After Thursday's close, Novavax (NASDAQ:NVAX) announced its Phase 3 trial of its vaccine to prevent RSV (respiratory syncytial virus) infection in older adults did not show efficacy. Due to the amount of additional material to cover related to the trial, I'm expanding the Novavax series to 4 articles in order to properly cover the results and the company. In this article, I will examine:
- the Phase 3 results and management's statement that the lack of demonstrated efficacy was likely due to the recent RSV season having the lowest attack rate ever recorded; and that trials of effective flu vaccines also fail to show efficacy during low attack rate seasons;
- the company's options for the RSV vaccine for older adults going forward; and
- what the information means for the stock.
In the next articles, I will examine:
- the company's other pipeline candidates;
- inaccurate claims that have been made about the company in articles and comments;
- what the company's valuation should be based on its pipeline candidates; and
- 6 potential catalysts in the coming months.
Phase 3 results
For starters, negative Phase 3 results do not necessarily mean the end of the road for vaccine and drug candidates. For example, sometimes trials fail because the trial methodology wasn't right, because two or more doses are needed instead of just one, or because the product needs to be combined with another drug for success. All are things that can be adjusted to result in success.
In a conference call, management said it thought the lack of efficacy of the trial was most likely due to the fact that this recent RSV season had the lowest attack rate ever on record. Attack rate is the percentage of people who get an infection. The attack rate related to the primary endpoint in the recent season in which the trial was done was 78% lower than the previous year: .4% in 2016 versus 1.8% in 2015. The attack rate for the secondary endpoint was 60% lower than last year, and was more than 33% lower than the lowest attack rate ever.
The company's President of Research and Development Gregory M. Glenn, M.D. said that clinical trials of flu vaccines that have taken place in low attack rate seasons also haven't show efficacy in such seasons.
CEO Stanley Erck said the same:
"The clear difference between [Phase 2] and [Phase 3] is the attack rate. The RSV attack rate in our study represents the lowest attack rate ever reported - by a lot. We had a bad RSV season - at least from a vaccine development point of view. Was it bad luck? Yes. Was it predictable? No. You've heard Greg lay out that a lower attack rate co-relates with lower efficacy, not only in our study, but in multiple reports of clinical trials with flu vaccines."
Erck also noted that 9 out of 9 RSV trials with humans showed efficacy, and many animal models showed the same:
"We have a preponderance of data that tell a consistent story of the effectiveness of our vaccine. Over many years, we have consistently shown in well accepted animal models that our vaccine stimulates a robust immune response in every way that we could measure... and shown in well accepted animal models that our vaccine protects these animals when vaccinated and then challenged with RSV. And in 9 clinical trials with pediatrics, women of child-bearing age, pregnant women, healthy adults and in the older adult population, we have consistently shown robust immune system responses by every measurement that we have. And finally, in 2 trials in the older adult population … we have shown that our vaccine provides protection from acute respiratory disease. There is a lot of data pushing us to what I believe is the correct analysis, and that is our vaccine works."
Given I own a sizable amount of shares and was considering buying more on the massive price drop after the announcement, I spent 10 hours researching whether their statement about lack of efficacy in low attack rate seasons was true because I think it's the key question related to the results. The data of the 9 trials and animal models are known facts, and are the main reason why almost all biotech analysts and experts thought the probability of Phase 3 success was high. It might even be "all" because I didn't find one who didn't think the probability was high, but I'll use "almost all" in case there's one or two I missed.
Phase 2 versus Phase 3: biotech drugs and vaccines
For some context about Novavax's Phase 2 trial, normally Phase 2 trials have between 100 and 300 participants as described here and here. Novavax's Phase 2 trial for the adult RSV vaccine had 1,600 participants. That's 8 times higher than the average Phase 2 trial. In fact, as the links above note, Phase 3 trials usually range between 300 and 3,000 participants or between 500 and 2,000. The mid-points of those ranges are 1,650 and 1,250. So Novavax's Phase 2 had as many participants as the average Phase 3 trial.
Also, Phase 2 trials are usually done at only 1 to 2 sites (such as a clinic or medical center) and Phase 3 trials are done at numerous sites. Novavax's Phase 2 trial was done at 10 sites around the U.S. so it was also similar in that way to a Phase 3 trial.
The largest ever study of clinical development success rates was published this year by the Biotechnology Innovation Organization (BIO) in partnership with Amplion and Biome. It analyzed 9,985 clinical and regulatory phase transitions at 1,103 companies. It found Phase 3 trials between 2006 and 2015 on average were successful 60% of the time. The study, and others before it, have found that cancer therapies have a much lower rate of success, so excluding those, the average rate is about 65%. Phase 2 trials only have a 31% success rate, so the average success rate of Phase 3 is roughly double.
But what about for vaccines? Maybe they fail more often in Phase 3? A comprehensive study by five researchers published in PLOS One examined over 900 vaccine candidates between 1998 and 2009. It found that on average only 18% of vaccines make it through Phase 1 and only 21% make it through Phase 2. Of the ones that make it through Phase 2, about 66% make it through Phase 3.
Success rates are different based on the type of infection, as this graph from the study shows. The 1st icon in each line is pre-clinical, the 2nd is Phase 1, the 3rd is Phase 2 and the 4th is Phase 3. When you look to the left of the icon, you see the percentage of candidates that made it through that phase. For example, if you look at the third square in the HIV/AIDS vaccines' line, you'll see that about 8% made it through Phase 2. For prepandemic flu, which is the type closest to RSV, 100% of the vaccines that made it through Phase 2 made it through Phase 3. For malaria and anthrax, 0% made it through Phase 3.
As this graph shows, vaccines for acute infections (i.e. ones that come and go like the flu and RSV) that make it through Phase 2 succeed in Phase 3 about 72% of the time. By contrast, vaccines for chronic infections only make it through Phase 3 about 28% of the time.
Another important data point on the chart is that prophylactic vaccines (to prevent or ameliorate the effects of a future infection) make it through Phase 3 about 87% of the time. All of Novavax's vaccines including its RSV ones are prophylactic. Therapeutic vaccines (to treat infections after they occur) only make it through Phase 3 about 13% of the time.
So the types that the RSV vaccine is most similar to succeed in Phase 3 about 87% (prophylactic), 72% (acute) and 100% (flu) of the time, respectively. Another factor is the size of the Phase 2 trial, and its p-value, which indicated a 96% probability that the vaccine is effective.
The key question becomes: is management's statement about low attack rate seasons correct?
Data on flu vaccine trials in low attack rate seasons
The first report I found was an article in the Journal of Infectious Diseases. In it, a team of nine researchers describe a flu vaccine trial by the University of Michigan School of Public Health in conjunction with a few other universities with support from the National Institute of Allergy and Infectious Diseases, and the federal Center for Disease Control (CDC). The trial was conducted over two years and tested two different flu vaccines.
In year one, the attack rate was 7.8%, and the vaccine efficacy was 75% for the inactivated vaccine and 48% for the live attenuated vaccine. In year two, the attack rate was only 1.8%, and "neither vaccine showed significant benefit over placebo" for any of the end points.
The researchers concluded: "Given the lower than expected attack rate, for these end points the study was underpowered to measure statistically significant vaccine efficacy." This was the case even though in year two "more participants than projected were enrolled."
In addition, the low attack rate also helped eliminate the efficacy differences between the two vaccines. In year one, the inactivated vaccine had efficacy between 36% and 58% higher than the live attenuated vaccine. But in year two, they had roughly equal efficacy: both low.
"In 2004-2005, the live attenuated vaccine performed less well than the inactivated vaccine when each was compared with the placebo group, but in the second year of the study there was little evidence of a difference. How can this be explained? … in contrast to the first year, the low attack rates in 2005-2006 made statistically significant reductions produced by either vaccine difficult to demonstrate."
So were these nine researchers low quality? I clicked the National Institutes of Health link for the article's lead author, Suzanne Ohmit, to check her background. She's written 62 articles in peer-reviewed medical journals, including the prestigious New England Journal of Medicine. Almost all of the articles are about flu vaccines, many of them have been cited in over a hundred other medical journal articles, and her articles have been used in numerous news stories including this Business Insider story (click the 3rd and 5th links and they go to an article she was the lead author on). She's an expert on vaccines.
In addition, the University of Michigan has turned up in numerous medical articles I've read about vaccines, including this NPR story. She is the main Research Scientist of a group at the university dedicated to respiratory infections particularly influenza, and her expertise is "influenza vaccine efficacy/effectiveness evaluations." So when an expert in efficacy evaluations states that a trial could not "measure statistically significant vaccine efficacy" due to "the lower than expected attack rate" and says the same thing two other times, I tend to believe her. But let's see what others have to say.
A report by 7 researchers published in BMC Infectious Diseases, a peer-reviewed journal with 12 vaccine experts on its editorial board, describes a flu vaccine trial with 6,374 participants over the course of two years that failed to meet its endpoints.
The researchers stated that:
"A key factor influencing efficacy findings in influenza vaccine studies is the disease attack rate, and the unpredictable nature of epidemic intensity poses a challenge when planning influenza vaccine efficacy trials … in Season 2 the attack rate was less than half of that predicted, and this had a large negative impact on the precision of the average efficacy estimate. … the result must be interpreted with some caution in view of the very low influenza attack rates in both seasons; a factor which has previously been associated with low efficacy estimates in influenza vaccine trials."
Moreover, the researchers stated:
"A negative impact of low attack rates in a given season on point estimates of efficacy has been well documented. Previous studies of a [trivalent influenza vaccine] product, involving the same investigators, methods, and source populations over successive seasons, have reported dramatic differences in efficacy estimates between seasons, without obvious explanation based on circulating strain match to the vaccine. In these studies, the low efficacy estimates were found in seasons with lower overall attack rates; a similar experience has occurred in pediatric trials."
These researchers and other before and after them state that no one has figured out the reason why low attack rate seasons don't show efficacy of vaccines that show efficacy in normal attack rate seasons. But they say they're fairly sure it's not simply due to drifts in flu strains from one year to the next. This is because they can measure how much drift occurred and whether it could have affected results in a significant way. For example, the BMC Infectious Diseases researchers noted: "While a clear reason for this phenomenon is not apparent … antigenic drift was modest and had little apparent impact." And numerous other researchers have said the same.
Another article by 14 authors in the Journal of the American Medical Association examined a two season flu vaccine trial in Pittsburgh. In the first season when the attack rate was 15.9%, vaccine efficacy was estimated to be 65%. In the second season when the attack rate was 3.3%, vaccine efficacy was estimated to be -7%. That's a much bigger swing than in Novavax's two seasons from a normal attack rate season to a low attack rate season.
Another study by 11 researchers published in 2009 in BMC Infectious Diseases of a flu vaccine trial with 6,200 participants in 2005-2006 concluded: "Due to the exceptionally low attack rate during the 2005/2006 season … our study was unable to demonstrate the efficacy of [trivalent influenza vaccine] in healthy adults." Note that in this study, there was some drift in flu strain, but the researchers said the low attack rate was a major reason for the inability of the trial to show efficacy.
The researchers emphatically stated that the inability of trials during low attack rate seasons to show flu vaccine efficacy did not mean flu vaccines don't work, and was not reason for people to not get flu shots:
"The low attack rates encountered in the adult population during some influenza seasons may raise the question about the need for adult vaccination. It should be kept in mind that currently the main target populations for adult vaccination (apart from the elderly persons, which are outside the scope of this study) are adults who have medical conditions that place them at risk for complications from influenza and also, in order to limit influenza virus transmission to those at risk in all age groups, health-care personnel or caregivers. As the attack rate for any coming influenza season is unpredictable at the time of annual influenza vaccination, it is not feasible to modify vaccination routines and put susceptible people at risk."
Other information from the conference call
If you haven't listened to the conference call, I recommend that you do here. It has valuable information and will show that a few claims made about it aren't true, including the claim that management gave investors nothing to go on during the call, or that they tried to move the focus away from the RSV vaccine and to the zika vaccine. They did touch upon the zika vaccine, but it was a fairly small part of the call. Most of the call was spent on the RSV vaccine for older adults. This includes the portion of the call where management spoke before Q and A, during which they could have talked about anything they wanted to, and they largely focused on the main topic of the call: the RSV vaccine for older adults.
In addition, management stated that governmental research has shown that RSV attack rates for children are fairly consistent, but there is less existing data on the consistency of attack rates for older adults. They also noted, though, that this attack rate was the lowest ever recorded for RSV by a significant amount.
Management noted that there is a chance that the results could possibly be due to a change in the vaccine from one trial to the next. They said that they have rigorous systems in place to prevent this from happening, but that they will examine this to be certain. Erck said, "It's possible that we've missed something, and that is of course a very high priority for us to analyze today, but to repeat, we have no data that steers us to the conclusion that the product has changed in a substantive way from last year's 201 trial."
They also noted, as is always the case with vaccines and drugs before approval, there is a possibility that this trial's results could be due to the vaccine not working. However, Novavax's RSV vaccines have worked in 9 out of 10 human trials and numerous animal models/tests, and the only clear difference between all of the successful trials and the one that didn't show efficacy is a low attack rate season.
An article earlier this week implies that because a Novavax executive said on the May 4th earnings call that he thought the recent RSV season was normal, that it's not right for the company to talk about the low attack rate being the likely cause. To me, that is a weak claim. The company only unblinded the results of the Phase 3 trial five days before the results were announced. The attack rate in the results is what matters.
No other RSV trials by any company were done during the past season, so Novavax's current data is more accurate than whatever anecdotal or less comprehensive information about the season was available in the spring. Whether an executive 4 1/2 months ago indicated that a normal RSV season occurred based on less comprehensive initial information doesn't matter for the key question at hand.
On that question, several dozen scientific vaccine experts agree that Novavax's explanation of the likely cause is reasonable. They say that lack of efficacy in trials is the norm in low attack rate reasons. I personally will go with them. I'm pretty sure the FDA will also go with the scientists, especially for an infection that kills 15,000 people and hospitalizes 200,000 per year. Note that researchers do not study the number of hospitalizations and deaths every year. They periodically do this and take the averages.
Options for the older adult RSV vaccine
Given all of the data above, the company is planning one or more additional trials for the RSV vaccine for older adults. Options include:
- Run a very similar Phase 3 trial given that the next season they test in will probably be a normal attack rate season.
- Adjust the Phase 3 trial somewhat to focus on parts of the country that tend to have higher attack rates. They discussed this option during the call.
- They also said they might do a trial in which people get a booster shot the first year, and then normal annual shots in ensuing years. This is due to the results of the other trial they announced last week which showed very good efficacy for people who received a shot the previous year. This trial used people who were in the Phase 2 trial to see how they would do with a follow-up shot.
- An analyst asked them about the possibility of taking a trivalent approach similar to flu shots which always include 3-4 strains of the flu in their shots to increase the range of protection. Management said they're considering that option.
- The company said it also will consider other ways in which to adjust the trial(s) for this vaccine.
The big picture
If you average the Phase 3 success rates of prophylactic vaccines (87%) and acute ones (72%), it's 79.5%. Before this, based on the large Phase 2 trial size and p-value, the societal impact of a disease (which the FDA considers), the safety profile, the FDA's modification of the endpoints to make Phase 3 easier to pass, and other reasons, I had the probability of Phase 3 success at 20% higher than normal: about 96% probable. All the scientific research and data says that results in low attack rate seasons are basically useless and don't apply. To be conservative, however, I am reducing the probability of Phase 3 success to be 25% below the average success rate: putting it at about 60% probable.
I encourage you to come up with your own probability as part of your investment decision. Based on the scientific consensus about data in low attack rate seasons, should the vaccine have an above average probability as before? Or should it be moved to average? Or below average? Should you factor in improvements the company can make to the trial?
While it can be hard to settle on a number, investing decisions in biotech need to be made primarily based on probabilities of trial success combined with the market size relative to the current share price.
Based on a 60% probability and large market size for this RSV vaccine, the recent low share price and all of the company's other pipeline products, I increased my number of shares five-fold in the last two days for a very low price.
Given the massive size of the completely untapped RSV market, even a 40% probability of Phase 3 success of this vaccine means the company is worth a lot more than its current share price based on the older adult RSV vaccine alone, not to mention its many other vaccine candidates. Over time, I'm confident that emotion will give way to information, and the stock price will most likely go up a lot. Institutions and wealthy biotech investors will probably lead the way. Studies show that retail investors on average react more heavily to news and tend to buy or sell in the first 8 hours after the news, while institutions wait for a day or more to act and react less heavily.
After reading a couple of articles, I had the impression it was too late for the company to do a trial this coming season for this RSV vaccine. However, I reviewed the transcription I made of the conference call and I found that CEO Erck states: "We'll have a better idea of the pathway and timing of a BLA when we decide the nature of our trials that will start this fall, and how we will expect to follow-up those trials next year. I can't say more than that right now, but I would expect to have more insight on that issue when we meet at our analyst day meeting."
A BLA is the application for FDA approval that companies submit after a successful Phase 3 trial, and I think he is referring to a BLA for the RSV vaccine for older adults. He's definitely not referring to the maternal RSV vaccine trial because that started about 10 months ago and will go until early 2018. It's not completely clear he's referring to trials for the older adult vaccine, but it seems he is. If yes, that is a positive for the company because it means new results might come as soon as next September.
Based on all of the data above, I believe the 85% sell-off Friday was significantly overdone. Moreover, the company has several other candidates in its pipeline, including two which have passed phases 1 and 2, which are well-documented to be the hardest stages to get through. As noted above, while nothing is guaranteed, vaccines that are prophylactic and for acute infections have high success rates in Phase 3. With a 79.5% average Phase 3 success rate for similar vaccines, the chances that none of their vaccines will pass Phase 3 is very small. If even one of its vaccines has positive Phase 3 results, the share price will skyrocket.
In the next articles, I will examine the other pipeline candidates and assess what the company should be valued at based on those candidates and the older adult RSV candidate. I will also examine some claims being made about the company (for example, about its cash), and give correct information on them. Finally, I will describe six potential catalysts that could move the stock price much higher in the coming months.
Disclosure: I am/we are long NVAX.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.