Biopharma Strategy Blog

Informational Portal Supporting Pharma and Biotech Commercialization and Marketing Efforts

The Payoff of Strategic Early-Stage Forecasting - Pharmacision Featured in the August 2010 Issue of PM360

Submitted by Tom Marten on Sun, 10/18/2009 - 20:26

Pharma Forecasting

Using analogs and rules of thumb may give you a gut feel for market potential but provides no real value when making new product planning decisions. Here are three essesntial elements for strategic early-stage forecasting.

As featured in the August 2010 issue of PM360, http://www.pm360online.com/f3_0810.

Early stage pharma forecasting can be a valuable exercise helping to drive clinical development and commercialization efforts, or end up as a “check the box” activity given merely an afterthought in strategic planning efforts. The difference is all in the approach. Early forecast numbers may be of little strategic value when projections are made without factoring market environmental trends and consumer behavioral tendencies.

Many early forecasts are derived from analogs used as comparables, and general “rule of thumb” assumptions based on industry statistics. Using analogs and rules of thumb may be straight-forward and inexpensive, and may give you a gut feel for market potential, but provide no real value when making new product planning decisions.

On the other hand, employing a strategic forecasting process for early stage products which involves:

· Learning your market environment

· Understanding behavioral drivers, and

· Linking these findings to an assumption driven forecast model

may require a little more time and up front investment, but will yield far greater returns in the long run by supporting new product planning decisions and optimizing product development efforts.

It’s the journey, or how you discover your markets and generate assumptions that’s most important, not the destination or the forecast projections you come up with. Focusing on the journey will uncover insights and opportunities enabling you to better predict future market outcomes and help you discover the keys to future commercial success.

I’ve heard a number of life science venture capital presentations over the past few years where the VC folks opine about the challenges involved with valuing early stage companies and predicting future success. Valuations are nothing more than an estimate of the combined value of companies’ future products’ financial performance. Given that life science companies have incredibly long product development times with a great deal of uncertainty, accurate long-term forecasting is virtually impossible. Within pharma companies themselves, whether small or big pharma, the same uncertainty holds for compounds in early development (i.e. pre-clinical through phase II) when predicting risks, clinical outcomes, and future sales.

Some of my entrepreneurial colleagues have gone so far as to say that 5-year forecasts in business plans of startups show the naïveté or inexperience of the company founders. Even within pharma, many new product planning groups and business development teams place little effort into new product forecasting until compounds reach later stage clinical trials when commercial and manufacturing teams need to start gearing up for larger clinical trials and commercial launches.

As a result, early forecasting often becomes a semi-mindless exercise, often involving Microsoft’s copy-paste function to copy the forecast of the closest analog to use as a forecast for the compound in question. Alternatively, early forecasting efforts often incorporate “rule of thumb” ratios or general industry statistics on patient penetration, margins, and growth rates without providing much in the way of rationale for using each ratio. Many early stage forecasting exercises involve identifying and sizing target markets, with little scrutiny beyond that. Often, early pharma forecasts are more political exercises to secure R&D budget funding. It’s intriguing how forecasts often go from $billions to $hundreds of millions or lower as compounds progress through clinical trial phases getting closer to launch and people finally become accountable for the numbers they come up with.

Perhaps, at this point, you’re thinking there’s too much uncertainty early on to develop an accurate forecast. You shouldn’t waste too much time in the forecasting process because you have no idea how your phase IIb and phase III trials will turn up and whether your competitors’ pipeline products will look any better. Maybe true, but does that mean that investing in early stage forecasting is not a good use of time or resources? Should you let your potential investors do it, because they won’t believe any numbers you come up with anyways? Is it better to not do it until you’re well into phase III trials? Should emerging pharmas wait and hire a consultant to create a good looking forecast when they’re ready to start shopping their compound around and seeking out-licensing partners? Buy a new dart board? Don’t do it at all? I would argue none of the above.

I get the sense that most people view early pharma project forecasting the wrong way. The destination or the end results, i.e. the numbers you come up with, should not be the focus. The journey, or how you get the numbers, should be the focus of any forecasting exercise and help lead the way for project planning teams as they make crucial product development decisions.

Ironically, the numbers you come up don’t matter a great deal. Let’s face it. They’re wrong. If you could accurately predict the future, you could make a lot more money betting on horses! What matters more are the assumptions on how you get the numbers. Forecasts derived from analogs and rules of thumb may give you a general gestalt for the level of market opportunity a new product may bring, but don’t provide any indication as to why or what will make your future product successful or not successful.

Instead, by truly learning your markets early on and understanding the behavioral drivers that impact prescribing decisions, not only will you be able to develop assumption driven strategic forecast models that can give you a general feel for the size of an opportunity, you will also gain a roadmap for optimizing product labels and enhancing commercialization plans. One should never dispute forecast projections, and should only dispute the assumptions used to create the forecast projections.

As stated earlier, Learning your market environment, understanding behavioral drivers, and linking these findings to an assumption driven forecast model are the three keys to developing successful strategic forecasts. Let’s take a look at some examples to see what I mean.

Exploring and Learning about market environments:

To have any sense of what a future market may be like, it’s important to understand the general market infrastructure and how this will impact your product concept. It’s even more important to get a sense of how, if at all, the market environment will evolve over time.

Case Study #1: Thrombolytic and the Impact of Reimbursement

When I was with a biotech company a few years back, we were evaluating the potential for a catheter directed thrombolytic we had in development. This was an agent that dissolved clots within blood vessels and had to be delivered via a catheter placed in the vessel that squirted out the drug at the site of the clot to break up the blockage. We needed to make some future projections to help justify the lead indication and product development strategy.

According to our clinical development and regulatory teams, pursuing a lead indication of hemodialysis graft occlusions appeared to be the quickest route to market and was identified as an area of high unmet need. Hemodialysis grafts consist of a Gortex tube that connects an artery to a vein to improve blood flow to the vein and enable hemodialysis machines to dialyze blood in patients with kidney failure. The problem is that these grafts occlude or clog up over time.

As this is a fairly large patient population with unmet need, it looked like a great lead indication. With our Frost and Sullivan reports and reports from a claims database, we could easily determine the number of graft placements, predict our small share of the market and we would be good to go with an early forecast projection.

Interestingly, however, when we talked to physicians in Europe, we discovered that they rarely, if ever surgically placed grafts for hemodialysis. Most of their patients with kidney failure had fistulas to create a vein large enough to support dialysis. Fistulas are created when a surgeon surgically connects an artery to a vein, and allows the connection to heal for at least 6 months. Upon further investigation, we discovered the reason why nobody used grafts in Europe. It was the same reason why we were developing this product in the first place – because grafts occluded frequently requiring a procedure to unclog or restore blood flow to the graft. They used fistulas because they didn’t occlude as frequently.

So why didn’t the US healthcare community use fistulas? As it turns out, the reason had nothing to do with medical practice and had everything to do with reimbursement. Many chronic kidney disease patients not at the point of requiring dialysis either did not have healthcare insurance or had health care insurance that would not reimburse for surgery to prepare a fistula. However, patients automatically qualified for Medicare when their kidney disease progressed to end stage renal disease, which is the condition when hemodialysis is required. Medicare would cover the cost of surgery to prepare a fistula, but unfortunately most patients couldn’t wait 6 months after fistula surgery to start hemodialysis. Instead, grafts were used because they were much quicker to place and use for dialysis. This created a constant cycle for patients with grafts undergoing hemodialysis. Patients received dialysis, which over time led to occluded grafts, which then required thrombectomy procedures to clear the occlusions before they could be used for dialysis again. Obviously, this was not the most cost effective health care system approach, but has been the case in the US the recent past.

Armed with this information, does the outlook for this indication change? Let’s assume this case study occurred at the time of this writing (Fall 2009). President Obama’s healthcare plan intends to place more emphasis on universal healthcare and improving healthcare efficiency. Bills are in Congress under debate with a great deal of uncertainty as to the final outcome of any final legislation and eventual new health care policies.

This leaves a number of questions for the product development team. Will any healthcare reform measures that are passed enable greater coverage for fistulas? If so, will the number of grafts and graft occlusions be reduced considerably? What will the impact be on actual practices and standards of care? Perhaps this introduces additional uncertainty, but allows the strategic planning team to create downside scenarios that weren’t envisioned before, and possibly wait until any reform measures are passed into law before reevaluating and making any decisions.

Either way, with a little digging, the team would be able to factor in evolving market environment scenarios when developing early stage forecasts and making key strategic decisions. With this knowledge, key decisions and expenses, such as Phase I/II clinical trial planning and funding, could be avoided if the team determined that an alternative lead indication would be more appropriate.

Case Study #2: Bad Blood and the Unaccepting Market

Understanding market environments is arguably more important for start up and emerging pharma and life science companies. Often early stage companies are science and technology driven and have had little exposure to healthcare providers and have limited budgets for market research and forecasting. Fortunately, understanding market environments and strategic forecasting can often be accomplished on a low budget with limited primary market research. Please note that I said limited and not no primary market research. Secondary data, market reports and literature findings are a great way to become oriented to a new therapeutic area or healthcare market environment, but nothing tops talking directly to future target customers.

I had the pleasure of working with an early stage company who had developed a blood gas analyzer that could measure the integrity of red blood cells in whole blood or packed red blood cell units. The theory behind this device was that red blood cells often degraded prior to expiration, which impacted the ability of transfusions with highly degraded RBCs to provide adequate tissue oxygenation. The company had a number of reports from the literature to support their theory and great proof of concept that their device worked. It was a portable unit that could fit on a small table top and did not require a great deal of training or special skills to operate.

The company had developed a 5 year forecast and budget that envisioned selling units at relatively low cost and placement of units in ICUs, hospital floors, and blood banks. Their business model envisioned that they would make most of their money on disposables and were counting on this technology becoming a standard practice given the relatively low cost, ease of use and direct impact on patient care.

One problem, when we talked to blood bank directors and hematologists, nobody had ever heard of a tissue oxygenation problem from transfused blood units. There was much greater concern over issues such as blood type matching, supply of rare blood types, and TRALI, an adverse event that could occur as a result of transfusions. The problem of tissue oxygenation did not exist in their minds. Blood bank directors actually became dismayed that a company would produce such a product given the clear FDA and American Blood Bank guidelines in place that called for a 42 day shelf life. After only a few phone calls, it was clear that the market was not ready for this product and that to be successful, the market environment would need to be radically different.

Fortunately, the company discovered these challenges early on and has been working with key opinion leaders to develop advocacy and support for running outcomes based clinical trials to increase awareness and demonstrate product value.

As such, basing business planning decisions on early top line forecasts without understanding the dynamics of the market environment would have lead to sure business failure.

Understanding Behavioral Drivers

In addition to understanding macro level market environmental factors related to future product potential, it’s important to understand the issues and beliefs that drive individual customer behaviors and how these behaviors will impact perceptions of new product concepts. By doing so, companies can discover levers or actions they can take when developing products to make them more readily accepted into the market or facilitate behavioral change to increase product adoption when launched. Here is another case study to illustrate these thoughts.

Case Study #3: AZT and the Blue Band of Death – Understanding the Impact of Patient Behaviors

Glaxo Wellcome in the mid 1990s faced significant challenges as a result of declining AZT market shares. AZT, a drug for the treatment of HIV/AIDS was also known as zidovudine and sold under the brand name Retrovir^®. AZT was facing considerable competition from a product called d4T or Zerit^® from BMS and the company had few answers on how to stop the decline.

The greatest challenge that AZT faced was patient perception. Clinically, AZT and d4T were very similar, and in fact d4T caused greater neurological side effects. At first glance, one would wonder how it ever became the favored drug. AZT’s problem was one of stigma. AZT was the first drug marketed for the treatment of HIV and was usually prescribed as patients’ immune systems weakened to the point of reaching AIDS stage. Getting a prescription for AZT signaled to both patients and caregivers that you didn’t have much longer to live. Additionally, AZT was initially prescribed at high dosages which created a great deal of nausea and other side effects making life miserable for patients taking the drug. As one of the only early options available, AZT was prescribed as monotherapy. The AIDS virus quickly became resistant to Retrovir when taken alone and stopped working after a short period of time fostering a sentiment that the drug was not very effective.

Retrovir was sold as a white capsule with a blue band around it. Over time, the white capsule with the blue band came to signal impending death for many HIV/AIDS patients! In hindsight, d4T’s greatest attribute was that it wasn’t AZT.

As the market evolved, more products became available and the medical community discovered combination regimens that were much more effective than monotherapy and were able to prolong life considerably. Glaxo and clinicians also realized that a much lower dose would work in combination therapies and that the lower dose was tolerated much better. However, the stigma remained.

From a market environment standpoint, compliance and adherence were the big issues of the day. Three drug regimens with two different classes of drugs were the standard of care and required complicated timing of pill taking as some drugs needed to be taken with food, some on an empty stomach, and many had different dosing schedules when pills had to be taken. As most regimens contained another Glaxo product called Epivir^®, alsoknown as lamivudineor 3TC, Glaxo embarked on a strategy to develop a combination pill combining AZT with 3TC. We understood that improving compliance would benefit patients and help increase sales as patients took more of their medicine.

I was tasked with revising an initial forecast for the new AZT/3TC combination pill that we intended to sell under the brand name Combivir^®. Combivir offered the benefits of combining two pills taken twice a day into one pill taken twice a day. This was hardly an improvement in compliance, but a step in the right direction. The key opinion leaders we worked with told us we were wasting our time and should be concentrating all research efforts on developing new AIDS drugs and searching for a cure.

Early forecasts took into account the AZT market share decline and predicted a modest conversion from Retrovir and Epivir prescribing over to Combivir. However, when we conducted market research with prescribers and patients, we learned differently. We heard from patients that zidovudine + lamivudine combination therapy sounded like a great option that would make their life a little easier when it came to taking all of their pills. Referring to AZT as zidovudine had an incredible impact on patients. Even when they learned that zidovudine was AZT, they still had a positive impression of zidovudine. Patients welcomed any opportunity to make pill taking easier, no matter how small of an improvement. The physicians we spoke with indicated that they would do anything possible to help improve compliance and would gladly switch AZT/3TC patients to Combivir or use Combivir for new patients.

We ran a quantitative forecasting study and learned that HIV prescribers anticipated that three quarters of all Retrovir prescribing would be converted to Combivir. When we added up the total anticipated prescribing for AZT (from both Combivir and Retrovir), it was much higher than the current AZT prescribing levels (from Retrovir alone). Could this simple combination really impact sales, drive up overall AZT market shares as well as improve compliance?

I proposed a revised forecast that significantly cut the Retrovir forecast and nearly doubled the Combivir launch forecast. I was skeptical of our findings and would have had to cut the Retrovir forecast much further and more than double the Combivir forecast if I were to believe our market research results were completely accurate.

As it turned out, our market research findings were correct. Retrovir sales dropped much further than initial estimates and Combivir sales were more than double the initial estimates. Combivir was a smashing success that saved the AZT franchise and stopped the declining AZT market shares.

With advance knowledge of customer behaviors and drivers of behavior, we were able to adjust our early forecasts and adjust production schedules to ensure that we could maintain product supplies and keep up with product demand to maintain the launch momentum.

Although this example may not be an early stage forecast, it illustrates the value of understanding customer behavioral drivers and the impact behavioral drivers can have on new product demand. In this case expectations were changed as greater than previously expected demand was anticipated. As stock-outs can negatively impact new product launches, changing production schedules long before the product launched enabled Glaxo to maintain adequate market supplies and enable patients to access Combivir without any delays.

For early phases of product development, developing early forecasts incorporating the impact of customer behavioral drivers in forecasting assumptions will not only provide a clearer picture of anticipated future product demand, it will also support new product development, commercialization planning, and investment decisions.

In Summary:

Employing a strategic forecasting process for early stage products involves:

· Learning your market environment

· Understanding behavioral drivers, and

· Linking these findings to an assumption driven forecast model

Although this may require a little more time and up front investment, it will yield far greater returns in the long run by supporting new product planning decisions and optimizing product development efforts. Maybe your forecast will not be perfect, but if you understand how your market works, not only will you make better projections, you will also make better strategic decisions on how to drive market levers and develop compelling products with high adoption and reimbursement potential.