Update on DrugDrug Interactios A Scientific Perspective药物间相互作用的科学的角度更新

1、1update on drug-drug interactionsa scientific perspectivekeith gottesdienermerck research laboratoriesjohn wagnerbarry gertztom bailliepeter honig juinn linnovember 3, 20042outline of this presentation approach to assessing drug interactions? areas of agreement with the “concept paper” areas for fur

2、ther discussion:- cyp induction- transporters- multiple inhibitors/multiple impaired (qtc) not covered today:- specific comments on the concept paper4study designs4tables- specific comments on the questions to the committee3approach to assessing drug interactions what is the primary question?- shoul

3、d (and how should) the dose of the substrate drug be adjusted in the presence of the interacting drug? which ddi to study and how to answer?- we are moving from the “past” where choice was largely determined by the likelihood of co-administration/clinical consequences of an interaction - goal: scien

4、ce-driven approach (where feasible)4preclinical in vitro studies to determine in vivo studies to conduct4in vivo studies using probe substrates4robust study designs are critical- but:4there are areas where the science is evolving4necessary tools (e.g. probes) are still lacking4approach to assessing

5、drug interactions (2) how do we use the data to answer the primary question?- pre-specified criteria to compare the pk (or pd) measures for the substrate drug in the presence of the interacting drug4based on the safety and efficacy profile of the substrate drug (e.g., therapeutic index), clinical co

6、ntext of the use of the drug, and concentration-response data for the substrate4often not clearly “positive” or “negative”4for the nce, often data is limited4concentration response info often primarily for efficacy4for probe substrates, still not much consensus (though much progress for cyp3a4 inhib

7、ition)4induction, in particular, has been problematic to interpret5areas of agreement integrated and scientific approach clarity on cyp interactions very useful- especially, in vitro/in vivo correlations- in vitro/in vivo clarity on substrates, inhibitors, inducers use of pk in poor metabolizers (wh

8、ere appropriate) robust study designs useful and consistent labeling language- still some discussion needed on how to label “moderate inhibitors” and how to define “sensitive” cyp substrates- the “devil is in the details”.6science-driven approach: where we are today with in vitro predictions of in v

9、ivo ddis? ready for “prime-time”:- cyp inhibition (1a2, 2c9, 2c19, 2d6, 3a4) almost ready for “prime time”:- p-glycoprotein (p-gp)?- ugt1a1 (and other ugts)- cyp2b6 and cyp2c8- cyp induction not ready for “prime time”:- most transporters- multiple paths of inhibition (interacting drugs/ genetics/ im

10、paired clinical conditions)7induction three major concerns for induction: - reduction in therapeutic efficacy (most common)- autoinduction (time-dependent pk)- imbalance between toxification/detoxification dose- and time- dependent dependent also on clearance and route of administration a concern bo

11、th with initiation and discontinuation of interacting drug8how to assess for the potential for cyp induction? animal models- previously, short-term, high dose studies in rodents- poor predictor? species differences in induction? in vitro models- hpxr assay- primary culture of human hepatocytes4enzym

12、e activity4mrna clearly very helpful in selection of drug candidates but is quantitative in vitro/in vivo prediction possible?9factors that complicate the in vitro/in vivo extrapolation interindividual variability- in vitro variability4rifampin fold-increases on cyp3a4 protein: 3-24- in vivo variabi

13、lity4rifampin fold-decreases on auc of midazolam in humans: 11-55 effect of plasma protein binding on cyp induction involvement of multiple mechanisms in cyp induction10some hypothetical examplesdrug5-day mousehpxr (ec50, um)mrna(% of rif1)enzyme act. (% of rif1)cmax conc.2a10 um30-50%30-50%15 umb5

14、um50-60%70-80%2 umc2.6 um30-50%40%200 nmd1 um60-70%0%(also a cyp inhibitor)400 nmen/an/a30%400 nm1 percent of 10 um rifampin control with drug tested at 1 and 10 um; 2 cmax in humans at clinical dose 11induction (cont) so which need an in vivo study? conclusion: - at present, we can only predict a l

15、ikelihood of cyp induction, e.g. “highly possible” or “less likely”- additional clinical data sometimes help:4evidence of autoinduction, or evidence of less accumulation than expected how do we interpret an in vivo study?- less consensus on use of probe substrates and their “clinical interpretation”

16、- e.g., how much decrease in midazolam is clinically relevant?- how do we extrapolate to other substrate drugs?12% baseline exposure020406080rifampinanticonvulsantssjwglucocorticoidsinduction of cyp3a4: effect on oral midazolam auc(percent of baseline exposure)13% baseline exposure020406080rifampina

17、nticonvulsantssjwglucocorticoidsinduction of cyp3a4: effect on oral midazolam auc14% baseline exposure020406080rifampinanticonvulsantssjwglucocorticoidsinduction of cyp3a4: effect on oral midazolam auc15transporters more recently, role of transporters has been recognized to a greater extent, and som

18、e clear examples of transporter-mediated drug interactions exist particularly, the understanding of p-glycoprotein has advanced greatly but, the in vitro methods are not standardized and/or are not readily available a quantitative in vitro prediction of in vivo relevance may not be possible at this

19、point- complicated by the fact that transporters also affect absorption, tissue distribution, and excretion16p-glycoprotein broad tissue distribution: gi, liver, renal, brain, etc. potential roles in drug absorption, distribution, and excretion some important in vitro methodologies exist, but- trans

20、genic mdr knockout mice are a powerful tool, but many caveats (e.g., human/rodent differences)- in vitro tools are becoming more sophisticated, but the interactions between p-gp substrates does not always follow simple kinetics- overlapping substrates between p-gp and cyp3a4, and many inhibitors aff

21、ect both- tools are most useful to identify p-gp substrates- in vitro/in vivo correlations are still limited17p-glycoprotein (cont) what might a clinical pharmacologist receive as an evaluation of the potential for p-gp mediated drug interactions?- “bidirectional transport was evaluated across monol

22、ayers of llc-pk1 cells overexpressing human (mdr1) and mouse (mdr1a) p-gp. drug f, at was not a substrate for human mdr1 (b/a ratio = 1.7), but was a moderately good substrate for mouse mdr1a (b/a ratio = 10.1).” - “brain penetration was evaluated in vivo in p-gp deficient and p-gp competent mice. c

23、omparison of brain concentrations indicates that p-gp activity resulted in a 6-fold reduction in auc in the brain; the auc ratios between brain and plasma were 1.0 and 0.17 in p-gp deficient and competent mice, respectively.” what should the approach be? if an in vivo study is indicated:- for assess

24、ing p-gp inhibitors, digoxin is a suitable probe4other probe p-gp substrates are less than ideal 18p-gp transporter based interaction if a nme is a substrate for p-gp and cyp3a, a clinical study with a dual inhibitor (e.g., ritonavir) may be appropriateritonavirketoconazole 200 mgerythromycinindinav

25、irvardenafil auc(co-administration)courtesy of shiew-mei huang, 200419p-gp transporter based interactionif a nme is a substrate for p-gp and not cyp3a4, a clinical study with a p-gp inhibitor (e.g., cyclosporine, verapamil) may be appropriatequinidineverapamilgrapefruit juicerifampinst johns wortapr

26、epitantdigoxin plasma auc or css(co-administration)ritonavircourtesy of shiew-mei huang, 200420p-glycoprotein (cont) if an in vivo study is indicated:- for assessing p-gp inhibitors, digoxin is a suitable probe4other probe p-gp substrates are less than ideal- no good p-gp inhibitor available to test

27、 for potential for interactions for a drug that is a p-gp substrate4ritonavir is a cyp3a4 and p-gp inhibitor, but also other effects (e.g., 2c9 inhibition, induction of ugts)4verapamil is similar on p-gp and cyp3a4, but also causes ecg pr-interval prolongation in healthy subjects4cyclosporine has my

28、riad effects, so interpretation is difficult4also carries significant risk to volunteers4mrl has not been willing or able to dose more than single doses of cyclosporin to volunteers limits usefulness21other transporters in vitro tools are far less standardized or available many cell-based systems co

29、ntain multiple transporters few well defined substrates and inhibitors in vitro/in vivo correlations are difficult:- active transport may not be relevant for compounds with high intrinsic permeability- transporters may be saturated at high intestinal concentrations even many of the well characterize

30、d clinical interactions cannot be linked to a single transporter mechanism and if an in vivo study were desirable, only a specific interaction could be studied; its less clear how to generalize to broader clinical practice overall: science doesnt support in vitro/in vivo correlation- need: more basi

31、c research on transporters- need: availability of better probes and inhibitors22multiple inhibitors/multiple impaired understanding of agencys desire for high exposures when evaluating qtc issues, but:- the new “hurdle” is very stringent: ci for any difference in qtc prolongation 8 msec- margins are

32、 critical, but for most drugs, are “extraordinary efforts” justified?23multiple inhibitors/multiple impaired (2) why might such requests be considered “extraordinary efforts”?- qtc effects of many inhibitors not well characterized- definitive qtc study results needed by end of phase iib to help guid

33、e study design for phase iii4in most cases, clinical adme study needed to define in vivo metabolic pathways4need clinical data on each inhibitor separately to assess usefulness to increase pk exposures4in most cases, need data on effect of concomitant administration of inhibitors before qtc study (t

34、olerability?)4special population pk may be needed in some cases (e.g., renal insufficiency, such as for telithromycin)- recruiting patients with genetic polymorphisms (pms)/impaired patients is time-consuming and difficult24multiple inhibitors/multiple impaired (3) what might a clinical pharmacologist receive as an evaluation for metabolic pathways: - “in vitro data indicate that cyp3a4 plays a major role (60%) in the drugs metabolism, and that cyp2d6 (10%), cyp2c19 (10%), cyp1a2 (10%) and cyp2c9 (10%) also contribute” - which inhibitors should we use? and, are we as “smart”