INTRODUCTION
This case is a patent law appeal from a district court order granting summary judgment of invalidity based on anticipation and obviousness. Ortho-McNeil Pharmaceutical (“Ortho”) brought suit against Teva Pharmaceuticals Industries (“Teva”) and Caraeo Pharmaceutical Laboratories (“Caraeo”) alleging infringement of U.S. Reissued Patent 39,221 (“RE'221”), directed to a combination tramadol and acetaminophen composition for use in prescription pain relief. We vacate-in-part, affim-in-part, and remand.
Acetaminophen is a popular non-opioid pain reliever, more commonly known by the brand name Tylenol®. Acetaminophen has a poorly understood mechanism of action, and interacts with other drugs in ways that could not be predicted in 1990. Nonetheless, it was a popular pain reliever and Ortho had a history of experimenting with combining acetaminophen and other drugs.
Tramadol is an atypical weak opioid with an unclear mechanism of action. Trama-dol and methods of making and using it were first described in U.S. Patent No. 3,652,589 to Flick. Flick mentions that tramadol can be combined with other analgesics, and often exhibits synergistic effects. The specification of Flick also includes 23 working examples. The first 22 examples describe different tramadol formulations. Example 23 discloses a combination of tramadol, p-acetamino-phenol, pentobarbital sodium, and ethoxy benzam-ide. In 1990, Tramadol was known to have several serious side effects, and for that reason was commonly administered through gradual dose titration.
Ortho first claimed a combination trama-dol and acetaminophen tablet in 1990, in U.S. Patent No. 5,336,691 (“'691 patent”). The patent described one benefit of the claimed composition as a synergistic effect between tramadol and acetaminophen. Ortho received FDA approval for such a combination tablet and marketed its product as Ultracet®. Ultracet® immediately became the fastest growing prescription pain reliever and enjoyed considerable commercial success. Several generic drug *597companies filed abbreviated new drug applications (“ANDAs”) to market generic versions of Ultracet®. Ortho brought multiple different lawsuits against these companies asserting the '691 patent against the ANDAs.
In pretrial motions, several of the drug companies alleged that the '691 patent was invalid for anticipation and obviousness based on Flick and other references. During the course of litigation, Ortho discovered that the p-acetamino-phenol mentioned in the Flick patent was actually an archaic name for acetaminophen. The ratio of tramadol to p-acetamino-phenol in the four-compound combination of Flick’s example 23 is 1:10, which falls within the scope of some of the claims of the '691 patent. In light of the cited references and the fact that p-acetamino-phenol is acetaminophen, Ortho sought reexamination and reissue of the '691 patent. Ortho submitted to the examiner the relevant prior art, a summary judgment motion submitted by Kali Laboratories, and a letter submitted by Barr Pharmaceuticals arguing anticipation and obviousness of the '691 patent.
Other prior art cited to the examiner during reissue proceedings and to the district court includes a series of German publications based on the WHO Cancer Pain Guidelines (hereinafter “the German references”). Collectively, these articles teach a method of managing cancer pain using a ladder approach aimed at customizing a co-administration of pain relievers and dosing regimens to meet the needs of each patient. The strengths of the medications and the dosages increase as the prescriber moves up the ladder. The middle rung of the ladder includes a non-opioid analgesic co-administered with a weak opioid. None of the German references mentions synergy.
In reissue proceedings for the '691 patent, Ortho canceled all but one of the asserted claims and redrafted them to narrow the scope of the invention and attempt to avoid the cited prior art references. Ortho also argued to the examiner that one of ordinary skill in the art at the time of filing would not understand the prior art to disclose the claimed invention as amended. The examiner ultimately allowed several redrafted claims and the one previously allowed claim to reissue as RE'221.
While the reissue proceedings were pending, Teva and Caraco filed summary judgment motions for invalidity and non-infringement of claim 6 of the '691 patent. Claim 6 was the one original claim allowed in RE'221, amended to independent form. Ortho then amended its complaints to assert the new reissue claims as well as RE'221 claim 6 against Teva and Caraco (among others) in the ongoing ANDA litigation between the parties. Ortho’s cases against Teva and Caraco were then consolidated.
Teva and Caraco filed summary judgment motions for invalidity and non-infringement, asserting that RE'221 was anticipated and rendered obvious by Flick and the German references. Ortho disputed Teva’s and Caraco’s interpretations of the prior art disclosures and submitted uncontroverted expert testimony explaining that one of skill in the art at the time of the original filing would not understand Flick, or the German references to disclose or render obvious the claimed invention.
On August 12, 2008, the district court granted summary judgment invalidating RE'221 claims 43-48, 51, 54, 67, 69, 72, and 74 as obvious over the prior art, and invalidating claim 6 as anticipated and obvious over the prior art.1 Ortho-McNeil timely *598filed its notice of appeal on August 25, 2008. This court has jurisdiction pursuant to 28 U.S.C. § 1295(a)(1).
DISCUSSION
This court reviews a grant of summary judgment de novo. Cross Med. Prods., Inc. v. Medtronic Sofamor Danek, Inc., 424 F.3d 1293, 1302 (Fed.Cir.2005). Obviousness is ultimately a determination of law, though it is based on questions of fact. Id. at 1302. Anticipation is a question of fact. Id. Because issued patents enjoy a presumption of validity, obviousness and anticipation must be proven by clear and convincing evidence. Impax Labs., Inc. v. Aventis Pharma., Inc., 545 F.3d 1312, 1314 (Fed.Cir.2008). When the patent examiner has considered the asserted prior art and basis for the validity challenge during prosecution, the burden of proving invalidity is especially heavy. Id.; Hewlett-Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464, 1467 (Fed.Cir.1990). Summary judgment is only appropriate where there are' no material questions of fact and the mov-ant is entitled to judgment as a matter of law. Celotex Corp. v. Catrett, 477 U.S. 317, 322-23, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986); Fed.R.Civ.P. 56(c).
Inventions in most instances rely upon building blocks long since uncovered, and combine elements that are in some sense already known. KSR Int’l Co. v. Teleflex, Inc., 550 U.S. 398, 418-19, 127 S.Ct. 1727, 167 L.Ed.2d 705 (2007). The combination of familiar elements according to known methods is likely to be obvious, however, when it does no more than yield predictable results. Id. at 1739. Each case must be decided in its particular context, including the characteristics of the science or technology, the nature of the choices available to one skilled in the art, the specificity of the prior art, and the predictability of results in the area of interest. Abbott Labs. v. Sandoz, Inc., 544 F.3d 1341, 1352 (Fed.Cir.2008).
I. Validity of the Reissue Claims
Each of claims 43-48, 51, 54, 67, 69, 72, and 74 recites a pharmaceutical composition with active ingredients consisting of tramadol and acetaminophen in a ratio of about 1:5 to about 1:19. A single tablet containing only tramadol and acetaminophen in a fixed dose ratio within the claimed range is not disclosed in the cited prior art. Flick discloses a four-compound combination comprising tramadol, acetaminophen, and two other active ingredients. Flick does not suggest that the pentobarbital sodium and ethoxy benzamide are merely optional for the combination to work as desired. The parties dispute whether Flick suggests that the composition in example 23 exhibits a synergistic effect.
Ortho submitted uncontroverted expert testimony that one of ordinary skill in the art would not find it obvious to try to remove two of the four active ingredients disclosed in Flick example 23 to arrive at the claimed composition. The expert explained that drug interactions are complicated and unpredictable. He testified that the interactions of tramadol and acetaminophen were especially poorly understood in 1990. The expert testified that one of skill in the art would not expect the combination of solely tramadol and acetaminophen to have any particular advantage unless Flick or some other reference specifically identified one. The ex*599pert testified that Flick’s broad statement that tramadol “often” displays synergistic affects when combined with other analgesics would not be enough give one of ordinary skill any expectations whether tra-madol combined only with acetaminophen in a 1:5 to 1:19 ratio would exhibit the synergistic effects discovered by Ortho.
Ortho’s expert also testified that no combination of the German references and Flick would suggest the claimed combination of tramadol and acetaminophen. The expert explained that the German references actually teach away from the claimed composition because they emphasize the importance of flexibility in choosing combinations and doses of medications based on individual needs. The expert testified that the claimed fixed-dosage combination tablet was disparaged in the German references and disfavored in the art at the time.
Ortho’s proffered reading of Flick and the German references, as well as expert testimony regarding the understanding of one skilled in the art, raises material questions of fact as to whether a skilled artisan would have found the claimed combination of tramadol and acetaminophen to be obvious. We therefore vacate the district court’s summary judgment invalidating RE'221 claims 43-48, 51, 54, 67, 69, 72, and 74 based on obviousness and remand the case for a trial on the merits.
II. Validity of Claim 6
Claim 6 recites a composition comprising tramadol and acetaminophen in an about 1:5 ratio. Unlike the other reissue claims, the open transitional term “comprising” causes claim 6 to read on compositions that include active ingredients other than tra-madol and acetaminophen. Flick’s example 23 thus differs from claim 6 only by the disclosed ratio of tramadol to acetaminophen (1:10 versus “about 1:5”). This court previously construed the claimed “about 1:5” ratio to read on a ratio from 1:3.6 to 1:7.1. Ortho-McNeil Pharm., Inc. v. Caraca Pharm. Labs., Ltd., 476 F.3d 1321, 1328 (Fed.Cir.2007).
The district court found that claim 6 was invalid for both anticipation and obviousness. We discuss each judgment in turn.
A
With respect to anticipation, Ortho contends that each element of claim 6 is not disclosed in Flick, and therefore is not anticipated. It is not disputed that example 23 of Flick discloses a pharmaceutical composition containing tramadol and acetaminophen, the first two limitations of claim 6. It is disputed whether example 6 discloses the claimed about 1:5 ratio, as that term is construed by this court.
The defendants argued to the district court that a paragraph at the end of example 22 teaches that tramadol can be administered in a 25 mg dose and that a preferred dose is 50 mg. Example 23 discloses 25 mg of tramadol in a four active agent combination. The defendants argue that these two teachings together suggest that the 25 mg dose of tramadol disclosed in example 23 can be modified to a 50 mg dose.
Ortho argued that example 22 discloses a tramadol-only formulation, and the suggestion to administer a 50 mg dose only applies to the administration of tramadol alone. Ortho argues that the specification does not teach increasing the dosage of tramadol past 25 mg when combining it with other active analgesics.
The district court agreed with the defendant’s interpretation of Flick and found that examples 22 and 23 anticipate claim 6 when read in combination. The district court improperly resolved disputed questions of fact in reaching this conclusion. *600What a reference discloses is a question of fact. Para-Ordnance Mfg., Inc. v. SGS Imps. Int’l, Inc., 73 F.3d 1085, 1088 (Fed.Cir.1995). The parties dispute whether and how Flick’s teachings about varying dosages apply to example 23. The parties also dispute whether one of skill in the art would have understood the teachings of one working example to apply to a different working example. Whether Flick discloses all of the limitations of claim 6 is a material question of fact over which Ortho has raised a genuine dispute. We therefore vacate the district court’s summary judgment that claim 6 is invalid for anticipation.
B
With respect to obviousness, there is no dispute that the only difference between claim 6 and example 23 of Flick is the weight ratio — the additional compounds in the example are irrelevant because of the transitional phrase “comprising” in the claim. Example 23 discloses a 1:10 ratio of tramadol to acetaminophen: 25 mg to 250 mg. Because the difference between 1:7.1 and 1:10 is so slight, Flick creates a prima facie case of obviousness with regard to claim 6. See Titanium Metals Corp. of Am. v. Banner, 778 F.2d 775, 782-83 (Fed.Cir.1985) (“The proportions are so close that prima facie one skilled in the art would have expected them to have the same properties.”); Haynes Int’l, Inc. v. Jessop Steel Co., 8 F.3d 1573, 1577 n. 3 (Fed.Cir.1993) (“[Wjhen the difference between the claimed invention and the prior art is the range or value of a particular variable, then a prima facie rejection is properly established when the difference in range or value is minor.”).
Ortho-McNeil can rebut the prima facie case if it can show that the prior art teaches away from the claimed range, or the claimed range produces new and unexpected results over the prior art range. See Iron Grip Barbell Co. v. USA Sports, Inc., 392 F.3d 1317, 1322 (Fed.Cir.2004). Before the United States Patent and Trademark Office, the district court, and now us, Ortho-McNeil principally relies on unexpected results in the form of synergy between tramadol and acetaminophen. As described in RE'221, the observed analgesic effect of tramadol-acetaminophen combinations exceeds what one would predict from merely adding the effects observed upon administering each drug individually. See, e.g., RE'221 fig. 1. Ortho-McNeil’s problem, however, is that its own evidence indicates no perceptible difference in synergy between a weight ratio of 1:7.1 and 1:10 — and indeed over a much broader range of ratios. Id. Ortho-McNeil does not dispute this. Thus, Ortho-McNeil has failed to create a material dispute of fact regarding whether “the claimed range achieves unexpected results relative to the prior art range.” In re Woodruff, 919 F.2d 1575, 1578 (Fed.Cir.1990) (emphasis added); cf. Ormco Corp. v. Align Tech., Inc., 463 F.3d 1299, 1311-12 (Fed.Cir.2006) (“Evidence of commercial success, or other secondary considerations, is only significant if there is a nexus between the claimed invention and the commercial success.”).
Ortho-McNeil also relies on the declaration of Dr. Stanski, which asserts that Flick teaches away from a ratio of about 1:5.2 Dr. Stanski averred that one of ordinary skill would not have increased the amount of tramadol in Flick’s example 23 *601because of concerns about the side effects of tramadol. However, Flick discloses higher doses than the 25 mg of example 23, including “preferably” tramadol-only formulations in doses of 50 mg and 75 mg, with no reservations about safety. See RE'221 col. 12 11.24-31. Setting that aside, and accepting Dr. Stanski’s statement as true, it is nonetheless insufficient to rebut the prima facie case. The question is not whether Flick teaches away from increasing the amount of tramadol; the question is whether Flick teaches away from lowering the ratio of tramadol to acetaminophen from 1:10 to 1:7.1. This could be accomplished by increasing the amount of trama-dol, which Dr. Stanski disparages, but it could equally be accomplished by decreasing the amount of acetaminophen. Dr. Stanski did not opine on that option, and therefore did not create a material issue of fact regarding whether the prior art teaches away from the claimed invention. We have considered Ortho’s other arguments and find them to be without merit.
For these reasons, we agree with the district court that Ortho has failed to raise a genuine dispute of material fact regarding the obviousness of claim 6.
CONCLUSION
We vacate the district court’s summary judgment order invalidating RE'221 claims 43-48, 51, 54, 67, 69, 72, and 74 based on obvious. We remand the case to the district court for proceedings consistent with this opinion. We vacate the district court’s summary judgment invalidating RE'221 claim 6 based on anticipation. We affirm the district court’s summary judgment invalidating claim 6 based on obviousness.
COSTS
Each party shall bear its own costs.
Opinion for the court filed by Circuit Judge PROST. Dissenting opinion filed by Circuit Judge MAYER.
. The district court identified the asserted claims as 43-47, 51, 67, and 69, although *598both parties identified the asserted claims as 43-48, 51, 54, 67, 69, 72, and 74. We will assume that the district court intended to grant the defendants' summary judgment motion with respect to all of the asserted claims and not solely the ones listed by the court in its opinion.
. We confine our analysis to Dr. Stanski’s August 25, 2005 declaration, as the supplemental expert testimony (submitted by Ortho when it was arguing the validity of the reissue claims) was not before the district court when determined that no material factual dispute existed regarding the obviousness of claim 6.