P. Abelson


 The full story of what happened has not been detailed. However, ru-
mors are that problems were encountered in the use of E. coli. The insulin
gene was well expressed, but the protein formed was retained within the
bacterial cells. Complete separation of the product from the host of other
E. coli proteins was absolutely essential to avoid pyrogenic activity, but
this was not easy. The initial sales price of Humulin was much greater than
that of animal insulin. Apparently, the production problems have been
overcome, possibly by using another organism such as yeast as an expres-
sion system. In any event, Humulin is now broadly accepted. At least half
of new diabetic patients in the United States are being treated with it.
Humulin now sells for less than animal insulin.
 In summary, experience with insulin demonstrates that important thera-
peutic agents can be produced by recombinant DNA technology. Costs can
be competitive. The time required to go from laboratory to widespread use
can be many years.


GROWTH HORMONE

 The second therapeutic recombinant DNA drug to be approved by the
FDA, a growth hormone for humans, also has a long history and favorable
circumstances for its acceptance. Genentech first produced its version of a
human growth hormone in 1979. Clinical trials began in 1981. At that
time, the human growth hormone being used for therapy was derived by
extraction of the pituitaries of human cadavers. The extracts were pooled
before final processing. Ultimately, concern was expressed that the prod-
uct contained harmful viruses, and in early 1985, the FDA cancelled ap-
proval of the human-derived growth hormone. This obviously had an ex-
pediting effect on the approval of the Genentech product which was
granted in October 1985. Some 4,000 children who would otherwise be
destined to be adult dwarfs are now being treated. It is estimated that
ultimately a total of the order of 10,000 to 15,000 children will receive the
drug. Costs are high. They range between $4,000 and $6,000 per patient
per year.
 Patients treated with the Genentech drug have responded well. How-
ever, the protein is not identical to the human hormone. Other small com-
panies have created the human variety by recombinant DNA technology,
and their products are now being tested clinically. Genentech has found it
prudent to create its own new comparable product and is subjecting it to
clinical trials. Clinical trials are expensive. Costs quoted range from $25
million to $60 million.