Problems in Vaccine Development

If it were as easy to know what were good to do,

chapels had been churches,

and poor men's cottages princes' palaces

—William Shakespeare

The Merchant of Venice

Hilary Koprowski was an optimist. He always saw obstacles as opportunities. I am grateful to him for teaching me the value of persistence, as well as receiving his constant encouragement. Hilary was responsible for two vaccines: oral polio (or at least the idea of it) and cell culture rabies. He also created the atmosphere at The Wistar Institute that facilitated the development of rubella and pentavalent rotavirus vaccines. I was reminded of Hilary when I was asked to consider the present state of HIV vaccine development, which has been marked by many failures and only one partial success. For after all, vaccine development has never been easy and I was moved to point out that the difficulties in developing an HIV vaccine are similar to those seen with vaccines against other viruses. Hilary was fond of putting quotations into his writings, so I will quote one from Emile Roux, the close collaborator of Louis Pasteur, who when inaugurating a monument to Pasteur said, “Science appears calm and triumphant when it is completed: but science in the process of being done is only contradiction and torment, hope and disappointment.”

That has been the case for three of the four phase 3 efficacy trials of candidate HIV vaccines conducted thus far, with complete failure to protect after immunization with the viral glycoprotein; complete failure to protect after immunization with a non-replicating adenovirus vector carrying the gene for HIV gag; complete failure to protect after prime-boost immunization with a DNA plasmid followed by an adenovirus vector, both carrying the genes for multiple HIV proteins; but 31% protection by a prime-boost regimen consisting of a canarypox vector followed by the viral glycoprotein.⁽²⁾ In fact the protection was higher than 31% during the early months of the trial, but faded later.

The great surprise from the canarypox/protein trial, conducted in Thailand, was that the correlate of protection appeared to be binding antibody with the function of antibody-dependent cellular cytotoxicity, rather than a neutralizing antibody. The protective antibody bound to the V1/V2 loops of the virus.^(3,4) This means that there are now at least four targets for induction of antibody against HIV: the V3 loop, the CD4 binding site, the membrane proximal region, and the V1/V2 loops. The first three targets are for induction of neutralizing antibodies.

But the result of the Thai trial also has an implication aside from the prevention of HIV infection. It reminds us that neutralization is not the only function of antiviral antibody. It was already known that non-neutralizing antibodies may be important for the control of Herpes simplex, cytomegalovirus, respiratory syncytial virus, lymphocytic choriomeningitis virus, and influenza.^(5,6) The case of influenza is particularly interesting, in that non-neutralizing antibody may play a major role in heterosubtypic immunity.

Coming back to HIV, an interesting report shows that immunization with a peptide from the gp41 portion of the viral envelope glycoprotein can protect against destruction of CD4+ T cells by natural killer cells.⁽⁷⁾

Virologists are also learning that structure of viral proteins is important for immunization. Dengue is a disease caused by flaviviruses of four serotypes, and it has long been thought that classical neutralizing antibodies against the envelope and premembrane proteins protect against infection. Recently, a vaccine composed of four chimeric viruses based on a yellow fever 17D attenuated virus as the vector for the E and PrM genes of each of the dengue viruses was tested in Thailand. The surprising result was that although there was some protection against types 1, 3, and 4 dengue, there was none against type 2.⁽⁸⁾ Table 1 lists some possible explanations for this failure. The last possibility is the most interesting,⁽⁹⁾ having shown that the conformation of the virus in mosquitoes is not the same as the virus in cell culture, possibly because the former is produced at 28°C whereas the latter is produced at 37°C. The 28°C virus is more compact and it may be that different epitopes are exposed on the type 2 virus injected by the mosquito. Elucidation of this phenomenon will be interesting, as our prediction that antibodies to the viral envelope that neutralize in vitro would be protective may be wrong.

OPEN IN VIEWER

Table 1.

Possible Explanations for Low Efficacy of Chimeric Dengue Vaccine

• Higher challenge dose of type 2, therefore more antibodies needed

• Dengue Type 2 infects monocytes rapidly and antibody thus not effective

• T cell response also needed

• Type 2 antibodies were heterotypic, not homotypic

• Chimera presents envelope protein in different conformation

Once again the lesson stated by the Spanish philosopher Miguel Unamuno is apt: “True science teaches, above all, to doubt and to be ignorant.” This was a lesson that Hilary never forgot, which permitted him to be creative.

There are two more obvious factors that influence efficacy of vaccines: challenge dose and duration of immune response. The first was demonstrated clearly in an experiment conducted in volunteers with an experimental attenuated virus vaccine against cytomegalovirus.⁽¹⁰⁾ As shown in Table 2, challenge with 1000 PFU could overwhelm natural immunity, even though seropositives were protected against challenge with a lower dose. This principle was also evident in the Thai HIV vaccine trial, in which efficacy was seen in low risk subjects but not in high risk subjects, who were presumably exposed more often.

OPEN IN VIEWER

Table 2.

Doses of Subcutaneous CMV Challenge Required to Infect or Cause Disease in 50% of Different Groups

	Infection	Disease
Seronegatives	<10 PFU	<10 PFU
Natural seropositives	≈500 PFU	1000 PFU
Vaccinees	100 PFU	>100 PFU

Memory after vaccination, both effector and central, plays an important role in protection. Two recent examples will serve: pertussis and mumps. Pertussis is resurgent in the United States and elsewhere where acellular vaccines are used. The reason is waning antibodies. While antibodies are found at high levels in the first two years after preschool booster vaccination, they fade thereafter and so does protection against pertussis.⁽¹¹⁾ Similarly, although the mumps vaccine is moderately effective, particularly after two doses, there have been many outbreaks in vaccinated populations living in close contact. Studies of memory B cells show that while good numbers are seen after measles and rubella immunization, mumps vaccinees have relatively few.⁽¹²⁾

Thus, there are many unsolved problems in vaccinology, among which are prolongation of immune memory, the multiplicity of virulence antigens, immaturity and post-maturity of the immune system, and the need for adjuvants that give selective stimulation without general excitation of immune cells. These are difficult problems to solve; as John Milton wrote, “long is the way and hard, that out of hell leads up to light.” Hilary would have agreed, and his optimism allowed him to traverse the path to light. Without him, there might never have been an oral polio vaccine, as it was he who had the courage to proceed when others were afraid.⁽¹³⁾