15 
ingly small. Although the probability of survival is greater at lower 
levels of biological containment, the experiments permitted at these 
levels generally involve organisms known to exchange genetic infor- 
mation by natural processes. 
On the other hand, neither the containment measures nor the assess- 
ments of their effectiveness are infallible. In the opinion of Jonathan 
King, even a small number of accidents might do irreparable harm if a 
deleterious combination of genes should establish itself in the environ- 
ment. He believes that monitoring of laboratory workers’ health and 
maintenance of medical records have been inadequate. In the absence 
of base-line health data, it is not possible to judge the effectiveness of 
physical containment methods which are, in any event, entirely de- 
pendent upon the training and behavior of laboratory personnel. 
Others have pointed out that the conduct of research with known 
pathogens may be quite different from the conduct of experiments for 
which no hazards have been demonstrated. Arthur Schwartz, a Uni- 
versity of Michigan mathematician, questioned the calculations of 
potential risks in recombinant DXA research. He stated in his testi- 
mony to the subcommittee that in his estimation the methods used by 
Curtiss and others to estimate probabilities are flawed because they 
do not take account of multiple factors and successive events. 
The majority of scientists, however, considers the present rules too 
stringent, particularly with regard to the use of K-12 bacterial hosts 
and plasmid vectors. Curtiss and others urged approval of the revi- 
sions of the guidelines recommended in September 1977 by the Recom- 
binant Advisory Committee ; these changes would generally reduce the 
physical and biological containment requirements for E. coli experi- 
ments. Nevertheless, Paul Berg observed that most scientists accept the 
present guidelines as "an interim solution to the anxieties that remain.” 
Xo testimony at the subcommittee hearings suggested that the rules 
per se have greatly impeded progress in recombinant DXA research, 
although they have added to its cost and inconvenience and un- 
doubtedly entailed some delays. Referring to the more than 300 XIH- 
fundecl research projects and the recent accomplishments in cloning 
the rat insulin gene and producing somatostatin, Marshall Shapo of 
the University of Virginia concluded that ‘‘research has indeed 
flowered. * * * That seems to me * * * to indicate that regulation and 
talk of regulation has not been either chilling or productive of ineffi- 
ciency, given the uncertainty that originally existed.” 
That most researchers have accommodated themselves to regulation 
and that regulation has permitted dramatic advances in research do 
not justify the continuation of unnecessary restrictions. The subcom- 
mittee inquired whether the current work in risk assessment is likely 
to remove all reasonable doubt about the safety of recombinant DXA 
experiments and thus the need for regulation. The responses suggest 
that most scientists expect further research to justify a selective relaxa- 
tion of the containment standai’ds, but they view this as a gradual 
process. They do not reject the possibility of discovering some-hazards. 
Roy Curtiss predicted that within 2 or 3 years scientists will have 
acquired “adequate information” on E. coli Iv-12 but cautioned that 
it will take longer to evaluate new host systems. 
Jonathan King rejected the notion of a short-term risk assessment 
program. He said, “Evolution goes on ; new niches occur in the environ - 
[ Appendix B — 275] 
