Sanofi’s premium price hands Novo insulin win


The hole blown in Sanofi’s profit outlook from the loss of Lantus sales makes a little more sense in the context of its US pricing strategy against its only rival, Novo Nordisk’s Levemir.

Data taken from EvaluatePharma’s Sales, Volume & Pricing module show that the annual price differential between the two long-acting insulins has grown substantially in the past four years and now stands at $551 (see charts below). With Novo smarting from its own loss of US reimbursement for short-acting insulins, its answer has been to take share from Sanofi in the long-acting space.

The data show that the annual discount for choosing Levemir has grown from $148 in 2011 to $551 today. It is not as though Novo has been holding the line on price – after all, it took a double-digit price increases in both 2013 and 2014. The difference has been that its increases have been smaller and have come from a lower base.

Lantus vs Levemir: the US pricing strategies
2011 2012 2013 2014
Cost per patient per year ($) Lantus  2,035 2,298 2,753 3,630
Levemir  1,887 2,034 2,404 3,079
Difference 148 264 350 551
Growth in cost per patient per year Lantus (7%) 13% 20% 32%
Levemir 11% 8% 18% 28%
Estimate of discount given to payers Lantus (20%) (21%) (24%) (33%)
Levemir (45%) (39%) (39%) (42%)
Share of global long-acting insulin market Lantus  79% 79% 78% 76%
Levemir 21% 21% 21% 23%
Source: EvaluatePharma

Sanofi has perhaps reasoned that it can charge a premium because Lantus has been widely recognised as the superior drug. Levemir was never properly trialled as a once-daily product in the way Lantus has been, even though, like Lantus, Levemir has a longer duration of action than human insulin or short-acting analogues.

Payers budgeting across thousands or tens of thousands of diabetics are acutely aware of the price, however, especially as the differential widened by $200 from 2013 to 2014, and have been willing to look past Lantus’s true 24-hour duration. To put the price differential in perspective, if every single one of the $4.7m US patients taking Lantus were to make the switch to Levemir and achieve the $551 cost reduction, payers could save $2.6bn.

Novo executives today claimed a two-point increase in Levemir’s market share, generating 46% growth in North American sales to DKr6.9bn ($1.2bn) for the first nine months of 2014.

This is small change when stacked up to Lantus, which is forecast to sell $5.7bn in the US this year, but still it represents formidable competition – sufficient to contribute to a 15% drop in Sanofi’s shares since the bad news about Lantus sales was disclosed Tuesday (Diabetes warning sends Sanofi stumbling, October 28, 2014).

In explaining Levemir’s performance today, Novo executives credited effective payer and provider relationships, a direct-to-consumer campaign introduced in August and introduction of the new FlexTouch device as well as pricing for the success in gaining share. It should be noted that Novo had prepared a sales push before the assumed US launch of Tresiba in 2013, and when this failed to materialise the resources were available to promote Levemir.

The Medicaid question

A further sign of where Levemir has been registering wins is the clear jump in Levemir Medicaid prescriptions and a corresponding fall for Lantus. If ever there was a value shopper in the US healthcare system, it is the Medicaid programmes for the poor and disabled, which are subject to state budget politics year on year.

In addition to having legal authority to demand discounts from healthcare providers and suppliers, these tend have much stricter preferred drug lists. EvaluatePharma’s data are derived from Medicaid claims sourced from the US government, and there is a lag on processing this, so the past two to three quarters are an evolving picture. The total number of prescriptions is likely to be lower; nevertheless, the data now available represent likely trends.

How is Medicaid shifting volume? One way is through the use of those preferred drug lists. This year, Texas removed Lantus in favour of Levemir. Texas has the third-biggest enrolment of any Medicaid programme at 4 million, so the decision in Austin no doubt reverberated back in Sanofi’s Paris headquarters.

Sanofi has signalled that it is moving to a fight back – starting by cleaning house at the very top – and has already cut its own price to gain 90% market access. Today, Novo's finance chief, Jesper Brandgaard, said the Levemir outlook in 2015 would be “modest positive to flat”, probably anticipating Sanofi’s next move. This price war is on.

This story has been corrected to state that the potential savings from a Lantus to Levemir switch would be $2.6bn, not $11bn.

EvaluatePharma's Sales, Volume & Pricing module derives data from a number of sources to make its calculations. The per patient per year cost is calculated at an Indication level, and weighted based on sales by indication:
average mg per patient multiplied by cost per mg. In primary care: the cost per mg for primary care products is based on the price the pharmacy pays for the product (NADAC), therefore the cost per patient excludes the pharmacy margin (~5% on branded drug). This also excludes payer rebates.

The discount estimate in primary care is calculated based on the pharmacy price paid to wholesalers tracked via NADAC divided by the USA government price (best commercial price) tracked via the Federal Supply Schedule (FSS). Given the FSS price is set over a contract life, a two year forward average price is used.

Quarterly Medicaid data is the number of prescriptions reimbursed by the Medicaid program only, to pharmacies for the drug (based on NDC number), as reported by state agencies.

To contact the writer of this story email Jonathan Gardner in London at or follow @JonEPVantage on Twitter

Share This Article