the microgram range. In addition, new ranges of therapeutic agent are being developed based on a rapidlyincreasing understanding ofthe functioning ofthe humancell and the relationship between the activity of genesand the functions of the proteins they manufacture(genomics and proteomics). Using these techniques,biological compounds such as antigens can be producedwhich modify cell function in a fundamentally differentway to traditional chemical therapeutic agents. Ourknowledgeofthehazardslinkedtothesetechniquesisstillin its infancy and understanding ofthese hazards is one of the major challenges for the future.As the drug development cycle progresses, pharma-ceutical agents are subject to comprehensive regulatorytesting, both for efficacy and safety, and detailed pre-clinical and clinical toxicology information is generated,upon which worker occupational exposure limits can bebased.However,despite comprehensive risk managementsystems, studies have shown that worker health effectshave been experienced in the industry, particularly inpersonnel working with potent compounds such assteroids, or compounds with the capacity to causecumulative damage, such as cytotoxic anti-cancer drugs.
To carry out this review, a systematic search of theMEDLINE database was undertaken (1966–November2002) using subject heading and key word searchesfor ‘pharmaceutical industry’ and ‘pharmaceuticalmanufacture’, together with terms for ‘occupationalexposure’,‘worker health’,‘overexposure’and ‘hazardoussubstance’. All of the abstracts were reviewed and majorarticles retrieved and examined. In addition, bibliog-raphies of existing reviews were studied and furtherreferences followed up. In particular, the reviews byTeichmann
.  and Huyart  were of value in thisrespect.
Acute pharmacological effects
Most harmful effects resulting from exposure to pharma-ceutical agents are the result of acute pharmacologicaleffects, although reports in the literature are relativelyrare. It is possible that this reflects conservativeoccupational exposure limit setting practices, togetherwith short-term exposure patterns and effective exposurecontrols. It is also possible that many mild cases are onlyreported internally and are not published. Among thosewhich have been published is the case presented byAlbert
. , in which an operator working on themanufacture of glibenclamide was admitted to hospital inhypoglycaemic coma and the report of a study by Baxter
.,in which operators were found to have absorbedsignificant levels of barbiturates.
Chronic effects from potent compounds
The problems of highly potent compounds arehighlighted by the case presented by Jibani and Hodges, in which an operator suffered severe effects from theexcessive absorption of vitamin D3. This case is unusual,however, as the most common problems reported in theliterature have been linked to exposure to two specifictypes of potent compound: steroid hormones and cyto-toxic anti-cancer drugs.
Oestrogens and progestagens
The first report of adverse effects in workers related tooestrogens dates back to 1942,when Scarff and Smith noted the development of gynaecomastia and loss of libido in men working with diethylstilboestrol.A series of reports throughout the 1970s and 1980shighlighted the problems of steroids, including those of Suciu
, showing an increase in psychologicaleffects, testicular discomfort and loss of libido in workersmanufacturing acetoprogesterone.An important studywasconductedbyHarrington
 looking at workers exposed to synthetic oestrogens.Of the 25 men studied, five were found to have effectslinked to steroid exposure, including gynaecomastia, lossof libido and galactorrhoea. Of the 30 women, 12 werefound to have menstrual breakthrough bleeding, a ratefour times more frequent than that in the controlpopulation. This study is interesting, as an attemptwas made to link clinical effects to exposure levels andbiological markers such as the plasma drug levels. Theinterpretation of the plasma drug levels proved difficult,as the relationship to exposure was not clear. To under-stand the true level of exposure, a series of samples isrequired over a defined period of time to determine thearea under the exposure versus plasma concentrationcurve.Itwas,however,notedthatthelevelswerehigherinthe workers with the higher exposure.Further studies by Shmunes and David investigating workers exposed to diethylstilboestrol andMills
 confirmed the prevalence of problems atvery low levels of exposure.Taskinen
 conducted a register-based,case-control study on the pregnancy outcome of femaleworkers in eight Finnish pharmaceutical factories todetermine whether they had a higher risk of spontaneousabortion than the general population or matchedcontrols. In a logistic regression model (which includedoestrogen exposure, solvent exposure frequency of usageand heavy lifting), the odds ratio (OR) was increased foroestrogens (OR= 4.2,
= 0.05),aswellasfor continuousheavy lifting (OR = 5.7,
= 0.02).However,it is difficult