Grace Rodela Fuja, a thirty-seven year-old woman suffering from breast cancer, sued her insurer, the defendant, Benefit Trust Life Insurance Company ("Benefit Trust"), for its refusal to pay for a "high-dose chemotherapy treatment with autologous bone marrow transplantation" ("HDC/ABMT"). Fuja sought injunctive and declaratory relief to prevent the defendant from denying coverage for the treatment under the Employee Retirement Income Security Act ("ERISA"), 29 U.S.C. Sec. 1132(a)(1)(B) and (a)(3). The court entered judgment for Fuja enjoining the defendant from denying coverage. The defendant Benefit Trust Life Insurance Company, subsequently paid for the treatment but now appeals the judgment. We reverse.1FACTS

In August, 1989, Fuja was diagnosed with breast cancer and underwent a lumpectomy and a modified radical mastectomy. Immediately thereafter, beginning in September 1989, she received six months of standard-dose chemotherapy treatment and seemed to remain in remission until February of 1992 when her oncologist observed that the cancer had spread to both her lungs. From February through December of 1992, Fuja responded to standard-dose chemotherapy but during this period her physician came to the conclusion that continued treatment with such standard dose chemotherapy offered her but a negligible chance of survival. Thus she prescribed a regimen of HDC/ABMT.

HDC/ABMT is a two-step procedure. Physicians first extract ("harvest") the bone marrow cells from the patient's body and place them temporarily in frozen storage. At this time, the patient undergoes a cycle of high-dose chemotherapy in hopes of killing the cancer cells. Because the high-dose chemotherapy also attacks the bone marrow cells, it is necessary to withdraw some of the bone marrow prior to undergoing the high-dose chemotherapy. Without initially removing a portion of the bone marrow cells, the high-dose chemotherapy would be lethal to the patient because of its myeloblative effect (it destroys bone marrow cells which produce blood cells (red and white) as well as platelets) rendering the patient highly susceptible to infection. After completing the administration of the high-dose chemotherapy, the patient's own ("autologous") stored marrow is reinfused intravenously into the bloodstream to relieve the patient from the toxic effects of the chemotherapy. HDC/ABMT in the past has proven effective in treating certain cancerous blood diseases such as leukemia and Hodgkin's disease but is not as yet universally accepted treatment for solid-type tumors including breast cancer.

Benefit Trust refused to cover the treatment because it did not fall within the parameters of procedures that are "medically necessary" as defined in the insurance contract. Fuja brought this suit in federal district court seeking to enjoin Benefit Trust from denying coverage. Following a hearing on December 17, 1992, the court issued a decision ordering the insurer to pay for the treatment, Fuja underwent the treatment in January of 1993. Unfortunately, the treatment was unsuccessful and she expired in April of 1993.

The defendant, Benefit Trust, maintains that the district court erred in ruling that the insurer was liable for coverage of the treatment arguing: (1) that the treatment was provided in connection with medical research, and (2) the treatment is not authorized for reimbursement by the Health Care Financing Administration ("HCFA").

We note at the outset that cases of this nature pose most difficult policy questions of who should bear the burden of paying for expensive medical treatments that are at the time of treatment of unknown efficacy. Although we fully realize the heartache Mrs. Fuja's family has endured, as judges we are called upon to resolve the legal question presented in this appeal, i.e., interpreting the Benefit Trust insurance contract.2Standard of Review

Under Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115, 109 S. Ct. 948, 956, 103 L. Ed. 2d 80 (1989), this court reviews Benefit Trust's denial of benefits de novo because the insurance plan in question does not grant Benefit Trust discretion to determine eligibility for benefits or construe terms of the insurance agreement. Under Federal Rule of Civil Procedure 52 the district court's findings of fact will not be set aside unless they are clearly erroneous. Rennie v. Dalton, 3 F.3d 1100, 1106 (7th Cir.1993), cert. denied, --- U.S. ----, 114 S. Ct. 1054, 127 L. Ed. 2d 375 (1994). Because contract interpretation is a question of law, we review the trial court's finding that the contract was ambiguous de novo. Central States Pension Fund v. Hartlage Truck Serv., 991 F.2d 1357, 1361 (7th Cir.1993). The parties have argued the question of who bears the burden of proving Fuja's entitlement to the insurance coverage. The trial court determined that because the "medically necessary" provision of the insurance contract is set forth in the contract benefits section, as opposed to the "exclusions" section, that Fuja bore the burden of establishing her entitlement to the insurance benefits. We agree with the trial court's determination on this issue. See Farley v. Benefit Trust Life Ins. Co., 979 F.2d 653, 658 (8th Cir.1992) (holding that the claimant had the burden of proof when seeking coverage under the same Benefit Trust policy at issue before us). Thus Fuja must demonstrate that the treatment was medically necessary.

To satisfy the plaintiff-appellee's burden and obtain coverage for the HDC/ABMT treatment, Fuja must demonstrate that the treatment is "medically necessary" under all five criteria enumerated in the insurance contract entered into between Benefit Trust and Fuja's employer, Emsco Management Services on behalf of their employees. The contract itself defines medically necessary as:

"required and appropriate for care of the Sickness or the Injury; and that are given in accordance with generally accepted principles of medical practice in the U.S. at the time furnished; and that are approved for reimbursement by the Health Care Financing Administration; and that are not deemed to be experimental, educational or investigational in nature by any appropriate technological assessment body established by any state or federal government; and that are not furnished in connection with medical or other research."

(Emphasis added). In this appeal, Benefit Trust is challenging the trial court's ruling on only two of the provisions: (1) whether the treatment is in connection with medical or other research and (2) whether it is approved for reimbursement by HCFA. If Benefit Trust establishes that the treatment is provided in connection with medical research or that it is not approved for reimbursement by HCFA we must reverse the trial court because Fuja has failed to carry her burden of establishing that the treatment is medically necessary under all five criteria. See supra (citing Farley, 979 F.2d at 658).

The district court determined that the phrase "in connection with medical or other research" was ambiguous and thus proceeded to interpret the clause "in an ordinary and popular sense as would a [person] of average intelligence and experience." Hammond v. Fidelity & Guaranty Life Ins. Co., 965 F.2d 428, 430 (7th Cir.1992) (quoting Allstate Ins. Co. v. Ellison, 757 F.2d 1042, 1044 (9th Cir.1985)). The court stated that it was "strictly constru[ing] [the ambiguous term] in favor of the insured," id. at 430, and determined that the phrase "in connection with medical or other research" implies "that the inherent nature of the treatment itself, and not the results of the treatment, must be part and parcel ('in connection with') of a medical research endeavor." Fuja v. Benefit Trust Life Ins. Co., 809 F. Supp. 1333, 1341 (N.D.Ill.1992). The trial court focused on the "inherent nature of the treatment" to preclude the insurer from denying coverage for any treatment that was part of a medical research study where the results of the treatment were being published as research findings. For reasons unexplained in the decision, the court expressed concern that Benefit Trust intended to use the phrase "in connection with medical treatment" "to exclude coverage for a commonplace medical treatment." Id. The court added that:

"Thus, even though Fuja signed an informed consent advising her that consent to treatment constitutes consent to participate in a research study, and even though Dr. Williams testified that the treatment will occur under a research protocol and that she may publish the results of the treatment as part of a research study assessing the efficacy of HDC/ABMT, Fuja has shown the inherent nature of the treatment itself is not part and parcel of a medical research endeavor. HDC/ABMT is a medically accepted, urgent and necessary treatment for Fuja's illness. Under the court's interpretation of 'in connection with medical or other research,' Fuja has satisfied the Fifth and final requirement of Benefit Trust's 'medically necessary' provision."

Id. at 1341-42.

The threshold question we are called upon to interpret is whether the phrase "in connection with medical or other research" is ambiguous. As stated above, because contract interpretation is a question of law, we review the issue of whether the phrase is ambiguous de novo. Central States Pension Fund v. Hartlage Truck Serv., 991 F.2d 1357, 1361 (7th Cir.1993). The district court found the phrase ambiguous and hypothesized that Benefit Trust might try to use this phrase to deny coverage for any medical procedure that was merely being reported in a study. First of all, we reject the trial court's unfavorable assessment of Benefit Trust's intentions in this case as the court has not delineated nor have we been able to find any evidence in this record that Benefit Trust exploited or expanded the use of the "in connection with medical or other research" contract term. In fact, the evidence is to the contrary as Benefit Trust had paid for all of Mrs. Fuja's standard cancer treatments prior to the HDC/ABMT.3 Moreover, we are of the opinion that to adopt the district court's construction of the term "in connection with medical or other research" would contravene the explicit language of the contract by requiring the insurer to pay for a treatment whose medical efficacy is questioned and thus still under investigation in medically recognized and accepted research studies. See Awbrey v. Pennzoil Co., 961 F.2d 928, 930-31 (10th Cir.1992) ("[a] court is without authority to alter or amend contract terms and provisions absent an ambiguity in the contract * * * [w]e will not read into the [contract] a requirement that, by its clear and unambiguous language, is absent"); Senn v. United Dominion Indust., Inc., 951 F.2d 806, 818 (7th Cir.) ("we are not permitted to allow our sympathies and desires to vitiate clear principles of contract and labor law, and in particular, we refuse to amend the clear terms of the health and welfare benefits contained in the [agreement]"), rehearing en banc denied, 962 F.2d 655 (7th Cir.1992), cert. denied, --- U.S. ----, 113 S. Ct. 2992, 125 L. Ed. 2d 687 (1993); Heller v. Equitable Life Assur. Soc., 833 F.2d 1253, 1257 (7th Cir.1987) ("In the absence of a clear, unequivocal and specific contractual requirement [placing a duty on a party,] we refuse to order the same.4 To hold otherwise and to impose such a requirement would, in effect, enlarge the terms of the policy beyond those clearly defined in the policy agreed to by the parties."). Like the Eighth Circuit in Farley, which construed the very same Benefit Trust contract language, we are convinced that the phrase "in connection with medical research" is unambiguous, see Farley, 979 F.2d at 661 ("[n]or do we consider the language of any of the criteria so vague as to be ambiguous or unenforceable"), and we refuse to "artificially create ambiguity where none exists." Hammond, 965 F.2d at 430 (citation omitted). This clause in the contract clearly excludes coverage for treatment that is of uncertain medical efficacy and subject to ongoing, recognized and accepted medical research procedures. Certainly it is conceivable that an insurer might attempt to exploit this phrase and deny coverage for a treatment whose results were merely being tabulated (as opposed to being collected to assess the very efficacy of the treatment itself) but there is no evidence in this record that Benefit Trust is guilty of exploiting the contract term "in connection with medical research." The uncontradicted expert testimony and documentary evidence presented to the trial court establishes that HDC/ABMT is still of uncertain medical value and is presently being researched at a number of leading medical colleges and oncology research centers throughout the country.

The evidence presented in the trial court convinces us that the treatment in question was provided "in connection with medical or other research." The most significant evidence is the testimony of the plaintiff's own expert, Dr. Stephanie Williams, who performed the procedure on Mrs. Fuja. In her deposition, Dr. Williams testified that a research protocol is the written document defining the objectives and methodology of the proposed study and must be approved by an Institutional Review Board. Dr. Williams further testified that the treatment would be provided under a protocol approved by the Institutional Review Board at the University of Chicago Medical Center. She testified that Mrs. Fuja was informed that "her treatment [would] be furnished in connection with medical research," and explained that the treatment was part of a "clinical trial" that involved human subject research and that such clinical trials are investigative. All clinical cancer trials occur in four phases (Phase I, Phase II, Phase III, and Phase IV), that are classified by the research objective and methodology. In Phase I, the researchers attempt to determine the patients' maximum tolerated dose of an agent (drug). In Phase II trials, the researcher is trying to assess the efficacy of a certain agent or combination of agents by analyzing the response of a statistically reliable number of subjects. Phase III involves randomized clinical trials in which some patients receive the experimental treatment (HDC/ABMT in this case) and others receive the conventional, nonexperimental treatment (standard chemotherapy). The responses of the two groups of subjects are documented, analyzed and compared to assess the efficacy of the experimental treatment. Phase IV occurs after the drug is approved by the Food and Drug Administration ("FDA") and seeks to determine if the drug is effective in other settings.

The FDA and the Department of Health and Human Services have promulgated, as well they should, strict requirements dealing with human subject research. Before such research may be performed the protocol must be approved by an Institutional Review Board and must be clearly explained to the plaintiff who is required to sign an informed consent. The informed consent includes advising the patient of the possible risks, any alternative treatments available and above all that the treatment is part of a research study. See 21 C.F.R. Parts 50 and 56; 45 C.F.R. Part 46.

The research protocol in which Mrs. Fuja participated was entitled "A Phase II Study Mitoxantrone, Vincristine, 5-Fluorouracil with Leucovorin (MVF) Chemotherapy Followed by Cyclophosphamide and Thiotepa High Dose Intensification Therapy in Breast Cancer." The Introduction to the protocol stated "Our studies have been one of the first in this area of investigation...." The protocol made clear that because of the "excessive" treatment related mortality rate under the previous protocol,5 the present protocol was being undertaken. As part of the protocol, the subjects signed an informed consent form labeled "Consent by Subject for Participation in Research Protocol." The form repeatedly stated that the subject was part of a "research project", "research study" or "research protocol" and that the "safety and efficacy of this two-step approach in the treatment of breast cancer will be evaluated." The informed consent also set forth that "all experimental procedures have been identified and no guarantee has been given about the possible results." Finally, each subject attested, "I, the undersigned, hereby consent to participate as a subject in the above described research project conducted at the University of Chicago Medical Center."

As is evident from the documents referred to above and the testimony of plaintiff's expert, Dr. Williams, the treatment Mrs. Fuja received was part of a Phase II clinical research study. In order to receive the treatment the patient had to sign an informed consent that clearly identified the treatment as "research" and "experimental." Moreover, the research protocol under which the treatment was provided stated that the University of Chicago Medical Center "studies have been one of the first in this area of investigation...." Finally, Dr. Williams testified that Mrs. Fuja was informed that her treatment [would] be furnished in connection with medical research." We are of the opinion that the contract term we are called upon to interpret is clear, definite and unambiguous and that the evidence is overwhelming and uncontradicted that Mrs. Fuja's HDC/ABMT was to be provided "in connection with medical or other research." Thus we hold that the district court committed error in ordering Benefit Trust Life Insurance Company to cover a procedure that the contract clearly exempts.

We point out that our conclusion is the same as that of one district court called upon to decide the identical question in Reger v. Espy, 836 F. Supp. 869 (N.D.Ga.1993). The court found:

"In late 1990 the Blue Cross Association in collaboration with the National Cancer Institute ('NCI'), developed and announced a project (the 'Demonstration Project') to support and complete four large, multicenter, randomized phase III trials testing HDC-ABMT in comparison to standard therapy in breast cancer. At this time OPM permits the Plan to make monetary contributions to the Demonstration Project.... Phase III trials are designed to compare the medical results of the experimental therapy to a standard, accepted therapy. The participants in a phase III trial are randomly selected to receive either the standard or experimental therapy.... Because physicians may not ethically allow a patient to receive a known inferior therapy, phase III clinical trials can be performed ethically only if it is unknown whether the experimental or standard treatment is superior.... To date no conclusions have been formed as to whether HDC-ABMT is better than, worse than, or equal to conventional therapy in terms of disease-free or overall survival."

Having concluded that Fuja has failed to satisfy her burden of proof concerning the "in connection with medical or other research" we need not address whether the treatment has been "approved for reimbursement by HCFA."

As stated above, cases of this nature pose troubling social as well as ethical questions that go well beyond the legal issues. As a court of law we are empowered to decide legal issues presented by specific cases or controversies. The greater social questions must be decided by the political branches of government which can engage in "legislative fact-finding" and "benefit from public hearings and constituent expression of opinion." Wangen v. Ford Motor Co., 97 Wis. 2d 260, 324, 294 N.W.2d 437, 469 (1980) (Coffey, J., dissenting); see also Welsh v. Boy Scouts of America, 993 F.2d 1267, 1270-71 (7th Cir.) ("[w]e must refuse[ ] to infringe on the legislative prerogative of enacting statutes to implement public policy.... The problems of public policy ... are for the legislature....") (quoting Wangen, 97 Wis. 2d at 324, 294 N.W.2d at 469) (Coffey, J., dissenting)), cert. denied, --- U.S. ----, 114 S. Ct. 602, 126 L. Ed. 2d 567 (1993). Chesterfield Smith, the former president of the American Bar Association once stated in a Law Day address: "courts are being asked today to solve problems for which they are not institutionally equipped.... The American public perceives the courts as a jack-of-all trades available to furnish the answer to whatever may trouble them." Wangen, 97 Wis. 2d at 323, 294 N.W.2d at 469 (Coffey, J., dissenting). The question of what procedures insurance companies should cover is just the type of problem to which Mr. Smith was referring.

In order to resolve the question of whether health insurance providers should cover treatments like HDC/ABMT, the prudent course of action might be to establish some sort of regional cooperative committees comprised of oncologists, internists, surgeons, experts in medical ethics, medical school administrators, economists, representatives of the insurance industry, patient advocates and politicians. Through such a collective task force perhaps some consensus might be reached concerning the definition of experimental procedures, as well as agreement on the procedures, which are so cost prohibitive that requiring insurers to cover them might result in the collapse of the healthcare industry. While such a committee would in no way be a panacea for our skyrocketing health care costs, it may help to reduce the incidence of suits in which one "expert" testifies that a procedure is experimental and another equally qualified "expert" testifies to the opposite effect. This so called battle of the experts occurs all too frequently in federal court.

Under the present state of the law, we are bound to interpret the language of the specific contract before us and cannot amend or expand the coverage contained therein. See supra at 1409 (citing Awbrey, 961 F.2d at 930-31; Senn, 951 F.2d at 818; Heller, 833 F.2d at 1257). The Benefit Trust insurance contract is unambiguous in that it denies coverage for treatment provided "in connection with medical or other research." Accordingly, the judgment of the district court is

REVERSED.