This appeal is from the decision of the Board of Appeals affirming the examiner’s rejection of claims 1 and 2 in appellant’s patent application1 for “4-aryl-4- (lower acyloxy)-l-(3-aryl-3-hydroxy-propyl)-piperidines and their preparation.” Appellant has withdrawn appeal of claim 2.
The appealed subject matter is represented by the chemical compound of claim 1:
1. 1 - (3-Hydroxy-3-phenylpropyl) - 4-phenyl-4-propionoxy-piperidine.
That compound has the structural formula
and is useful as an analgesic.
The references are:
Elpern II 2,846,437 August 5, 1958.
Elpern I 2,850,500 September 2,1958.
Pohland II 2,904,550 September 15,1959.
Pohland I 2,951,080 August 30, 1960.
Cutler et al. 2,962,501 November 29, 1960.
The disclosure of those references was reviewed by this court in In re Caraba-teas, 345 F.2d 1013, 52 CCPA 1386, and need not be further detailed here beyond noting that Cutler and Pohland I disclose an analgesic of the structural formula
The examiner found appellant’s piper-idinol ester obvious from a consideration of the Cutler, Elpern and Pohland patents, employing reasoning similar to that used by the examiner in Carabateas.
Appellant contends that an affidavit showing the compound of claim 1 to be nineteen times more active as an analgesic than the “reverse ester” compound of Cutler and Pohland I when administered subcutaneously is persuasive evidence of unobviousness and patentability of the compound.2 As reference to our decision in Carabateas will establish, we thought the Elpern patents collectively suggested that piperidinol esters having the structure of appellant’s compounds *1000there claimed would possess greater analgesic activity, perhaps twice as great, than the “reverse esters” which Cutler obtained from the requisite piperidine carboxylic acid. We also referred to a Braenden publication in that record, not relied on by the examiner in his rejection, which showed that in changing from the carboxylate structure of meperidine O (rmg-¿-0-C2H5), to the “reverse” struc-O II ture (ring-0-C-C2H5), the substituent being in the 4-position of the ring, and the remainder of the molecule being kept the same, a five to ten-fold increase in analgesic activity is observed.3
We have re-examined the Elpern references on which heavy reliance was placed in our original Carabateas opinion. While we are of the view that those references, together with Cutler, collectively establish obviousness of the structural formula and analgesic activity both of appellant’s compounds claimed there and the present compound, we think we erroneously concluded in Carabateas that the Elpern patents would suggest to one of ordinary skill that any improvement in analgesic activity would necessarily inhere in compounds derived 0 from piperidinol, ring-0-C-C2H5, when compared to otherwise identical compounds of the carboxylate structure, O ring-$-0-C2H5.4
In the absence of a suggestion in the present record that the nature of appellant’s improvement would be expected, *1001we think appellant has presented clear and convincing evidence of the greater effectiveness of the claimed compound over the prior art, sufficient here to establish patentability of the compound. In re Lohr, 317 F;2d 338, 50 CCPA 1274. Under the circumstances, we think the record fails to support a finding that the subject matter as a whole would be obvious to one of ordinary skill in the art.
The decision of the board is reversed.
Reversed.
. Serial No. 85,195, filed January 27, 1961.
. That figure was established by a finding that appellant’s compound and that of Cutler were 8220 times and 173 times, respectively, more potent than meperidine, a well known analgesic, on a molar basis.
. That appears consistent with the discussion in In re Lunsford, 327 F.2d 526, 51 CCPA 1000, wherein it was observed that the analgesic prodine, having the formula
is more potent than the analgesic meperi-dine, having the formula
the Braenden reference is not in the present record.
. We said in Garabateas:
Referring to the Elpern patents, it will be seen that the Elpern II “l-(3phenoxypropyl) -4-acetoxy-” compound, which has the same sort of ester structure as appellant’s compounds, possesses tivice the analgesic activity of the Elpern I ethyl (3-phenoxypropyl) piper-idine-4-earboxylate. Compared with the Elpern I ethyl (2-phenoxyethyl) piperi-dine-4-carboxylate and ethyl (4-phe-noxybutyl) piperidine-4-carboxylate esters, this same Elpern II compound, albeit not the exact “reverse ester” but instead a homolog thereof, is, respectively, four and eight times more effective as an analgesic. We believe this amply suggests that esters having the structure of appellant’s compounds will possess greater analgesic activity than their “reverse esters.”
In drawing that conclusion, we assumed that the “4-acetoxy” compound of Elpern II was identical to (e. g. the exact “reverse ester”) the eí/ií/i-4-carboxylate compound of Elpern I but for reversal of the ester linkage. In fact it is not, as comparison of the respective structures (ring-0-C-CH3 vis a vis ring-C-0-C2H5) 11 IL o o demonstrates. We think such a comparison of ester homologues to be too tenuous to suggest that any improvement would be obtained solely because of reversal of the ester linkage. That comparison being the controlling evidence in Garabateas, we think the decision there rendered necessarily is incorrect.