Concern with FDCs

To our knowledge, this is the first empiric study into the concerns described about FDCs by the 59th Report. It examines the regulation and use in India of oral FDCs in four therapeutic areas: analgesia (NSAIDs), diabetes (metformin), anxiety/depression (benzodiazepines/anti-depressants), and psychosis (anti-psychotics). We chose these areas because the drugs are commonly used and/or are associated with considerable potential for interactions and adverse events. Examples of both NSAID and anti-psychotic FDCs were highlighted in the 59th Report.

India’s FDC Regulatory Framework

The much-amended Drugs and Cosmetics Act, 1940, and Drugs and Cosmetics Rules, 1945, govern the regulation of drugs [15] (S1 and S2 Texts). The 1940 law, passed under British colonial rule, placed responsibility for imports on central government, with the states being responsible for manufacture, distribution, and sale. Following independence in 1947 and subsequent adoption of the Constitution, “drugs” became a matter contained in the “Concurrent List” so that both the national Parliament and the state legislatures had, and have, power to make laws in relation to them. In 1952, national rules introduced the concept of a “new drug” along with the requirement for prior central approval for import (S3 Text). This was followed in 1961 by the requirement for prior central approval for manufacture, along with an obligation on state license applicants to produce evidence that the drug had been approved (S4 and S5 Texts). FDCs were not specifically mentioned, but they were regarded as new drugs, with recorded central approvals for FDC formulations dating (continuously) from 1961.

Increased central control of drug regulation has occurred incrementally ever since, whilst the states have retained their licensing powers over the manufacturing and sale of most drugs. A 1988 amendment inserted a new Part XA into the national rules entitled “Import or manufacture of new drugs for clinical trials or marketing” (S6 Text). Part XA included (and includes) requirements for pre-manufacturing central approval before a state manufacturing license is granted and for license applicants to produce evidence of that approval, whilst expressly including FDCs in the definition of a new drug and setting out specific data submission requirements for FDCs. After September 1988, FDCs combining drugs for the first time that had been individually approved previously, or previously combined FDCs with new claims, were expressly included within the definition of a “new drug” under Rule 122E(c). Those FDCs therefore required central approval prior to manufacturing under Rules 122B or 122C, and applicants had to submit evidence to state authorities of that prior approval. This is reflected in the heading of Rule 122D: “Application for permission to import or manufacture fixed dose combination of drugs”.

In 2001, the rules were amended again to impose the legal duty on the CDSCO to be satisfied when approving new drugs for import or manufacture that they are safe and effective. The duty was imposed for FDCs as well, with the amendment further stating that FDCs needed prior approval even though they fell within the definition of new drugs and so were covered as far as the “safe and effective duty” was concerned, whilst the post-1961 provisions and the 1988 amendments covered them as far as the requirement for prior central approval was concerned. An amendment in May 2002, inserting Rules 69(6) and 75(6), essentially duplicated the requirement to produce evidence of prior approval of “new drugs” that had been in the rules since 1961, and extended it to require evidence of approval in favour of the applicant. The 59th Report (Section 9.2 of [6]) noted “some ambiguity” until May 2002. We identified no ambiguity in the rules. Our detailed analysis of the rules (S1 Text) leads us to consider that an FDC needed prior central approval for manufacture—and the submission to states of evidence of that approval—from 1961 if it fell within the three different definitions of a “new drug” applying from 1961–1988, 1988–1999, and 1999 onwards (see Box 1). Rules 69(6) and 75(6) are not relevant to determining that question, but they imposed an additional requirement of producing evidence of approval in favour of the applicant.

Further amendments in 2005 removed references to minimum numbers or ranges of participants and sites in “new drug” clinical trials and gave the CDSCO discretion to override data submission rules. For 4 y after approval, or after inclusion in the Indian Pharmacopoeia if earlier, companies wanting to market new drugs—including FDCs—must obtain approval of their own formulation from the CDSCO. After 4 y, new drugs cease to be deemed “new” drugs, and applications for manufacturing/distribution licences can be made to state licensing authorities without prior CDSCO approval.

Objectives of This Study

We aimed to determine if evidence existed to support the concerns of the 59th Report. Our objectives were to (i) determine the numbers of approvals granted by the CDSCO (1961–2013) for oral FDC formulations in the four therapeutic areas of interest; (ii) determine for each therapeutic area the numbers of FDC formulations on the market (2011–2012), their CDSCO approval status, the numbers of branded products marketed, and the relative contributions to FDC sales (2011–2012) of formulations with and without a record of CDSCO approval (“approved” and “unapproved”); and (iii) evaluate the impact of the May 2002 amendment to the rules by determining the proportions of new formulations launched on the market before and after 1 May 2002 that had CDSCO approval, the numbers of products arising, and their sales volumes (2011–2012).

Finally, we wished to determine if FDC formulations available in India were approved by United Kingdom (UK) and/or United States of America (US) regulators or included drugs banned, restricted, or unapproved internationally, and to apply our findings to make recommendations for rationalising the regulation of, and hence the use of, FDCs in India.

FDC Approvals in India

Using publicly accessible records available from the CDSCO for the period 1961–2013, we collated information on FDC approvals granted annually in each area [16]. The CDSCO listed approvals chronologically in a portable document format (pdf) that included the drugs comprising individual FDC formulations, indication, and the date of approval. Relevant information was extracted into an Excel spreadsheet. We focussed on original formulation approval, that is, the first approval granted for the drug combination in the FDC being examined. We categorised a formulation as “approved” if the combination of drugs, irrespective of dose amounts or modified release variations, was ever recorded as approved by the CDSCO. We categorised a formulation as “unapproved” if it was not included in the list of CDSCO approvals, 1961–2013 (Box 2). We assumed the CDSCO approval records were complete. No information was available publicly on the clinical evidence that was provided to support approvals. State drug authority records of FDC manufacturing/distribution/sale licences were unavailable, but from the list of 294 FDCs banned by the CDSCO in 2007, we identified FDCs in the study categories that had state licenses only [14].

FDC Approvals in the UK and US

To determine approvals in the UK and US, we searched the Medicines and Healthcare Products Regulatory Agency (MHRA) and Food and Drug Administration (FDA) approval listings for 2014 [17,18]. The FDA index (the Orange Book) lists all approved FDCs and single drug formulations (SDFs) alphabetically by generic name [18]. The MHRA publishes no index of generic name FDC approvals, and its list of approvals does not include medicines licensed centrally by the European Medicines Agency [17], so to minimise the risk of overlooking FDCs approved for use in the UK, we also examined listings in the British National Formulary and in the Monthly Index of Medical Specialties (MIMS) [12,19].

FDC Sales in India

We determined for each therapeutic area the numbers of FDC formulations and branded products marketed, and we undertook a time-trend analysis of annual sales volumes over 5 y, 2007–2012. Data were obtained from PharmaTrac, a commercial database comprising monthly audits of pharmaceutical product sales through multiple supply routes (some 5,000 pharmaceutical companies, 18,000 distributers and stockists, and 32,000 sub-stockists) to over 500,000 retailers, hospitals, and dispensing doctors in 23 regions of India [20]. The audits capture 35% of national sales. The sampling data are projected to estimate national sales totals. The data largely represent private prescription sales; 85% are based on retailer sales, and 15% on drugs supplied through hospitals and dispensing doctors. They include generic and branded product names, manufacturers, sales volumes and value, and date of market launch. Each formulation has a code that identifies it in the database according to therapy type and treatment group, so, for example, diclofenac + paracetamol is M1A14, a musculoskeletal therapy in the anti-rheumatic systemic non-steroidal group; metformin + gliclazide is A10B10, an alimentary tract and metabolism therapy in the oral anti-diabetic group. Sales volumes are reported as the numbers of units of ten tablet/capsule doses; we converted units to the numbers of tablets/capsules. Value is reported in rupees; we used current exchange rates to report in US dollars. We examined monthly data available for the 5-y period November 2007–October 2012. We adhered to SAMPL guidelines for reporting [21].

There was a large amount of data; in presenting the results, we focus on the most recent 12-mo period, November 2011–October 2012. For each therapeutic area, we categorised sales according to whether FDC formulations had a record of approval by the CDSCO. Taking into account the 59th Report note on “ambiguity” in the national rules (Section 9.2 of [6]) until 1 May 2002 regarding the powers of state drug authorities to grant manufacturing licenses for FDC formulations that did not have prior CDSCO approval, we also examined CDSCO approvals for FDC formulations marketed before and after 1 May 2002, anticipating that all new formulations marketed after 1 May 2002 would have CDSCO approval. (S1 Table includes the individual formulation and product data for each of the 5 y examined.)

FDC Formulations: CDSCO Approvals and Formulations Marketed and Approved in India, the UK, and the US

CDSCO approvals for FDC formulations.

Between 1961 and 2013, the CDSCO approved some 1,125 oral FDC formulations including 67 NSAID FCD formulations (of which 52 were original formulations and 15 were variants where, for example, the dose of one component was changed or a modified release dose form used), 41 metformin FDC formulations (21 original, 20 variants), nine original benzodiazepine/anti-depressant formulations, and five original anti-psychotic drug formulations (Table 1). While some older formulations had become obsolete by 2011–2012 (e.g., indomethacin + dexamethasone, approved November 1971; trifluoperazine + analgin + ergotamine + caffeine, approved March 1976), 60 of these 87 approved original formulations were available on the market in 2011–2012.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Formulations, CDSCO approvals, products, and sales of NSAID, metformin, anti-depressant/benzodiazepine, and anti-psychotic FDCs.

https://doi.org/10.1371/journal.pmed.1001826.t001

In each therapeutic area, more new (original) formulations were launched on the market after 1 May 2002 than before (Table 1). Overall, the proportions of formulations with CDSCO approval after 1 May 2002 were unchanged compared with before 1 May 2002. The proportions of formulations with CDSCO approval increased for NSAID FDCs (26% to 28%) and anti-psychotic FDCs (0% to 38%) and decreased for metformin FDCs (100% to 75%) and anti-depressant/benzodiazepine FDCs (20% to 18%) (Table 1).

With the exception of the NSAID FDC aspirin + caffeine, none of the FDCs on the market in the therapeutic areas examined were included in the Indian Pharmacopoeia [22].

FDC Products Arising from Unapproved Formulations

Individual formulations gave rise to multiple branded products. On average, the ratio of formulations to products was around 1:20 (Table 1), but several formulations had many more products (Tables 2–4 and S2). The 124 NSAID FDC formulations gave rise to 2,739 products (excluding “other”), 25 metformin formulations to 536 products, 16 anti-depressant/benzodiazepine formulations to 301 products, and ten anti-psychotic formulations to 211 products. Overall, unapproved formulations gave rise to substantial proportions of branded products for all groups except metformin FDCs: 43% of NSAID FDC products, 76% of anti-depressant/benzodiazepine FDC products, 78% of anti-psychotic FDC products, and 2% of metformin FDC products (Table 1 and Fig 1)

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Proportions of formulations on the market, products on the market, sales volume, and sales value in each therapeutic area arising from formulations unapproved by the CDSCO.

*Excludes “other” formulations.

https://doi.org/10.1371/journal.pmed.1001826.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Metformin FDC formulations marketed in India.

https://doi.org/10.1371/journal.pmed.1001826.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Anti-depressant/benzodiazepine FDC formulations marketed in India.

https://doi.org/10.1371/journal.pmed.1001826.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Anti-psychotic FDC formulations marketed in India.

https://doi.org/10.1371/journal.pmed.1001826.t004

The 19 NSAID formulations launched on the market before 1 May 2002 gave rise to 1,895 NSAID branded products, of which 31% (n = 583) were from unapproved formulations. Of 844 products arising from the 105 formulations marketed after 1 May 2002, 69% (n = 582) were from unapproved formulations (Table 1). Among metformin FDCs, no products marketed before 1 May 2002 were from unapproved formulations, compared with 9% (n = 28) after 1 May 2002. For antidepressant/benzodiazepine FDCs and anti-psychotic FDCs, 87% and 100%, respectively, of products marketed before 1 May 2002 were from unapproved formulations, compared with 62% and 57% after 1 May 2002 (Table 1).

FDC Sales: Sales Trends for 2007–2012 and Sales Volumes (2011–2012) Arising from Unapproved Formulations

The total oral drug market in each therapeutic area was composed of FDCs and SDFs (Table 1). Sales volumes of NSAIDs and metformin/anti-diabetes drugs were substantial compared with anti-depressants/benzodiazepines and anti-psychotics. During the 5 y examined, FDCs comprised increasing annual proportions of total sales volumes (Fig 2), and by 2011–2012 they accounted for over half of all NSAIDs and oral anti-diabetic drugs sold, one-third of anti-psychotic drugs, and almost one-fifth of anti-depressants/benzodiazepines (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. FDCs as proportions of total annual oral drug sales volumes in each therapeutic area, November 2007–October 2012.

Metformin FDCs are expressed as a percent of total oral diabetes drugs. Numerator and denominator data are included in S3 Table.

https://doi.org/10.1371/journal.pmed.1001826.g002

Overall, across the four therapeutic areas, approved and unapproved formulations contributed varying proportions of FDC sales volumes: almost all metformin FDC sales were from CDSCO-approved formulations, but formulations unapproved by the CDSCO accounted for over two-thirds of anti-depressant/benzodiazepine FDC sales (69%), almost half of anti-psychotic FDCs (43%), and more than a quarter of NSAID FDCs (28%) (Table 1). Sales values followed generally similar patterns. Fig 1 summarises the proportions of formulations and products on the market and sales volumes and value in 2011–2012 arising in each therapeutic area from unapproved formulations.

Formulations marketed before 1 May 2002 accounted for most of the NSAID FDC sales volume (5,397.3 million tablets/capsules); 14% of this volume arose from unapproved formulations. In contrast, of the sales volume arising from formulations marketed after 1 May 2002 (1,743.8 million tablets/capsules), 74% was from unapproved formulations. Five metformin FDC formulations marketed after 2002 had no record of CDSCO approval, but their contribution to overall sales volume was low (0.4%). For anti-depressant/benzodiazepine and anti-psychotic FDCs, sales arising from formulations marketed before 1 May 2002 were comprised almost entirely of products arising from unapproved formulations; of sales arising from formulations marketed after 1 May 2002, however, 25%–29% were from unapproved formulations.

Formulation Composition, CDSCO Approval, Sales Volumes, and Numbers of Branded Products Marketed

FDC Formulations with State Licenses but Not Approved by the CDSCO

The list of 294 FDCs with state licenses but unapproved by the CDSCO included 72 NSAID FDC formulations, of which 33 were on the market; 15 anti-depressant/benzodiazepine FDC formulations, with five on the market; and one metformin formulation not marketed [14]. No anti-psychotic formulations were listed.

FDCs Including Drugs Banned, Restricted, or Never Approved Internationally

Multiple formulations, most without CDSCO approval, included drugs banned, restricted, or never approved internationally owing to adverse effects (Table 5) [23–29]. Sales volumes were generally high. None of these drugs had MHRA (UK) or FDA (US) approval.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Drugs withdrawn or restricted internationally that were included in FDC formulations marketed in India.

https://doi.org/10.1371/journal.pmed.1001826.t005

Many NSAID FDC formulations available in India (both approved and unapproved) included muscle relaxants or proteolytic enzymes, of which several were approved as SDFs in the UK and/or US but none were components of NSAID FDCs in either country (Table 6). FDCs including two NSAIDs (diclofenac + nimesulide, mefenamic acid + ibuprofen) were not approved in the UK or the US.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Muscle relaxants and proteolytic enzymes included in NSAID FDC formulations marketed in India.

https://doi.org/10.1371/journal.pmed.1001826.t006

Factors Potentially Contributing to the Large Numbers of FDCs

In the four therapeutic areas examined, most of the FDC formulations available in India were unavailable in the UK and US. FDCs were sold in large volumes. They dominated oral NSAID and diabetes drug sales and comprised substantial sales proportions in the CNS areas. In contrast, National Health Service data for publicly subsidised prescribed medicines that were dispensed in England show that metformin FDCs accounted for <5% of oral diabetes drug use in 2012 [38]. There was no FDC dispensing recorded in the other therapeutic areas [38].

Problematic production and use in India of FDCs will not be rectified without addressing the contributing causes. A commonly cited driver, though not discussed in the 59th Report, is manufacturers’ desire to avoid government price controls imposed on drugs included on the National List of Essential Medicines of India [39]. Price control has been evaded by including these drugs in FDC formulations [39,40] (though amendments in the 2013 National Pharmaceutical Pricing Policy may prevent this) [41]. However, price control could also be evaded by manufacturing SDFs of different strength to those on the National List of Essential Medicines of India [39], and the evasion argument does not explain why prescribers prefer FDCs. Neither does it explain the large numbers of FDC products on the market.

Attempts been made to improve India’s drug regulatory situation following the 59th Report. The Drugs Controller General (India) (DCG[I]) requested that states ask manufacturers to prove the safety and efficacy of their FDCs licensed without central approval within a period of 18 mo (January 2013) [42], policy guidelines for approval of FDCs were published (June 2013) [43], and the Chaudhury committee made recommendations on new drugs, clinical trials, and banning of drugs (July 2013) [44]. Legislative underpinning for the 18-mo safety and efficacy review, however, was lacking, and following industry lobbying, the DCG(I) excluded from it FDCs licensed before 21 September 1988 (June 2014) [45]. The president of Indian Drug Manufacturers’ Association subsequently made “an earnest appeal” to the Ministry of Health and Family Welfare to “maintain status quo” for FDCs licensed by states or approved by DCG(I) up to 30 September 2012 [46].

In our view, significant factors contributing to the large number of FDCs on the market include structural problems with provisions of the Drugs and Cosmetics Rules, 1945, stemming partly from the aged Drugs and Cosmetics Act, 1940, and amendments to the rules (2005) [47]. These include: the blurring over time of the initial clarity of responsibility between central government (responsible for imported drugs) and the states (responsible for manufacture, distribution, and sale); the absence of a duty on the regulator in the Drugs and Cosmetics Act, 1940, to be satisfied of a new drug’s safety and effectiveness; the apparently random interchangeability of the different concepts of efficacy and effectiveness; the unclear and contradictory scope of the FDCs that are covered by the 1988 rules, and their inconsistency with WHO guidelines [8]; and the downgrading in 2005 of data submission requirements [40]. The Drugs and Cosmetics (Amendment) Bill introduced in Parliament in 2013 and the proposed version put out for 12 d of consultation by the Ministry of Health and Family Welfare on 31 December 2014 are not aimed at these issues [48,49].

Recommendations

First, there should be an immediate ban on sales and manufacturing of FDCs not approved by the CDSCO. Withdrawal from the market should be staged, prioritising FDCs that include drugs banned or unapproved internationally and potentially most likely to harm patients. Prescribers need to review the needs of patients taking these FDCs, taper off the drugs, and substitute appropriate alternatives prescribed as SDFs with monitoring of benefit/need. Approved FDC formulations containing drugs that have been withdrawn or restricted by other regulators should be reviewed with the intention of applying similar restrictions in India unless a strong evidence base dictates otherwise. All other approved FDCs should be reviewed for continuing approval. In all cases, the evidence and rationale for decisions need to be made freely available. This means that the Ministry of Health and Family Welfare should exercise its statutory powers to prohibit the import, manufacture, distribution, and sale of FDCs under Section 10A or 26A of the Drugs and Cosmetics Act, 1945 [15], in line with any CDSCO recommendations or, in the event of it not doing so, should explain and publish its reasoning.

Second, transparency and accountability of regulatory decisions must be improved. Complete details of approvals, including the clinical evidence base, need to be made available by the CDSCO for public scrutiny. Support needs to be provided to the CDSCO in terms of planning, information systems, and skilled personnel in order to achieve this.

Third, the CDSCO and the state authorities should establish transparent and public procedures for ensuring that notice of all new drug approvals given by the CDSCO is immediately given to each state licensing authority, and vice versa for all manufacturing, distribution, and sale licences granted by each state. Such notices should be published on the websites of the CDSCO and the state licensing authorities.

Finally, following the DCG(I)’s worrying restriction of the FDC safety and efficacy review, and the 2014 parliamentary elections, the new Parliament should monitor closely the activities of the Ministry of Health and Family Welfare and the CDSCO. If the proposed Drugs and Cosmetics (Amendment) Bill, 2015, reaches Parliament [49], opportunity exists for legislators to propose amendments on safety, effectiveness, rationality, need, and, perhaps most importantly, transparent regulation. Requiring a review of the safety and effectiveness of all currently available FDCs, not unlike the legally required review mandated by the US Congress in 1962, rather than the administrative review that has been shown to be susceptible to industry lobbying, makes eminent sense [50].