A transformative genetic treatment for a rare, inherited form of blindness will come with a price tag of $425,000 per eye, or $850,000 for both, said Spark Therapeutics Inc., the Philadelphia-based biotechnology company that is bringing the therapy to market.

Since Spark's Luxturna was approved by the U.S. Food and Drug Administration last month, speculation over the price has grown as it became clear the therapy would be one of the first in a wave of medicines that yield remarkable results after a single treatment — and would carry a commensurate cost.

In a novel arrangement, Spark will offer discounts based on whether the drug works initially and remains effective for the estimated 1,000 to 2,000 patients in the United States with a type of inherited retinal disease caused by a mutant gene.

"We believe that this price reflects not only the breakthrough, life-altering value of one-time Luxturna, but it will enable us to continue to invest and build on the revolutionary science that supports not only Luxturna but the rest of our pipeline," chief executive officer Jeff Marrazzo said in a phone interview. Spark shares rose  $2.45,  or 4.6 percent, to close at $55.76 Wednesday.

The company's effectiveness-based discount is sharply different from how most drugs are sold. Health insurers are also used to paying for medicine over the course of a disease or over a patient's life, in the case of some chronic conditions.

A onetime treatment presented a challenge, because the cost would be paid for by one insurer or government, only to have others reap the benefits when the patient changes coverage.

To help mitigate that dynamic, Spark is rolling out several programs to spread out the cost over the years or give rebates to payers if the benefits wane with time.

For example, the company said it was discussing a program with the U.S. Centers for Medicare & Medicaid Services that would spread payments for Luxturna over several years, even though the therapy would be given only once. It didn't say how many installments would be made, or how long it would take to pay the full cost of the drug.

In an agreement with the Boston-area insurer Harvard Pilgrim Health Care, Spark will get the full price of treatment up-front. If patients don't get an immediate benefit, measured at 30 days, or a long-term one, measured at 30 months, Spark will have to give back some of the money in a rebate.

Spark has also proposed selling the gene therapy directly to insurance companies or specialty pharmacies. That would sidestep the current process that requires hospitals or health-care providers to buy expensive therapies up-front. Spark is working with Express Scripts Holding Co. on such an arrangement, and said it's talking with other drug plans.

Express Scripts has been a frequent critic of costly drugs but said the Spark treatment was an exception.

"Many people were anticipating this would be more than a million dollars," said Steve Miller, the St. Louis-based company's chief medical officer. "In the end, this is a revolutionary product, and I think in most plans this will be covered."

Effectiveness-based or outcomes-based pricing agreements are being explored as an alternative to traditional pricing structures, especially for ultra-expensive onetime treatments. In September, for example, Novartis announced that it would charge $475,000 for Kymriah, its onetime, potentially curative T-cell therapy for a childhood leukemia. But the company also said it was working with private insurers and the federal Centers for Medicaid and Medicare to develop a novel payment policy based on whether patients responded within a month of treatment.

Spark's discounted model "wouldn't be the first time some fairly innovative payment structure was attempted in the U.S.," said Gbola Amusa, head of health-care research for the investment bank Chardan.

Amusa said the advent of sticker-shock drugs has raised the philosophical question of what a life is worth. "Spark's drug is different because it isn't life-saving," he said. "What's the price of blindness?"

Ultimately, Spark's biggest challenge may be finding patients to treat.

Of the few thousand people believed to have the disease, only a few have actually been tested and confirmed to have it, since there was no cure and thus little use in diagnosis. Many with more advanced forms of the disease won't qualify for treatment, according to the company.

Staff writer Marie McCullough contributed to this article.