COMPARATIVE ASSESSMENT OF THE PHYSICOCHEMICAL AND IN-VITRO BIOAVAILABILITY EQUIVALENCE OF CO-TRIMOXAZOLE TABLETS MARKETED IN TIGRAY, ETHIOPIA

COMPARATIVE ASSESSMENT OF THE PHYSICOCHEMICAL AND IN-VITRO BIOAVAILABILITY EQUIVALENCE OF CO-TRIMOXAZOLE TABLETS MARKETED IN TIGRAY, ETHIOPIA

Gebremedhin Solomon Hailu*, Girma Belachew Gutema, Adissu Alemayehu Asefaw, Dagim Ali Hussen and Mussie Ghezu Hadera

Department of Pharmacy, College of Health Sciences, Mekelle University, Mekelle 1871, Ethiopia

ABSTRACT

The availability of numerous brands of co-trimoxazole in our local market today places health professionals in a difficult situation of choice of a suitable brand or the possibility of alternative use. The aim of the study was to predict the physicochemical and in-vitro bioavailability equivalence of six brands of co-trimoxazole tablets marketed in Tigray, Ethiopia. Weight uniformity, friability, hardness, disintegration, assay and dissolution profile were performed using methods described in British and United State Pharmacopoeias and all these tablets passed compendial specifications. Statistical analysis (ANOVA) of thickness, weight and disintegration data showed a significant difference (p<0.05) between generic and innovator brands. Similarly, a significant difference was found between the mean % release of trimethoprim (p<0.0002) and sulfamethoxazole (p<0.0001) of the generic and innovator brands at the pharmacopoeia specified time, 60 min. These statistical results indicated that the brands were not equivalent with respect to their in-vitro release profile. All products showed a trimethoprim % release of greater than 85% within 15 min. Model independent approach of similarity factor (f₂) showed; generic brands C (f₂=33.71), D (f₂=31.85) and E (f₂=36.33) were not bioequivalent with the innovator and so may probably not be used interchangeably. However, generic brands B (f₂=52.02) and F (f₂=77.55) were similar with the innovator and so may probably be used interchangeably. Results have shown that more than cost consideration and company reputation is required for day-to-day rational decision making in drug products sourcing.

Dissolution Profile

INTRODUCTION: Generic drug prescribing and dispensing is frequently promoted as a means of lowering healthcare costs. The issue of bioequivalence of generic products to innovator drugs and the overwhelming proportion of substandard and counterfeit drugs in the market tends to affect this practice. Post-market monitoring, in which data about a product is collected and assessed after it had been granted marketing authorization of approved medicines has been employed to assess the quality, therapeutic effectiveness and safety of medicines. The data and information obtained could be used for product improvement, development of standards and regulations ¹.

The importance of dissolution in quality assurance and regulatory science is well documented ². It therefore becomes apparent that sensitive and reproducible dissolution data derived from physicochemically and hydrodynamically defined conditions are necessary in order to compare various in- vitro dissolution data and be able to use such results as a surrogate for possible in- vivo bioavailability, bioequivalence testing, and in- vitro-in- vivo correlations ³.

Though the problem is worldwide, Southeast Asia and Africa seem to be particularly plagued by counterfeited pharmaceuticals ⁴. For example, a study done in Southeast Asia in 2001 showed that 38% of antimalarials on sale in pharmacies did not have any active ingredients ⁵. In a WHO survey, 20-90% of antimalarial ⁶ and 28% of antibiotic ⁵ drugs failed quality specifications.

Penicillin, tetracycline, co-trimoxazole and chloramphenicol, are among the favored counterfeited antimicrobials ⁴. There is no sufficient data to contemplate the nature and extent of the problem in Ethiopia. But few studies indicate the occurrence of poor quality paracetamol ⁷ and ciprofloxacin ⁸ dosage forms marketed in the capital, Addis Ababa.

Generic products are usually far cheaper than their branded (innovator) versions as generic manufacturers do not have the investment costs for the development of a new drug ⁹. Like many other countries, there is a widely held belief that cost of pharmaceuticals is strongly correlated with their quality in Ethiopia.

However, the validity of this argument needs to be ascertained with data derived from research. In the local market, the price of the cheapest generic is 43 times lower than the most expensive brand of co-trimoxazole tablets. There is a growing concern about this situation. How can a patient know if buying a cheaper brand would be therapeutically effective or not?

The combination of trimethoprim (TMP) and sulfamethoxazole (SMX) in a 1:5 ratio, known as co-trimoxazole; has been used for the treatment of a wide variety of infections due to Gram-positive and Gram-negative organisms. The synergistic bactericidal effect of the combination is due to sequential blockade of bacterial enzyme systems associated with tetrahydrofolate synthesis ¹⁰.

Co-trimoxazole is reported to have a very high number of generics with very wide price margins amongst the antibiotic family of drugs marketed in Ethiopia. For the health care providers to use these brands interchangeably, the bioequivalence of these brands have to be ascertained ¹.

Thus, the present study was carried out to evaluate the physicochemical and bioavailability equivalence of different brands of co-trimoxazole tablets marketed in Tigray, Northern Ethiopia; sourced from different pharmaceutical manufacturers in order to determine the appropriateness of their interchangeability.

MATERIALS AND METHODS:

Materials:

Instruments: A High Performance Liquid Chromatography (Shimadzu, LC/20A Series,  Japan), a Pharma Test dissolution tester (Apparatus 2) (Pharma Test®, PTWS610, Germany), a Pharma Test hardness tester (Pharma Test®, PTB 311, Germany), a Pharma Test friability tester (Pharma Test®, PTFE, Germany), a basket-rack assembly LOGAN DST-3 disintegration tester (LOGAN Instruments Corp.), and an Electronic micro balance (Sartorius, Germany) provided by Addis Pharmaceutical Factory were used for the study.

Chemicals and Reagents: Acetonitrile HPLC grade (Sigma-Aldrich, Germany), Triethylamine HPLC grade (Fluka, Switzerland), Methanol (Sigma-Aldrich, Germany), Chloroform (Sigma-Aldrich, Germany), Isopropyl alcohol (Sigma-Aldrich, Germany), Diethylamine (Sigma-Aldrich, Germany), Glacial acetic acid (Sigma-Aldrich, Germany), Sodium hydroxide (Sigma-Aldrich, Sweden), Hydrochloric acid (Sigma-Aldrich, Germany) and  double distilled and deionized Water were used in the study. TMP working reference standard, Batch number 20/001544 with purity of 99.5% and SMX working reference standard, Batch number 200911/01 with purity of 99.93% obtained from Addis Pharmaceutical Company were also used for the study.

Methods:

Sampling: Six brands of a double strength tablets containing 480 and 960 mg of one of the most widely used antibiotic, co-trimoxazole was evaluated.

The study was designed to collect samples at various levels of drug distribution chains, such as hospital pharmacies, private sector pharmacies and drug stores which are available in Tigray. Detail information on co-trimoxazole tablets included in the study is given in Table 1.

The drug samples were anonymously purchased in their original package as supplied by the manufacturers. Then the collected samples were taken to Addis Pharmaceutical Factory; and the physicochemical parameters of the tablets were analyzed as described in BP and USP ^11-13.

TABLE 1: DETAIL INFORMATION ON CO-TRIMOXAZOLE TABLETS INCLUDED IN THE STUDY

*= innovator product

Uniformity of Weight Test: The test for uniformity of weight for each brand of co-trimoxazole tablets was carried out as described in BP ¹¹. The weights of 20 tablets were determined individually using an electronic micro balance. The average weight for each brand as well as the percentage deviation from the mean value was calculated.

 Hardness and Thickness Testing: Using forceps, four tablets were individually placed between the platens of integrated hardness, thickness and diameter tester. The four tablets were randomly selected from each brand and the resulting visual readings of tablet hardness and thickness were recorded. The thickness of product C (oblong tablet) was measured using Digital Caliper.

 Friability Test: Ten tablets from each brand were dusted and weighed on the analytical balance. The tablets were placed in the drum of the friability tester and rotated at 25 rpm for four minutes (100 times) ¹². Thereafter they were removed; reweighed and the percent friability was determined ¹⁴.

Disintegration Test: Disintegration test was carried out according to specification given in BP ¹¹. A 900 ml beaker with 37 ± 0.5^oC water was prepared. Six tablets were placed into the basket-rack assembly and connected to the disintegration apparatus.

Assay of the Active Ingredients:

Chromatographic System: High performance liquid chromatography method was used for the assay according to USP specification ¹³. A stainless steel column of Nucleosil 100 C18 (25 cm x 4.6 mm) packed with stationary phase of 5 μm particle size and UV detector at 254 nm was employed. A mixture of water, acetonitrile and triethylamine (700:200:1( v/v)) the pH of which was adjusted to 5.74 with  sodium hydroxide and dilute glacial acetic acid was used as a mobile phase with flow  rate of 1.5 ml/min; injection volume of 20 μl; and an ambient oven temperature.

Preparation of Samples: Twenty tablets of the sample was weighed and crushed to powder. A portion of the powder containing 32 mg of TMP and 160 mg of SMX was transferred to a 100 ml volumetric flask and the solution was then diluted with methanol to volume and filtered. Five ml of filtrate was transferred to a 50 ml volumetric flask and diluted with mobile phase to volume to get a working sample solution of 0.032 and 0.16 mg/ml of TMP and SMX, respectively. A similar standard working solution was also prepared. The content of both TMP and SMX in co-trimoxazole tablets were calculated from peak areas of the chromatograms of the test and reference standard solutions.

Dissolution Studies: The dissolution studies of all brands were carried out according to the USP specifications using a type II (paddle), operated at 75 r/pm. Six tablets selected at random from each sample were used simultaneously for the study.  The dissolution medium was 900 ml 0.1N hydrochloric acid maintained at 37±0.5^oC. Five ml volume of leaching fluid was withdrawn at 10, 30, 45 and 60 minutes time intervals by replacing with the same volume of dissolution medium. After filtration and appropriate dilution the concentration of each component was determined by using HPLC in similar fashion to the assay. The percentage dissolved of both TMP and SMX was tested statistically to ascertain differences among brands using one-way analysis of variance (ANOVA, p=0.05) while Dunnett’s test was employed to ascertain where the difference arose with regard to the innovator products ¹.

 Data Analysis: Analytical data obtained from the experiments carried out were analyzed using GrahPad Instat 3 and GraphPad Prism 5 software programs for statistical and graphical comparisons. To compare the dissolution profiles of all generic and innovator brands, a one way ANOVA and model independent approach of similarity factor (f₂) was employed. The similarity factor (f₂) is a logarithmic reciprocal square root transformation of the sum of squared error and is a measurement of the similarity in percent dissolution between the two curves and is calculated as ^10,15:

Where, n is the number of time points, R_t and T_t are the dissolution value of reference and test product at time t.

RESULTS AND DISCUSSION:

Uniformity of Weight Test: The results of weight variation test are depicted in Table 2. According to the specification outlined in BP ¹¹, the test for weight variation where the strength is greater than 250 mg; the tablet passes the test if not more than two of the individual weights deviate from the average weight by more than ±5% and none deviated by ±10%.

All brands of co-trimoxazole tablets showed acceptable uniformity of weight as none had percentage deviation in weight greater than 5% as stipulated by the pharmacopoeia which agrees with similar works conducted in Turkey ¹⁰. Application of ANOVA revealed significant differences among samples mean weights (p<0.0001), with the post hoc identifying for mean tablet weight difference between innovator and generic brands; similar to results reported for nifedipine tablets in Nigeria ¹⁶.

Friability Test: Adequate tablet hardness as well as reasonable friability is required for consumer acceptance ¹⁷. The pharmacopoeia states that the friability value of tablets should be less than 1% ¹³ and as such all the brands of co-trimoxazole passed this friability specification (Table 2). Tablets with higher crushing strength showed low friability value similar to those ciprofloxacin hydrochloride brands conducted elsewhere ¹⁸.

Thickness and Hardness Test: Examination of tablet thickness gives insight as to the tablet tooling used by various manufacturers and results are indicated in Table 2. Similarly, mean hardness values of tablets of co-trimoxazole in this study are shown in Table 2. A force of about 5 kg/cm² is the minimum requirement for a satisfactory hardness of tablets ¹⁹. Generally, all the studied tablets passed hardness and friability test specification of tablet dosage forms.

However, statistical analysis (ANOVA) of thickness (p<0.0001) and hardness (p<0.0091) data showed a significant difference among different brands. These differences in tablet sizes (weight, diameter and thickness) may actually have some negative psychological effects on clinicians and their patients since they could raise doubts on the general equivalence of the brands ¹⁶.

Disintegration Test: Disintegration is a crucial step for immediate release dosage forms because the rate of disintegration affects dissolution and subsequently therapeutic efficacy of medicine ⁹.

The mean disintegration times of different brands of co-trimoxazole tablets included in the study are shown in Table 2. Product B showed the highest disintegration time (6.24 min) which correlates with its low friability and high hardness values.

Differences between the means of disintegration times of generic and innovator brands, were statistically significant (p <0.05) similar to earlier reports on artesunate tablets ¹⁹.

Assay of Active Ingredients: The results for actual content of different brands of co-trimoxazole 480 and 960 mg tablets included in the study are depicted in Table 3. It is described that co-trimoxazole tablets should contain not less than 93.0% and not more than 107.0% of the stated amount ¹³. All brands of co-trimoxazole tablets passed as per the USP specification. The highest and lowest TMP content was obtained 99.93 and 98.59% for product A and C, respectively. Similarly, the highest and lowest SMX content was obtained 101.25 and 100.01% for product D and C, respectively. Statistical comparison for mean difference of drug content indicated that with 95% confidence interval, there was no significant difference in the drug content among the different brands (p >0.05); unlike the works reported on ciprofloxacin ⁸ and; lamivudine and zidovudine combination tablets ¹⁷.

Dissolution Studies: The official Pharmacopoeial specification is that not less than 70% of the stated drug amount should be contained by samples taken at 60 min ^{12, 13}.

From the dissolution test results (Table 3), it can be observed that all samples showed a mean release of more than 70% before 60 min. However, the mean is not the only factor to be considered in the analyses of the results obtained from a dissolution test of solid dosage forms ¹⁷.

TABLE 2: OFFICIAL AND NON-OFFICIAL TABLET PARAMETERS OF DIFFERENT BRANDS OF CO-TRIMOXAZOLE TABLETS

^a (n=4), ^b (n=10), ^c (n=6), ^d (n=20), *=p<0.01, ^#=p<0.05, SD=standard deviation

TABLE 3: ASSAY, % DISSOLVED AND F₂ VALUE AT 60 MIN OF CO-TRIMOXAZOLE TABLETS

RSD = relative standard deviation, TMP = trimethoprim, SMX = sulfamethoxazole

The dissolution profile of each brand of co-trimoxazole is given in Figure 1(A-F). Similarly, the trends of percent release (dissolution profile) of TMP and SMX for the six brands of co-trimoxazole tablets are shown in Figure 2(A-B).

There was an immediate release and super imposable dissolution profile of TMP for all brands compared to SMX release and dissolution profile (Figure 2(A-B)). It was observed that products A and F with the smallest disintegration time 0.57 and 1.04 min, respectively showed fast percentage drug release that agrees with earlier works on ciprofloxacin tablets ⁸.

One-way analysis of variance was undertaken for time points 10 and the pharamacopoeially specified time, 60 minutes. The results of ANOVA as shown in Table 4 and 5 indicated that the percent dissolved was significantly different at the two time points at 0.05 level of confidence for both TMP and SMX. These statistical results indicated that the products of co-trimoxazole tablets were not equivalent with respect to their in vitro release profile; similar observation was made on ciprofloxacin ^{1, 8} and   artesunate ¹⁹ in other works.

TABLE 4: RESULTS OF ANALYSIS OF VARIANCE AT TWO TIME POINTS FOR TRIMETHOPRIM RELEASE

Df = Degree of freedom, Sign. = Significance

TABLE 5: RESULTS OF ANALYSIS OF VARIANCE AT TWO TIME POINTS FOR SULFAMETHOXAZOLE RELEASE

In order to ascertain the source of differences, pair-wise comparisons of the generic brands against innovator product, were performed by multiple comparisons using Dunnett’s test and the outcomes at 0.05 levels are as shown in Table 6. Values above the critical value (2.66) indicate that the mean % dissolved difference was significant while values below the critical value indicate that the differences were not significant. Consequently, it can be inferred that the difference in TMP % dissolved for product B, C, D and E at 10 min were significantly different from the innovator product, while product F was not significantly different from the innovator product. Similarly, products C, E and F were significantly different from the innovator product, while product B and D were not significantly different from the innovator product in TMP % dissolved at 60 min.

TABLE 6: DUNNETT’S TEST FOR TMP AND SMX RELEASE OF THE PRODUCTS AT 0.05 LEVELS WITH CRITICAL VALUE 2.66^a

^a = Critical Value is obtained from a table of Dunnett’s test; ^b = Mean difference is obtained by subtracting mean % dissolved of innovator from mean % dissolved of other brands

In the same way, the differences in SMX % dissolved for products B, C, D and E at 10 min were significantly different from innovator product, while product F was not significantly different from innovator product. At 60 min, only product C showed significant difference in SMX % dissolved from the innovator product. These results showed that product F had the least departure in SMX % dissolved from the innovator product, at the two time points. At the pharamacopoeially specified time, product C showed a significant difference from the innovator product, in both TMP and SMX % dissolved. It is worthy to note that the difference identified by ANOVA and the comparison performed by Dunnett’s tests are statistical and not bioavailability equivalence ¹. To prove the bioequivalence of two or more drug products, a similarity in the rate and extent to which the drug in the dosage form becomes available for absorption need be demonstrated ¹⁹. In this study f₂ derived from dissolution profiles of the six brands of co-trimoxazole were used as estimators for bioavailability of TMP and SMX, hence their bioequivalence. Similarity factor f₂ is included as guidance on dissolution testing of immediate release solid oral dosage forms ¹⁵.

The high rate of dissolution of TMP (Figure 2A) precludes any possibility of bioavailability problem resulting from drug dissolution and hence justifies interchangeability among the six brands with regard to TMP. In cases where greater than 85% of drug is dissolved within 15 min, dissolution profiles are usually accepted as similar without further mathematical evaluation ¹⁹.

That means, all the six brands have the same bioavailability with regard to TMP. The rate of dissolution of SMX, however, did not meet the criterion of 85% dissolution and as such were subjected to further mathematical (f₂) evaluation to demonstrate bioequivalence (Table 3). On the contrary, another study ¹⁰ showed that more than 85% of the labeled amounts of both TMP and SMX in test and reference products were dissolved in 15 min and hence profile comparison with f₂ was unnecessary.

For two dissolution profiles to be considered similar and bioequivalent, f₂ should be between 50 and 100 ^{9, 15}. Therefore, the SMX dissolution profiles of products C (f₂ =33.71), D (f₂=31.85) and E (f₂=36.33) (as shown in Table 3) using the model independent approach (f₂) were not similar with innovator product and so may probably not be used interchangeably; similar results were reported on one brand of artesunate (f₂ = 23.8) ¹⁹, three brands of ciprofloxacin (f₂ = 17.6-25.8) ¹ and three brands of methformin hydrochloride (f₂ = 24.5-39.4) ²⁰.

On the other hand, the dissolution profiles of products B (f₂=52.02) and F (f₂=77.55) were similar and most probably bioequivalent with innovator product, and so may be used interchangeably as indicated by other works on two brands of ciprofloxacin in Nigeria (f₂ = 53.9 and 53.3) ¹, six brands of ciprofloxacin in India (f₂ = 65.35 -77.33) ⁹.

Interestingly from this study, it was understood that price may not necessarily indicate the quality and effectiveness of a drug product as earlier pointed out in other studies ¹. Products B and F were sold at the price of 1.50 and 12.00 Birr per 10 tablets, respectively; but both are bioequivalent to the innovator product which was sold at about 65.00 Birr per 10 tablets at the local market.

On the other hand, products C, D and E, which were sold at a price of 2.50, 2.00 and 3.00 Birr per 10 tablets, respectively, were not bioequivalent with the innovator brand.

Therefore, our observation here was that the cheaper brands may or may not be bioequivalent with the innovator brand which further highlights the need to really characterize multisource medicines.

Results have shown that more than cost consideration and company reputation is required for day-to-day rational decision making in drug products sourcing.

CONCLUSION: Six brands of co-trimoxazole tablets have been subjected to analysis according to the monograph of BP and USP. Our results indicated that all brands of co-trimoxazole tablets included in this study were chemically equivalent. Statistical analysis (ANOVA) showed that there was significant difference in the mean drug release between the generic and innovator products; and the similarity factor (f₂) also showed that three of the generic products were not bioequivalent with the innovator brand.

This is significant in therapy where drugs are expected to not only conform to their label claims but also have satisfactory bioavailability. These findings support the need for activation of the regulatory rules with emphasis on post marketing evaluation of pharmaceutical products in order to monitor the safety, quality and efficacy of essential drugs in the country.

ACKNOWLEDGEMENTS: The authors acknowledge Mekelle University for financial support. Similar thanks also go to Addis Pharmaceutical Factory SC for allowing accesses to their laboratory facilities during this study.

REFERENCES:

1. Ngwuluka NC, Lawal K, Olorunfemi PO and Ochekpe NA: Post-market in vitro bioequivalence study of six brands of ciprofloxacin tablets/caplets in Jos, Nigeria. Sci Res Essays 2009; 4: 298-305.
2. Nithyanandan P, Hauck WW, Munoz J, Deng G, Brown W,  Manning RG and Wahab S: Dissolution variability: comparison of commercial dosage forms with US Pharmacopeia Lot P Prednisone Reference Standard Tablets—A Technical Note. AAPS PharmSciTech 2008; 9: 238-242.
3. Pillay V and Fassih R: Unconventional dissolution methodologies. J Pharm Sci 1999; 88: 843-851.
4. Kelesidis T, Kelesidis I, Rafailidis PI and Falagas ME: Counterfeit or substandard antimicrobial drugs: a review of the scientific evidence. J Antimicrob Chemoth 2007; 60: 214–236.
5. Ahmad K. WHO fights fake pharmaceuticals. Lancet Infect Dis 2006; 6: 195.
6. World Health Organization (2007): Survey of the quality of antiretroviral medicines circulating in selected African countries. http://apps.who.int/prequal/info_general/ documents/ARV_survey.pdf
7. Hundie K, Debella A and Hymete A: Comparative quality evaluation of some brands of paracetamol tablets, suppositories and syrups. Eth Pharm J 2008; 26: 59-64.
8. Kahsay G, Debella A and Asres K: Comparative in vitro quality evaluation of ciprofloxacin tablets from drug retail outlets in Addis Ababa, Ethiopia. Eth Pharm J 2007; 25: 1-8.
9. Nayak K and Pal D: Comparative in vitro bioequivalence analysis of some ciprofloxacin HCl generic tablets. Int J Pharm Sc Res 2010; 1: 1-7.
10. Sayar E, Şahin S, Cevheroğlu S and Hincal AA: Comparison of dissolution profiles of two commercially available co-trimoxazole tablets. FABAD J Pharm Sci 2008; 33: 87–94.
11. British Pharmacopoeia. The Her Majesty’s Stationery Office: London; 2001; Vol. I.
12. British Pharmacopoeia. The Her Majesty’s Stationery Office: London; 2009; Vol I-III.
13. US Pharmacopoeia National Formulary, USP 30/NF 25. United States Pharmacopoeia Convention Inc., Rockville, MD, USA; 2007.
14. Ibezim EC, Attama AA, Obitte NC, Onyishi VI and Brown SA: In vitro prediction of in vivo bioavailability and bioequivalence of brands of metronidazole tablets in Eastern Nigerian drug market. Sci Res Essays 2008; 3: 552-558.
15. Food and Drug Administration, Center for Drug Evaluation and Research, 1997: Guidance for industry: Dissolution testing of immediate release solid oral dosage forms (http://www.fda.Gov/cder/guidance/1713bp1. pdf).
16. Okoye EI and Iwuagwu MA: Physicochemical equivalence of some brands of Nifedipine retard tablets available in Nigeria. Afr J Biotechnol 2010; 9: 1274-1279.
17. Beck RC, Athayde ML and Cardoso SG: HIV/AIDS treatment and physicochemical quality control of medicines: evaluation of non-generic lamivudine and zidovudine tablets manufactured in Brazil. Braz J Infect Dis 2007; 11:540-543.
18. Mu’az J, Gazali LK, Sadiq GU and Tom GM: Comparative in vitro evaluation of the pharmaceutical and chemical equivalence of multi-source generic ciprofloxacin hydrochloride tablets around Maiduguri metropolitan area. Nig Journ Phar Sci 2009; 8: 102-106.
19. Esimone CO, Okoye FB, Onah BU, Nworu CS and Omeje EO: In vitro bioequivalence study of nine brands of artesunate tablets marketed in Nigeria. J Vector Borne Dis 2008; 45: 60–65.

Hamdan II and Jaber AK: Pharmaceutical evaluation of metformin HCl products available in the Jordanian market. Jordan J Pharm Sci 2010; 3: 1-7