Ill . 
Biological Containment 
Certification of Host-Vector Systems 
Certification of Host-Vector systems must, be undertaken with deliberate 
caution due to the many unknown factors inherent in a new host or vector. 
We can learn from our experience with £ . Coli , a system about which we 
know a great deal, that many experiments and tests must be done before 
any claims about safety can be legi tamatoly made. In the guidelines, 
the criteria developed for approving new host-vector systems are too 
vague and lack sufficient inputfrom relevant disciplines such as epi- 
demiology or microbial ecology. More importantly, the process by 
which H-V systems are approved may act to weaken these standards. 
The NIH creates a dangerous precedent by mandating initial review of 
HV systems to a subcommitte of an already too narrowly focussed Advisory 
Committee. While this subcommittee may be strengthened by the inclusion 
of special members capable of evaluating the sufficiency of the data 
and assessing the impact of widescale use of the system, it may just 
as well be weakened by its vulnerability to 'experts' who may have a 
strong interest in using the HV system they are supposed to review. 
By relying on vaguely constituted subcommittees, the NIH skirts the 
important issues of who should determine what will be sufficient data 
to merit certification and on what grounds such decisions should be 
made. The NIH should make an aggressive effort to disseminate all such 
information to the public and relevant professionals before any 
decisions are made. One way to do this would be to require, not merely 
encourage as the guidelines now state, investigators to publish their 
data on these systems. By circulating the data and by making it 
explicit in the guidelines that these comments will be evaluated by 
a diversely composed group representative- of the various interest in- 
volved, a more responsive and responsible policymaking process will 
ensue . 
IV. Experimental Guidelines 
Exemptions 
Creating exemptions to the prohibited production of large-scale 
experiments or the deliberate release of any recombinant DNA organism 
into the environment plaices the very rais on d' etre of these guidelines 
in je pardy. If exceptions are made to the 10 liter limit, for example, 
what happens to the efficacy of cont nment? Will adjustments have 
to be made? Where are the provisions and standards for such adjustment 
in these guidelines? 
While exemptions may be necessary in order to pursue risk-assessment 
studies, exemptions to private industry raise a myriad of difficult 
questions. However, requests for exemption may well come from private 
industry, as it has already expressed interest in changing the ID liter 
[Appendix A — 189] 
