Lantus numbers. US only.

[tripoley]

Not sure if anyone has done this before but I thought I would break down the US launch of Lantus.  Data is from the SNY annual reports.  Exchange rates are from Forex yearly averages.  Lantus was launched in 5/2001 so it was basically a half year.  I just went up to $1B US which it surpassed in 2009.  FYI Lantus was launched in Germany in 6/2001, US 5/2001, UK 8/2002 and 40 total countries in 2003.  Total Lantus sales in 2009 were 3080M Euros (2214.5M Dollars) worldwide.

Year	Euros (M)	Exchange Rate	Dollars (M)
2001	58	1.117	64.8
2002	239	1.064	254.3
2003	347	0.883	306.4
2004	495	0.805	398.5
2005	717	0.804	576.5
2006	1006	0.797	801.8
2007	1200	0.731	877.2
2008	1452	0.683	991.7
2009	1909	0.719	1372.6

[mnholdem]

Does Sanofi break down their individual drug sales by quarter? It would be interesting to look at the first 6 quarters after launch. It might give us a peek at how Sanofi grows their diabetes drugs. 2001 monthly sales of Lantus would be awesome, but where to look? I suppose I could run custom date ranges on the internet later today, after my grandkids are done wearing me out and go home (chuckle). Good thing I exercise 5 days a week or I wouldn't being able to manage the exertion.

[tripoley]

I was just looking for annual sales so I didn't look specifically at quarterly results.  I suspect that info is in the 10Qs though.  I do know that worldwide sales (Germany basically) for 2000 were 10M Euros for a half year.  I'll see if I can find the US data for the first few quarters today.

[alethea]

Feb 28, 2015 9:09:43 GMT -5 tripoley said:

Not sure if anyone has done this before but I thought I would break down the US launch of Lantus.  Data is from the SNY annual reports.  Exchange rates are from Forex yearly averages.  Lantus was launched in 5/2001 so it was basically a half year.  I just went up to $1B US which it surpassed in 2009.  FYI Lantus was launched in Germany in 6/2001, US 5/2001, UK 8/2002 and 40 total countries in 2003.  Total Lantus sales in 2009 were 3080M Euros (2214.5M Dollars) worldwide.

Year	Euros (M)	Exchange Rate	Dollars (M)
2001	58	1.117	64.8
2002	239	1.064	254.3
2003	347	0.883	306.4
2004	495	0.805	398.5
2005	717	0.804	576.5
2006	1006	0.797	801.8
2007	1200	0.731	877.2
2008	1452	0.683	991.7
2009	1909	0.719	1372.6

One euro has pretty much always been worth more than one dollar.  I think after year 2002 in your table the conversions are backwards.  For example in 2009, 1,909 euros should convert to $2,655 dollars.  And 1 billion in US dollars is first acheived in 2006 where 1,006 euros converts to $1,262 dollars.

[tripoley]

Geez, let me redo that:

Years	Euros (M)	Exchange Rate	Dollars (M)
2001	58	0.896	52
2002	239	0.942	225.1
2003	347	1.134	393.5
2004	495	1.244	615.8
2005	717	1.246	893.4
2006	1006	1.256	1263.5
2007	1200	1.370	1644
2008	1452	1.471	2135.9
2009	1909	1.394	2661.1

[tripoley]

And quarterly US sales:

Quarter	Euros (M)	Exchange Rate	Dollars (M)
2Q01	8	0.876	7
3Q01	18	0.844	15.2
4Q01	32	0.883	28.2
1Q02	40	0.872	34.9
2Q02	58	0.989	57.4
3Q02	66	0.981	64.7
4Q02	75	1.045	78.4
1Q03	70	1.083	75.8
2Q03	83	1.144	94.9
3Q03	90	1.163	104.7
4Q03	104	1.249	129.9
1Q04	106	1.217	129
2Q04	123	1.208	148.6

[jpg]

Thank you Tripoly,

Nice work. Your breakdown would probably work for ever blockbuster drug I can think of (but have not checked). Many have unrealistic expectations. Sanofi and Mannkind know what they are doing.

[bobw]

Tripoly, thanks for digging up the quarterly numbers.  I added a few columns, to look at it in terms of the number of Afrezza prescriptions that have to be filled to match Lantus revenue.  I couldn't get my spreadsheet to format correctly in the post, so I put it in as a jpg.

The columns were calculated as follows:

Quarterly Increase: The difference between the current quarter and the previous quarter.

Average Increase:  The average quarterly increase

Average Script Qty Increase: If you assume that annual Afrezza revenue per patient is $2,000 and one script per quarter, then dividing the average increase by $500 gives you the average number of new patients using Afrezza.

Average New Scripts/Week:  Divide the quarterly numbers by 12

The conclusion is that to match Lantus revenue, we need over 1,000 new patients per week for several years.  Of course, even a fraction of Lantus sales would be enough to make Afrezza successful.

It is striking that the Lantus ramp up is pretty linear.  I would have expected an S-curve.

[tripoley]

It doubled from 2Q01 to 3Q01 and nearly doubled from 3Q01 to 4Q01. The Lantus "available now" message was on 05/22/2001 so the first quarter of sales only had five to six weeks in it. Pretty rapid early adoption in the first couple quarters. Surprisingly so.

[dreamboatcruise]

At first glance I was thinking... that's basically the same as Exubera, but then realized that Exubera plateaued at 1900 TRx whereas this is showing 1700 - 1900 NRx per week. Big difference.

Hopefully Afrezza will do better than this since if the thesis about it allowing primary care docs to get patients on insulin earlier and retain them in the primary care setting is true, it would mean that the number of prescribers would be an order of magnitude higher than for Lantus.

I do believe not too much can be read into the first couple of months of data, but at some point in the not too distant future I will be looking for a signal in the data that Afrezza is indeed being written by a much larger pool of doctors.

I suppose that Afrezza could still be a blockbuster even if the thesis of many primary care doctors that currently refer to endos for insulin deciding to write their own Afrezza prescriptions turns out to be largely false, and then I'd expect that the Afrezza launch might look the same as Lantus.

I think you'd see that a launch for something like lipitor probably had NRx plateau at much higher level. I'd hope that Afrezza would be somewhere between the NRx plateau of Lantus and Lipitor, representing that it is being written by a sizable proportion of primary care docs.

jpg... would you not expect to see a meaningful difference in NRx between Afrezza and Lantus appear within 6 months? 12 months? 18 months? or unrealistic?

[tripoley]

Exuberas second quarter 2007 sales (near the end) was about $4M.


www.bloomberg.com/bw/stories/2007-07-24/pfizers-exuberant-ad-rolloutbusinessweek-business-news-stock-market-and-financial-advice

[jpg]

Feb 28, 2015 15:40:49 GMT -5 dreamboatcruise said:

jpg... would you not expect to see a meaningful difference in NRx between Afrezza and Lantus appear within 6 months? 12 months? 18 months? or unrealistic?

I will tell you with deep conviction that I am 100% certain that I don't know...

Maybe I am just not a control freak or simply don't get the 'precise' modelling so many are attempting to do? At this time point there are too many uncertainties in all this to be precise. Models for Afrezza ranging from a 'total flop' to 'the best selling drugs of all time' all have, or can have, mechanistically explanations and numbers to justify them (including curves, uptake models, GPs vs specialists prescribing, renewal/ refill %ages vs dropouts etc as inputs). We all want simple answers but at the same time these models obviously can't all be right! It's like people are trying to model the future with poor and very limited data/ simplistic information. This is just a personal observation and I am not saying that to say we shouldn't attempt to do it but simply to point out the huge limitations to doing this haphazardly. The pitfall to flawed analysis is being completely wrong... Selling to early (or to late) because of a flawed emotion driven model can be painful.

I for one think Afrezza/ Mannkind now has plenty of 'safety margin'. This doesn't mean I know what the stock price will do in a month, 6 months or a day (as I said I think time sensitive synthetic 'investment' instruments are tools of financial destruction for retail 'investors': the vast majority should stick to owning unleveraged time insensitive stock). As early as we are from the launch I more and more think that with time Afrezza will do very well and will find enough niches (to borrow a term from our negative friends) to become a blockbuster. I think our first niche will be type 1 early adopters with CMGs (because they are the ones who like technology and know the power of technology to manage their disease well). They will lead the way. The vast majority of type 2's will not care about Afrezza. They don't care about insulin. Why should they? No one cares about a blood pressure or cholesterol medication until they have to take it... Once you take it and start having side effects/ risks and see your quality of life change then you care about better alternatives. Same with type 1s and insulin but multiplied by 100. They live or die by their insulin and the vast majority have no choice but to structure their lives around their disease. As I have repeatedly (and boringly: sorry) pointed out the needles pain is only a small fraction of the equation. The HbA1C is only a small piece of the puzzle (which is probably only valuable when very high and will be shown to be obsolete with more widespread use of CMGs). The 'big picture of 'life as a type 1' is what counts. Give type 1s a better mousetrap and many will become our best salespeople (as we are seeing early signs of). Once we have this 'mind space' then we have a big chunk of a lot of stuff. I certainly would like to see a rise in NRx but am way more interested in seeing acceptance and embrace by those who live and die by insulin: type 1s, and especially type 1s with CGMs. Our first 'niche'. We could easily do very well with only type 1s (the vast majority of those on prandials).

As for your Lantus question: Afrezza isn't Lantus obviously. Lantus was a better mouse trap for all MDs (GPs and specialists) dealing with diabetics. It was a relatively easy insulin to prescribe to type 2s on a daily basis (although we know it should ideally have been given BID) and solved a few headaches. I don't know the breakdown of PCP vs endos prescribing Lantus but from my experience it is often prescribed by PCP because it isn't that complicated to prescribe and is at relatively low risk of causing complications (hypo). Could Afrezza be the same type of product? Yes easily. It would be interesting to know if Lantus first sold to type 1s and then to the much bigger type 2 market. It would also be interesting to know who first started prescribing Lantus (relative to their numbers obviously): GPs or endos?

What I am trying to say (I think?) is that if we do want to analyze these things well (and not simply with silly 'back of the envelope calculations' with weird numbers and very partial understandings of comparators like seems to be the trend of many internet pundits (PA, AF and sadly so many others)) even with the valuable information provided above on Lantus original sales it is very hard to extrapolate from one drug to the next without a lot of work and analysis. If I had the time and money I think the way to do this would be to look at multiple drug launches (mealtime analogues, Lantus, other diabetic pills and as many blockbusters and flops as we can reasonable manage to look at: 'positive and negative controls') and get a 30000 ft view of numbers and time. Then break it down to pcp vs endo and type 1s vs types 2 (for diabetic products obviously). Reimbursement rates, coverage, pharma reps selling etc. This would be a lot of work and something I presume some more health care focused and sophisticated investor groups (hedge funds etc) and obviously big pharma/ biotech would routinely do. It gives you a road map (which is good for every drug hitting the market) to be a useful measuring tool. If I get rich of Mannkind and a few of my other biotechs I will quite my job as an MD and get a job setting up this type of info. Opps... I won't need a job anymore!

These comparators would not be 'designed' to avoid the optimistic but (sorry) naive 'management has got to get as many lines running as fast as possible because we are going to run out of product in 1 month' but more to understand the dynamics of good marketing efforts and to extrapolate once real meaningful data (from samples if you are the pharma who has this closely guarded info) does start coming in. For us with only NRx data points (not directly in the loop with samples) it probably takes at least 2-3 months minimum to even be able to have a feel for what is happening? The first 2 months are usually just noise. Even more so for a disruptive consumer drug (i.e.; not a binary oncology drug) like Afrezza.

Obviously looking only at Exubera doesn't make much sense. One interesting and highly relevant comparator could be the introduction of the current rapid acting analogues and their original market uptake compared to what was prescribed before then (not certain it would be easy to find that older data though?). I do vividly recall people making fun of the first rapid analogues as 'weak, no big advantage and barely non inferior to the regulars'. Does that sound familiar? 


How to answer a simple question with as much complexity as possible... I should have been a lawyer or economist!

[mnholdem]

Continuous growth with the exception of one quarter. I expect we'll see this kind of sales growth pattern with Sanofi's global sales force behind Afrezza, except for the one flat quarter, of course. Frankly, I think Afrezza offers more selling points than Lantus did when it launched.

[tripoley]

Feb 28, 2015 17:20:56 GMT -5 jpg said:

Obviously looking only at Exubera doesn't make much sense. One interesting and highly relevant comparator could be the introduction of the current rapid acting analogues and their original market uptake compared to what was prescribed before then (not certain it would be easy to find that older data though?).

I'll see if I can dig up some sales data on Humalog.  If Aventis had this early data 2001 until SNY took them on, I'm sure LLY will have it too.

[tripoley]

Here's what I found on Humalog.  I believe they started sales sometime in 1996 but the first quarterly data that listed Humalog sales was the first quarter of 1998 of 25M and increase of 13M over the first quarter of 1997.  I'm assuming sales were small before then.  These are worldwide sales.  So...

Quarter	Sales (M)
1Q97	12
2Q97	?
3Q97	?
4Q97	?
1Q98	25
2Q98	30
3Q98	32
4Q98	42.6
1Q99	42.5
2Q99	50.4
3Q99	58.7
4Q99	72.9

Humulin R sales were 209.8M in the first quarter of 1997 which is what Humalog was intended to replace.