## Facts

- Wellcome discovered that AZT (zidovudine), an old compound synthesized in 1964, could be used to treat and prevent AIDS/HIV.
- Beginning in 1983, Wellcome researchers searched for an anti-AIDS drug by screening nucleoside analogues expected to act as chain terminators in viral replication; AZT emerged as a candidate.
- Wellcome lacked facilities for human cell-line testing and arranged external testing, including blind in vitro HIV assays conducted at the U.S. National Institutes of Health (NIH).
- In February 1985, NIH scientists reported that AZT inhibited HIV replication in the human-cell assay.
- Relying on NIH in vitro results, Wellcome’s own animal and toxicity data, and its mechanistic rationale, Wellcome filed a patent application on March 16, 1985 claiming AZT’s use for treatment and prophylaxis of AIDS (priority date for the Canadian patent).
- Generic manufacturers Apotex and Novopharm challenged the patent’s validity and defended infringement actions, alleging lack of utility, overbroad (“covetous”) claims (especially prophylaxis), and improper inventorship for not naming NIH scientists.

## Issues

1. Whether the patent satisfied the statutory utility requirement where clinical efficacy in humans had not yet been demonstrated at filing, and whether utility could be established by the doctrine of sound prediction.
2. Whether claims covering both treatment and prophylaxis were impermissibly broader than the disclosed invention (“covetous claiming”).
3. Whether NIH scientists were co-inventors (not merely verifiers) and, if omitted, whether the omission was wilfully made for the purpose of misleading under Patent Act § 53(1).
4. What standard of review applied to trial findings on technically complex patent questions involving mixed fact and law.

## Decision

- The Supreme Court of Canada dismissed the appeals and upheld the patent as valid and infringed.
- The Court confirmed that utility may be established by sound prediction when full proof is not available at filing, provided specific requirements are met.
- The Court rejected the covetous-claiming attack and sustained claims for both treatment and prophylaxis on the evidence.
- The Court held the NIH scientists were verifiers, not co-inventors, and there was no wilful misleading omission under § 53(1).
- The Court applied appellate deference to trial-level findings grounded in expert evidence and found no palpable and overriding error.

## Legal Principles

- Utility can be satisfied by “sound prediction” where full demonstration is not yet available at filing.
- Sound prediction requires: (1) a factual basis at the filing date, (2) a sound line of reasoning from those facts to the predicted utility, and (3) proper disclosure of that basis in the specification.
- Sound prediction does not include a lucky guess or mere speculation; whether it is met is largely a fact-driven inquiry.
- Patent claims must not exceed the invention as disclosed; breadth (including prophylaxis) may be valid if supported by the same factual and scientific foundation.
- Inventorship turns on conception of the inventive idea; third parties who conduct confirmatory testing at the patentee’s request may be verifiers rather than inventors.
- Under Patent Act § 53(1), an incorrect statement or omission invalidates a patent only if wilfully made for the purpose of misleading.

## Conclusion

The Court upheld Wellcome’s AZT “new use” patent and confirmed that pre-clinical evidence can satisfy utility through sound prediction when grounded in a factual basis, supported by a coherent scientific rationale, and properly disclosed, while also clarifying limits on covetous claiming and distinguishing inventors from verifiers.


Apotex, Inc. v. Wellcome Found. Ltd., 2002 SCC 77 (Can. 2002)