The Department of Dermatology at Nippon Medical School reported many adverse drug reactions (ADRs) to the Japanese Ministry of Health and Welfare (MHW)'s “Adverse Drug Reaction Monitoring” system. We have assessed causalities of these ADRs (1, 167 cases of 1, 531 ADRs over 10 years), and the suspect drugs. Using theseanalyses, we investigated whether the items on the MHW's report form are sufficient to assess causalities.
We investigated 4 aspects of causalities: 1) whether ADRs reappeared after drug reintroduction, 2) whether ADRs are well known, 3) whether ADRs occurred after drug administration and 4) whether ADRs abated after curtailing drug use.
As for the first aspect, in 45 cases tested by reintroduction of a drug (po 43, iv 2), the results were positive. In cases where skin tests such as patch tests, scratch-patch tests and other tests were done instead of reintroduction tests, 313, 324 and 22 tests were positive, respectively. We assessed the total 704 ADRs (46%) as “definite”. In most of the patients with rash (74%), contact dermatitis (84%), shock (73%) and photosensitivity reaction (88%), skin tests etc verified the causality.
In the second aspect in cases where ADRs are well known, skin tests were not done, because ADRs such as rosacea-like dermatitis and steroid acne occur only after long term use of steroids and because candidiasis, mycosis and futoraful dermatitis were frequent. We assessed these 581 ADRs (38%) as “definite”.
Among the rest of the 246 ADRs, 183 occurred after drug administration (aspect 3) in which dechallenge was positive in 128 (8%). We assessed causality of these ADRs as “probable”. Dechallenge was not positive in 55 (4%, including 7 recoveries in which it is unknown whether they came before or after drug curtailment). We assessed causalities of these ADRs as “possible”. In 63 ADRs (4%), it was unclear whether ADR occurred before or after drug administration. We assessed these causalities as “doubtful”.
Analysis of MHW's ADR report form revealed cases where the date of onset of administration or the date of ADR occurrence was unknown, and thus whether or not “ADR occurred after drug administration” could not be assessed. The same situation was found in cases in which ADRs abated but the dates of recovery or dates of drug curtailment were unknown, and consequently whether or not “event abated after curtail ment of drug” could not be assessed. The USA's FDA MEDWatch form has a check box for event abated after curtailment of use. MHW's report form has a check box for event recurrence after reintroduction but does not have a check box for whether skin tests were positive or drug concentrations were monitored. These analyses suggest that the MHW's report form needs to be improved.
Checking a box indicating that the event occurred after drug administration ranks the causality assessment from “doubtful” up to “possible” or “probable”. Checking a box for event abated after curtailment of drug use changes the causality assessment from “ossible” to “probable”. To improve the assessment of causality, we propose that the MHW form include 4 check boxes for probability of causality: whether the event recurred after reintroduction or not, whether skin tests were positive or negative, whether the event occurred after drug administration or not, and whether the event abated after curtailment of use or not.

View full abstract

A new policy on the utilization of foreign clinical data was adopted by the Ministry of Health and Welfare in the summer of 1998. It is based on the ICH Guideline, an international agreement, that was achieved through a lengthy process. Now many pharmaceutical companies will be forced to drastically reexamine their new drug development strategies.
The guideline emphasizes that the influence of intrinsic and extrinsic ethnic factors on the drug's efficacy and safety should be evaluated through the development process. And in most cases, sponsors need to conduct a “bridging study” to determine whether or not the foreign clinical data can be extrapolated to the domestic population. “Bridging studies” can be classified roughly into the following three categories, 1) no need for bridging study, 2) controlled pharmacodynamic study, and 3) controlled clinical study. Among these categories a dose-response study using a clinical endpoint could be recommended as a bridging study for the time being.
In order to render the guideline meaningful, and the new drug development more efficient, further discussion is needed on the appropriate method or way of thinking of (1) assessing the pharmacokinetic profile, (2) deciding the sample size of the bridging study and evaluating its similarity to the foreign data, and (3) conducting post marketing surveillance for drugs approved mainly based on foreign clinical data.

View full abstract