conclusion of the Ashby Working Party that genetic manipulation using 
restriction enzymes offers opportunities of great potential value and we are 
anxious to provide guidance on the appropriate conditions for such work as 
quickly as possible. It should be recognised moreover that the techniques we 
have been asked to consider are not the only ones that may have associated 
hazards. However, a variety of other genetic techniques leading to the intro- 
duction of new characteristics into cells have been employed for many years 
and such epidemiological surveys as have been conducted show no evidence 
of hazard. 
1.4 The underlying reason for the present review, and of course for the 
establishment of the Ashby Working Party, was the concern of the scientific 
community that some experiments involv ing the techniques of genetic manipula- 
tion might lead, perhaps inadvertently or in an unpredictable manner, to the 
release of harmful products into man, animals or plants. It should be stressed 
at the outset that most of the hazards that may be involved are conjectural. 
At present there is no experimental evidence that some of the most serious 
hazards that can be envisaged — for example those due to the incorporation of 
deoxyribonucleic acid (DNA) from oncogenic viruses, or from bacteria carrying 
mammalian DNA, into the human body — are real; but equally there is no 
proof that they are not. It is clear however that the introduction of drug 
resistance into a drug-susceptible pathogenic bacterium could have serious 
consequences if the organism should escape. Until further knowledge is 
gained of the use of the novel genetic techniques under discussion, it seems to 
us essential that rigorous precautions, based on the best estimate of possible 
hazard, should be observed by all laboratory workers using these techniques. 
We therefore consider that work in this field should be done only under appro- 
priate containment conditions and to this end we have devised a set of guide- 
lines for categorising experiments (Section 2 below) that take account both of 
the conjectured hazards and of methods available for biological containment; 
and w? have drawn up a code of practice (Section 3 below and Appendix II) 
for the operating procedures and physical containment levels appropriate to 
each category of experiment. 
1.5 In the present state of knowledge of the field, containment measures 
should allow a suitable margin of safety until any areas of doubt can be clarified 
by further experimental evidence. We see a need for a flexible approach and 
consider this will be better met by requiring those who plan to work in the field 
of genetic manipulation to submit their experimental protocols to a central 
advisory group — referred to below as the Genetic Manipulation Advisory 
Group (GMAG) — for advice on the appropriate safety precautions, rather 
than by imposing rigid guidelines. We define certain cases in which work 
should not proceed until the GMAG has given approval and we can envisage 
circumstances in which the group could advise against the conduct of a particular 
experiment because of the possible hazard, or could propose modifications of 
the experimental procedures or safety precautions. As work proceeds, the 
experience gained should quite quickly build up into a body of “‘case law” on 
which future experimental protocols could draw, and the GMAG should gain 
valuable expertise not only in the design of safe experiments but also in related 
matters such as training, health monitoring and epidemiological studies and 
the role of safety committees and safety officers. 
1.6 Conscious of the restraint already shown by the scientific community, and 
of the importance of work in this field, we consider that certain types of 
experiment (ie those falling into categories I and II defined below), which would 
IV- 6 
[ 446 ] 
