BRORBY, Circuit Judge.

This diversity case presents the questions of whether Oklahoma either has or would adopt the tort theories of alternative liability or market share liability in the context of diethylstilbestrol (DES) litigation. We conclude Oklahoma has not adopted either theory and would not apply either theory to the facts of this case. We therefore affirm the judgment of the district court.

BACKGROUND

Procedurally, this case comes to us as a result of the district court granting a summary judgment in favor of three defendants who are drug manufacturing companies. A fourth defendant, another drug manufacturer, was not dismissed under the summary judgment order and is not a party to this appeal.

The record on appeal reveals, after an examination that gives plaintiffs the benefit of all reasonable inferences, the relevant facts are simple and undisputed. The plaintiffs, Mrs. and Mr. Wood, are parents of a son whose premature birth and resulting permanent injury to his digestive system costs them significant medical expense. The Woods contend their losses were ultimately caused by exposure to a synthetic estrogen, DES. In 1959, while the son's grandmother was pregnant with his mother, the grandmother was prescribed and took DES. The grandmother purchased this medication from a military pharmacy in Florida. Twenty-nine years later, the mother gave birth to her son prematurely. The Woods allege injuries to their son were the result of the mother's in utero exposure to DES while she was carried by the grandmother.

The Woods originally named twenty-seven drug manufacturer defendants and claimed all were liable under market share nonidentification theory. By November 1991, when the Woods filed their third amended complaint, twenty-four of the original twenty-seven defendants had been dismissed, although the Woods added another defendant during the interim. The four remaining defendants filed motions for summary judgment based on the inability of plaintiffs to identify the specific manufacturer of the DES ingested by the grandmother. The Woods moved to amend to assert the additional nonidentification theory of alternative liability, and their motion was denied. Nonetheless, the Woods asserted alternative liability in their opposition to defendants' summary judgment motions.

The district court granted summary judgment to three of the drug manufacturers, now appellees, on the grounds that the Woods failed to establish causation under established Oklahoma law and that Oklahoma would not recognize either alternative liability or market share liability theories. Order, No. CIV-90-1307-A, 1992 WL 687292, Slip Op. at 3-5 (W.D.Okla. June 22, 1992). As to the fourth defendant, the district court found the Woods had presented sufficient evidence to meet the current Oklahoma standard for causation and therefore presented a triable claim. Slip Op. at 7-8. The plaintiffs' case against this fourth defendant was subsequently stayed due to the filing of a bankruptcy petition.

An order granting a motion for summary judgment receives de novo review from a court of appeals using the same legal standard employed by the district court. See Thrasher v. B B Chem. Co., 2 F.3d 995, 996 (10th Cir. 1993). We must determine whether there is a genuine issue of material fact and whether the movant is entitled to judgment as a matter of law. A federal court sitting in diversity must apply the law of the forum state, in this case Oklahoma, and thus must ascertain and apply Oklahoma law with the objective that the result obtained in the federal court should be the result that would be reached in an Oklahoma court. See Allen v. Minnstar, Inc., 8 F.3d 1470, 1476 (10th Cir. 1993). If a federal court cannot ascertain the law of the forum state, we must in essence sit as a state court and predict how the highest state court would rule. See Adams-Arapahoe Sch. Dist. No. 28-J v. GAF Corp., 959 F.2d 868, 870 (10th Cir. 1992). A federal district court's state-law determinations are entitled to no deference and are reviewed de novo. Salve Regina College v. Russell, 499 U.S. 225, 111 S.Ct. 1217, 113 L.Ed.2d 190 (1991).

DISCUSSION

The Woods do not appeal the district court's finding that they failed to make a sufficient showing of causation under common principles of Oklahoma tort law. To prevail against a manufacturer in a products liability action, a plaintiff "must prove that the product was the cause of the injury; the mere possibility that it might have caused the injury is not enough." Kirkland v. General Motors Corp., 521 P.2d 1353, 1363 (Okla. 1974). We have interpreted the current Oklahoma causation standard in a products liability case to require a "significant probability" that defendant's acts are related to plaintiff's injury. See Dillon v. Fibreboard Corp., 919 F.2d 1488, 1491 (10th Cir. 1990) (interpreting Case v. Fibreboard Corp., 743 P.2d 1062, 1067 (Okla. 1987)); see also Blair v. Eagle-Picher Indus., 962 F.2d 1492, 1496 (10th Cir.) (same standard under Case and Dillon), cert. denied, ___ U.S. ___, 113 S.Ct. 464, 121 L.Ed.2d 372 (1992). Accepting the district court's conclusion that plaintiffs failed to establish causation, Slip Op. at 3, we do not further address an unappealed issue. Therefore, the Woods' case hinges on whether Oklahoma has or would adopt another tort theory that relieves plaintiffs of their burden of establishing causation.

I [10] Alternative Liability Doctrine

The Woods first challenge the district court's refusal to extend the alternative liability doctrine to their case. Although the district court could have ignored the Woods' argument as beyond the scope of their third amended complaint, the court concluded Oklahoma has yet to extend the doctrine to products liability actions and would not do so here. We agree.

Alternative liability first attracted attention in the California Supreme Court case of Summers v. Tice, 33 Cal.2d 80, 199 P.2d 1 (1948). Under this theory, when more than one negligent defendant cannot be identified as the specific source for an injury, a plaintiff need not prove causation as to each individual defendant, but need only prove that the one who caused the injury is among the group of negligent defendants. The court will presume each individual defendant caused the whole injury unless the defendant can prove otherwise. "The Summers burden shift seems intuitively fair where all those who could have caused the injury are before the court and the odds are equal and significant that each is liable and to the same degree." Menne v. Celotex Corp., 861 F.2d 1453, 1466 (10th Cir. 1988). Oklahoma adopted the theory of Summers in the dogbite case of Hood v. Hagler, 606 P.2d 548 (Okla. 1979), but has not recognized the theory in a products liability context.

Specifically, the Oklahoma Supreme Court rejected alternative liability, and other nonidentification theories, in Case v. Fibreboard Corp., 743 P.2d 1062. Having examined the decisions of other state supreme courts that reject alternative liability in DES litigation, the Oklahoma Supreme Court found:

The applicability of alternative liability theory, concert of action theory and enterprise liability theory were all considered in connection with the DES cases and rejected as inapplicable. The factors which rendered these theories inapplicable in the field of DES litigation are clearly present in the area of asbestos litigation and also render the theories inapplicable.

743 P.2d at 1067 (emphasis supplied) (footnote omitted). Although Case factually dealt with asbestos litigation, we read this holding as also rejecting alternative liability in DES litigation. Despite Hood, we must apply the most recent statement of state law by the state's highest court. Grain Dealers Mut. Ins. Co. v. Lower, 979 F.2d 1411, 1416 (10th Cir. 1992) (quoting Southwest Forest Indus., Inc. v. Sutton, 868 F.2d 352, 354 (10th Cir. 1989), cert. denied, 494 U.S. 1017, 110 S.Ct. 1320, 108 L.Ed.2d 496 (1990)). Oklahoma would not apply alternative liability to the Woods' claim.

II [15] Market Share Liability

The Woods next argue Oklahoma law would recognize a market share theory in DES cases. Under this theory, causation and damages are apportioned to defendants based on the percentage of the product sold by each defendant within the entire production of the product. See Sindell v. Abbott Lab., 26 Cal.3d 588, 163 Cal.Rptr. 132, 607 P.2d 924, 937, cert. denied, 449 U.S. 912, 101 S.Ct. 286, 66 L.Ed.2d 140 (1980); Martin v. Abbott Lab., 102 Wn.2d 581, 689 P.2d 368, 380 (1984). A plaintiff must meet a threshold requirement of joining a number of defendants to represent a "substantial share" of the market before the burden of disproving causation shifts to the defendants. Sindell, 607 P.2d at 937. We agree with the district court. Oklahoma would not adopt a market share theory here.

Again, we believe Case to be dispositive. Although the specific holding of Case rejects market share theory in asbestos litigation, the Oklahoma Supreme Court made several strong statements of policy suggesting market share theory would not be welcome in products liability litigation in general.

[T]he public policy favoring recovery on the part of an innocent plaintiff does not justify the abrogation of the rights of a potential defendant to have a causative link proven between that defendant's specific tortious acts and the plaintiff's injuries where there is a lack of circumstances which would insure that there was a significant probability that those acts were related to the injury. We therefore would not recognize the applicability of a market share theory of liability to asbestos injury litigation under Oklahoma law.

Case, 743 P.2d at 1067. The Oklahoma Supreme Court further opined "[t]he creation of a program of compensation for victims of asbestos related injuries as a matter of policy is a matter for the legislative body and not for the courts." Id. (emphasis in original).

The Oklahoma Supreme Court's concern for lack of a clear-cut connection between cause and effect in asbestos litigation further supports our rejection of market share theory in the case at hand. One factor for rejecting market share liability is the presence of an identified tort-feasor. Id. at 1065. Although the Woods may not be pleased that the remaining defendant is in bankruptcy, they have located a triable defendant. Second, there is a less obvious link between the injury and the risk to which the plaintiff was exposed, than in the original rounds of DES litigation. Market share theory relies on a product whose use carries a "singular risk factor" where the injury is known to "arise at a specific time which can be related to a known ingestion of a dangerous substance." Id. at 1066. In the initial DES cases, the risk of cervical cancer to plaintiffs who were exposed in utero to DES was clear cut. Id. Here, the Woods allege the exposure to DES caused malformed reproductive organs in Mrs. Wood which led to the premature birth of her son causing permanent damage to his digestive tract. Although certainly a plausible scenario, there is less certainty that a known ingestion of DES presents a consistent singular risk of a digestive tract injury to the grandchild of person ingesting the drug. Based upon this reasoning we conclude the Oklahoma Courts would not recognize the applicability of a market share theory of liability to this type of DES injury.

CONCLUSION

The law in Oklahoma requires the plaintiff to prove cause as an essential element of a negligence action. We do not believe the Oklahoma courts would allow the abrogation of this element under the facts of this case. We therefore conclude Oklahoma would apply neither the alternative liability doctrine nor the market share theory to this case. We see no reason to certify either question to the Supreme Court of Oklahoma given the clear pronouncements of the Oklahoma Supreme Court in Case.

The judgment of the district court is AFFIRMED.