2022 Volume 45 Issue 4 Pages 477-482
Early access to novel drugs, regardless of regional differences, is significant for patients worldwide. Although various efforts have been made to reduce the drug lag, it still exists in some regions, including Japan. In this study, we focused on the drug lag of first-in-class drugs in Japan and obtained fundamental information because we considered that first-in-class and me-too drugs are essentially different and should be treated separately. We analyzed 97 first-in-class and 176 me-too drugs in new molecular entity (NME)-approved drugs in Japan and the United States during the fiscal years between 2009 and 2019. Since government policy and the Evaluation Committee on Unapproved or Off-labeled Drugs with High Medical Needs (the Committee) have a huge impact on drug lag, we distinguished NMEs developed at the Committee’s request. First-in-class drugs were developed at the Committee’s request significantly more than the me-too drugs (p = 0.0034). Although it was not statistically significant, the approval lags were 498.0 d for first-in-class drugs and 535.0 d for me-too drugs. Multiple regression analysis showed that multi-regional clinical trial (MRCT) development strategy (p = 0.0043) and foreign origin drugs (p = 0.0072) were a reducing factor and a prolonging factor of drug lag, respectively. In conclusion, the drug lag for first-in-class drug approval was one year. Global drug development using MRCT is one of the most effective development strategies for reducing drug lags.
Early patient access to new innovative drugs is a fundamental social demand worldwide. Regulatory authorities and pharmaceutical industries collaborate to achieve early patient access through the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) frameworks for global development. The ICH E17 Guideline on General Principles for Planning and Design of Multi-Regional Clinical Trials (MRCTs) was finalized in 2017, and was adopted by the regulatory authorities of ICH regions.1) MRCT can reduce the development lag between regions.2) However, a time lag in accessing new drugs between regions sometimes exists because drug development is mostly for commercial purposes by private companies. For instance, quick recovery of the investment is a major concern for such developers; thus, drug price and sales forecast have a huge impact on the development strategy of new drugs. Accordingly, the U.S. is prioritized as the first target region of drug development because it has a value-based drug pricing system, and its pharmaceutical market is the largest in the world. Other countries, including Japan, are set aside at the beginning of drug development, resulting in drug lag from the U.S.
In Japan, drug lag has been discussed for more than a decade. The Ministry of Health, Labour and Welfare (MHLW) and the Pharmaceuticals and Medical Devices Agency (PMDA) have worked on this issue by increasing the review staff and implementing various regulatory measures including establishment of Evaluation Committee on Unapproved or Off-labeled Drugs with High Medical Needs (hereinafter, referred to as “the Committee”). The Committee reviews the medical needs and requirements for approval of development requests for unapproved drugs and indications in Japan that are approved in other developed countries. It has been reported that drug lag has decreased significantly because of these efforts. PMDA estimated that the drug lag was 0.6 years in 2019, and development lag and review lag, which are the main components of the drug lag, were 0.5 and 0.1 years, respectively.3) However, substantial drug lag in Japan still exists, compared with that in the U.S. where more than half of the drugs are first approved (zero drug lag).4)
Although drug lag has gradually decreased in Japan, most reports on drug lag are between the U.S. and Japan. Both countries accept first-in-class drugs and me-too drugs as new drug applications (NDAs). Here, we adopted Aronson and Green’s definition of the first-in-class drugs and me-too drugs.5) These drugs should be treated separately in the discussion of drug lag as their impact on patients is distinctive. When first-in-class drugs are approved, the situation is serious for most patients, and it opens the door for novel promising treatments for the disease. In contrast, me-too drugs may be significant for some patients because of different safety profiles, pharmacokinetic profiles, off-target effects, or other characteristics. Since only superiority over the placebo group or non-inferiority over the active group can be proved based on the current drug approval criterion, it is uncertain whether the me-too drugs are more effective or safer than the first-in-class drugs during approval. The distinctive characteristics of me-too drugs may be found after clinical data accumulation in the post-marketing period.
There are other reasons to distinguish drug lags of first-in-class drugs from those of me-too drugs because these drugs are fundamentally different with respect to drug development. Although some argued that neglecting the innovative value of me-too drugs may not be appropriate, and that first-in-class drugs were just a winner of the drug development race,6) me-too drugs are still cheaper and easier to launch for a pharmaceutical company because the systems for that class of drugs are in place.5) Since pharmaceutical companies of the me-too drugs can focus more on the marketing strategy, it is reasonable to assume that the characteristics and causes of drug lags for the first-in-class drugs are different from those for the me-too drugs. In this study, we revealed the current drug lag situation in Japan against the U.S. by focusing on first-in-class drugs as innovative drugs, which is fundamental to discussing drug development strategies and policies.
Although our aim is to exclude me-too drugs from the discussion, it does not mean that me-too drugs are meaningless. First-in-class drugs are not always best-in-class drugs, and me-too drugs can be the best in the class. Since best-in-class drugs are usually beneficial, especially in clinical settings, some may argue that time lags of best-in-class drugs should be examined for discussing drug lags that are truly meaningful for patients. However, the main purpose of our study is to discuss drug development strategies and policies. Whether best-in-class or not cannot be determined until clinical data are accumulated in the post-marketing period. Furthermore, effects of bringing first-in-class drugs to the market are more remarkable and fundamental for the entire society, including patients, healthcare workers, regulatory authorities, and pharmaceutical industries.
We surveyed approved drugs as new molecular entity (NME) in Japan between the fiscal years (FY) 2009 and 2019. We only analyzed NMEs also approved in the U.S. Drugs used for diagnosis and vaccines were excluded. We classified drugs as first-in-class or me-too using the Japan National Health Insurance (NHI) new drug price determination method,7) and information on which method was applied was found in records and materials of the Central Social Insurance Medical Council (Chuikyo) meetings when discussing new drug prices.8) The NHI new drug price determination method for NMEs has three distinct calculation methods: cost accounting system, price determination by comparable drugs (I), and price determination by comparable drugs (II). Price determination by comparable drugs (II) and cost accounting system has never been used for first-in-class drugs and me-too drugs, respectively. Price determination by comparable drugs (I) was used for both first-in-class and me-too drugs. In price determination by comparable drugs (I), a certain corrective premium is given when having a novel mechanism of action that is clinically useful. Thus, we categorized first-in-class or me-too drugs based on this information. The first-in-class drugs were defined as drugs priced with the cost accounting system or price determination by comparable drugs (I) when the innovativeness or usefulness premium was given based on the satisfaction of the requirement of a novel mechanism of action that is clinically useful. Others were considered me-too drugs, including drugs priced with price determination by comparable drugs (II) or some with price determination by comparable drugs (I) when the above requirement was not satisfied. In short, all new drugs priced by cost accounting system were classified as first-in-class drugs. All new drugs priced by price determination by comparable drugs (II) were classified as me-too drugs. When new drugs were priced by price determination by comparable drugs (I), the category would be determined depending on whether the corrective premium criterion was satisfied. If certain corrective premiums were given with the satisfaction of “new action mechanisms,” the drugs were classified as first-in-class drugs. If certain corrective premium were given without the satisfaction of that requirement or no corrective premium was given, the drugs were classified as me-too drugs.
We excluded drugs not listed on the NHI drug price listing. We also excluded drugs with no clinical trials in Japan and drugs with no information available on the particular NDA date in the U.S. All information regarding approval and price were acquired from publicly available data, namely, websites of the MHLW,9) PMDA,10,11) and the U.S. Food and Drug Administration (FDA).12)
NDA lag was calculated as the date of NDA or biologics license application in the U.S., subtracted from the NDA date of the same drug in Japan. Similarly, the approval lag was calculated as the approval date in the U.S., subtracted from the approval date in Japan. We also analyzed factors associated with drug lag using multiple regression analysis. We set the approval lag of first-in-class drugs, excluding the drugs requested by the Committee, to the response variable. The role of the Committee is to review the medical needs and requirements of unapproved drugs and indications in Japan that are approved in other developed countries. Since a drug lag is a premise for drugs reviewed by the Committee, we distinguished those (see DISCUSSION for detail).
We selected factors from the literature search13–16) that considerably affect the drug lag as independent variables: oncology or not, biological products or not, MRCT as development strategy or not, foreign origin or Japan origin of the drug, and two types of regulatory measures (priority review or others). Moreover, estimated peak sales were added to the factors since marketability is a crucial determinant for private companies to decide on investment in drug development. We set 10 billion Japanese yen (¥10 B) of the peak annual sales of NHI new drug price determination method as a boundary for high marketability because ¥10 B is a criterion for repricing for market expansion.17)
Drug lag between first-in-class and me-too drugs was compared using the Wilcoxon rank-sum test or the Chi-square test. Multiple regression analysis was used to explore the factors associating with drug lags. Statistical significance was set at p < 0.05. All statistical analyses were performed using EZR (ver. 1.54).18)
During our survey period, 273 NMEs, including 97 first-in-class drugs, were identified for analysis (Fig. 1). Figure 2 shows the transition in the number of annually approved first-in-class and me-too drugs. No typical trend change was observed when 21.9–46.7% of approved new drugs were first-in-class drugs.
NME: new molecular entity; FY: fiscal year; U.S.: United States; NHI: National Health Insurance; NDA: new drug application; * Evaluation Committee on Unapproved or Off-labeled Drugs with High Medical Needs.
White and gray columns represent the annual approved number of me-too and first-in-class drugs, respectively. The line graph shows percentage of first-in-class drugs in total annual approved drugs.
Table 1 shows the characteristics of the analyzed drugs. With regarding to estimated annual peak sales, 24 of 97 first-in-class drugs and 89 of 176 me-too drugs were above ¥10 B. The ratio above ¥10 B estimated peak sales for first-in-class drugs (24.7%) was statistically significantly less than that for the me-too drugs (50.6%) (p < 0.001). The ratio of drugs with foreign average drug price adjustment in first-in-class (12.4%) and me-too drugs (20.5%) was in the same range (p = 0.13). Of the 97 first-in-class drugs, 17 (17.5%) were developed at the Committee’s request. In contrast, 10 (5.7%) of the 176 me-too drugs were developed at the Committee’s request. The ratio of first-in-class drugs was 3.1 times more than that of me-too drugs, and the difference was statistically significant (p = 0.0034). The 10 me-too drugs were requested by the Committee in Japan because one of them was evidently more effective than the drugs previously approved in Japan, and the rest were expected to show more benefits based on clinical results and situations in other countries.
| First-in-class drugs | Me-too drugs | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Committee’s* request | Including | Excluding | Including | Excluding | |||||
| n | % | n | % | n | % | n | % | ||
| Total | 97 | 80 | 176 | 166 | |||||
| ATC classification | A | 19 | (19.6) | 15 | (18.8) | 15 | (8.5) | 15 | (9.0) |
| B | 9 | (9.3) | 8 | (10.0) | 22 | (12.5) | 19 | (11.4) | |
| C | 3 | (3.1) | 3 | (3.8) | 7 | (4.0) | 7 | (4.2) | |
| D | 1 | (1.0) | 1 | (1.3) | 3 | (1.7) | 3 | (1.8) | |
| G | 1 | (1.0) | 1 | (1.3) | 4 | (2.3) | 4 | (2.4) | |
| H | 1 | (1.0) | 1 | (1.3) | 3 | (1.7) | 3 | (1.8) | |
| J | 10 | (10.3) | 10 | (12.5) | 13 | (7.4) | 13 | (7.8) | |
| L | 32 | (33.0) | 25 | (31.3) | 62 | (35.2) | 59 | (35.5) | |
| M | 4 | (4.1) | 4 | (5.0) | 2 | (1.1) | 2 | (1.2) | |
| N | 12 | (12.4) | 8 | (10.0) | 27 | (15.3) | 24 | (14.5) | |
| P | 0 | (0.0) | 0 | (0.0) | 1 | (0.6) | 0 | (0.0) | |
| R | 0 | (0.0) | 0 | (0.0) | 10 | (5.7) | 10 | (6.0) | |
| S | 0 | (0.0) | 0 | (0.0) | 5 | (2.8) | 5 | (3.0) | |
| V | 5 | (5.2) | 4 | (5.0) | 2 | (1.1) | 2 | (1.2) | |
| Bio product | 29 | (29.9) | 26 | (32.5) | 47 | (26.7) | 45 | (27.1) | |
| Oncology | 30 | (30.9) | 24 | (30.0) | 47 | (26.7) | 46 | (27.7) | |
| Development strategy | MRCT | 33 | (34.0) | 33 | (41.3) | 78 | (44.3) | 77 | (46.4) |
| Other | 64 | (66.0) | 47 | (58.8) | 98 | (55.7) | 89 | (53.6) | |
| Origin | Japan | 13 | (13.4) | 13 | (16.3) | 21 | (11.9) | 21 | (12.7) |
| Foreign | 84 | (86.6) | 67 | (83.8) | 155 | (88.1) | 145 | (87.3) | |
| Regulatory measures | Priority review | 53 | (54.6) | 42 | (52.5) | 53 | (30.1) | 47 | (28.3) |
| Other | 44 | (45.4) | 38 | (47.5) | 123 | (69.9) | 119 | (71.7) | |
| Estimated peak sales above ¥10 B | 24 | (24.7) | 24 | (30.0) | 89 | (50.6) | 88 | (53.0) | |
| Applied foreign average drug price adjustment | 12 | (12.4) | 10 | (12.5) | 36 | (20.5) | 34 | (20.5) | |
* Evaluation Committee on Unapproved or Off-labeled Drugs with High Medical Needs; ATC: Anatomical Therapeutic Chemical; MRCT: Multi-Regional Clinical Trial; ¥10 B: 10 billion Japanese yen.
Figure 3 shows the median approval lag for first-in-class and me-too drugs for each fiscal year. The median and interquartile range (IQR) of NDA lag and approval lags in first-in-class and me-too drugs during the survey period were 766.0 [IQR: 1875.0] and 831.0 [IQR: 1778.0], and 742.0 [IQR: 1649.5] and 690.0 [IQR: 1571.3] days, respectively. When drugs developed at the Committee’s request were excluded, the median of NDA lag and approval lags decreased to 399.0 [IQR: 1360.3] and 498.0 [IQR: 1460.8] days for first-in-class drugs, respectively and 565.5 [IQR: 1367.0] and 535.0 [IQR: 1369.5] days, respectively for me-too drugs. It was not statistically significant in all comparisons of lags between first-in-class and me-too drugs (p ≥ 0.26). The lags reduced when only drugs not developed at the Committee’s request were considered in both first-in-class and me-too drugs; the reduction degrees were noticeably different, involving twice as many days in first-in-class drugs (367 and 333 d) as in me-too drugs (176.5 and 155 d).
Gray horizontal solid and dotted lines indicate total median approval lag of first-in-class drugs (solid, 831.0 d) and me-too drugs (dotted, 690.0 d), respectively. Black horizontal solid and dotted lines indicate total median approval lag of first-in-class drugs excluding the drugs requested by the Committee* (solid, 498.0 d) and me-too drugs (dotted, 535.0 d), respectively. * Evaluation Committee on Unapproved or Off-labeled Drugs with High Medical Needs.
Further focusing on the first-in-class drugs excluding the drugs developed at the Committee’s request, the multiple regression analysis revealed that MRCT strategy for clinical development was a reduction factor of drug lag and foreign origin product was a prolongation factor (Table 2). The median and IQR of the approval lag for first-in-class drugs developed by MRCT strategy or non-MRCT strategy were 155.0 [IQR: 304.0] or 1227.0 [IQR: 1877.0] days, respectively (p < 0.001).
| Coefficient | 95% Confidence interval | p-Value | |
|---|---|---|---|
| Intercept | 683.8 | [−148.2–1515.8] | 0.11 |
| Oncology | −260.0 | [−854.8–334.8] | 0.39 |
| Bio product | −232.1 | [−812.1–347.8] | 0.43 |
| MRCT | −821.4 | [−1376.6–−266.3] | 0.0043 |
| Foreign origin | 1002.6 | [279.6–1725.6] | 0.0072 |
| Priority review | −402.1 | [−975.2–171.0] | 0.17 |
| Estimated peak sales above ¥10 B | 70.6 | [−531.4–672.6] | 0.82 |
* Evaluation Committee on Unapproved or Off-labeled Drugs with High Medical Needs; MRCT: Multi-Regional Clinical Trial; ¥10 B: 10 billion Japanese yen.
Our results show that drug lags between first-in-class and me-too drugs are comparable. We presumed that drug lag would occur depending on the development strategy of pharmaceutical companies, which is not specific to first-in-class drugs. The ratio of employing MRCT as a development strategy is not very different between first-in-class and me-too drugs (Table 1). Notably, the median of estimated peak sales above ¥10 B in first-in-class drugs was less than half of that in me-too drugs (24.7 vs. 50.6%) with statistical significance. This estimation is utilized by the government to adjust the drug price for excessive market expansion, as well as by pharmaceutical companies to predict market size to plan drug development strategy. As drug development requires huge costs and developers need to recover their investment, estimated market size is a significant factor in drug development strategy. To compensate for the relatively small market size estimated, MRCT could have been employed more in first-in-class than me-too drugs to maximize profitability globally.
From the perspective of recovering investment, not only market size but also drug prices are important. In Japan, drug prices are determined by the government and not by developers; therefore, there is an adjustment system such that much higher and much lower drug prices, distinctive from other countries, are adjusted to a similar price range. In this situation, drug prices in several countries, including the U.S., are used as a reference for the adjustment, and it may be a valid reason to develop drugs in those countries before Japan. It is reasonable to develop drugs, especially in the U.S., where developers can determine the drug prices to obtain a certain profit. However, less than 21% of the approved drugs were adjusted for price using this scheme, and the proportion is not different between first-in-class and me-too drugs. It is considered that the economic aspects of drugs, such as market size, drug price, and price determination method, are not vital factors for drug lags in Japan.
Since the purpose of the Committee is to issue the request to develop unapproved or off-labeled drugs with high medical needs, these drugs are usually approved or commercially available in foreign countries, including the U.S. Alternatively, a drug lag is a premise. Thus, we compared drug lags by both including and excluding those drugs developed at the Committee’s request. Regardless of the development strategies employed for first-in-class or me-too drugs, drug lag was longer for more than 155 d than the median value if developed at the Committee’s request. Moreover, in this study, we included drugs with a drug lag of more than 15 years, such as anagrelide hydrochloride hydrate (median approval lag: 6405 d) in first-in-class drugs and glatiramer acetate (median approval lag: 6856 d) which were developed at the Committee’s request. These substantially increased the variations in the discussion of drug lag. We further discussed the exclusion of drugs developed at the Committee’s request. The median lags of first-in-class drugs were 1.09 years in NDA and 1.36 years in approval. We found there still exists over a year of drug lag for first-in-class drugs that are comparable with those of me-too drugs.
The meaning of a one-year drug lag depends on the context. If the target disease of the drug is critical and life-threatening, one year is incredibly long, leading to severe consequences. In contrast, this one year can be fruitful for the follow-up countries to observe and accumulate information on benefits and risks from post-marketing data. Okubo and Ono reported that local clinical studies might play a substantial role in optimizing post-marketing drug use.19) If we can use and analyze various post-marketing data with the data of Japanese population in foreign countries during the drug lag period, local clinical trials may operate at a low cost. It would be better to utilize real-world data, instead of risking patients’ health in clinical trials. We should note here that we are not in the position of judging whether a one-year drug lag for first-in-class drugs is intolerable or whether a one-year drug lag for me-too drugs is acceptable. We should treat those separately in the discussion on the drug lag issue; moreover we should have a different standard for that issue between first-in-class and me-too drugs. The impact of these factors on patients is distinctive. We further investigated the factors affecting drug lags specific to first-in-class drugs.
The results of multiple regression analysis revealed that the MRCT strategy reduces drug lag, and foreign-origin drugs development increases drug lag. In our analysis, 41.3% of first-in-class drugs were developed using the MRCT strategy, and the median drug lag (155 d) was only 12.6% of the lag (1227 d) with a non-MRCT strategy. MRCT has already been confirmed to shorten drug lag in several disease areas, including orphan cancers,15) and is a vital development strategy to reduce drug lag even in first-in-class drugs in Japan. However, we identified two first-in-class drugs, belimumab and vortioxetine hydrobromide, developed using MRCT, that are still located in the upper quartile of median drug lag for first-in-class drugs. The MRCT strategy was unsuccessful in reducing drug lags in these problematic areas. These MRCTs are not available in North America for clinical trial in operational regions, indicating that these MRCTs are follow-up development posterior to the U.S. It may be obvious that simultaneous drug development reduces drug lag between the U.S. and Japan. Unless thoughtfully designed, the MRCT strategy would not necessarily improve the drug lag situation.
In this study, the development of foreign-origin drugs can be classified into two types depending on the developer; Japanese company or foreign company. In the former, an extra working period for licensing before initiating a clinical study in Japan is required, which increases drug lags. In contrast, there are various reasons why foreign companies delay the initiation of clinical development in Japan, including marketability and regulatory aspects. The recently introduced Sakigake designation system can help resolve this situation20); however, more data are needed for further discussion.
There are several limitations to this study. We used the NHI new drug price determination method to identify the first-in-class drugs. The use of different classifications and definitions of first-in-class drugs may lead to different results. We set the threshold of estimated peak sales to ¥10 B in the multiple regression analysis. However, the range of estimated peak sales was quite wide, from less than ¥0.1 B to more than ¥90 B in the surveyed drugs. Moreover, we only selected oncology as an explanatory factor for the disease. We do not exclude the possibility that other factors of diseases may be associated with results.
In this study, by focusing on first-in-class drugs, we revealed the current drug lag situation in Japan against the U.S., which is truly meaningful for patients. The results indicated that there still exists a one-year drug lag. Further data accumulation and discussion are required to judge the meaning of a one-year drug lag, but first-in-class and me-too drugs should be treated separately in discussions on drug lag issues. Meanwhile, MRCT is a key factor in reducing drug lag.
The authors declare no conflict of interest.
