L'Oreal, S.A. appeals the judgment of the United States District Court for the District of Columbia, 40 USPQ2d 1425, 1996 WL 679430 (D.D.C.1996), reversing a decision of the United States Patent and Trademark Office Board of Patent Appeals and Interferences and awarding Estee Lauder Inc. priority to an invention for sunscreens utilizing a copper compound to boost their Sun Protection Factor. Because the court erred in holding that Estee Lauder's inventors did not have to establish that they recognized the success of their invention before the date L'Oreal's inventors reduced their invention to practice, we reverse.

This case involves an interference between U.S. patent application serial number 07/135,666, filed by Walter P. Smith and Edward Pelle, which is assigned to Estee Lauder Inc., and U.S. patent application serial number 07/180,556, filed by Michel Hocquaux and Georges Rosenbaum, which is assigned to L'Oreal, S.A.1 The applications are directed to a composition for protecting skin from damage caused by ultraviolet (UV) light exposure or to its method of use. The sole count provides:

A composition for protecting skin from damage caused by exposure to ultraviolet light said composition comprising a sunscreen agent and copper 3', 5' diisopropylsalicylate [ (CuDIPs) ] in a cosmetic carrier or a method for protecting skin from damage caused by exposure to ultraviolet light comprising apply [sic] to the skin an effective amount of the composition comprising a sunscreen agent and copper 3', 5' diisopropylsalicylate in a cosmetic carrier.

The '666 application was filed on December 21, 1987, while the '556 application was filed on April 12, 1988. Because L'Oreal had applied for a patent for the same invention in Luxembourg on April 13, 1987, however, it was given the benefit of that filing date and deemed the senior party. See 35 U.S.C. § 119 (1994). Consequently, Estee Lauder was required to prove priority by preponderant evidence.

Estee Lauder first became interested in the potential use of CuDIPs to boost the Sun Protection Factor (SPF) in sunscreens in the summer of 1986. Dr. Walter Smith, Director of Estee Lauder's Biological Research Department (BRD), requested that Edward Pelle, a senior scientist in the BRD, attend a conference on copper complexes in August 1986, which he did. The conference was held by Dr. John R.J. Sorenson of the University of Arkansas, who is reputed to be the world's leading expert on CuDIPs.

Pelle spoke with Sorenson about the potential use of CuDIPs to reduce skin inflammation and to minimize damage caused by exposure to the sun. Upon his return, Pelle discussed the conference and Sorenson's qualifications with Smith, and they agreed that Sorenson was their best source of CuDIPs. In September 1986, Pelle telephoned Sorenson and ordered two grams of CuDIPs for $20.00.

Later that month, Pelle received from Sorenson two grams of material he believed to be CuDIPs (the "material"). He used the material to perform several experiments to test whether CuDIPs inhibited inflammation after UV light exposure. First, in November 1986, Pelle applied a template with two square-shaped openings to his arm. On one opening he applied isopropyl alcohol (IPA), while on the other he placed a solution of IPA mixed with the material. Pelle then exposed his arm to UV light, after which he applied the same solutions to the same openings a second time. Pelle found that the site treated with IPA only was red, whereas the site treated with IPA and the material was not. He then took a photograph of his arm, which he taped into his laboratory notebook. Pelle showed his arm to Smith and to Ken Marenus, who began working as laboratory manager and Pelle's supervisor in the BRD that day, and the three discussed the results. Pelle then repeated his arm test, but this time he did not apply the material until after the UV exposure. He found that the material effectively reduced redness.

Pelle and Marenus then added the material to one of Estee Lauder's SPF 6 sunscreen products. They mixed two different concentrations of the material to the sunscreen samples, with the lower (.3 millimolar) concentration labeled BRD 461, and the higher (3.0 millimolar) concentration labeled BRD 462. Marenus and Marie Randazzo, an Estee Lauder employee whose duties included obtaining and sending samples to contract laboratories for SPF testing, next filled out requests for biomedical testing on BRD 461 and 462. They sent the samples to Harrison Research Laboratories, an independent laboratory which both Estee Lauder and L'Oreal use, indicating the SPF range on which the test should focus.

In January 1987, Dr. Lynne Harrison, president and principal investigator of the SPF testing laboratory at Harrison Research, told Marenus the results of the testing on BRD 461 and 462, which were confirmed in a written report received later that month. Pelle and Marenus analyzed the results and determined that the addition of the material they believed was CuDIPs increased the SPF. Both Pelle and Marenus memorialized this determination in separate memoranda to Smith.

Smith, Marenus, and Pelle then decided to expand testing to see whether CuDIPs was effective in boosting SPF in a broader range of sunscreens. Marenus telephoned Sorenson and ordered 100 more grams of CuDIPs for $500. It is undisputed for purposes of this appeal that this material was, in fact, CuDIPs. In late February 1987, Pelle prepared samples incorporating .3 millimolar of CuDIPs into SPF 15 (BRD 498), 25 (BRD 499), and 20 (BRD 500) sunscreens. Pelle also prepared a control sample (BRD 497) containing only SPF 15 sunscreen, which was necessary because Marenus' earlier testing on SPF 15 had not shown an SPF boost.

On or about February 26, 1987, a request for biomedical testing was completed and given to Randazzo for BRD 497-500. On or about March 9, 1987, she sent the samples to Harrison, accompanied by a cover letter, requesting SPF testing and telephonic notification of the test results. Pelle subsequently received a verbal report of partial results from Randazzo, who would have received them from Harrison. The partial results are reflected in an undated page of notes in Pelle's handwriting. The final written results were dated and received after April 13, 1987.

Pelle offered varying and indefinite testimony as to when he received the partial, verbal results. At trial, he claimed that he received them "in the early part of April"; more specifically, "around the first week in April of 1987." He also testified that he had requested the results by March 12, 1987, but they actually arrived "a little bit after that." At his deposition, taken two years before trial, however, Pelle testified that his best recollection was that he received the partial results sometime in "the spring of 1987." Ultimately, the court found Pelle's testimony lacking in credibility. Randazzo, on the other hand, simply did not recall when Harrison telephoned her with the partial results. Nor did any documentary evidence of hers reveal such date.

For her part, Harrison testified that all SPF testing was completed on BRD 497-500 by April 10, 1987. She finished the BRD 497 and 498 testing before 2:00 p.m. on April 7, 1987. Under her normal practice, Harrison testified, she would have called Randazzo with the partial results on April 7, 1987. While she offered long distance telephone records showing a telephone call to Randazzo's extension at 2:36 p.m. on that date, she did not recall making the call to Randazzo with the partial results. Harrison conceded, however, that she was working on other SPF testing for Estee Lauder during that period, so she could have called Randazzo with results from those tests. Harrison also testified that she would have notified Randazzo of the remaining results on BRD 499 and 500 by Friday, April 10, or at the latest by Tuesday, April 14, after the critical date.

Pelle testified that upon receiving the partial results from Randazzo, he calculated the means of the BRD 497 and 498 results, compared them, and concluded that CuDIPs was effective at boosting SPF. He also compared the results of BRD 499 against its nominal SPF and determined that they showed a higher SPF. He did not calculate a mean for the BRD 499 or 500 results. Pelle could not pinpoint when he performed these calculations and comparisons, however. Nor did any documentary evidence answer this question. Moreover, although Pelle shared these results with Marenus, Marenus could not say when this occurred.

Estee Lauder then filed the '666 application. The United States Patent and Trademark Office declared an interference between the Estee Lauder and L'Oreal applications on December 6, 1989. Before the Board of Patent Appeals and Interferences, Estee Lauder relied solely on the experiments leading up to and culminating in the BRD 461-62 tests to establish its reduction to practice. It did not rely at all on the BRD 497-500 tests, which it viewed as cumulative. Further, Estee Lauder did not attempt to establish that it was first to conceive the invention and then diligently reduce it to practice. The board found that Estee Lauder had failed to prove that the initial 2-gram sample it obtained from Sorenson was actually CuDIPs. Consequently, it held that Estee Lauder had not reduced its invention to practice prior to April 13, 1987, and awarded priority to L'Oreal.

Estee Lauder then filed an action in the United States District Court for the District of Columbia, pursuant to 35 U.S.C. § 146 (1994). Section 146 actions have been described as a hybrid of an appeal and a trial de novo. Case v. CPC Int'l, Inc., 730 F.2d 745, 752, 221 U.S.P.Q. (BNA) 196, 202 (Fed.Cir.1984). At trial, Estee Lauder relied on both the BRD 461-62 tests and the BRD 497-500 tests to establish its reduction of the invention to practice. The court agreed with the board that Estee Lauder failed to show that BRD 461-62 contained CuDIPs. But it found that the BRD 497-500 samples did contain CuDIPs and did meet the elements of the count. It then held that these tests established a successful reduction to practice even though neither inventor, nor any other Estee Lauder employee or agent, received and analyzed the BRD 497-500 test results or concluded that the tests were successful until after the critical date of April 13, 1987. Consequently, it reversed the board and awarded priority to Estee Lauder. L'Oreal appeals.

The primary question before us is whether Estee Lauder established by a preponderance of the evidence that it successfully reduced its invention to practice before the April 13, 1987, critical date. If it did, then it is entitled to priority. 35 U.S.C. § 102(g) (1994); Mahurkar v. C.R. Bard, Inc., 79 F.3d 1572, 1577, 38 USPQ2d 1288, 1290 (Fed.Cir.1996). Whether Estee Lauder was the first to conceive the invention and was reasonably diligent in reducing it to practice is not before us because Estee Lauder failed to raise this issue before the board or on appeal. The only issue before the board was priority of reduction to practice. Therefore, although the district court did not abuse its discretion in allowing Estee Lauder to introduce evidence of the BRD 497-500 tests insofar as this evidence was relevant to its reduction to practice, Case, 730 F.2d at 752, 221 USPQ at 202, the district court could not examine, absent compelling circumstances, whether this evidence established any other issue not raised before the board, such as priority based on prior conception coupled with diligent reduction to practice. See Conservolite, Inc. v. Widmayer, 21 F.3d 1098, 1102, 30 USPQ2d 1626, 1629 (Fed.Cir.1994).

Reduction to practice is ultimately a legal question, which is based on underlying factual determinations. Fujikawa v. Wattanasin, 93 F.3d 1559, 1564, 39 USPQ2d 1895, 1899 (Fed.Cir.1996); Conservolite, 21 F.3d at 1100, 30 USPQ2d at 1628. Accordingly, we review the court's judgment for legal error and clearly erroneous findings of fact. Conservolite, 21 F.3d at 1100, 30 USPQ2d at 1628.

To prove actual reduction to practice, an inventor must establish that he " 'actually prepared the composition and knew it would work.' " Hahn v. Wong, 892 F.2d 1028, 1032, 13 USPQ2d 1313, 1317 (Fed.Cir.1989) (quoting Mikus v. Wachtel, 542 F.2d 1157, 1159, 191 U.S.P.Q. (BNA) 571, 573 (CCPA 1976)); see also Burroughs Wellcome Co. v. Barr Lab., Inc., 40 F.3d 1223, 1228, 32 USPQ2d 1915, 1919 (Fed.Cir.1994) (reduction to practice requires "the discovery that an invention actually works" (emphasis added)); see also Standard Oil Co. (Indiana) v. Montedison, S.p.A., 494 F. Supp. 370, 206 U.S.P.Q. (BNA) 676 (D.Del.1980), aff'd, 664 F.2d 356, 212 U.S.P.Q. (BNA) 327 (3d Cir.1981) (reduction to practice requires a showing of three elements: (i) production of a composition of matter satisfying the limitations of the count, (ii) recognition of the composition of matter, and (iii) recognition of a specific practical utility for the composition).

Here, the court found that prior to the critical date, Estee Lauder had conceived its invention, prepared the composition of the count, and sent the composition to be tested. The court also found that Harrison had completed all of the testing prior to the critical date. The rub appears, however, in the court's finding that Estee Lauder failed to establish that the test results were reviewed or analyzed before April 13, 1987:

L'Oreal argues that Estee Lauder cannot show reduction to practice by April 13, 1987, because it cannot show that the test results for BRD 497-500 were received and analyzed by Estee Lauder prior to the critical date. Were the Court to agree that receipt and analysis of the results by the inventors was necessary to establish reduction to practice, L'Oreal would be correct. While Estee Lauder plainly had a well-supported hypothesis that CuDIPs boosted SPF, and prepared test samples meeting the elements of the count well before April 13, 1987, the evidence does not show that Estee Lauder's scientists received or analyzed the results of BRD 498, 499, or 500 by that date.

40 USPQ2d at 1435. The court held that if the law required proof of receipt and recognition of success prior to the effective date, Estee Lauder could not prevail. But it concluded that there is no such requirement. It held that "the patent law is more concerned with whether the results of the tests submitted by the inventor in fact show success than whether this success was recognized by the inventor prior to the critical date." Id. at 1436. Thus, because the tests were completed by April 13, 1987, and because they ultimately demonstrated that CuDIPs actually worked for its intended purpose, the court held that Estee Lauder reduced its invention to practice before L'Oreal.

On appeal, L'Oreal does not challenge the board's findings. Nor does it challenge the adequacy of the testing or the ultimate success of the testing. Indeed, L'Oreal accepts that Estee Lauder's test results would be adequate to establish a reduction to practice, had they been received, analyzed, and success determined prior to the critical date. L'Oreal contends, rather, that it was burned by the trial court's erroneous conclusion of law that Estee Lauder did not have to prove receipt and analysis of the BRD 497-500 test results before the critical date.

Because we see no clear error in the court's thorough and well-supported factual findings,2 we are left to answer a single question of law: where testing is required to establish utility, must there be some recognition of successful testing prior to the critical date for an invention to be reduced to practice, or is it only necessary that the testing be completed before the critical date and ultimately prove successful, regardless of when that success is appreciated or recognized? We hold that the law requires the former.

Hahn requires that in addition to preparing a composition, an inventor must establish that he " 'knew it would work,' " to reduce the invention to practice. 892 F.2d at 1032, 13 USPQ2d at 1317 (quoting Mikus, 542 F.2d at 1159, 191 USPQ at 573). This suggests that a reduction to practice does not occur until an inventor, or perhaps his agent, knows that the invention will work for its intended purpose. Indeed, we agree with Standard Oil that the "utility requirement is satisfied when an inventor has learned enough about the product to justify the conclusion that it is useful for a specific purpose." 494 F. Supp. at 381, 206 USPQ at 691. But until he learns that threshold information, there can be no reduction to practice. Moreover, Burroughs Wellcome states that a reduction to practice requires "the discovery that an invention actually works." 40 F.3d at 1228, 32 USPQ2d at 1919. This suggests that until that "discovery" is actually made, there is no reduction to practice. These cases trumpet, therefore, the principle that a reduction to practice does not occur until the inventor has determined that the invention will work for its intended purpose.

"It is well-settled that conception and reduction to practice cannot be established nunc pro tunc. There must be contemporaneous recognition and appreciation of the invention represented by the counts." Breen v. Henshaw, 472 F.2d 1398, 1401, 176 U.S.P.Q. (BNA) 519, 521 (CCPA 1973) (emphasis added) (holding no reduction to practice during lab experiments because there was no "indication in the contemporaneous record" that utility "was recognized at that time" (emphasis added)).

Similarly, Langer v. Kaufman, 59 C.C.P.A. 1261, 465 F.2d 915, 175 U.S.P.Q. (BNA) 172 (1972), involved a count for a composition, which required, in part, a gamma X-ray diffraction pattern of cocrystallized titanium trichloride. In 1956, prior to the critical date, the appellants had samples of the composition prepared and the X-ray diffraction patterns photographed. The samples were then sent to a separate "analytical section" for X-ray diffraction pattern determinations. During the interference proceeding, after the critical date, one of the appellants identified, for the first time, the diffraction patterns of the samples in the 1956 photographs as being of the gamma form. The court agreed with the board that the appellants had failed to establish a reduction to practice because there was " 'no evidence of any contemporaneous interpretation or evaluation of the patterns either by [one of the appellants] or by anyone else.' " 465 F.2d at 919, 175 USPQ at 174; cf. Knorr v. Pearson, 671 F.2d 1368, 1375, 213 U.S.P.Q. (BNA) 196, 202 (CCPA 1982) ("Where a count to a mechanical invention includes a structural feature recited as a positive limitation, conception and reduction to practice of that invention require a contemporaneous recognition of that feature." (emphasis added)); Heard v. Burton, 51 C.C.P.A. 1502, 333 F.2d 239, 244, 142 U.S.P.Q. (BNA) 97, 100 (1964) (Rich, J.) (no conception even though appellant had in fact prepared the catalyst in the count because there was no "recognition and appreciation of the invention" prior to the critical date).

In ruling that no recognition of successful testing is required before the critical date, the district court here relied on Scott v. Finney, 34 F.3d 1058, 32 USPQ2d 1115 (Fed.Cir.1994), and Reese v. Hurst, 661 F.2d 1222, 211 U.S.P.Q. (BNA) 936 (CCPA 1981). It is true that Scott states that in deciding the sufficiency of testing necessary to establish a reduction to practice, courts examine "the record to discern whether the testing in fact demonstrated a solution to the problem intended to be solved by the invention." 34 F.3d at 1063, 32 USPQ2d at 1119 (emphasis added). That is an accurate statement, but it pertains only to how a court determines whether the testing is sufficient to show a reduction to practice. That question is not before us. L'Oreal does not challenge the sufficiency of the testing; it argues that an actual reduction to practice does not occur until the success of the testing is known to someone. Scott sheds no light on this issue.

Reese, on the other hand, is a more formidable obstacle. There, Hurst had his laboratory assistant, Howerton, perform certain tests on June 13, 1969. Howerton sent the test samples for analysis and received the successful results June 27, 1969. The board held that reduction to practice occurred on June 13, the test date, not June 27, the date Howerton received the results. 661 F.2d at 1227, 211 USPQ at 941. Reese contended that there was no evidence of how the results were analyzed at the time of testing or any suggestion that Hurst considered the tests successful. The court rejected these arguments, stating that the test results themselves demonstrated a successful reduction to practice and that there was no rule that the inventor or tester must expressly state that the tests were a success. Id. at 1228, 211 USPQ at 942. It then affirmed the board's June 13 reduction to practice date. We interpret Reese to say that when the test results themselves show success, there is no requirement for further analysis. Nor is it necessary for the inventor to articulate, verbally or in writing, that the tests were successful because the results themselves say that. But we do not see Reese as imposing a rule that the successful testing date is the date of reduction to practice, even when success is not known until later. Indeed, to the extent Reese can be read as establishing such a rule, it stands alone among our precedent. Reading Reese to create such a rule would be especially strained because whether reduction to practice occurred on June 13 or June 27 was entirely immaterial to the decision. The critical point was that reduction to practice occurred before July 2, 1970, the filing date of the senior party.

We agree with L'Oreal, therefore, that when testing is necessary to establish utility, there must be recognition and appreciation that the tests were successful for reduction to practice to occur. Here, that time did not arrive until Pelle received the test results, calculated the mean of the BRD 497 and 498 results and compared them, as well as compared the BRD 499 and 500 test results to their nominal SPFs. Only then did Estee Lauder determine that CuDIPs successfully boosted SPF values. Estee Lauder has not established that these events unfolded before April 13, 1987.

Estee Lauder argues, however, that there was no need for it to receive and analyze the BRD 498-500 test results and determine whether they were successful because they already fully recognized the utility of the compositions. For this proposition, Estee Lauder relies on Pelle's arm tests and the BRD 461-62 test results. But Estee Lauder failed to establish that CuDIPs was the material used in these tests. It cannot rely upon tests performed on a composition that failed to meet the elements of the count to demonstrate that the composition works for its intended purpose.

Perhaps recognizing the flaws in this argument, Estee Lauder argues alternatively that the "activities and knowledge of Dr. Harrison and Randazzo inure to the benefit of the inventors and prove reduction to practice prior to the critical date." Estee Lauder cites no evidence that either Randazzo or Harrison was in a position to judge the success of the test results; it relies only on the arguments of counsel. But, "arguments of counsel cannot take the place of evidence lacking in the record." Knorr, 671 F.2d at 1373, 213 USPQ at 200. First, as the court found, Estee Lauder did not present sufficient or persuasive evidence that Harrison actually phoned Randazzo with the test results prior to the critical date. So, Randazzo cannot be credited even with having knowledge of the BRD 498-500 results before the critical date. Even had she known the results, however, Estee Lauder points to no evidence that she had any basis upon which to determine whether they were successful. While she knew the expected SPF range of the samples, there is no evidence that she either calculated the mean test results for any sample or compared the sample results to the SPF of the same sunscreen product not containing CuDIPs. Indeed, there is no evidence that she had any idea how to determine success. The same is true of Harrison. In sum, there is simply no basis upon which to conclude that either Randazzo or Harrison determined, or even knew how to determine, whether the BRD 498-500 test results demonstrated that CuDIPs was effective in boosting the SPF in sunscreens.

Accordingly, the judgment of the United States District Court for the District of Columbia is reversed, and priority is awarded to L'Oreal.

Each party shall bear its own costs.

REVERSED.