Concordance Between Pharmaceuticals and Medical Devices Agency Review and Ministry of Health, Labour and Welfare Decision Among New Drug Applications in Japan

Abstract

New drug applications (NDAs) in Japan are reviewed by the Pharmaceuticals and Medical Devices Agency (PMDA). Those that pass the review are subsequently subject to deliberation by the Ministry of Health, Labour and Welfare Pharmaceutical Affairs and Food Sanitation Councils (MHLW‐PAFSC), and the MHLW legislatively grants approval based on its positive opinions. However, little is known regarding the relationship between the PMDA review and the MHLW decision. We retrospectively assessed the MHLW decision of NDAs that passed the PMDA review at the initial MHLW‐PAFSC deliberation from 2002 to 2022. The reasoning behind a non‐supportive opinion from the MHLW‐PAFSC and the sponsor's actions to overcome unresolved issues were also documented. A total of 2,117 of 2,153 (98.3%) NDAs that passed the PMDA review were approved at the initial MHLW‐PAFSC deliberation with a positive opinion. The remaining 36 applications were not approved at the initial deliberation and subjected to continued deliberation because of a non‐supportive opinion from the councils, although 29 were finally approved through revision of the application document (24 applications), re‐analysis of the data (1 application), or additional clinical trials (4 applications). Seven applications have not been approved, of which one was refused, four were withdrawn, and two were unknown. The MHLW approves NDAs that passed the PMDA review at the initial deliberation at a high rate, suggesting that NDAs only suitable for approval passed the review and reached the MHLW‐PAFSC deliberation. Only a few NDAs were not approved at the initial deliberation; however, most of them were finally approved.

Study Highlights.

• WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

New drug applications (NDAs) in Japan are reviewed by the Pharmaceuticals and Medical Devices Agency (PMDA) and subsequently deliberated by the Ministry of Health, Labour and Welfare Pharmaceutical Affairs and Food Sanitation Councils (MHLW‐PAFSC) before the approval decision by the MHLW.

• WHAT QUESTION DID THIS STUDY ADDRESS?

How did the PMDA review correspond with the MHLW decisions? What were the reasons for not being approved by the MHLW among NDAs that passed the PMDA review?

• WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

We analyzed 2,153 NDAs that passed the PMDA review and underwent deliberation by the MHLW‐PAFSC between 2002 and 2022. A total of 2,117 NDAs were approved at the initial MHLW deliberation; however, the remaining 36 underwent continued deliberation because of a non‐supportive opinion, whereas 29 were finally approved with additional actions. Together these findings indicate that passing the PMDA review almost guarantees the MHLW approval, that scientific discussion is completed during the PMDA review, and that the deliberations by the MHLW‐PAFSC are formalistic and repetitive and look not a value‐added exercise in the drug review process.

• HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

There is room for improvement in the drug review process to achieve a more rapid drug approval in Japan.

In Japan, new drug applications (NDAs) for regulatory approval are sequentially examined by the Pharmaceuticals and Medical Devices Agency (PMDA) and the Ministry of Health, Labour and Welfare (MHLW). Submitted NDAs are reviewed by the PMDA review team for efficacy, safety, clinical meaningfulness, data integrity, and product quality based on the Pharmaceutical and Medical Device Act (PMD Act) in Japan. The PMDA drafts the review report and holds the PMDA Expert Discussion at the end of the review process to hear opinions of external experts from academia. Those that pass the PMDA review are subject to deliberation by the MHLW Pharmaceutical Affairs and Food Sanitation Councils (PAFSC), which comprises the Committee on Drugs and the Pharmaceutical Affairs Council, with members from medical and non‐medical fields. After these procedures, the MHLW legislatively grants approval based on a positive opinion from the MHLW‐PAFSC.

The approval rates of submitted NDAs vary among regulatory agencies. A previous study reported that the approval rate at the initial review was 85.0% by the U.S. Food and Drug Administration (FDA) and 91.6% by the European Medicines Agency (EMA) for 107 NDAs submitted to both agencies from 2014 to 2016. ¹ Another study reported that, of 912 new drug/biologics license applications approved by the FDA from 2013 to 2018, 117 (12.8%) were not approved at the initial review, received complete response letters, and went through multiple review cycles. ² The EMA published 1,694 public assessment reports for NDAs from 2017 to 2022, of which 1,335 (78.8%) were approved, 53 (3.1%) were refused, and the remaining 306 (18.1%) were withdrawn during the review cycle. ³ The PMDA annual reports indicated that the approval rate of NDAs was approximately 90% in Japan, since there were 2,190 NDAs, 1,952 approvals, one refusal, and 174 withdrawals between April 2004 and March 2023, whereas prior to the establishment of the PMDA in April 2004, there was 140 NDAs, 109 approvals, and 31 withdrawals. ⁴

The FDA holds advisory committee meetings as needed for the regulatory review of NDAs prior to making a decision on whether it is suitable for approval, whereas the EMA makes a decision based on the committee's recommendation for every NDA, ⁵ , ⁶ and the concordance between approval decisions and the committees' opinions differs between the two agencies. A previous study indicated that the FDA's approval decisions do not always coincide with the recommendations of the FDA Advisory Committees from 2010 to 2021. ⁷ The FDA conclusions contrast with the advisory committee recommendations in approximately 83 of 376 (22.1%) cases from 2008 to 2015. ⁸ In addition, NDAs with negative recommendations from advisory committees have been approved by the FDA approximately once a year. ⁹ On the contrary, the EMA Committee for Medicinal Products for Human Use (CHMP) is responsible for conducting the initial assessment of NDAs and recommends whether approval should be granted in the European Union. ⁶ The decisions by the EMA are made based on the opinions of the CHMP.

The correlation between the PMDA review and the MHLW decision has not been thoroughly examined for NDAs that pass the review. The added value of the MHLW deliberations in the regulatory review process has also not been well evaluated, although a previous study revealed that there were 246 discussions regarding drug labels for 229 new drugs, and 76 of those were required label revisions in the MHLW‐PAFSC meetings from April 2012 to March 2015. ¹⁰ In addition, some drugs that have passed the PMDA review were considered multiple times by the MHLW‐PAFSC but have not yet been approved in Japan; however, the reasons for the non‐supportive opinions from the councils and the consequences of those extended drug development are unclear.

The objective of this study was to assess the association between the PMDA review and the MHLW decisions among NDAs that passed the PMDA review and underwent deliberation by the MHLW‐PAFSC.

METHODS

We examined NDAs that passed the PMDA review and underwent deliberation by the MHLW‐PAFSC from January 2002 to December 2022. The NDA database was obtained from the PMDA website and included new molecular entities, new indications, new dosages, new dosage forms, new combinations, and new administration routes. Generic medicines, biosimilars, and regenerative medicines were excluded. The meeting minutes of the MHLW‐PAFSC, which were available since January 2002, including those of the Committee on Drugs and the Pharmaceutical Affairs Council, for their recommendations, were collected. ¹¹ Detailed information was collected from the review reports by the PMDA, ¹² whereas we analyzed those reports written by the services of the Pharmaceuticals and Medical Devices Evaluation Center of the National Institute of Health Sciences, which was the predecessor of the PMDA, before the PMDA was established in April 2004. Further information was supplemented using drug labels and its supplemental documents and sponsor press releases.

The main outcome measure of this study was the MHLW decision of NDAs that passed the PMDA review at the initial MHLW‐PAFSC deliberation. We defined “continued deliberation,” that is, postponing the committee's judgment, a term noted in the meeting minutes by the MHLW‐PAFSC in many cases, as applications that were evaluated by the councils, but had not received positive opinion for approval at that time. The therapeutic area was assessed based on the submitted indication for drugs that had been under continued deliberation by the MHLW‐PAFSC. The approval status for these drugs in the United States and Europe was collected from the FDA ¹³ and EMA ¹⁴ websites.

The reasoning behind a non‐supportive opinion from the MHLW‐PAFSC and the sponsor's actions to overcome unresolved issues were also assessed in this study. Based on the Points to Be Considered by the Review Staff by the PMDA, ¹⁵ the reasons for the non‐supportive opinions from the MHLW‐PAFSC were classified into five main categories as follows: efficacy (Is the efficacy in the study population considered to be more effective than placebo according to the results of properly designed clinical studies?), safety issues (Are there any unacceptable risks as compared to the benefits?), unclear clinical meaningfulness (Do the obtained results have clinical significance?), data integrity (Has the reliability of the conducted studies and submitted documents been ensured?), and product quality (Can the drug be supplied continuously with stable efficacy and safety from a quality assurance standpoint?). For instance, if the prespecified primary endpoint was met but the effect size was too small, the application was classified as having met efficacy, although clinical meaningfulness was unclear. The efficacy issues were further divided into six subcategories, including endpoint disagreement, inconsistent results, indication disagreement, dosage/administration disagreement, inappropriate study design, and small sample size. The actions needed to overcome unresolved issues were classified into revision of the application document, re‐analysis of the clinical data, or addition of new clinical data for applications that have been finally approved.

The difference between the initial MHLW‐PAFSC deliberation date and the approval date was assessed for all NDAs, those did not go through continued deliberation, those for new molecular entities, and those for other new drugs. The difference was also calculated for revision of the application document, re‐analysis of the clinical data, or addition of new clinical data among applications that underwent continued deliberation and eventually approved by the MHLW.

RESULTS

During the study period, 2,153 NDAs that passed the PMDA review were considered by the MHLW‐PAFSC (Figure 1 ). The MHLW decisions always aligned with the recommendations of the MHLW‐PAFSC regardless of the positive or non‐supportive opinions. A total of 2,117 applications (98.3%) were approved by the MHLW based on a positive opinion from the MHLW‐PAFSC during the initial deliberation. The remaining 36 applications underwent continued deliberation because of a non‐supportive opinion from the committees (Table 1 ). The top three therapeutic areas for drugs that had been under continued deliberation were cardiology (8 applications), oncology (5 applications), and infectious diseases (5 applications), accounting for 50.0% of the total (Table 2 ). Continued deliberations have increased in recent years along with an increase in NDAs, whereas the proportion of continued deliberations per NDAs has remained unchanged. Most of them were approved after the second deliberation, and the maximum number of deliberations was four. Of the 36 applications undergoing continued deliberation, 22 (61.1%) were approved by the FDA or EMA prior to the initial decision by the MHLW.

Figure 1.

Flow diagram of outcomes for new drug applications deliberated by the MHLW Pharmaceutical Affairs and Food Sanitation Councils from 2002 to 2022. During the study period, 2,153 new drug applications that passed the PMDA review and were subsequently deliberated by the MHLW Pharmaceutical Affairs and Food Sanitation Councils. A total of 2,117 (98.3%) applications were approved with a positive opinion from the committees during the initial deliberation. The remaining 36 applications were under continued deliberation because of a non‐supportive opinion from the committees, whereas 29 were finally approved through additional actions of the sponsors. Seven applications had not yet been approved, of which one was refused, four had been voluntarily withdrawn by the applicants, and the remaining two are unknown. MHLW, Ministry of Health, Labour and Welfare; PMDA, Pharmaceuticals and Medical Devices Agency.

Table 1.

New drug applications under continued deliberation in Japan from 2002 to 2022

Generic Name	Brand Name	Sponsor	New drug category	Indication	Initial MHLW‐PAFSC deliberation date	MHLW Approval datea	FDA approvalb	EMA approvalb
Azelnidipine	Calblock	Sankyo	NME	Hypertension	8/30/2002	1/31/2003
Interferon alfa	IFN‐α Nisseki	Japanese Red Cross /Fuso/Japan Chemical	NME	Suppression of proliferation of hepatitis B virus or C virus	10/11/2002	Withdrawn
Interferon alfa	IFN‐α Nisseki	Japanese Red Cross /Fuso/Japan Chemical	NME	Cutaneous T‐cell lymphoma	11/22/2002	Withdrawn
Letrozole	Femara	Novartis	NME	Post‐menopausal breast cancer	11/22/2002	1/23/2006	X	X
Infliximab	Remicade	Centocor	NI	Rheumatoid arthritis with inadequate response to existing treatments	2/28/2003	7/17/2003	X	X
Clavulanate/Amoxicillin	Augmentin ES	GlaxoSmithKline	NC	Pediatric acute otitis media	7/27/2005	10/11/2005	X	X
Methylphenidate	Concerta	Janssen	NI, ND	Pediatric attention‐deficit/hyperactivity disorder	8/29/2007	10/26/2007	X	X
Ramosetron	Irribow	Astellas	NI, ND	Diarrhea‐predominant irritable bowel syndrome in men	4/25/2008	7/16/2008
Omalizumab	Xolair	Novartis	NME	Bronchial asthma (limited to patients with refractory asthma whose symptoms cannot be controlled by existing treatments)	10/27/2008	1/21/2009	X	X
Amlodipine/Atorvastatin	Caduet	Pfizer	NC	Comorbidity of hypertension or angina pectoris and hypercholesterolemia or familial hypercholesterolemia	4/24/2009	7/7/2009	X	X
Sibutramine	Obescare	Eisai	NME	Obesity	7/24/2009	Withdrawn	X	X
Paclitaxel	Abraxane	Taiho	NDF, ND	Breast cancer	11/30/2009	7/23/2010	X	X
Alogliptin/Pioglitazone	Liovel	Takeda	NC	Type 2 diabetes mellitus	8/26/2010	7/1/2011
Acetaminophen	Calonal	Showa Yakuhin Kako	NDF	Osteoarthritis	11/29/2010	9/26/2014		X
Recombinant human erythropoietin	Epogin	Chugai	NI, NDF	Anemia in cancer chemotherapy	6/13/2011	Refused	X	X
Bevacizumab	Avastin	Chugai	NI, ND	Unresectable or recurrent breast cancer	8/1/2011	9/26/2011	X	X
Anti‐Hib vaccine	Vaxem‐Hib	Takeda	NME	Prophylaxis of Hemophilus influenzae type b infection	9/5/2014	1/22/2016	X	X
Rivaroxaban	Xarelto	Bayer	NI, ND	Treatment and prevention of recurrence of deep vein thrombosis and pulmonary thromboembolism	2/20/2015	9/24/2015	X	X
Ibrutinib	Imbruvica	Janssen	NME	Relapsed or refractory chronic lymphocytic leukemia/small lymphocytic lymphoma	8/31/2015	3/28/2016	X	X
Duloxetine	Cymbalta	Shionogi	NI	Chronic musculoskeletal pain	2/5/2016	3/18/2016	X
Telmisartan/Amlodipine	Micatrio	Boehringer Ingelheim	NC	Hypertension	4/20/2016	9/28/2016
Ticagrelor	Brilinta	AstraZeneca	NME	Old myocardial infarction at especially high risk of developing atherothrombosis; Acute coronary syndrome for which percutaneous coronary intervention is indicated	5/27/2016	9/28/2016	X	X
Rifaximin	Rifxima	Aska	NME	Improvement of hyperammonemia in patients with hepatic encephalopathy	5/30/2016	9/28/2016	X	X
Semaglutide	Ozempic	Novo Nordisk	NME	Type 2 diabetes mellitus	12/4/2017	3/23/2018
Lisdexamfetamine	Vyvanse	Shionogi	NME	Pediatric attention‐deficit/hyperactivity disorder	12/3/2018	3/26/2019	X	X
Anamorelin	Adlumiz	Ono	NME	Cancer cachexia in the following malignancies: non‐small‐cell lung cancer, gastric cancer, pancreatic cancer, and colorectal cancer	8/29/2019	1/22/2021
Diclofenac	Joyclu	Seikagaku	NME	Knee and hip osteoarthritis	12/4/2020	3/23/2021
Favipiravir	Avigan	Shionogi	NI	SARS‐CoV‐2 infection	12/21/2020	Withdrawn
Vericiguat	Verquvo	Bayer	NME	Chronic heart failure	2/25/2021	6/23/2021	X
Sacubitril/Valsartan	Entresto	Novartis	NI, ND	Hypertension	8/30/2021	9/27/2021
Andexanet alfa	Ondexxya	Alexion	NME	Reversal of anticoagulant effects in life‐threatening or uncontrolled bleeding in patients treated with a direct‐acting factor Xa inhibitor	8/30/2021	3/28/2022	X	X
Prasugrel	Efient	Daiichi Sankyo	NI, ND	Prevention of recurrence after ischemic cerebrovascular disease	11/5/2021	12/24/2021
Aducanumab	Aduhelm	Biogen	NME	Alzheimer's disease	12/22/2021	Not yet approved	X
Finerenone	Kerendia	Bayer	NME	Chronic kidney disease associated with type 2 diabetes mellitus	1/28/2022	3/28/2022	X
Ensitrelvir	Xocova	Shionogi	NME	SARS‐CoV‐2 infection	6/22/2022	11/22/2022
Dexmedetomidine	Precedex	Pfizer	NI, ND	Sedation of non‐intubated pediatric patients during non‐invasive procedures and examinations	10/28/2022	Not yet approved

EMA, European Medicines Agency; FDA, U.S. Food and Drug Administration; MHLW, Ministry of Health, Labour and Welfare; NC, new combination; ND, new dosage; NDF, new dosage form; NI, new indication; NME, new molecular entity; PAFSC, Pharmaceutical Affairs and Food Sanitation Councils; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2.

^a Approval date or status at the end of 2022.

^b Approval status by the FDA or EMA at the initial MHLW decision for the same indication in Japan. Date is presented as m/d/y.

Table 2.

Characteristics of 36 new drug applications under continued deliberation in Japan

Characteristic	No. (%)
Medical specialty of the subject drugs
Cardiology	8 (22.2)
Oncology	5 (13.9)
Infectious disease	5 (13.9)
Endocrinology	3 (8.3)
Gastroenterology	3 (8.3)
Anesthesiology	3 (8.3)
Psychiatry	2 (5.6)
Others	7 (19.4)
Initial MHLW decision year of the subject drug
2002–2008 (476 applications)	9 (25.0)
2009–2015 (822 applications)	10 (27.8)
2016–2022 (855 applications)	17 (47.2)
Expedited designation
Priority review	1 (2.8)
Orphan disease	3 (8.3)
Sakigake	0 (0)
None	32 (88.9)
New drug category
New molecular entity	19 (52.8)
Others	17 (47.2)
Nationality of industries
Japanese	16 (44.4)
Non‐Japanese	20 (55.6)
Number of MHLW‐PAFSC deliberations
4	5 (13.9)
3	19 (52.8)
2	9 (25.0)
1	3 (8.3)
Approval outside Japan at the initial MHLW decisiona
FDA approval	21 (58.3)
EMA approval	18 (50.0)

EMA, European Medicines Agency; FDA, U.S. Food and Drug Administration; MHLW, Ministry of Health, Labour and Welfare; PAFSC, Pharmaceutical Affairs and Food Sanitation Councils.

^a Same indication for new drug application in Japan.

The reasons for the continued deliberation of 36 applications involved the following concerns: efficacy (23 applications), safety (20 applications), unclear clinical meaningfulness (17 applications), data integrity (5 applications), or product quality (3 applications) (Figure 2 ). Efficacy concerns included dosage/administration disagreement (10 applications), indication disagreement (7 applications), and inconsistent results in the clinical trials (6 applications), whereas those related to study design, such as inappropriate study design (3 applications), endpoint disagreement (2 applications), and small sample size (1 application), were less common. Four applications underwent continued deliberation at the MHLW‐PAFSC because of additional safety information (ibrutinib), a lack of data integrity (anti‐Hib vaccine, andexanet alfa), or concerns regarding product quality (paclitaxel) that sponsors informed to the PMDA after those reviews were completed.

Figure 2.

The reason for continued deliberation and the actions to overcome unresolved issues for drugs that were finally approved. The main reasons for continued deliberation for 36 applications included concerns regarding efficacy (23 applications), safety (20 applications), and clinical meaningfulness (17 applications). Dosage/administration disagreement (10 applications) was the most common reason among the six efficacy subcategories. Revision of the application document (24 applications) was the most common action to overcome unresolved issues for regulatory approval.

Twenty‐nine applications finally received a positive opinion by the MHLW‐PAFSC and were approved by the MHLW through revisions to the application document (24 applications), re‐analysis (1 application), or additional clinical trials (4 applications). Continued deliberation because of issues related to efficacy, safety, unclear clinical meaningfulness, data integrity, or product quality was finally approved in 78.3%, 75.0%, 64.7%, 80.0%, and 66.7% of the applications, respectively. Of the 29 applications that were finally approved, 22 (75.9%) were approved for the same indication either by the FDA or the EMA prior to their approval in Japan. By the end of 2022, seven drugs had not been approved in Japan, of which only recombinant human erythropoietin for the treatment of anemia in cancer chemotherapy was refused by the MHLW because of the risk of tumor progression and a worse prognosis in cancer patients, ¹⁶ four applications were voluntarily withdrawn, whereas the details of the remaining two applications were unknown. Of the seven drugs that have not been approved in Japan, interferon alfa for the treatment of suppression of proliferation of hepatitis B virus or C virus, interferon alfa for the treatment of cutaneous T‐cell lymphoma, and favipiravir for the treatment of severe acute respiratory syndrome coronavirus 2 were neither approved by the FDA nor the EMA, whereas dexmedetomidine was approved by the FDA after the MHLW initial decision.

The difference between the initial MHLW‐PAFSC deliberation date and the approval date, which is described as median (interquartile range), was 28 (21–51) days, 28 (21–50) days, 41 (28–60) days, and 26 (21–36) days for all (2,146 applications), those did not go through continued deliberation (2,117 applications), those for new molecular entities (725 applications), and those for other new drugs (1,421 applications), respectively, among NDAs that passed the PMDA review and eventually approved by the MHLW. The difference was 111 (63–159) days, 210 days, and 733 (192–1,337) days, for revisions to the application document, re‐analysis, or additional clinical trials, respectively, among 29 applications that underwent continued deliberation but were approved through additional actions.

DISCUSSION

The present study revealed that the MHLW decision based on the recommendations of the MHLW‐PAFSC for the initial deliberation was consistent with the PMDA review results. Most NDAs that passed the PMDA review were approved by the MHLW. There were a small number of applications that were not approved at the initial deliberation and underwent continued deliberation, which often resulted in approval with minor revisions to the application documents. A high approval rate by the MHLW indicates that NDAs only suitable for approval pass the PMDA review and reach the deliberation by the MHLW‐PAFSC, and passing the PMDA review almost guarantees the MHLW approval.

The MHLW‐PAFSC followed the PMDA review results but requested minor revisions in some cases. The very high approval rate by the MHLW strongly suggested that NDAs with defects in efficacy, safety, clinical meaningfulness, data integrity, or product quality were voluntarily withdrawn by the applicants during the PMDA review because it would not be possible to achieve regulatory approval. Most applications undergoing continued deliberation were finally approved with minor revisions of the submission documents. Together these results indicate that the scientific discussion is fully completed during the PMDA review. Most of the concerns raised at the MHLW‐PAFSC are made by stakeholders other than scientific members with relevant expertise, and such arguments do not take place at the PMDA Expert Discussion because of its membership. In this context, enhancing the participants and transparency of the PMDA Expert Discussion whose roles are similar to the FDA Advisory Committee and the EMA CHMP (Table 3 ), of which only a few summary sentences are presented in the review reports, with the abolishment of the repetitive process by the MHLW‐PAFSC may be helpful to boost scientific discussion and achieve a more rapid drug approval through more simple and reasonable review process. Future studies are needed to address the association between the PMDA review results and the recommendations from the PMDA Expert Discussion that is held at the end of the review process, and it may be of interest to compare the association between advisory committee's opinions and regulatory agencies' decisions among different jurisdictions.

Table 3.

Advisory committee meetings of the pharmaceutical regulatory agencies among Japan, the United States, and Europe

Feature	PMDA Expert Discussion 24	FDA Advisory Committee 25	EMA CHMP 6
Role	Provide scientific evaluations, advice, safety monitoring, decision support, policy development, and training	Offer expert advice, review applications, hold public hearings, provide recommendations, and monitor post‐market safety	Conduct scientific evaluations, issue marketing authorization opinions, provide scientific advice, and develop guidelines
Participants	Scientific experts with relevant expertise, agency reviewers, and others (as needed)	Academicians/practitioners, consumer representatives, industry representatives, patient representatives, agency reviewers, and others (as needed). Sponsor representatives submitting the product may give a presentation at the meeting	Chairperson (elected by serving CHMP members), representatives from each of the EU member states and the EEA‐EFTA states, co‐opted experts for specialized knowledge, agency reviewers, and others (as needed)
Frequency	As needed based on the evaluation and regulatory processes	As needed based on the evaluation and regulatory processes	Monthly meetings, with additional meetings as needed
Review items	Drugs, medical devices, clinical trial results, safety data, and others All products approved by the agency are reviewed by the committee at the end of regulatory review process	Drugs, medical devices, clinical trial results, safety data, and others. The committee meetings not always convened for every decision or product review but are held where additional expert input is deemed necessary or beneficial	Drugs, medical devices, clinical trial results, safety data, and others All products approved by the agency are reviewed by the committee at the end of regulatory review process
Transparency	Limited transparency with published summaries, but discussion content may be limited and detailed proceedings are not always made public	High transparency with public meetings which is available by webcast, published materials, and detailed records	Moderate transparency with published agendas, minutes, and highlights of the meetings, and scientific assessments/opinions

CHMP, Committee for Medicinal Products for Human Use; EEA‐EFTA, European Economic Area and European Free Trade Association; EMA, European Medicines Agency; EU, European Union; FDA, U.S. Food and Drug Administration; PMDA, Pharmaceuticals and Medical Devices Agency.

The reasons for continued deliberation mainly involved efficacy, safety, or unclear clinical meaningfulness, whereas concerns regarding safety and clinical meaningfulness were often associated with conducting additional clinical trials or not yet approved. Our results are compatible with a previous report that efficacy is the most common cause of the initial withdrawal during the PMDA review among NDAs approved in Japan from 2001 to 2011, although in their study, safety and clinical meaningfulness were not major reasons for initial failure. ¹⁷ A previous FDA report indicated that, of the 96 NDAs that required multiple review cycles before approval between 2001 and 2011, 74 (77.1%) had safety issues and 43 (44.8%) had concerns regarding efficacy. ¹⁸ Of the new molecular entities that received complete response letters, the primary reasons for rejection were safety (53.8%) and efficacy (76.3%). ¹⁹ Another study analysis of 61 new molecular entities that received complete response letters indicated that 42 applications (68.9%) had safety concerns and 41 (66.1%) had efficacy concerns. ²⁰ The FDA's evaluation focusing on applications with efficacy concerns revealed that additional clinical trials were required for those with inconsistent study results or unclear clinical meaningfulness. ² These results suggest that efficacy, safety, and clinical meaningfulness are the major reasons for not receiving approval from regulatory agencies.

The decisions by the FDA or EMA of those times overlapped with the continued deliberation may not affect the MHLW approval process. In this study, 61.1% of NDAs under continued deliberation have been approved by the FDA or the EMA before the initial decision by the MHLW, indicating a relatively low rate of global approval for the continued deliberation applications, since a previous study reported that 80% of new drugs in Japan have been first approved outside Japan. ²¹ Furthermore, 12 of 36 (33.3%) continued deliberation applications were finally approved by the MHLW before approval by the FDA or the EMA. Our study indicated that approval status outside Japan did not influence the MHLW approval process, though such discussions may be completed during the PMDA review process.

We found that there were some discordances in approval decisions among the PMDA/MHLW, the FDA, and the EMA for the NDAs of continued deliberation in Japan. For instance, recombinant human erythropoietin for the treatment of anemia in adults receiving chemotherapy to reduce blood transfusions has been approved by the FDA and the EMA but refused by the MHLW‐PAFSC with a negative opinion regarding the risk of tumor growth by its nature of one‐kind growth factor. Aducanumab for the treatment of Alzheimer's disease has been approved by the FDA; however, the application to the EMA has been withdrawn by sponsors based on the negative opinion from the CHMP, and the application to the PMDA/MHLW underwent continued deliberation by the MHLW‐PAFSC with concerns regarding efficacy, safety, and unclear clinical meaningfulness. Differences in the regulatory attitudes for risk–benefit assessment were discussed elsewhere. ²²

Regulators commonly set a target period for review and complete the review within their own time frame, that is, in Japan, the target review period is typically 12 months for standard reviews, 9 months for priority reviews including orphan designations, and 6 months for Sakigake designations ²³ ; however, there are differences in review cycles and withdrawals for NDAs that are not appropriate for approval among agencies. The FDA issues a complete response letter with critical comments on what is insufficient for approval when the review cycle is completed. The EMA publishes an assessment report for every NDA with a summary of what the agency thinks regardless of the conclusions. In these cases, a re‐application is required for further consideration in both the FDA and the EMA. In Japan, almost all NDAs that are not suitable for approval are voluntarily withdrawn by sponsors before the completion of the PMDA review, and detailed information on withdrawn applications does not open to the public except for sponsors' press releases, although the number of withdrawn applications was available in the PMDA annual reports. ⁴

Availability of application documents and regulators' assessment regarding NDAs is important for new drug development by improving the transparency of the review process and facilitating regulatory science communication. The regulators should inform their assessment to the public when the review was completed even if it was a negative conclusion. On the other hand, publicizing of not yet approved NDAs, whether refusal or withdrawal, will face a dilemma between confidentiality of intellectual property vs. transparency from both applicants' and regulators' views. Regardless of such conflict, in general, transparent manner will be desirable from the perspective of the facilitations of both further drug development and regulatory science.

This study had several limitations. First, we could not find detailed information on NDAs withdrawn with its reasonings during the PMDA review because these are not open to public. Second, for NDAs under continued deliberation, we could assess only the meeting minutes from the MHLW‐PAFSC. The review reports by the PMDA only become available after regulatory approval, and information on items under continued deliberation is highly confidential and limited information is available. Third, our study did not take into account the differences in the documents submitted to the PMDA, the FDA, and the EMA.

CONCLUSION

Most NDAs that passed the PMDA review were approved by the MHLW in Japan. Some NDAs passed the PMDA review but were not approved during the initial MHLW's deliberation, although most were finally approved through revision of the application document, clinical data re‐analysis, or further data collection. The high approval rate by the MHLW suggests that NDAs only suitable for approval passed the PMDA review and reached deliberation by the MHLW‐PAFSC; in other words, scientific discussions are completed during the PMDA review and the MHLW‐PAFSC supports the PMDA's view. Integrating the MHLW‐PAFSC deliberations into the PMDA review would improve patient access to new drugs and transparency of regulatory activities in Japan.

AUTHOR CONTRIBUTIONS

M.M., M.T., and M.I. wrote the manuscript. M.T. designed the research. M.M. performed the research. M.M., M.T., R.T., M.E., Y.Ta., H.M., M.I., and Y.To. analyzed the data.

FUNDING

This study was supported by JSPS KAKENHI Grant Number JP21K18094 (M.T.) and JP24K13311 (H.M.). The views expressed in this article are those of the authors and do not necessarily reflect the official views of the PMDA.

CONFLICT OF INTEREST

The authors declared no competing interests for this work.