In asking nearly half a million dollars for the first CAR-T therapy, Kymriah, Novartis is swinging for the fences – this is most likely the biggest list price ever in oncology.
However, the Swiss group is showing a pragmatic stripe and acknowledging the many unknowns for a treatment that is a scientific achievement but is unproven commercially; the benefit will vary by indication, patient and payer. A proposal to set reimbursement based on outcomes and indication will take time to evolve, but will almost surely bring the cost down below the proposed $475,000.
And, while this amount is undoubtedly high, it has come in below many analysts' expectations, and below Novartis's own cost-effectiveness calculation of up to $750,000 (Vantage Point – CAR-T value calculation in the firing line, June 27, 2017).
In truth, with a therapy so complex and costly, Novartis probably knew that it had to put itself at economic risk even to get payers to listen – otherwise payers could simply play hardball and push for an unacceptably low price. Amgen, for example, knew that this was the case when its cholesterol-lowering drug Repatha hit the market, and it has continued to evolve its position (ACC – Aggressive payers block access to PCSK9s, March 20, 2017).
What is more surprising is Novartis's proposal for “indication-based” pricing, an idea that has been met with scepticism throughout the sector. This means that Novartis realises that outcomes will differ between diseases – for example, paediatric acute lymphoblastic leukaemia, yesterday’s approval, versus diffuse large B-cell lymphoma, an upcoming filing – and therefore the clinical and economic benefit will differ.
Bernstein analyst Tim Anderson wrote yesterday that the small patient population – around 650 a year – for paediatric ALL meant that if payers accept the $475,000 price tag and Novartis is reimbursed for every patient meeting the one-month successful response criterion, its revenue would be $300m a year. More people develop diffuse large B-cell lymphoma, however, which would necessarily mean a lower price in that indication, Mr Anderson wrote.
Germany’s cost-effectiveness agency, IQWIG, is known to price drugs by indication, for example, although as a practical matter there is a single reimbursement rate based on weighted average because when drugs are billed the payer does not know for which disease this is.
If this is the case for Kymriah in the US it would almost certainly lead to an erosion of its initial price point. In paediatric ALL the value calculation would logically lead to a higher price because of the number of life years gained than it would with a disease affecting adults likely to gain fewer life years.
However, some experts believe that this will call for a slightly different approach. “Pricing in Europe is a series of mathematic averages. That’s traditional pricing for traditional medications,” Miguel Forte, chief commercialisation officer for the International Society for Cellular Therapy, told EP Vantage today. “Here we’re going to see a different approach with risk sharing.”
A month, give or take
Yet to be determined is what constitutes a reimbursable response. Novartis has proposed one month in the case of this paediatric ALL population, most of which did not progress to additional stem cell transplant. “Payers may say, “One month is not enough. I want to see two months,’” Mr Forte said.
Moreover, the number of different commercial payers in the US will make the value calculation more complex, if the risk-sharing agreements are rolled out to them, a difficulty acknowledged even by Dr Bruce Levine, a University of Pennsylvania oncologist and part of the team that developed Kymriah.
“If a kid gets CAR-T on Aetna, and then the parent’s job changes and they go to Kaiser, and then they get a stem cell transplant – that’s a question,” Dr Levine said.