How can we expedite HSV-2 vaccine discovery?



My earlier posts emphasize some of the reasons that past attempts to obtain a HSV-2 vaccine have not succeeded.

I now turn my attention to outlining two steps that, if implemented, would expedite the rate of HSV-2 vaccine development efforts, and these are:

Step 1.  Appropriate Use of Small Animal Models (i.e., De-Risking HSV-2 Vaccine Development)

Step 2.  Streamline the Path to Phase I Human Clinical Testing of HSV-2 Vaccines

There is too much ground to cover here in a single post.  However, I hope to provide a reasonably concise overview of these two critically important issues that, if addressed, would greatly accelerate the rate of HSV-2 vaccine discovery.  In subsequent posts, I will tie up the loose ends that are not fully covered herein.

The crux of the issue is that, for the past 30 years the FDA’s favorite HSV-2 vaccine approach, the subunit vaccine, has not made one iota of a difference in our ability to treat or prevent HSV-2 genital herpes.  During this period of time, tens of millions of people have continued to suffer with HSV-2 genital herpes, and hundreds of millions more have acquired HSV-2 genital herpes.

As we look down the road 20 years, to 2033, I would suggest that it is time for the FDA to become a little bit less singleminded in their concern about the ABSOLUTE RISK of a Phase I Clinical Trial of a new HSV-2 vaccine candidate in n=10 human volunteers.  Rather, I would suggest that it is time for the FDA to weigh the RELATIVE RISK of a clinical trial of a new type of HSV-2 vaccine versus the status quo that has allowed a vaccine-preventable disease, HSV-2 genital herpes, to continue spreading unchecked through the American public for decades.

I would suggest that the manageable risk of a live-attenuated HSV-2 vaccine would be vastly preferable to the HSV-2 genital herpes epidemic, which will still be ongoing in 2033, if the FDA continues to promote and foster a culture in which it is perfectly acceptable for scientists to say that “a live-attenuated HSV-2 vaccine would be too dangerous to test in humans.”  There is simply not a shred of evidence to support such baseless claims.  This common misconception about the “dangers” of a live-attenuated HSV-2 vaccine has been perpetuated by nothing more than ignorance of the facts.   To the contrary, there is a mountain of evidence that demonstrates that HSV-2 (or any other viral pathogen) may be very stably attenuated by using a combination of (1) the right kinds of stable genetic modifications (i.e., large deletions) placed (2) into one or more strategically-chosen viral genes.

I elaborate, as follows.




One of the primary tools by which new vaccine candidates are evaluated are small animal models, such as mice or guinea pigs.  The tests that scientists run on HSV-2 vaccines in mice or guinea pigs are an important component of “pre-clinical testing” in which we attempt to gauge how safe and effective a HSV-2 vaccine candidate would be in humans.

In recent years, increasing numbers of scientists have begun to question if mice and/or guinea pigs are really a good model for evaluating the effectiveness of new HSV-2 vaccine candidates.  Thus, some investigators have suggested that we should explore new animal models for pre-clinical testing of HSV-2 vaccines, such as mice or rabbits that have been “humanized” (i.e., to contain human immune cells) or primate models such as monkeys or apes.

I understand the desire to blame the Herpevac vaccine failure in human clinical trials on a bad animal model that “lied to us,” and told us that Herpevac would work in humans.  However, I note that in my own studies in mice and guinea pigs, I find that Herpevac-like, glycoprotein subunit vaccines are ineffective and elicit only a small fraction of the protection against HSV-2 genital herpes that is possible (Halford, et al., 2011,  If mouse HSV-2 vaccine-challenge models were intrinsically flawed, I would expect that the mice would lie to me, in the same way that other investigators have claimed that their mice and/or guinea pigs lied to them.

Animal models of HSV-1 and HSV-2 infection are the primary area of my research expertise. During my graduate training with two herpes immunologists, Dan Carr and Bryan Gebhardt (1992-1996), I became quite familiar with the literature on herpes immunology which dates back to the 1970s and earlier.  Over four decades, herpes immunology studies in mice have repeatedly informed us of new principles about the response of the verterbrate immune system to HSV, and these findings have generally translated well to the human condition.  There are nuances about the mouse model that don’t correlate with the human situation; for example, mice catch on to HSV-1 and HSV-2 reactivation REALLY FAST, as several viral immune evasion mechanisms that work well in humans don’t do their job in the mouse (e.g., ICP47 and gE-gI).  However, setting these minor differences aside, I see no hard evidence to support claims from other investigators that mice and guinea pigs are not a perfectly effective model for screening HSV-2 vaccines to figure out which approaches work (live-attenuated HSV-2 vaccine) and which are hopelessly ineffective (gD-2 subunit vaccine) at eliciting a protective immune response against HSV-2 genital herpes.

I once heard a vaccine researcher from Italy (working on a meningitis vaccine) express the same sentiment in a talk, but he was far more eloquent in his choice of words.  Apparently a similar dialogue about the reliability of mice was playing out in his field.  What he said went something like this: “Many people in my field think that vaccine studies in mice are unreliable and have no value in predicting which meningitis vaccines will work in humans.  Personally, I have not found that the mice lie to me, but I do find that it is extremely important to ask them the right questions.”

One of the most fundamental limitations of HSV-2 vaccine-challenge studies performed in small animal models is the CHRONIC ABSENCE OF A POSITIVE CONTROL, which may be used to empirically define what “100% protection” against HSV-2 genital herpes should look like.  Rather, most HSV-2 vaccine studies only compare naïve animals (0% protected) versus vaccinated animals, and conclude that the vaccinated animals were better protected than the naïve animals.  This is a good starting point, but the next, essential step is to measure whether the HSV-2 vaccine candidate elicits 0.3% protection against HSV-2 genital herpes (i.e., statistically significant, but useless) or alternatively elicits something close to complete (~100%) protection against disease.

To define how good a HSV-2 vaccine is on a scale of 0 to 100% protection, an investigator must include a positive control to define what one should consider as “100% protection against HSV-2.”  Such a positive control group may be established in a vaccine study by infecting a control group of animals with wild-type HSV-2 under conditions that limit the virus’s capacity to spread and cause disease.  A simple way to achieve this goal is to provide animals with oral acyclovir (anti-herpes drug) in their drinking water during the first 3 weeks of exposure to wild-type HSV-2, which limits but does not completely prevent viral replication and spread.  It is well established that any species of animal (mouse, guinea pigs, or rabbits) that survives a 1st exposure to HSV-2 will become latently-infected and uber-resistant to exogenous challenge with a 2nd dose of wild-type HSV-2.  This is one of many ways that a researcher may create a group of “HSV-2 latently infected animals” that would serve as a positive control that roughly approximates what “100% protection against HSV-2 vaginal challenge” might look like.

This technology has been around since the 1980s.  The chronic absence of “HSV-2 latently infected animals” in HSV-2 vaccine studies means that most HSV-2 vaccine researchers either (1) don’t know how to run a properly controlled animal experiment (which I doubt), or (2) don’t want to include a positive-control group that might reveal that their test HSV-2 vaccine is not terribly effective.

I close by noting that the mouse and guinea pig vaccine-challenge studies that initially served as proof that a Herpevac-like, glycoprotein D subunit vaccine “should be effective” did not include a positive-control group of HSV-2 latently infected animals.   Had such a positive control group been included, I am confident that a Herpevac-like, glycoprotein D subunit vaccine would have never advanced to human clinical trials because it would have been painfully apparent that this approach only elicits 2 to 5% of the protection against HSV-2 that is possible.

Flawed experimental designs, not bad rodents, are the real culprit that explains why small animal models have not served as a more realistic gauge of HSV-2 vaccine efficacy in the past.

A recent publication from my lab ( elaborates on another misconception propagated by inappropriate use of animal models in the past; namely, the erroneous belief that there are no “correlates of immunity” available to quickly and easily differentiate a robust versus an anemic HSV-2 vaccine candidate.

BOTTOM LINE:  If the pharmaceutical industry wishes to “de-risk” HSV-2 vaccine development, then an important step in that direction will be to quit making multi-million dollar decisions based on subpar animal studies.  As with all other areas of scientific inquiry, the inclusion of a positive control group in HSV-2 vaccine-challenge studies would be a huge step in the right direction.




The FDA regulatory process is intended to protect the public against potentially harmful products that could be sold or peddled as “medical treatments.”  In the complete absence of any safety regulations or oversight, all forms of snake oil and worse could be sold under the guise of medical treatments.  As a consumer, I greatly appreciate what the FDA does for us U.S. citizens in terms of ensuring that the food and drugs we ingest are safe.  This is a critical public service whose value cannot be overemphasized.

In the realm of vaccines, the FDA applies the exact same logic to vaccines that they apply to food.  This is an error in logic.  Food items such as jelly beans are either safe, or they are not.  In contrast, when one obsesses over the safety of a HSV-2 vaccine and keeps it out of clinical trials for years to decades, this action has the consequence of allowing millions of people per year to continue to be infected with disease-causing strains of wild-type HSV-2.  Therefore, I would suggest that it is inappropriate to singlemindedly obsess about whether or not a new HSV-2 vaccine poses ANY RISK.  All vaccines pose a risk in the absolute sense of the word, but good vaccines pose a risk that is vanishingly small.  Therefore, a vaccine’s risk should be weighed in terms of RELATIVE RISK.  That is, what is the risk of adverse events associated with injecting people with a new HSV-2 vaccine candidate versus the risk of doing nothing and letting 20 million people per year continue to be newly infected with wild-type HSV-2?

Right now, the FDA regulatory approach towards new HSV-2 vaccine candidates focuses on assessing whether there is ANY RISK AT ALL that a HSV-2 vaccine might cause overt symptoms or disease if injected into a million vaccine recipients.  Importantly, the FDA regulatory process seeks to address this hypothetical question BEFORE A SINGLE HUMAN SUBJECT CAN BE IMMUNIZED in a clinical trial.  Of course, in the absence of the FDA allowing small-scale testing of new HSV-2 vaccines in humans (e.g., 10  subjects), it is impossible for a scientist to counter these hypothetical concerns with data that directly demonstrates safety in humans.

Hence, the FDA has created a “Catch-22″situation, which goes something like this:

1.  The FDA won’t allow a new HSV-2 vaccine to proceed to Phase I clinical trials in 10 – 20 human subjects until a laundry list of hypothetical concerns are addressed through pre-clinical testing in animal models.  No matter how strong the animal data, the pre-clinical data may always be met with a new set of “concerns” from the FDA that “OK, aspect A of the safety of the HSV-2 vaccine candidate looks good in animals, but what about aspects B, C, D, etc. which you have not addressed.”  Moreover, no matter how many iterations of tests you run in animals, the FDA may still tell you 5 years later that, “Well your HSV-2 vaccine is clearly very safe in animals, but how can we know that it will be safe enough in human recipients?”

2.  The formula for sufficiently addressing all of the FDA’s concerns about the safety of a new HSV-2 vaccine candidate may easily require 10 years and $30 million dollars in legal and paperwork filing fees, which a sponsoring company may or may not recover.  If all goes swimmingly well, then a company that backed a given HSV-2 vaccine might break even on their investment in 15 or 20 years.  This is obviously a huge disincentive for companies to get involved with sponsoring a new HSV-2 vaccine approach, as it will take 10 years just to figure out whether or not the approach has the potential to fly in humans.

3.  Companies are reluctant to invest 10 years and $30 million in a fundamentally new HSV-2 vaccine candidate that differs from the types of HSV-2 vaccines that the FDA has approved in the past.  Thus, even if HSV-2 subunit vaccines are lame, most companies would rather invest in a lame HSV-2 subunit vaccine that the FDA is likely to approve for clinical trials, rather than support a new type of HSV-2 vaccine that is more effective, but may never get the green light from the FDA to advance to human clinical trials.

BOTTOM LINE:  The FDA regulatory process with the respect to vaccines has grown so onerous, that it has stymied the interest of most companies in pursuing HSV-2 vaccine research.  While I appreciate the FDA’s desire for safety, the potential risks of new HSV-2 vaccines should be weighed against the risks of continuing the status quo of focusing only on “uber-safe” HSV-2 subunit vaccines that are unlikely to be effective, and thus permitting the HSV-2 genital herpes epidemic to continue to spread in an unchecked manner.

If a vaccine is well designed, then the odds of serious adverse event approach the odds of winning the lottery (i.e., one-in-a-million).  In contrast, the continued, unchecked spread of HSV-2 genital herpes in the human population means that tens of millions of people will continue to live with recurrent HSV-2 genital herpes, and hundreds of millions more will be newly infected with disease-causing strains of wild-type HSV-2 by 2033.

HSV-2 genital herpes is almost certainly a vaccine-preventable disease.  However, to achieve this goal, and bring the power of vaccinations to bear, the FDA will need to re-think its position on the pros and cons (RELATIVE RISK) of testing new classes of live-attenuated HSV-2 vaccines, which will likely be at least 100 times more effective than the HSV-2 subunit vaccines that we have been exclusively testing in human clinical trials for the past 25 years.

I would suggest that small-scale “Compassionate Use Trials of HSV-2 Vaccines” are a compromise that could be struck to achieve a more appropriate balance between (1) ensuring the continued safety of HSV-2 vaccine trials versus (2) offering people with HSV-2 genital herpes some real hope that a therapeutic HSV-2 vaccine may be identified in their lifetime.  Specifically, Compassionate Use Trials of a therapeutic, live-attenuated HSV-2 vaccines would achieve the dual goal of (1) timely testing of a new HSV-2 vaccine modality in humans and (2) testing of a HSV-2 vaccine candidate that offers the greatest odds for success of a therapeutic vaccine capable of reducing the frequency and duration of genital herpes outbreaks in those already infected with HSV-2.

In a subsequent post, I will elaborate more fully on the concept of Compassionate Use Trials, and how this might be coupled with human trials of a therapeutic HSV-2 vaccine to greatly accelerate human testing of new classes of HSV-2 vaccine.

– Bill H.