68 



gram, in my opinion and in the opinion of my industry colleagues who have contrib- 

 uted to the success of this program, has been the consensual adoption of specific 

 tests that together define a sufficient screening information data set (SIDS) upon 

 which reasonable, objective decisions can be based. This SIDS package has been 

 internationally validated and found to be feasible, replicable and parsimonious in 

 its use of resources — time, money, and experimental animals. 



To date, SIDS assessments have been completed for nearly 100 chemicals at the 

 cost of about $60,000 per chemical. For this modest investment, we now know that 

 it is possible to assemble a broad (if not deep) dataset that can serve the purposes 

 of decisionmakers in several regards: does the chemical need further testing? should 

 actions be taken to restrict exposures to the chemical? can the chemical be removed 

 from a priority list of consideration? All these decisions can and have been sup- 

 ported by the SIDS undertaking. Moreover, the SIDS approach is embedded in a co- 

 herent approach to prioritization and action. It has already moved us significantly 

 towards action on the basis of information rather than ignorance. We can no longer 

 invoke the tired dictum of what you don't know can't hurt you. We need to know, 

 and we need to know in order to know when to act — and when not to act. 



The SIDS database is an evolving concept, and part of its success has been its 

 openness to innovation and incorporation of advances in toxicology methods. It is 

 a screening or first-tier approach to information acquisition: It may be followed by 

 further, more refined and focussed testing. Alternatively, it may provide information 

 suflBcient for decisionmaking by either industry or government— for instance, indus- 

 try may determine that a positive result in a short term genotoxicity assay is suflB- 

 cient information to disinvest in a certain chemical. 



2. New Chemicals 



Another disappointment of TSCA has been its failure to establish a rational, pre- 

 cautionary approach to the evaluation of new chemicals. Catching up with the uni- 

 verse of existing chemicals is a very difficult task, and some sort of prioritization 

 is necessary even over the long term. However, TSCA should function as a vigilant 

 gatekeeper over the entrance of new chemicals into the environment, in order to 

 prevent an increasing list of unknowns, as new chemicals join existing chemicals in 

 the black hole of ignorance. 



This objective has not been accomplished. The new chemicals notification system 

 has been implemented by EPA to operate largely in the absence of real data on tox- 

 icity. Certainly, EPA has developed ingenious and often valuable approaches based 

 on structure-activity relationships (SAR) and other nondata-based methods of analy- 

 sis, and the experience and knowledge gained fi-om this approach should be pre- 

 served. However, as a joint EPA-EEC study demonstrated, SAR analysis works best 

 when combined with a data-based evaluation system, such as the new chemicals 

 program of the European Community. Based upon the success of the SIDS program 

 described above, there can be no rational objection to the requirement of at least 

 the SIDS tests as a prerequisite to the premanufacture notification (PMN) program 

 under section 5 of TSCA. EPA's review of PMN submissions will then be based upon 

 actual data, and decisions as to the need for fiirther testing will be rationally fo- 

 cussed on those endpoints or data gaps revealed by such data. 



The EEC approach to new chemicals, with mandatory testing, does not seem to 

 have impeded innovation in the chemical industry. Several of the world's largest 

 chemical companies are still European. A gutless TSCA has not paid off in any com- 

 petitive advantage for American industry. 



3. Significant New Use Rules (SNURs) 



The third leg of TSCA's approach to chemicals was the mandate to review signifi- 

 cant changes in the use or production patterns of already approved chemicals. In 

 this way, it was to be possible for earlier decisions, based as they were upon a com- 

 bined analysis of toxicity and exposure, to be reconsidered if one variable — expo- 

 sure — were to change substantially. Thus the SNUR approach could be considered 

 a kind of safety net for chemicals, whose current level or range of use was not of 

 concern, but whose inherent toxicity was such that changes in levels or ranges of 

 use might elicit control. 



