216 • Impacts of Applied Genetics — Micro-Organisms, Plants, and Animals 
Evaluation of the Guidelines 
Two basic issues must be addressed. The first 
is how well the Guidelines confront the risks 
from genetic engineering, which may not have a 
definitive answer in view of the uncertainty 
associated with most of the risks. Consequently, 
it is also necessary to consider a second issue— 
whether confidence is warranted in the deci- 
sionmaking process responsible for the Guide- 
lines. 
THE PROBLEM OF RISK 
The Guidelines are designed to address the 
risks to public health and the environment from 
either rDNA molecules or organisms and vi- 
ruses containing them. The underlying premise 
is that research should not be unreasonably 
restricted. This is essentially a risk-benefit ap- 
proach; at the time that the original Guidelines 
were drafted, it represented a compromise be- 
tween the extremes of no regulation and of no 
research without proof of safety. Physical and 
biological containment levels were established 
for various experiments based on estimated 
degrees of risk. The administrative mechanism 
created by the Guidelines is that of a Federal 
agency— NIH— advised by a diverse body of 
experts— RAC. Scientific advice on the technical 
aspects of risk assessment is provided by techni- 
cal experts on RAC; public input is provided by 
experts in nontechnical subjects and by the 
right of the public to comment on major actions, 
which are published in the Federal Register. 
Compliance is accomplished by a combination 
of local self-regulation and limited Federal over- 
sight, with the ultimate enforcement resting in 
the Federal funding power. 
Since their initial appearance, the Guidelines 
have evolved. As scientists learned more about 
rDNA and molecular genetics, two trends oc- 
curred. First, containment levels were progres- 
sively lowered. Major revisions were made in 
1978 and 1980; minor revisions were often 
made quarterly, as proposals were submitted to 
the RAC at its quarterly meetings, recom- 
mended by RAC, and accepted by the Director. 
By now, approximately 85 percent of the per- 
mitted experiments can be done at the lowest 
physical and biological containment levels. Se- 
cond, the degree of centralized Federal over- 
sight has been substantially reduced to the point 
where almost none remains. Under the 1976 
Guidelines, all permitted experiments ultimately 
had to be reviewed by the IBC and ORDA before 
they could be started; the 1978 Guidelines no 
longer required preinitiation review of most 
experiments by ORDA, although ORDA con- 
tinued to maintain a registry of experiments 
and to review IBC decisions. Under the 
November 1980 revision to the Guidelines, there 
will be no Federal registration or review of ex- 
periments for which containment le\els are 
specified in the Guidelines. About 97 percent of 
the permitted experiments fall into this 
category. 
Preinitiation review of experiments by RA(’ 
has been an important part of the oversight 
mechanism. Expert review encourages experi- 
mental design to be well thought out and pro- 
vides a means for catching potential pi’ohlems, 
e.g., one application re\iewed by B,\(; ne\('r 
mentioned that the s[)ecies to he used as a DN.\ 
donor was capable of manufacturing a potent 
neurotoxin; it was turned down aft(M' a RAC 
member familiar with the species brought this 
fact to the Committee’s attention.'-’ 
The burdens imposed on rDN.A acti\ ities by 
the Guidelines appear to he reasonable in \ iew 
of continuing concerns about risk. I,(‘ss than 15 
percent of permitted expcM'iments re(|uii’(' pre- 
initiation appro\ al by the local IBC's, which usu- 
ally meet monthly. Preinitiation approx al of e.\- 
periments by NIH is retjuired only for: I) e.xperi- 
ments that have not been assigned containment 
levels by the Guidelines; 2) expei-iments using 
new host-vector systems, which must he cei ti- 
fied by NIH; 3) certain experiments re(|uiring 
case-by-case approval; and 4) i'(*{|uests for ex- 
ceptions from Guideline re(|uii'ements. 1 he low- 
est containment levels place minimal burdens 
on the experimenter, (see table 35). For in- 
dustrial applications, NIH approval must hi- 
received not only when th(‘ pi’oje('t is .scaled-u|) 
beyond the 10-1 limit, hut also for each addi- 
tional scale-up of the same project. Many re|)ia‘- 
sentatives of industry consider these suhse- 
’’R. M. tlenig. "Irmihic on llu- H V( ( omniitti'r S()lii> (hi-i 
Downgrading of £. ro/i Cont.iinnicnt. HioSrirntr. \n\ p|i 
762, December 1979. 
