The Pharmaceutical Industry at a Glance Drug Lag: Analysis of Factors Contributing to the Long-term Lag of Japanese Approved Products

Printable PDF

Hiroshi Azuma, Senior Researcher, Pharmaceutical Industry Policy Institute
Chie Yoshiura, Senior Researcher, Pharmaceutical Industry Policy Institute
Yuki Miura, Senior Researcher, Pharmaceutical Industry Policy Institute
Akiko Yoshida, Senior Researcher, Pharmaceutical Industry Policy Institute
Shinichiro Iida, Senior Researcher, Pharmaceutical Industry Policy Institute

SUMMARY

1. Introduction

The National Institute of Biomedical Innovation Policy (NIBIO) has conducted a series of analyses of the current status of drug ^lag1). In "Drug Lag: Analysis of the Actual Lag in Japan " ²⁾, we reported that although the overall drug lag period has been shortening, there are still some items with long-term drug lags, and that many of the items with long-term lags are approved by different companies in Europe, the U.S. and Japan. The long-term drug lag is a direct disadvantage to patients suffering from diseases, and is recognized as a very serious issue.

Imai et al. of Kitasato University reported that there are multiple factors behind the drug ^lag3). They cite the difference between originator companies and Japanese development companies in their analysis, which focused on domestically approved drugs, and the small projected sales as factors.

In this paper, we focus on the most recently approved drugs and analyze in detail the differences between the companies that obtained approval, i.e., the difference between in-house and acquisition/partnership development methods and their backgrounds, as well as analyze and discuss the degree of influence of additional Japanese studies, market size forecasts, and product characteristics on the occurrence of long-term lags.

Survey Methodology

This report is based on a more detailed analysis of the factors behind the "Drug lag: Analysis of the actual lag situation of products approved in Japan" (hereafter referred to as "Drug Lag"), which appeared in Policy Research Institute News No. 70. The survey covered drugs that were approved in Japan between 2019 and 2022 as drugs with new active ingredients and that were approved later in Japan than in the U.S. and Europe, resulting in a drug lag. The drug lag period was calculated by identifying the date of application for approval and the date of approval from the data published by the review authorities in each country. Specifically, analysis was conducted based on information published on the websites of the Pharmaceuticals and Medical Devices Agency (PMDA), the U.S. Food and Drug Administration (FDA), and the European Medicines Agency (EMA). If the product was approved in either Europe or the United States, the information from the region where the product was approved at the earlier date was used. The main basic statistic was the median value, since there are some products with significantly long approval periods or with short approval periods due to special exceptions. In Europe, the approval information is based on the central review system, which may differ from the approval status in each country.

Items for which the Ministry of Health, Labour and Welfare (MHLW) solicited development companies or requested development by the "Study Council for Unapproved and Off-label Drugs of High Medical Need " ⁴⁾ (hereinafter referred to as "Study Council") were excluded from the analysis in this survey because of the influence of the factor of unapproved drugs not developed in Japan.

The definition of "in-house" or "acquisition/partnership" in terms of development type is defined in this report as follows. In-house" refers to products developed by a single company (including parent-subsidiary companies) in both the U.S., Europe, and Japan. Acquisitions and alliances were defined as those in which a company that had obtained approval in Japan acquired a company or in-licensed a developed product. Details of the logistic regression analysis reported in 3-7 are provided in Appendix 1.

Results

3-1. influence of difference in development mode

In ISIR News No. 70, we reported that the median lag period was longer for the cases in which the Japanese and European firms were different from each other than for the cases in which the Japanese and European firms had the same approval status. In order to explore this factor in more detail, we decided to investigate the status of corporate alliances related to Japanese development rights in the process of obtaining approval in Japan. Of the drugs approved in Japan from 2019 to 2022, 114 had a drug lag; the median drug lag for the 114 drugs was 20.2 months, the minimum was 0.03 months, the maximum was 257.1 months, and the mean was 35.6 months. Of the 114 products, 63 (55%) were in-house developed products and 51 (45%) were acquired through acquisitions or partnerships. The median lag periods for the in-house and the acquired/partnered products were 10.5 months and 44.8 months, respectively. These results were compared with those in Policy Research Institute News No. 70, and confirmed a similar trend of longer lags for acquired/partnered products, even though the time periods studied were different.

3-2 Effect of additional Japanese participation study

Information on Japanese subjects is important in the review process when applying for approval in Japan. In some cases, additional clinical trials are conducted to obtain information on Japanese subjects. Imai et al. also reported that differences in development strategies affect the length of development delays. In this report, we surveyed the studies in which Japanese participants were identified among the evaluation studies listed in the "Clinical Efficacy and Clinical Safety Data and Summary of Review by the Agency" section of the PMDA's review report. The timing of the start of each study was compared with the U.S. and European approvals, and the presence or absence of additional studies was determined. Information that was not disclosed in terms of the timing of the trials was considered to be of unknown timing because the content could not be confirmed.

As a result of the survey, 38 (33%) of the products had additional studies with Japanese participants after the approval in the U.S. and Europe, 75 (66%) had not, and 1 (1%) was of unknown timing. The median lag period by the presence of additional trials was 53.1 months for those with additional trials and 9.8 months for those without additional trials. This indicates that more than 30% of all drugs in which Japanese participants start trials after approval in the U.S. or Europe, but it also reveals a long-term drug lag, with a median difference of less than 4 years.

3-3. effect of difference in corporate classification of companies approved in the U.S. and Europe

In recent years, the remarkable rise of emerging biopharmaceutical companies (hereafter referred to as "start-ups") in drug development around the world has been reported in various ^{fields5, 6)}, and we investigated their impact on drug lag for products approved in Japan. We categorized the companies that obtained approval in Europe and the U.S. according to whether they are emerging companies or not. The definition of start-ups in this paper is the same as in Policy Research Institute News No. 70, i.e., companies that have been in business for less than 30 years and had sales of less than US$500 million in the year prior to approval, while pharmaceutical companies are defined as non-start-ups.

The median lag time was 42.8 and 15.3 months for 33 (29%) and 81 (71%) of the approved products in the U.S. and Europe, respectively. The median lag periods were 42.8 months and 15.3 months, respectively.

3-4. effect of nationality of domestic development companies

In the Policy Research Institute News No. 70, we analyzed the difference in the nationality of the domestic development firms, whether they were Japanese or foreign, and we also examined the effect on the long-term lag in this report. The median lag periods were 47.5 and 14.4 months for 38 (33%) and 76 (67%) items, respectively, for which the domestic developer was a Japanese firm and an overseas firm. The median lag periods were 47.5 months and 14.4 months, respectively. In the case of Japanese firms developing only for the domestic market, it is assumed that the time taken to introduce a product from a U.S. or European approved company through an alliance had an impact on the lag time.

3-5. relationship with market size forecast

This paper examines the relationship between the market size forecast at the time of the NHI drug price listing in Japan and the lag period, even for the most recent data from 2019 to 2022.

In the Chuikyo document "Calculation of NHI Drug Prices for New Drugs," 35 items (31%) had a market size forecast (at peak) of 10 billion yen or more, while 70 items (61%) had a market size forecast of less than 10 billion yen. The median lag period (9.8 months) for products with a forecasted market size of 10 billion yen or more and 28.9 months for those with a forecasted market size of less than 10 billion yen were not available as of November 2023, including COVID-19-related vaccines. Thus, long-term lags were observed for items with relatively small projected market size, and the number of items accounted for more than half of the total.

3-6. classification by other items

In addition, the survey also examined whether or not small molecule drugs were applicable, whether or not orphan drugs were applicable, and whether or not preferential measures were taken by the pharmaceutical affairs bodies at the time of approval in the U.S. and Europe. Figure 6 shows the median number of applicable items and the median lag period for each category.

Logistic regression analysis

Statistical methods were used to estimate the impact of acquisitions and alliances and other factors on long-term drug lag, and to compare the impact of these factors (Table 1).

In a logistic regression analysis (Appendix Table 1) conducted on the probability of the lag period exceeding 24 months using data for 114 products, two influencing factors were statistically significant: "acquired/partnered products" was significant (P value = 0.003) and "products for which additional studies with Japanese participants were conducted after approval in the US and Europe" was significant (P value = 0.000). Two factors were statistically significant (P value = 0.000).

At the same time, this analysis estimated a 5.9 times higher odds ratio that the lag period would be 24 months or longer for acquired/partnered products compared to in-house products. The odds ratio of having a lag period of 24 months or longer was estimated to be 41.7 times higher for products that underwent additional studies with Japanese participants after approval in the U.S. and Europe than for products that did not undergo additional studies.

Comparing the marginal effects, the odds of lag period of 24 months or longer were estimated to be 21% higher for acquired/partnership products compared to in-house products, while the odds of lag period of 24 months or longer were estimated to be 45% higher for products for which additional Japanese participation studies were conducted after US/European approval compared to products for which no such studies were conducted. This result indicates that the impact of additional Japanese studies conducted after US/EU approval is particularly large.

To check the robustness of the analysis, a Cox proportional hazards model was also used. The results showed that "acquired or affiliated products" was significant (P value = 0.080), and "items for which additional Japanese participation studies were conducted after approval in the U.S. and Europe" was significant (P value = 0.000), both of which were statistically significant.

On the other hand, none of the other analyzed influence factors (Appendix Table 1), including whether or not the company that obtained approval in the U.S. and Europe in 3-3 was an emerging company, were statistically significant in any of the models. In order to evaluate the influence of other explanatory variables on the "items for which an additional study with Japanese participants was conducted after approval in Europe and the U.S.," which had the largest influence in the logistic regression analysis, the influence of each explanatory variable was evaluated with the additional study with Japanese participants after approval in Europe and the U.S. as the dependent variable (Appendix Table 1). Of the 114 products analyzed in this paper, 105 products for which market size forecast data at the time of NHI drug price listing were publicly available were included. As a result, statistical significance was found for "sales forecast of less than 10 billion yen" (P value = 0.036), "Japanese national of domestic development company" (P value = 0.002), "emerging company approved in the US or Europe" (P value = 0.034), and "not classified as an orphan drug" (P value = 0.007). Molecular drugs" and "Not receiving preferential treatment at the time of approval in Europe and the U.S." were not statistically significant (P-values = 0.788 and 0.960, respectively).

4. characteristics by development type

4-1 Relationship between development type and the presence or absence of studies with Japanese participants after approval in the U.S. and Europe

To gain a better understanding of the background and factors that influence long-term lag, the analysis was conducted separately for in-house and acquired/partnership products. In the Policy Research Institute News No. 70, a difference in the lag period between Japan and the U.S. and Europe was also examined in detail, since the difference in the lag period was observed in cases where the company acquiring approval differed between Japan and the U.S. and Europe. Of the 63 in-house products, 9 (14%) had undergone additional Japanese participation studies after approval in the US and Europe, and 54 (86%) had not. Of the 51 acquired/partnership products, 29 (57%) had undergone additional Japanese studies after approval in the U.S. or Europe, and 21 (41%) had not. The lag periods for in-house products and those acquired/partnership products are shown in Figure 7, with a median of 51.3 months for in-house products and 54.8 months for those acquired/partnership products. The survey results showed a significant difference in the distribution of products, with most of the in-house products having completed the necessary studies prior to approval in the U.S. and Europe, while more than half of the acquired/partnered products had additional studies conducted after approval in the U.S. and Europe. The groups of acquired/partnered products that showed a long-term lag had additional studies. Although the percentage of companies conducting additional studies in-house was low, the lag was long, suggesting that additional studies are a factor in the occurrence of long-term lag, regardless of whether the company is in-house or in an acquisition alliance.

4-2 Relationship between development type and company classification of companies approved in the U.S. and Europe

We investigated the relationship between the type of development and the type of company approved in the U.S. and Europe by company classification. Of the 63 in-house products, 8 (13%) were from startups and 55 (87%) were from pharmaceutical companies. Of the 51 acquired/partnered products, 25 (49%) were from start-ups and 26 (51%) were from pharmaceutical companies. The median lag periods are shown in Figure 8. While the median lag periods for in-house products were 18.3 months and 9.3 months for both start-ups and pharmaceutical companies, the median lag periods for acquired/partnered products were 43.1 months and 48.4 months for start-ups and pharmaceutical companies, respectively.

Although half of the acquired/partnered products were from startups, many of them were from pharmaceutical companies that obtained European approval. The lag period may have been influenced more by the development mode of the acquisition or alliance than by the difference between startups and pharmaceutical companies.

5. impact of timing of acquisition of Japanese rights for acquired/partnered products

We investigated the impact of acquisitions and alliances on the lag period from a different angle. We categorized the timing of when the companies that obtained approval in Japan acquired the rights to develop the drug in Japan, using the U.S. and European approvals as the starting point for the drug lag period. The results showed that 38 (75%) of the 51 acquired/partnered products acquired the rights to develop in Japan prior to the US/EU approval, and 13 (25%) acquired the rights to develop in Japan after the US/EU approval, with a median lag period of 30.6 months and 87.8 months, respectively. The median lag time for each was 30.6 months and 87.8 months, respectively.

Of the 13 products for which Japanese development rights were obtained after approval in the U.S. or Europe, 11 (84%) had additional studies with Japanese participants.

6. discussion and summary

This report, a follow-up to Policy Research Institute News No. 70, examines additional factors that induce a long drug lag for Japanese-approved products, focusing on the differences in the type of development in Japan, such as whether the product is developed in-house or through acquisition or tie-ups.

Suggestions for Item Characteristics and Factors of Long-Term Lag

Analysis of the characteristics of drugs with drug lag in recent years showed that the median lag period was long for the following categories: acquisition/partnership, additional Japanese studies, startups, Japanese nationals, and market size of less than 10 billion yen, although the percentage of products in these categories was low (except for products with a market size of less than 10 billion yen). Logistic regression analysis on the probability that the lag period between US/EU approval and Japan approval exceeded 24 months showed significant effects on the lag period for "acquired/partnership products" and "products for which additional studies with Japanese participants were conducted after US/EU approval. A comparison of the marginal effects of the two influencing factors that were found to be significant indicated that the influence of conducting clinical trials in Japan after approval in the U.S. and Europe was greater than that of acquisition/alliance on the probability that the lag period would be 24 months or longer. The significant effect of acquisitions and alliances in addition to the presence of additional trials suggests that changes in development policy and strategy for the Japanese business, as well as the length of transaction procedures related to acquisitions and alliances, may have an impact on long-term lag. In the subsequent detailed analysis, the proportion of Japanese additional study items was high in acquisition/partnerships and low in in-house studies, but these lags were long, suggesting that conducting additional Japanese studies is a factor in the long-term lag, regardless of the development mode. In addition, the proportion of cases in which an additional Japanese study was conducted was larger in the case of acquisitions and alliances, suggesting that for many products, development and business expansion into Japan were not anticipated prior to the acquisition or alliance.

On the other hand, the effect on the probability of the lag period exceeding 24 months was not significant for the emerging companies, Japanese nationals, and orphan drugs, but the effect on the probability of conducting clinical trials in Japan after approval in the U.S. and Europe was significant, suggesting that the addition of trials may have an impact and be related to the long-term lag. Similarly, the impact on the probability of conducting clinical trials in Japan after approval in the U.S. and Europe was significant for products with a peak year market size forecast of less than 10 billion yen, suggesting that market size is a factor in the decision to conduct additional clinical trials in Japan.

The effect of whether the drug was a small molecule or a biotech drug, and whether the drug received preferential treatment at the time of approval in the U.S. or Europe, was not statistically significant for both the prolonged lag and Japanese clinical trials after approval in the U.S. or Europe, suggesting a low level of influence.

This paper discusses the following possibilities that may have led to different results from previous studies. The fact that the market size is less than 10 billion yen has no statistically significant effect on the long-term lag (Cox proportional hazard model). However, the difference is that the target years of the study, 2010-2020 and 2019-2022, are different for the target products, and this paper excludes the study conference products from the study. Compared to the previous study, the percentage of items with a market size of less than 10 billion yen is higher in this paper. In addition, the higher percentage of items with a market size of less than 10 billion yen and a drug lag of 24 months or less (data omitted) may have been the reason for the difference. Alternatively, the original model may have been too strongly influenced by the presence or absence of additional studies.

Direction of Solution

The most influential factor on the occurrence of long-term lag was the presence of additional Japanese studies after approval in the U.S. and Europe. Factors that increased the probability of the occurrence of long-term lag included the presence of an acquisition or alliance, the fact that the Japanese application was filed by a Japanese company, and the fact that the U.S. and European development companies were emerging companies. It is assumed that many of the products approved in the U.S. and Europe by emerging companies overseas are developed in Japan through acquisitions and alliances by Japanese companies, and that studies targeting Japanese patients are added after introduction, and this business development pattern may be one of the reasons for the long-term lag. Although further detailed investigation and analysis is needed to understand the background behind the addition of Japanese studies, it is easy to imagine that the Japanese region was not included in the pivotal studies conducted by emerging companies in the US and Europe. One of the directions to solve this problem is for emerging companies to incorporate Japan into pivotal studies conducted by emerging companies in the U.S. and Europe.

In the direction of eliminating drug lag/loss, the Ministry of Health, Labor and Welfare (MHLW)-led "Expert Panel on Comprehensive Measures to Achieve a Rapid and Stable Supply of Pharmaceuticals ^"7) will be launched in 2022, and the "Study Group on the Ideal Pharmaceutical Regulations to Strengthen Drug Discovery Capability and Ensure Stable Supply " ⁸⁾ will be launched in 2023, Discussions have been held by industry, government, and academia. In order to promote the incorporation of Japanese regions into pivotal studies by European and U.S. companies, a notice was issued allowing the omission of Japanese Phase 1 prior to pivotal ^studies9), and the establishment of PMDA's Washington Office to promote pharmaceutical consultation for development for ^{Japan10) .} The establishment of an "additional fee for expedited introduction " ^{11), which} provides incentives for early application and approval in Japan, are expected to prove effective in the future as useful measures under the NHI drug price system.

Another useful measure would be to incorporate requirements for early deployment in Japan through acquisition and in-licensing by major pharmaceutical companies and Japanese pharmaceutical companies in the early clinical phase. The risk of not achieving a product in the early clinical phase is high, making acquisitions and in-licensing difficult in the pharmaceutical industry, but we expect the industry to continue its efforts in early acquisitions and in-licensing, taking advantage of scientific and technological trends.

Acknowledgments

We would like to thank Professor Junichi Nishimura of Gakushuin University, Department of Economics, for his kind guidance and support in statistically evaluating the impact on the lag period. We would like to take this opportunity to express our deepest gratitude.

Details of Statistical Analysis

The impact on the drug lag period was estimated by logistic regression analysis using the statistical analysis software Stata/IC 14.0 for Windows (Stata Corp LP, CollegeStation, TX, USA). The 114 items studied in this paper were included in the estimation, and the explained variable was a dummy variable that took 1 if the drug lag period was 24 months or longer and 0 if it was less than 24 months. As an explanatory variable, a dummy variable was used in which a factor that was thought to influence the increase in the drug lag period was assigned a value of 1. The dummy variable was a dummy variable that took 1 if the product was an acquisition or partnership, a dummy variable that took 1 if the domestic developer of the product was a Japanese national, a dummy variable that took 1 if the company that obtained approval in Europe or the U.S. was an emerging company, a dummy variable that took 1 if the product was a small molecule drug rather than a biotech drug, and a dummy variable that took 1 if the product was a biotech drug rather than a small molecule drug in Europe or the U.S. A dummy variable that takes 1 for products for which additional studies with Japanese participants were conducted after approval in the U.S. and Europe, a dummy variable that takes 1 for products that were not designated as orphan drugs, and a dummy variable that takes 1 for products that did not receive preferential treatment at the time of approval in the U.S. and Europe.

Survival time analysis using the Cox proportional hazards model was conducted to confirm robustness.

The probability of the lag period being 24 months or longer was strongly influenced by the additional Japanese studies conducted after the approval in the U.S. and Europe. A logistic regression analysis was conducted using this variable as the explained variable to estimate the factors influencing the probability of this variable taking the value of 1. The variables that were assigned to sales forecast of less than 10 billion yen, Japanese nationality for domestic development companies, emerging companies for companies approved in Europe and the U.S., and non-applicability of orphan drugs showed statistically significant effects with a P value of <0.1, while the variables assigned to small molecule drugs and non-applicability of preferential treatment in Europe and the U.S. did not show statistically significant effects.

Appendix Table 1

Details of the results of the statistical analysis are presented below. 4.

Return to News List

TOP