This is an appeal from the decision of the Trademark Trial and Appeal Board, 161 USPQ 826 (1969) (report of result only), sustaining an opposition to the registration of MYOCHOLINE 1 for a medicinal preparation-for treatment of dysphagia, abdominal distention, gastric retention, and urinary retention, filed by appellee, the owner of the prior registration of MYSOLINE 2 for an anti-con-vulsant drug, on the ground that MYOCHOLINE so resembles MY-SOLINE “as to be likely, when applied to the goods of the applicant, to cause confusion, or to cause mistake, or to decieve.” Lanham Act, § 2(d), 15 USC 1052(d). We affirm. .
*950Appellant’s MYOCHOLINE is a pharmaceutical dispensed on prescription in tablet form through the normal drug channels. It was marketed in 1964 and advertised in medical journals, at professional conventions and by direct mail. Opposer’s MYSOLINE is also dispensed only on a prescription basis and may be procured in tablet form as well as in liquid suspension. Sales of MYSOLINE during the period 1957 through 1966 were found by the board to exceed $17,000,-000, and advertising expenditures during the same time period surpassed $1,000,000. Opposer’s drug is a suppressant of convulsions and is indicated for treatment of epilepsy, petit and grand mal, and other psychomotor defects. Appellant’s drug is contraindicated for use by those afflicted with the disorders for which MYSOLINE is prescribed.
Appellant asserts that there is no likelihood of confusion, that the marks are dissimilar in structure, and that when broken down into its constituent parts, each mark can be seen to be suggestive of the nature of the product it identifies. MYSOLINE is a three-syllable word whereas MYOCHOLINE is composed of four syllables. The marks are different in the middle syllables — “O-CHOL” as contrasted with “SOL”. These differences are asserted to be significant as well as audibly and visually distinguishing. Appellant states that “CHOL” connotes the product as a choline derivative thereby indicating the chemical origin of the drug. On the other hand, “SOL” must, it is urged, suggest “solace” — the relaxing quality of opposer’s anticonvul-sant. According to appellant, “MY” and “MYO” mean muscle, while “INE” means “characterized by,” and a number of third-party registrations of marks applied to drugs which utilize these word segments were introduced to show the etymology of the two marks here in issue. It would follow, according to appellant, that MYOCHO-LINE suggests a medicine for treatment of a muscle disorder characterized by derivation from a choline compound whereas MYSOLINE would indicate a drug characterized by affording solace and relaxation to an affected muscle. It appears to us that appellant’s analysis of opposer’s mark yields a definition which would be equally applicable to the drug identified by MYOCHOLINE.
We'regard appellant’s approach as too myopic. It is the entirety of each mark as applied to the respective drug which must be considered. Although appellant does contend that the segment of the public which would have the responsibility of distinguishing between the drugs identified by these marks, i.e., physicians and pharmacists, is a discriminating class; nevertheless, we must still look to the whole of the marks, and we are satisfied that even within this class of persons there is a likelihood of confusion. We agree with the reasoning of *951the board expressed in its unpublished opinion in relevant portion as follows:
The question of likelihood of confusion in this case depends upon a consideration of the marks reviewed in their entireties. The existence of third party registration of similar marks or the listing of such marks in a directory has very little weight on this question. * * * [citation omitted]. Applicant’s own witness, an independent pharmacist, stated that neither the prefix “my” or “myo” nor the suffix “ine” has any meaning to a pharmacist.
SjC i\i S}S $ #
In considering the marks “MYOOHOLINE” and “MYSOLINE,” we find substantial similarities therebetween. These similarities are apparent in appearance and especially in sound. And in our opinion these similarities are such as to be likely to cause confusion or mistake.
Applicant emphaizes the opinion of its witness that it is highly impossible there would ever be any confusion arising from the dispensing of the products of the parties. Much of this opinion is based on matters other than the mark and that pharmacists never make mistakes. Notwithstanding, this witness indicated that “HYCMOINE” and “HYOODAN” were examples “where there.might be a possibility for confusion.” The testimony of this witness is wholly unpersuasive that there would be no likelihood of confusion resulting from the contemporaneous use of “MYOOHOLINE” and “MYSOLINE.”
The fact that confusion, as to prescription drugs could produce harm in contrast to confusion with respect to nonmedicinal products was an additional consideration of the board as is evident from that portion of the opinion in which the board stated:
The products of the parties are medicináis and applicant’s product is contraindicated for the disease for which opposer’s product is indicated. It is apparent that confusion or mistake in filling a prescription for either product could produce harmful effects. Under such circumstances, it is necessary, for obvious reasons, to avoid confusion or mistake in the dispensing of the pharmaceuticals. * * * [citations omitted].
The board’s view that a higher standard be applied to medicinal products finds support in previous decisions of this court, Clifton v. Plough, 52 CCPA 1045, 1047, 341 F.2d 934, 936, 144 USPQ 599, 600 (1965) (“[I]t is necessary, for obvious reasons, to avoid confusion in the dispensing of pharmaceuticals.”); Campbell Products, Inc. v. John Wyeth & Bro., Inc., 31 CCPA 1217, 1220, 143 F. 2d 977, 979,62 USPQ 302, 304 (1944) (“[I]t seems to us that where ethical goods are sold and careless use is dangerous, greater cafe should be taken in the use and registration of trade-marks to assure that no harmful confusion results.”). As Judge Each points out in dissent, this is a doctrine which has been adopted by other Federal courts in which injunction against the use of a mark has been sought on the ground of likelihood of confusion. Judge Rich, however, would distinguish between actions brought to enjoin use and inter partes contests arising in the Patent Office (as welí, presumably, ás ex parte actions arising in the Patent *952Office) wherein it is registration which is sought to be prevented. Because use or nonuse is independent of registration, and since a judgment of this court can only affect registration, Judge Rich concludes that any special harm which might result from confusion of goods in use stemming from similarity in the identifying marks is a consideration irrelevant to any trade-mark proceeding before us. We cannot agree with this view. Instead, we reaffirm our position as accurately presented by the board in this case.
Judge Rich’s theory respecting the distinction between registration and use is at first blush tempting even without the support of prior decisions of this court. It fails, however, when the role of the Patent Office in the statutory registration of trademarks is fully appreciated. By its nature, likelihood of confusion relates to use. Confusion is determined in part by a comparison of marks, but the ultimate question is whether in use there would be a likelihood of confusion as a result of trademark similarity. Congress has seen fit to prohibit registration for this reason, and that legislative action compels the conclusion that some impediment to use was hoped for by denying registration. Section 32(1) (a) of the Lanliam Act, 15 USC 1114 (1) (a), imposes civil liability for the “use in commerce [of] any reproduction, * * * copy, or colorable imitation of a registered mark in connection with the sale, offering for sale, distribution, or advertising of any goods or services on or in connection with which such use is likely to cause confusion, or to cause mistake, or to deceive * * In other words, the initial agency determination of likelihood of confusion is similar to that required to establish infringement. Under such circumstances, the Patent Office conclusion of likelihood of confusion operates as a potent impediment to use — the result intended by Congress. The converse is equally true. A Patent Office finding of no likelihood of confusion is an incentive to use. Indeed, registration on the principal register is prima facie evidence of registrant’s right to use the mark in commerce. Lanham Act, § 7 (b), 15 USC 1057 (b).
In effect, we are saying that the Patent Office functions as a guardian of the public interest in interposing an obstacle to the use of marks which when applied to goods to be consumed by the public would be likely to confuse. This was the thesis advanced by Professor Deren-berg, The Patent Office As Guardian of the Public Interest m TradeMark Registration Proceedings, 14 Law and Contemporary Problems 288 (1949), and he found a specific legislative intent to confer that responsibility in the § 2(d) prohibition against registration of marks which are likely to cause confusion, mistake, or to deceive. Once the transition is made from pure registration considerations to a consideration of the marks in use, it follows immediately that the nature of the goods becomes a relevant factor. The goods here involved being drags, confusion as to which could unquestionably give rise to serious *953consequences, we think due consideration ought properly be given that fact in the determination of likelihood of confusion.
With regard to Judge Bich’s ultimate conclusion that no public interest would be served by denying registration in this case, we have little doubt that our contrary conclusion is colored by our differing views as to the actual likelihood of confusion stemming from contemporaneous use of the marks involved.
The decision of the Trademark Trial and Appeal Board sustaining the opposition and refusing the registration of MYOCHOLINE is affirmed.
Application Serial No. 210,558, filed January 25,1965.
Reg. No. 569,143, registered January 13,1953.