Table A1: Review of bioequivalence trials of rifampicin-containing solid oral dosage forms

This table summarizes bioequivalence trials of rifampicin containing solid oral dosage forms published since 1970, in peer-reviewed journals.

Bioequivalence trials of oral modified release formulations of anti-TB drugs are not listed in this table.

The reported bioequivalence trials are arranged in descending chronological order.

Unless otherwise stated, 2-FDC, 3-FDC and 4-FDC are the combinations of RH, RHZ and RHZE, respectively.

R: Rifampicin, H: Isoniazid, Z: Pyrazinamide, E: Ethambutol, FDC: Fixed-dose combination, PAS: p-amino salicylic acid, TB: tuberculosis

*vs: versus

Since the 1980s, combinations of isoniazid with p-amino salicylic acid (PAS), thioacetazone, ethambutol and rifampicin have been marketed for convenient administration and to avoid monotherapy with isoniazid which was a tempting choice for patients because of its small bulk. Rifampicin-containing FDC preparations in combination with isoniazid and pyrazinamide were first developed at the Lepetit Research Center, Italy, and the plasma concentrations of the three initial combination preparations (Rifater 1, 2 and 3) were found to be closely similar to the corresponding separate formulations. The problem of rifampicin bioavailability as a consequence of the manufacturing process was identified in the early 1980s when, in a further Lepetit preparation (Rifater 4), the order of mixing of the three component drugs was changed, resulting in an alarming reduction in the absorption of rifampicin^1,2. Since then altered bioavailability of rifampicin from various preparations has been reported and efforts made in both industry and academia to elucidate the underlying causes of this problem. However, the studies were hindered by lack of information in public domain regarding the changes made in the formulations and their effects on rifampicin bioavailability.

It is apparent from the excellent reviews by Fox^1,2 that much of the information regarding the development of FDCs and rifampicin bioavailability has not been published. Complete information regarding the excipients used, the change in the manufacturing process, etc., was not disclosed^3,4 and remained in the company’s drug master files. This lack of information has delayed progress in industry as well as academia with regard to understanding and addressing the problem of rifampicin-bioavailability in FDCs.

Even four decades after the discovery of rifampicin, the cause of altered bioavailability of rifampicin from some of the formulations is not yet clear and the reasons are only speculative. Hypotheses put forward in the literature include raw material characteristics⁴, changes in the crystalline habit of rifampicin^5,6, excipients⁷, manufacturing and/or process variables³, degradation in gastro-intestinal (GI) tract^8,9, inherent variability in absorption¹⁰ and metabolism¹¹, as well as others. As mentioned earlier, there is evidence that particle size, excipients and manufacturing process are causative factors for reduced bioavailability; however, complete information regarding these variables is not reported in the literature⁴. Rifampicin being the only water-insoluble component, formulation and manufacture of rifampicin-containing FDCs with the other highly water-soluble component drugs is the most critical process. Hence, it is necessary to address the issue of variable bioavailability of rifampicin from the perspective of raw material characterization and the manufacturing process. In this regard, further studies are necessary to identify/specify optimum particle size range, physicochemical properties, excipients that may interact with rifampicin and the critical manufacturing variables which have an effect on rifampicin bioavailability. Once these parameters are optimized, good manufacturing practices (GMP) should produce batch-to-batch uniformity and reproducibility to ensure acceptable bioavailability.

It was considered that the variable bioavailability of rifampicin was largely confined to FDC formulations; however, reduced plasma concentrations following administration of rifampicin-only formulations were also reported by Zak and colleagues¹² as early as 1981. In recent years, the problem of bioavailability associated with generic formulations of rifampicin was again highlighted by McIlleron et al. ¹³, who found that two rifampicin capsule formulations showed reduced blood concentrations and were responsible for the failure of TB treatment. In this regard, reduced blood concentrations from the capsule ‘rifampicin-only’ formulations indicate that apart from the manufacturing variables, the raw material also needs to be optimized.

Polymorphism of rifampicin is always regarded as a probable reason for the variable bioavailability of rifampicin from solid oral dosage forms. Based on the first report of rifampicin polymorphism¹⁴, it was assumed that impaired bioavailability may result from changes in the rifampicin crystalline form during the tableting process⁵. The biopharmaceutic and clinical relevance of polymorphism is important only when the solubility of physical forms differs significantly¹⁵. Although in the original report it was stated that the crystalline form of rifampicin is affected by grinding, the effects on solubility were not studied for the different physical forms and requires further investigation.

In a few of the recent reports, it was found that in-vitro degradation of rifampicin is catalyzed by isoniazid in acidic medium and hence this was considered as the reason for poor bioavailability from FDCs^9,16. Although this might explain the reduced bioavailability of rifampicin in the presence of isoniazid when compared to rifampicin alone⁸, this mechanism does not provide justification for reduced, or increased, bioavailability of rifampicin from FDCs when compared to the individual drugs given in combination at the same dose levels. In addition, as evident from Figure A1, similar or increased bioavailability of rifampicin in the presence of isoniazid compared to that of rifampicin alone remains unanswered by this mechanism. Thus, degradation of rifampicin in presence of isoniazid does not explain the anomalous behaviour of rifampicin from solid oral dosage forms.

The other probable reasons such as inherent variation in the absorption of rifampicin and extent of metabolism¹¹, in our opinion may not be the contributory factors for the altered bioavailability of rifampicin when determined by controlled bioequivalence trials. In the randomized, two-way crossover study design, which is adopted for most of the trials listed in Figure A1, every volunteer acts as their own control and hence, gastric emptying time, pH of the stomach, rate of metabolism and other individual variations have only a minor role. However, in order to explain the variations in absorption from the different dosage forms (syrup > capsules > tablet)¹⁷, it is necessary to determine the effect of pH on the solubility and subsequently on absorption of rifampicin from the different segments of the GI tract. In other words, detailed information about the biopharmaceutic properties of rifampicin and in-vitro/in-vivo variables affecting its solubility and permeability is necessary in order to understand the in-vivo behaviour of rifampicin-containing dosage forms.

On the other hand, in-vitro dissolution tests do not guarantee in-vivo bioavailability of rifampicin. It is reported that formulations showing poor dissolution had good bioavailability and vice versa ¹⁸. However, this report does not mention the medium, pH and dissolution conditions used. As rifampicin is a zwitterion, it might be possible that dissolution of rifampicin from either FDCs or separate formulations is a function of pH and hence selection of a discriminatory dissolution medium is important. In addition, with the increased understanding of the complex absorption procedure and the factors affecting it, in present context, it may be possible to develop a dissolution test as a surrogate for costly bioequivalence trials using appropriate dissolution medium, pH and hydrodynamic conditions ¹⁹.

Rationale for biopharmaceutic and pharmacokinetic studies of rifampicin in FDC product development

One of the most significant tools developed to facilitate product development in recent years has been the Biopharmaceutic Classification System (BCS), which is based on the two fundamental tenets of drug absorption, i.e. solubility and permeability²⁰. According to BCS, drug molecules are divided into four categories based on their high or low solubility and permeability. Realization of these important properties has resulted in number of guidelines to reduce the regulatory burden and to hasten the product registration process ²¹. In BCS, solubility limits are set on the basis of the largest dose of the drug soluble in 250 ml of buffer solutions in the pH range of 1-8 and at a temperature of 37^oC. On the other hand, permeability limits are based on the criterion that more or less than 90% of drug is absorbed. Thus, bioavailability of a compound is a function of absorption, dissolution and dose, described by Absorption number (An), Dissolution number (Dn) and Dose number (Do). The anti-TB drugs like isoniazid, pyrazinamide and ethambutol, by virtue of their high solubility and bioavailability, may be considered as BCS class I drugs and hence do not possess any bioavailability problem. However, the BCS class of rifampicin cannot be judged from the literature because of its zwitterionic nature and variable bioavailability. Better appreciation of the biopharmaceutic and pharmacokinetic properties of rifampicin alone and in combination with other anti-TB drugs will help to predict the physicochemical, pharmaceutic, manufacturing and physiologic variables which affect the absorption of rifampicin from various dosage forms²². In addition, by understanding the relationship between the drug’s absorption, solubility and dissolution characteristics, it is possible to define the dissolution conditions to use as a surrogate for in-vivo bioequivalence assessments¹⁹. Rifampicin, a zwitterionic molecule with two pKa values (1.7 and 7.9), shows a highly pH-dependent solubility and lipophilicity profile, especially in the pH range that exists across the GI tract (pH 1.2 to 7.4). Further, rifampicin absorption may be complicated because of its high molecular weight and hydrogen bonding capacities. These fundamental physicochemical properties that determine the intestinal absorption are complex for rifampicin and understanding of these may help in reducing the variability in bioavailability. Recent findings²³, using in-vitro, in-situ and in-vivo absorption models, indicated that permeability of rifampicin is not a rate-limiting step. In these studies, it was clearly demonstrated that the rate and extent of drug release from the formulations is ultimately deciding the overall bioavailability. However, simulation of in-vivo dissolution conditions and their applicability to predict bioavailability is difficult for rifampicin, leading to poor in-vitro/in-vivo correlations. Thus in the process of developing a dissolution test as a surrogate marker for bioavailability of FDCs, we proposed a decision tree that can predict the bioavailability.

To address the issue of minimum registration requirements in terms of sample size and sampling time for bioequivalence estimations of rifampicin-containing products, a thorough understanding of pharmacokinetics of rifampicin is necessary. The number and frequency of samples taken during bioavailability studies is determined by pharmacokinetic parameters such as absorption rate constant (k_a) and elimination rate constant (k_el) which ultimately affect C_max, T_max and blood concentration-time profile. In addition, the minimum sample size to acquire statistically significant results is determined by variability in these pharmacokinetic measures²⁴. Thus, better understanding of the biopharmaceutics and pharmacokinetics of rifampicin will help in elucidating the rifampicin bioavailability problem and will provide the scientific evidence to recommend and implement FDCs in TB programmes.

The 4D approach, BCS and tuberculosis

All through evolutionary history to the present, unraveling the mysteries of the genome on one side and the vastness of the universe on the other, man has been kept engaged in a seemingly never-ending fight by the tiny co-inhabitants of this planet -those causing diseases like tuberculosis, malaria and AIDS. As we change the weapon in the form of more potent and effective drugs, the enemy changes shape, countering with the phenomenon of drug resistance. The successful eradication of smallpox has proved that the correct strategy, based on sound scientific principles, properly implemented, can help us to emerge conquerors in this battle.

The 4D approach²⁵ of disease management proposes such a strategy to envisage a world free from such infections. The first D, denoting disease, may be acute like malaria or chronic like TB and may or may not have a cure, the therapy only serving the purpose of prolonging the life of the patient as in the case of AIDS. First and second line anti-TB drugs constitute the second D, the latter being reserved for cases of resistance and toxicity associated with the former. The delivery system/mode forms the third D and has to ensure that the drugs are available, at their optimally-effective concentrations at the desired site(s) of action or the “destination”, comprising the fourth D.

This link between the drug, delivery and destination is provided by the BCS (Figure A2). Today BCS finds an integral role in every stage of the life cycle of a drug molecule, beginning with judging the drug candidate’s suitability for a purpose, proposing techniques for development and deliverability²⁶ and its ultimate evaluation based on clinical and regulatory standards^27,28. In cases where more than one drug is available for the same indication, it helps to decide upon the drug candidate to take forward to the delivery state and the approach to be employed. For TB, WHO and IUATLD have proposed the use of a cocktail of drugs to be incorporated into FDCs to increase patient compliance and prevent the emergence of resistance. These FDCs, though simple in idea, represent a very effective delivery system in order to overcome the emergence of drug resistant strains, and improve patient compliance. To date, biowaivers are granted only for class I drugs, but extension of this umbrella to include other classes of drugs, especially when they do not deviate greatly from class I inclusion criteria, is a very promising and desirable possibility²⁹. The applicability of BCS to many drugs simultaneously when they are a part of FDCs, and especially when they belong to different classes, is a challenging task. Nevertheless, once accomplished, this approach will have far-reaching consequences on the regulatory front. Regulatory agencies have shown concern about the quality of FDCs, mostly regarding the bioavailability of rifampicin from these formulations. The BCS, provides an opportunity to develop and adopt a surrogate marker for bioavailability assessment of these FDCs, so that with the aid of a resulting biowaiver²¹, the quantum of monetary, human and material effort can be channeled towards the implementation of DOTS and related policies for an assured and speedy eradication of TB from the globe. This amalgamation of BCS with the 4D approach can similarly be applied as a platform for uprooting or controlling other infectious diseases like malaria and AIDS which are being addressed by FDCs.

Figure A2. BCS and 4D approach for disease management

Delivery is dependent on the nature of the drug, the disease and the destination (the route). The drug may be required to be localized at a specific site, or to be delivered into the systemic circulation, as determined by the disease condition. In some cases existing technologies may be readily used for delivering molecules. However, in many cases it is necessary to develop a delivery system in order to meet the destination. The rate and extent to which the drug is absorbed from the drug product and reaches its destination is governed by two tenets of biopharmaceutics. These two properties, solubility and permeability, are of profound importance in drug development and delivery and form the basis for determining bioequivalence of oral immediate-release drug products. Drugs other than class I, when evaluated in vitro may not correlate to the in-vivo performance because of highly complex and multi-faceted cascade phenomena. This may be further complicated when a formulation contains a combination of drugs. The true understanding of solubility, permeability, dissolution and pharmacokinetics of a drug product is needed to define dissolution test specifications that can predict the in-vivo performance. Use of such surrogate dissolution would help in product evaluation for number of drugs, especially drugs for AIDS and cancer, where performing in-vivo biostudies in normal healthy volunteers is not possible.