Level A
- Level A correlation is one in which there is a predictive mathematical model for the relationship between the entire in vitro dissolution / release-time course and the entire in vivo response-time course (e.g., concentration-time profile or amount of drug absorbed-time profile)
- AXIS: % dissolved and % absorbed
- Strongest level
Level B
- Level B correlation is one in which there is a predictive mathematical model for the relationship between summary measures that characterize the in vitro and in vivo time courses (e.g., mean dissolution times and mean absorption times)
- AXIS: MDT or MRT or MAT
- Medium level
Level C
- Level C provides a mathematical model of the relationship between amount dissolved in vitro at a particular time and a measure of bioavailability
- AXIS: % dissolved @ 15 min and another single point (Cmax, Tmax)
- Weakest level
Noyes-Whitney Eq.
= (D*A)/h * (Cs-C)
- D = diffusion rate constant
- A = surface area
- Cs = conc. of drug in stagnant layer (solubility)
- h = thickness of stagnant layer
- SINK conditions: C remains below Cs/3
Agitation, Temp., Solubility, Salt form on RoD
- Agitation and Temp. both directly proportional to RoD (AFFECTS H)
- More soluble, quicker RoD (vice versa)
- Salt = ions, increase ability to go into solution, so increase RoD (AFFECTS Cs)
Zero order vs First order
- ZERO order if straight line on regular graph paper (linear scale)
- FIRST order if straight line on semilog graph paper (most drugs follow first order dissolution profiles)
IVIVC
- For rate and extent of abs. in vivo
- Can be surrogate to bioequivalence studies
- SUPAC (scale up post approval): minor post approval changes like site of manufacture, form change, new strength
IVIVC IR formulations
- Not successful for obtaining good correlations, generally, bc in lab it can be a few minutes for dissolution but in body it can take hours due to permeability, food, gastric emptying
- A difference in sensitivity: dissolution test in vitro is more sensitive than in vivo profile
IVIVC ER formulations
- Great success with IVIVC
- Release is prolonged so higher probability of RoD in vitro being correlated to in vivo
- Permeability, GE, motility contribute less to overall input
Three KEY factors of ER products not behaving in vivo as predicted in vitro
- Transit times
- Permeability
- Drug degradation
Transit Times and ER
- Variable between stomach/food/colon
- Colon has less SA/transporters/mucus
- Worry about safety/efficacy bc 1 patient can be getting a too low amount for it to work while another can be getting too much leading to toxicity
- Need a wide therapeutic window
Permeability and ER
- Has to be permeable and abs. in colon for ER
Drug Degradation and ER
- Colon is a rich region for microbiome, has bacteria that can break down drug
- F will decrease
MDT Calc
- ZERO order: time at which 50% dissolved, or triangle area (B*H/2) / 100%
- FIRST order: MDT = 1/Kdiss (use half life off graph @ 50%, to find k)
Ideal behavior characteristics in vivo
- Be readily absorbed in large intestine (non polar/unionized)
- Not be degraded by microflora in large intestine
- Remain in GI tract > 24 hr
- Not have physiological effect on large intestine
- Show linear elimination in vivo
Briefly comment on gastrointestinal transit times and the use of 24-hour release products
The transit through the GI tract (large bowel/stomach) is highly variable. In many conditions, the mean transit time through the alimentary canal is less than 24 hours. Even under “normal” circumstances a shortened transit can occur. For these reasons, the ideal of 24-hour release products is not always realized in vivo.
Discuss the consequence of a patient chewing your product, instead of swallowing it whole. The drug is then released quickly. What dose would the patient receive?
The product contains 4 times as much drug (100mg vs 25mg) as the conventional product. Rapid release of the total content could give rise to increased incidence of adverse events. This phenomenon is called “dose dumping”.
Why do statisticians state that more subjects are needed in a bioequivalence trial when the 90% confidence interval is 0.78-1.29 in a trial with 24 subjects?
A larger number of subjects reduces the 90% confidence interval. Both the lower (0.78) and the upper (1.28) limits are outside the criteria used for bioequivalence (0.8/1.25). The ratio of the means of the measure may be one. To have statistical assurance this is so a larger number of subjects is needed.
When is it likely that testing the same product (lot, etc.) against itself will not give a bioequivalent outcome? Do you think that you could conduct such a trial and that the outcome could show the products to be not equivalent?
When the intrasubject variability is high, declaration of bioequivalence may not occur unless a large number of subjects is used. It is extremely unlikely
that a test result would show the product to be not equivalent to itself in a well-controlled study.
