs of each clinical trial were summarized on a phase-by-phase basis. When multiple clinical trials were conducted in the same phase, their costs were combined if they were initiated before the first clinical trial in the subsequent phase or before the application for approval was submitted. Costs for each phase were calculated on the basis of out-of-pocket expenses, excluding capital costs. Additionally, these costs were adjusted to the 2023 reference pricing using gross domestic product (GDP) deflators in Japan to account for inflation.

The durations of each phase were calculated on the basis of the phase start and end date. The start date of a phase was defined as the earliest date of submission of the clinical trial notification for that phase in any country. The end date of a phase was defined as the date of submission of the clinical trial notification for the subsequent phase in any country, the date of submission of the approval application in any country, or the date of the decision to discontinue the project. If multiple dates existed, the earliest date was used.

The distribution of each parameter was estimated using kernel density estimation with a Gaussian kernel. Spearman’s rank correlation coefficient was calculated to analyze correlations, considering the distribution of each parameter. The box plot displays only the quartile range, as individual values cannot be shown to protect trade secrets. Analyses were performed using the statistical software R version 4.3.0.

Results

A total of 631 projects with clinical trials conducted between 2012 and 2022 by 22 Japanese pharmaceutical companies were available for analysis. The analysis population for trial cost comprised 618 projects, while the analysis population for trial duration comprised 352 projects (Fig. 1). The population for the trial duration analysis was smaller than that for the trial cost analysis because of missing end-date information for certain phases, owing to inaccessible data on subsequent phases. The characteristics of each population are presented in Table 1.

Fig. 1.

Flowchart of the analysis population

Table 1.

Characteristics of each population

Characteristics		Trial cost (n = 618) n (%)	Trial duration (n = 352) n (%)
Drug modality	Small molecule	391 (63.3)	218 (61.9)
	Antibody	138 (22.3)	85 (24.1)
	Others	89 (14.4)	49 (13.9)
Product origin	Self-originated	320 (51.8)	177 (50.3)
	Product licensing	228 (36.9)	143 (40.6)
	M&A	70 (11.3)	32 (9.1)
Development regiona	Japan	234 (37.9)	167 (47.4)
	USA/Europe	320 (51.8)	151 (42.9)
	Others	51 (8.3)	32 (9.1)
	No answer	13 (2.1)	2 (0.6)
Therapeutic area	Cancer	204 (33.0)	103 (29.3)
	Neurology	72 (11.7)	44 (12.5)
	Gastrointestinal	40 (6.5)	18 (5.1)
	Immunology	32 (5.2)	23 (6.5)
	Cardiovascular	30 (4.9)	15 (4.3)
	Others	205 (33.2)	140 (39.8)
	No answer	35 (5.7)	9 (2.6)
Rare disease applicability	Rare disease	86 (13.9)	45 (12.8)
	N/A	455 (73.6)	240 (68.2)
	No answer	77 (12.5)	67 (19.0)
Estimated annual sales in Japan	> 10 billion JPY	79 (12.8)	62 (17.6)
	5–9.9 billion JPY	79 (12.8)	58 (16.5)
	2–4.9 billion JPY	56 (9.1)	33 (9.4)
	< 2 billion JPY	55 (8.9)	48 (13.6)
	No answer	349 (56.5)	151 (42.9)

For reference, the median annual average exchange rates during the study period were 109 JPY per USD and 130 JPY per EUR

^aJapan, projects targeting approval in Japan only; USA/Europe, projects targeting approval in the USA or Europe, including multi-regional trials; Others, projects targeting approval in Asia and other regions, excluding the USA or Europe

The distributions derived from kernel density estimation and the median costs and durations for each phase are shown in Fig. 2. The median costs in the analysis population were 859 million JPY for Phase 1, 1,739 million JPY for Phase 2, and 4,183 million JPY for Phase 3. For reference, the median annual average exchange rates during the study period were 109 JPY per US dollar (USD) and 130 JPY per euro (EUR). The median durations were 20 months for Phase 1, 29 months for Phase 2, and 31 months for Phase 3. The cost distributions were positively skewed, with a rightward tail, a trend that was more pronounced in the later phases.

Correlation coefficients between trial costs of each phase and project characteristics indicated that the development region was most strongly associated with increased trial costs (Fig. 3). Correlation coefficients between projects targeting approval involving the USA or Europe and trial costs ranged from 0.28 to 0.42 per phase, indicating a trend towards increased costs for all phases compared with projects targeting approval in Japan only. In contrast, project characteristics were not strongly associated with trial duration. The box plot of trial costs by development region for each phase indicated the largest difference in costs for Phase 3 (Fig. 4). The median Phase 3 cost for projects targeting approval only in Japan was 3,050 million JPY. In contrast, for projects targeting approval involving the USA or Europe, the median Phase 3 cost was 8,083 million JPY, resulting in a 2.65-fold difference, which amounts to 5,033 million JPY. The difference in median Phase 1 cost was 2.19-fold, amounting to 610 million JPY, and the difference in median Phase 2 costs was 2.39-fold, amounting to 1,499 million JPY, with median costs for all phases being more than twice as high. Aside from the development region, the subgroup most associated with trial cost or duration showed a marginal negative correlation coefficient (− 0.29) between an annual sales forecast for the Japanese market of less than 2 billion JPY and Phase 2 costs. Moreover, a marginal positive correlation coefficient (0.28) was observed between cancer in the therapeutic area and Phase 3 costs. This suggested that projects expected to have lower sales were allocated relatively less funding for clinical trials, whereas trial funding for cancer agents was relatively higher.

Fig. 3.

Correlation coefficients between trial parameters and project characteristics

For phases in which both trial cost and duration data were collected, the correlation coefficients between costs and durations are presented in Table 2a. Costs and durations were positively correlated. Specifically, the correlation coefficient between costs and durations in Phase 1 was 0.46, indicating a stronger correlation in earlier phases. These results were based on one-to-one correlations that were not adjusted for other variables, including sample size, which is considered to impact costs. Correlation coefficients between phases for cost and duration are presented in Table 2b for projects in which data from both phases were collected to identify the relationship between costs and durations across phases. Phase 3 costs, the most substantial, showed a correlation coefficient of 0.48 with Phase 1 costs and 0.52 with Phase 2 costs, indicating an association across phases. This suggests that projects with higher overall trial costs tend to have relatively higher costs across all phases.

Table 2.

Correlation coefficient between trial parameters

(a) Correlation between cost and duration
Phase	na	Correlation
Phase 1	169	0.46
Phase 2	124	0.26
Phase 3	147	0.13

(b) Correlation between phases
Correlation coefficients (nb)	Phase 1	Phase 2	Phase 3
Trial costs
Phase 1
Phase 2	0.15 (80)
Phase 3	0.48 (43)	0.52 (42)
Trial durations
Phase 1
Phase 2	− 0.01 (46)
Phase 3	0.27 (34)	0.16 (30)

^aPhases in which both trial cost and duration data were collected were included

^bProjects in which data from both phases were collected were included

n, The number of projects

Discussion

This study quantified the costs and durations of clinical trials conducted by Japanese pharmaceutical companies, using the most recent actual data from each company. In drug development, the probability of success from the start of Phase 1 to approval is approximately 10% [1, 3]. Bringing a single new drug to market is estimated to cost billions of dollars, including investments in failed projects during the research and development process [2]. This study focused on actual phase unit costs, differentiating them from the estimated overall cost accounting for the probability of success. Challenges in accurately collecting data for some projects in this survey, including those still determining whether to proceed to the next phase, prevented precise estimates of success probability for phase progression. Limited data indicated an overall success probability from Phase 1 to approval of 14.6%. These results provide insights into the investment risks faced by Japanese pharmaceutical companies in research and development aimed at bringing new drugs to the market.

The cost findings of this study are consistent with or slightly higher than those reported in previous studies using data from USA or global pharmaceutical companies, considering a median annual average exchange rate of 109 JPY per USD during the study period [7, 8]. Among them, clinical trials targeting global approval, including those in the USA or Europe, had a median Phase 3 cost of 8,083 million JPY, identifying them as a higher-cost subgroup. Factors influencing this result may include relatively large-scale projects targeting the USA or Europe among those undertaken by Japanese pharmaceutical companies, inherently high site payments in North America, and increased outsourcing costs associated with accessing foreign sites from Japan. The estimated costs per site for clinical trials are quantitatively higher in North America than those in other regions [15]. Pharmaceutical companies must consider overall clinical trial costs when making investment decisions. Assuming all phases from Phase 1 to Phase 3 are conducted, the estimated overall cost for the clinical trial is 6,781 million JPY, on the basis of the sum of the median costs for each phase in this study. Phase 3 costs correlated with Phase 1 and Phase 2 costs, indicating that clinical trials with higher Phase 1 or Phase 2 costs tend to have higher Phase 3 costs. Projects incurring high costs in any phase are likely to observe escalating costs in subsequent phases, thereby increasing the overall clinical trial cost. The median annual research and development expenditure of 21 major Japanese pharmaceutical companies in 2022 was approximately 63 billion JPY [16]. Given this limited budget, funds must be allocated to appropriate projects. Specifically, for global projects, investment decisions for clinical trials are made cautiously, considering the probability of success, as the total cost may surpass 10 billion JPY when all phases from Phase 1 to Phase 3 are conducted. Approximately 10% of new drugs approved in Japan were not approved internationally, with most of these drugs being produced by Japanese pharmaceutical companies [17]. If Japanese pharmaceutical companies hesitate to invest in global development, essential drugs may not be marketed in other countries. A marginal correlation was observed between higher trial costs and projects with annual sales forecasts exceeding 10 billion JPY for the Japanese market. Conversely, projects with annual sales forecasts below 2 billion JPY showed lower costs, suggesting that decisions to undertake costly projects with significant costs may be influenced by expected returns, as projects with higher sales forecasts justify larger budgets.

Regarding durations, the findings of this study largely align with those reported in studies using global data, although Phase 3 durations were slightly shorter [9]. No project parameters had a strong impact on the overall phase duration. The correlation coefficients between projects targeting approval involving the USA or Europe and phase durations of each phase fell within the range from 0.08 to 0.15, showing no meaningful relationships. In contrast, correlation coefficients with costs for each phase ranged from 0.28 to 0.42, indicating a stronger relationship between costs and phase durations. Several studies have linked trial duration to therapeutic areas, particularly in oncology [9, 18]. While this study also identified a trend of longer oncology trials compared with others, no notable differences were observed. Given that this study did not focus on a specific therapeutic area, the limited number of projects in certain therapeutic areas may have influenced the results. The correlation between cost and duration, without accounting for other parameters, showed a stronger relationship in the earlier phases. Factors such as extended observation periods for participants, delays in case enrollment, and prolonged preparation for subsequent phases or applications contribute to increasing clinical trial costs [10, 19, 20]. If the impact of duration on costs remains constant across phases, these correlation coefficients suggest that cost variations in later phases may be influenced by more significant factors.

A strength of this study is the large-scale, questionnaire-based survey that quantified clinical trials conducted by Japanese pharmaceutical companies. Limitations include the lack of access to more detailed clinical trial information, such as the number of patients enrolled, number of sites, number of countries, frequency of visits, and size of the conducting company. This information was restricted to protect trade secrets, and the absence of these parameters may have impacted cost and duration estimates. Additionally, confidential information, such as sales forecasts, was often missing, making it difficult to estimate these values. The limited data also complicated efforts to impute missing values. Once all relevant parameters are available, the marginal effects on trial costs and durations can be accurately estimated using multivariable regression. The survey was conducted retrospectively using each company’s database. As these databases are optimized independently and not in a common format, this may have led to variations in the information provided. To minimize potential bias and standardize answers across companies, we conducted a question-and-answer session with respondents from each company. Although this study focused exclusively on clinical trials, research on drug discovery also presents investment risks for pharmaceutical companies. Further research is necessary to understand the overall burden of drug discovery and development on companies in Japan.

Conclusions

This study quantified the costs and durations incurred by Japanese pharmaceutical companies in conducting clinical trials, highlighting the substantial investments required by the Japanese pharmaceutical industry to bring new drugs to the global market. The findings provide critical insights into the scale of clinical trials and the factors influencing them, assisting government agencies in the formulation of pharmaceutical policies and helping pharmaceutical companies strategically allocate their resources. Furthermore, the study suggests the need for a system that encourages Japanese pharmaceutical companies to invest in research and development without hesitation, ensuring the timely availability of essential drugs in the global market.

Declarations

Funding

The authors received no external financial support for the research, authorship or publication of this article.

Conflict of interest

N.O. is an employee of Chugai Pharmaceutical Co., Ltd. Y.T. is an employee of Nippon Shinyaku Co., Ltd.

Availability of data and material

The datasets analyzed during the current study are not publicly available as they contain information that could reveal nonpublic details from individual companies. However, they are available from the corresponding author on reasonable request.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Code availability

Analyses were performed using the statistical software R version 4.3.0. The code is available from the corresponding author on reasonable request.

Author contributions

N.O. designed the study, analyzed and interpreted data, and wrote the manuscript. Y.T. designed the study, collected and interpreted data, and wrote the manuscript.

Additional information

The views expressed in this article are those of the authors and do not necessarily represent the official views of the Office of Pharmaceutical Industry Research.