Updated from 2:07 p.m. EDT

Last week, reader J.D. emailed to ask about Dendreon ( DNDN). Specifically, he wanted to know how I came up with a valuation for the company's prostate cancer vaccine Provenge.

"I'd like to learn more about how to value a company's pipeline," he wrote. "I'd like to gain some insight into how you might determine how much a successful drug could actually add to the bottom line."

Great question, J.D., and timely with Dendreon set to go in front of a Food and Drug Administration advisory panel on March 29. Prompted by J.D.'s email, I sat down this week and built a revenue model for Provenge.

Based on my conservative assumptions, if Provenge is approved (and that's an "if" still very much up in the air), peak sales in the U.S. could top $1 billion.

Based on this Provenge sales forecast, Dendreon would be worth about $27 a share today in my valuation model. Upside to my conservative sales forecast would, naturally, lead to a higher stock price.

Now, I'm not teaching J.D. or anyone very much by simply giving you my estimates. So let's walk through the process of building a Provenge revenue model. It's a really helpful skill and one that can be used for any drug or biotech company, in case you don't give a hoot about Dendreon.

Before I get started, I want to make one thing perfectly clear. I remain on the fence about Provenge's chances at next week's advisory panel meeting, so please don't construe that this Provenge revenue model has me somehow leaning toward a positive outcome.

This revenue forecast, then, is helpful if (pay attention to that little word) Provenge is approved.

Let's start with a chart of my first of three Provenge revenue models:

Prostate cancer -- Revenue Model 1
2006 2007 2008 2009 2010 2011 2012
Newly diagnosed PrC patients 219,000 222,285 225,619 229,004 232,439 235,925
% patients with AIPrC 35% 35% 35% 35% 35% 35%
Total AIPrC patients 76,650 77,800 78,967 80,151 81,354 82,574
% patients choosing immunoTx 65% 68% 70% 72% 75% 75%
Eligible Provenge patients 49,823 52,904 55,277 57,709 61,015 61,930
Provenge penetration 2% 10% 20% 40% 50% 55%
Provenge patients 996 5,290 11,055 23,084 30,508 34,062
Cost/year 30,000 31,200 32,448 33,746 35,096 36,500
Provenge sales $29,893,500 $165,059,950 $358,723,818 $778,975,944 $1,070,686,207 $1,243,237,996

To start a drug revenue model, you first need to define the target patient population that will be eligible to receive the drug. With Provenge, it's a bit tricky. For starters, we're obviously talking about prostate cancer, so that's not very hard. According to the American Cancer Society, there will be 219,000 men diagnosed with prostate cancer in 2007. That's the incidence figure, or the number of men who will newly diagnosed this year. (I have that number growing 1.5% each year to account for an aging population.)

Disease Prevalence

Now, with many diseases, including most cancers, you can also look at disease prevalence, which is the total number of patients at any one time. Prevalence numbers will be larger than incidence numbers because it takes into account more than one year.

For the purposes of my model, I found finding prevalence figures for prostate cancer difficult, so I chose to use incidence. That's me being conservative, a stance I'll take throughout this model because I'd rather underestimate than overestimate.

Provenge, at least initially, will not be given to all prostate cancer patients. Instead, use will be restricted to patients who have metastatic, androgen-independent (hormone-refractory) disease, which means their cancer has spread beyond the prostate and it is no longer being controlled by hormone treatments. These patients will also be asymptomatic, which means they won't have any outward symptoms, such as bone pain, which could signal new growth of their cancer.

Here comes the hard part -- figuring out the percentage of prostate cancer patients who fit that category, which is labeled "AIPrC" in my model above. After a lot of Googling and research, including looking back at some old analyst notes, my best guess is that 35% of prostate cancer patients are no longer responding to hormone treatments. That's what I used in my revenue model to come up with a total population of AIPrC patients -- the target group for Provenge.

Not all eligible prostate cancer patients will choose to take Provenge. How many will? Frankly, this is an educated guess, especially since Provenge is a first-in-class cancer immunotherapy without precedent.

But again, after some research, I learned that about 65% of hormone-refractory prostate cancer patients choose to move onto chemotherapy treatment for their disease. Provenge, if approved, would slot between hormone therapy and chemotherapy, so for my modeling purposes, I decided to assume that the same 65% of patients would choose Provenge. Given the excellent side-effect profile of Provenge, this percentage might actually end up higher, but again, I like to be conservative.

Applying this 65% to the total AIPrC patient pool gets me to a total number of eligible Provenge patients. I estimate that number to be 49,823 in 2007, growing to 61,930 in 2012.

Marketing a new drug is not like turning on a light in a darkened room -- results aren't immediate, and not every eligible patient will take a drug, for a variety of reasons. Again, calculating a penetration rate, or market share, is a mix of art and science. In the real world, the best drug with the best marketing will likely top out at 80% market share.

In this case, I start Provenge at a 2% market share because approval, if it happens at all, would come late in 2007. This is a new kind of therapy for cancer and until additional clinical data is gathered and disseminated, I think some doctors might be reluctant to try it. For these reasons, I have peak market share topping out at 55%. I hate to repeat myself, but this is fairly conservative modeling.

In my model, then, I assume 996 Provenge patients in 2007, growing to 34,062 patients in 2012.

Product Pricing

Next consideration is pricing. Dendreon hasn't publicly discussed its thinking on Provenge pricing, but various analysts and commentators have pegged the price in the range of $20,000 to $45,000 a year. I'm using a $30,000 annual price (with a 4% hike per year) in my model, which is in line with other targeted cancer therapies.

That gets us to the end: My 2007 Provenge sales estimate is just under $30 million, rising to $1.24 billion in 2012.

Now, at the top I said my peak sales forecast for Provenge was $1 billion. The lower, final estimate results from the average of three revenue models, each looking at the Provenge market in a slightly different way. I don't always build multiple sales models, but in this case -- given all the unknown variables -- I thought it prudent.

Here's model No. 2:

Prostate cancer - Revenue Model 2
2006 2007 2008 2009 2010 2011 2012
Stage III 24,213 24,939 25,688 26,458 27,252 28,070 28,912
Stage IV 35,749 36,821 37,9268 39,064 40,236 41,443 42,686
% patients failing hormone Tx 85% 85% 85% 85% 85% 85% 85%
Total AIPrC patients 50,968 52,497 54,072 55,694 57,365 59,086 60,858
% patients choosing immunoTx 65% 65% 68% 70% 72% 75% 75%
Eligible Provenge patients 33,129 34,123 36,769 38,986 41,303 44,314 45,644
Provenge penetration 0% 2% 10% 20% 40% 50% 55%
Provenge patients 0 682 3,677 7,797 16,521 22,157 25,104
Cost/year $0 $30,000 $31,200 $32,448 $33,746 $35,096 $36,500
Provenge sales $0 $20,473,725 $114,718,376 $253,001,257 $557,516,461 $777,619,314 $916,284,390

In this model, I start with the number of prostate cancer patients diagnosed with Stage III and Stage IV disease. These numbers aren't easy to find, but are derived from database estimates collected by the National Cancer Institute. These patients more closely represent likely Provenge patients because their disease is more advanced at the time of diagnosis.

I assume that 85% of these patients fail hormone therapy, which makes them eligible for Provenge, based on its likely initial approval. In this model, then, my "total AIPrC patient" estimate is lower than in model No. 1.

After that lower patient base, everything else in the model remains the same. At the end, I derive an estimated U.S. 2007 Provenge sales of $20.5 million, rising to $916.3 million in 2012.

Lastly, let's look at my third model:

Prostate cancer - Revenue Model 3
2006 2007 2008 2009 2010 2011 2012
Annual prostate cancer deaths 27,000 27,540 28,091 28,653 29,226 29,810 30,406
% patients choosing immunoTx 0% 25% 50% 65% 75% 80% 90%
Eligible Provenge patients 0 6,885 14, 045 18,624 21,919 23,848 27,366
Provenge penetration 0% 25% 40% 50% 60% 75% 85%
Provenge patients 0 1721 5618 9312 13,152 17,886 23,261
Cost/year n/a $30,000 $31,200 $32,448 $33,746 $35,096 $36,500
Provenge sales $0 $51,637,500 $175,286,592 $302,159,027 $443,811,179 $627,726,532 $849,012,689

This model is the most conservative because it starts with only the number of prostate cancer patients who die each year, as estimated by the American Cancer Society. This is a much sicker and narrower patient population than will likely receive Provenge, but it does provide a minimum revenue floor for modeling purposes, which is why I like it.

You'll also notice that because this model is based on prostate cancer deaths, I assume a much higher penetration rates for Provenge, maxing out at 85% to 90%. This model comes up with 2007 Provenge revenue of $51.6 million, rising to $850 million in 2012.

When you average all three revenue models together, you get the following Provenge U.S. revenue forecast, with sales topping out at $1 billion:

2007 2008 2009 2010 2011 2012
Average of three models $34,001,575 $151,688,306 $304,628,034 $593,434,528 $825,344,018 $1,002,845,025

In one of my previous Dendreon columns, I said that it was easy to get to peak Provenge sales of $500 million. Obviously, I've easily surpassed that guess with this more rigorous forecast.

Briefly, let me discuss valuation. In the beginning of this article, I peg Dendreon with a $27 share price based on my revenue projections (and assuming, of course, that Provenge is approved.) This valuation underscore why Dendreon, currently at $3.74 and significantly shorted, is a great story to follow.

I derive my valuation using a fairly simple, discounted revenue valuation model. Peak Provenge sales of $1 billion in 2012, risk adjusted by 85% and discounted back just under five years gets to a net present value for Provenge of $355 million. At a reasonable six-times-sales multiple, you get an enterprise value for Dendreon of $2.1 billion. With 82 million shares outstanding and its cash on hand, Dendreon would be worth about $27 per share.

Of course, all these projections are moot if Provenge can't get a positive nod at the FDA advisory panel meeting. We'll see what happens next week.

As originally published, this story contained an error. Please see Corrections and Clarifications .

Adam Feuerstein writes regularly for RealMoney.com. In keeping with TSC's editorial policy, he doesn't own or short individual stocks, although he owns stock in TheStreet.com. He also doesn't invest in hedge funds or other private investment partnerships. Feuerstein appreciates your feedback; click here to send him an email.

More from Stocks

How Small-Cap Stocks Can Protect Your Portfolio From a Trade War

How Small-Cap Stocks Can Protect Your Portfolio From a Trade War

Week Ahead: Trade Fears and Stress Tests Signal More Volatility To Come

Week Ahead: Trade Fears and Stress Tests Signal More Volatility To Come

3 Great Stock Market Sectors Millennials Should Invest In

3 Great Stock Market Sectors Millennials Should Invest In

Why Millennials Are Ditching Stocks for ETFs

Why Millennials Are Ditching Stocks for ETFs

Trump's 'Space Force' Could Launch a $1 Trillion Industry, Morgan Stanley Says

Trump's 'Space Force' Could Launch a $1 Trillion Industry, Morgan Stanley Says