Presented in part: 15th International Papillomavirus Workshop, Gold Coast, Australia, December 1996 (abstract 293).
A fusion protein vaccine consisting of human papillomavirus 6 L2E7 with Alhydrogel was developed for the treatment of genital warts. Twenty-seven subjects with genital warts received 3 immunizations over 4 weeks in an open-label study. The vaccine was well-tolerated, and all subjects made serum IgG antibodies, predominantly IgG1, against L2E7. Nineteen of 25 tested persons made antigen-specific T cell proliferative responses to L2E7, and peripheral blood mononuclear cells when cultured with L2E7 in vitro produced both interferon-γ and interleukin (IL)-5, although IL-5 predominated after the final vaccination. Five subjects completely cleared warts within 8 weeks. Subjects whose warts were not cleared by 8 weeks were offered conventional therapy. Recurrence of warts was not seen in any of the 13 persons whose warts cleared by vaccine alone or with conventional therapy. While these preliminary results of the use of this therapeutic immunogen are encouraging, proof of efficacy will require randomized double-blind trials.
Genital warts are benign neoplastic tumors caused by certain genotypes of human papillomavirus (HPV), principally HPV types 6 and 11 [1]. They are one of the most common sexually transmitted diseases worldwide, occurring each year in 1.5%–2.5% of European men and women aged 20–24 years [2, 3]. Similar incidences are estimated for the United States, where 500,000–1,000,000 new cases of genital warts are reported annually [4]. A wide variety of treatments are in use, but failure of treatment and recurrence after initial clearance is seen with all treatment modalities [5]. Data from England illustrate the frequency of recurrence and show that total prevalent cases exceed incident cases by a factor of 1.8 [6]. The economic burden of genital warts is substantial and estimated to exceed $3.8 billion in the United States for 1997 [4].
The concept of immunotherapy as a treatment for genital warts was first described by Biberstein [7] in 1925, although he used poorly characterized autogenous vaccine preparations. Such approaches to treatment using autogenous vaccine preparations remained in use for many years, until a double-blind controlled cross-over study showed no differences in outcome for such therapy versus placebo [8]. A number of animal papillomavirus model systems have been used for vaccine research, including the cottontail rabbit papillomavirus (CRPV) and the bovine papillomaviruses (BPV). The development of recombinant proteins lead to experiments using the BPV-2 and -4 models, which demonstrated successful prophylactic and therapeutic vaccination against these papillomavirus diseases [9, 10]. We therefore proposed to explore a similar immunotherapeutic strategy as a potential treatment for genital warts. Given that HPV-6 and -11 show a high degree of sequence homology [11] and that most genital warts are caused by HPV-6 [1], we developed a therapeutic vaccine based on a fusion protein of HPV-6 L2 and E7 proteins. Preclinical studies demonstrated that specific antibody lymphoproliferative and delayed hypersensitivity responses were elicited in mice immunized with an L2E7/Alhydrogel adjuvanted protein. Alhydrogel (Superphos, Copenhagen, Denmark) is the only adjuvant widely licensed for pharmaceutical use at present [12]. A phase I clinical trial evaluated the safety and immunogenicity of the vaccine in healthy male volunteers [13]. This trial revealed no serious adverse events after vaccination with the HPV-6 L2E7 fusion protein adjuvanted with Alhydrogel (TA-GW), demonstrated that the vaccine was immunogenic, and determined a dose and immunization schedule for further studies. The phase IIa trial reported here was initiated primarily to determine the safety and immunogenicity of the vaccine, TA-GW, in patients infected with genital warts, and to obtain preliminary clinical efficacy data.
Twenty-seven men with genital warts were recruited at the Genitourinary Medicine Clinic, General Infirmary, Leeds, United Kingdom. Sixteen subjects had persistent relapsing warts, defined as a history of lesions for 11 month and a history of previous failed therapy (but no treatment within the prior 2 weeks); 11 subjects had new warts, which were defined as a history of lesions for ⩽3 months and no history of previous therapy. Exclusion criteria included age < 18 years, human immunodeficiency virus (HIV) infection or other immunodeficiency, or any other significant medical disease.
All subjects were vaccinated intramuscularly with TA-GW into the deltoid at 0, 1, and 4 weeks after study entry. They remained in the clinic for 30 min after vaccination for observation and were given diary cards to record any adverse events in the following week. Subjects with new warts could opt at entry to receive conventional therapy using podophyllotoxin 0.15% cream (Warticon; Perstorp Pharma, Basingstoke, UK) and vaccination with TA-GW. Subjects were followed for 8 weeks, and all subjects with persisting warts were then offered conventional therapy as deemed appropriate by the physician.
For each patient at each visit, the extent and location of the genital warts were mapped and photographed. The two maximum perpendicular diameters of each wart were measured, and the sum of these products for all warts was defined as the “wart area.” A 100% reduction in wart area was defined as a complete response, a 25%–99% reduction as a partial response, and < 25% reduction as no response.
Noninvasive sampling for HPV DNA was done at each visit for up to 8 weeks by firm application of a cotton wool swab to the lesions. Samples were suspended in normal saline. Peripheral blood mononuclear cells (PBMC) and serum were separated from blood samples taken at weeks 0, 1,4, and 8 after the first vaccination. All blood samples were assayed for antigen-specific in vitro T cell proliferation and cytokine production within 18 h of sampling. Serum samples were frozen and later tested for anti-HPV-6 L2E7 antibodies.
HPV-6 DNA was isolated from a vulval condyloma by phenol extraction, and the L2 and E7 genes were amplified by polymerase chain reaction (PCR). A recombinant L2E7 gene fusion was constructed, expressed in Escherichia coli and purified from solubilized inclusion bodies under reducing conditions. The solubilized product was purified by successive chromatography on anion and cation exchange columns, followed by size exclusion and removal of urea by G25 gel filtration. The final product formed a discrete 0.22-μm filterable stable aggregate. A formulation of 300 μg of the fusion protein adsorbed onto 1.2 mg of 2% Alhydrogel (Superphos, Denmark) was the final vaccine preparation [13].
DNA was extracted from the samples and amplified by MY09/MY11 primers using standard methodology. HPV typing was performed using restriction analysis [14].
PBMC were separated from fresh 60-mL samples of citrated blood using ficoll-hypaque density gradients. T cell proliferation assays were set up in round-bottom 96-well plates. Cells, 4 × 105 per well in RPMI 1640 containing 10% human AB serum, were cultured with L2E7 (15 ng–100 μg) or control antigens (tetanus toxoid, influenza virus, inactivated herpes simplex virus, and phytohemagglutinin [PHA]), and incubated at 37°C in a 5% CO2 humidified atmosphere. Wells were pulsed with 0.5 μCi of [3H]thymidine for the final 18 h of a 6-day culture, harvested, and the incorporated radioactivity was determined by counter (1450 Microbeta Plus; Wallach, Turku, Finland). For each patient, the peak proliferative response to L2E7, with the background subtracted (Δ counts per minute [cpm]), was calculated for each visit. A “responder” was defined as a PBMC sample that produced a 3-fold increase over the corresponding week 0 Δcpm value. Stimulation indices were also determined.
PBMC cultures (4 × 105 cells/well in 2 mL) were incubated with L2E7 (25 and 100 μg/mL) with PHA and medium controls. Culture supernatants were harvested after 6 days and analyzed for interleukin (IL)-5 or interferon (IFN)-γ production in specific capture ELISAs. ELISA plates were coated with an optimum dilution of either anti-IL-5 (PharMingen, San Diego) or anti-IFN-γ (Endogen, Woburn, MA) capture antibodies in carbonate-coating buffer, pH 9.6, 100 μL per well, and incubated overnight at 4°C. The plates were washed twice with PBS and then blocked with 2% bovine serum albumin (BSA)—PBS, 300 μL per well, for 1 h at 37°C. The plates were washed twice with PBS containing 0.05% Tween 20 (PBST), and 100-μL test samples were added in triplicate. Standards and positive controls were prepared from a stock cytokine solution, diluted in RPMI containing 10% human AB serum, and then added in triplicate. The plates were incubated for 1 h at 37°C and then washed twice. Optimum dilutions of anti—IL-5 or anti-IFN-γg biotinylated detection antibody in PBST were added, 100 μL per well, and incubated for 1 h at 37°C. The plates were washed twice and incubated with horseradish peroxidase-streptavidin (Zymed, South San Francisco) for 1 h at 37°C. The plates were then washed twice and developed with o-phenylenediamine (OPD; Sigma, St. Louis) for 10 min at room temperature, and the reaction was terminated with 1 M H2SO4. The absorbance of each well was read at 492 nm. Cytokine concentrations for the test samples (means of the triplicates) were calculated from known standard curves for each of the samples and controls. Individual cytokine values were determined as the highest sample value (i.e., either 25 or 100 μg/mL), the medium control value for each patient at each time point. Group mean cytokine values are presented. The detection limit of both assays was 100 pg/mL. All PHA control wells produced high levels of cytokine.
ELISA plates were coated overnight (pH 9.6, 4°C) with L2E7 in 100 mM carbonate buffer. Wells were blocked with 0.05% Triton-X100 in PBS for 2 h. Serum samples and a positive control were titrated using 2-fold dilutions from 1/30-1/1920 in 1% BSA-PBS and 0.5% Tween. These were added to the plates in triplicate and incubated at 37°C for 1 h. After washing with PBST, the biotinylated detection reagent (mouse monoclonal anti-human IgG; Sigma) was added and incubated for 1 h. Plates were washed and incubated with horseradish peroxidase-streptavidin (Zymed) for 30 min, developed with OPD at room temperature, and the absorbance was measured at 492 nm. Titers were recorded as the log10 of the reciprocal of the dilution that gave an absorbance of 0.8. All titration curves were linear and parallel.
Twenty-seven men were enrolled into the clinical trial: 11 with new warts and 16 with persistent or relapsing warts. Of the 11 subjects with new warts, 8 opted to receive concomitant podophyllotoxin therapy and 3 were treated with vaccine alone. All subjects had 3 vaccinations. Only 3 persons did not attend the 8-week follow-up.
The vaccine was well-tolerated with no systemic adverse events reported or detected. There was only mild local tenderness at the site of vaccination in some subjects: Any local intolerance was reported by 22, 16, and 9 subjects at week 0, 1, and 4 vaccinations, respectively. Elevated post-vaccination body temperature (38.1°C) was recorded by 1 subject after the first immunization.
The HPV typing and clinical data are summarized in table 1. Nine subjects had HPV-6 and 4 had HPV-11. In 6 patients, there was an amplified product indicating HPV positivity, but this was insufficient for HPV typing. Eight patients were negative by PCR for HPV and also β-globin, indicating insufficient DNA in the sample.
T cell proliferation values. Each point represents counts per minute (cpm) or stimulation index (SI) for each subject (background subtracted) when peripheral blood mononuclear cells were cultured with L2E7.
Group mean cytokine concentrations produced when peripheral blood mononuclear cells were cultured with L2E7. IFN, interferon; IL, interleukin.
Figure 1 shows the level of [3H]thymidine incorporated into proliferating T cells. After vaccination, 19 of 25 subjects made antigen-specific T cell-proliferative responses. Three patients with persistent warts had higher peak proliferative responses than other patients (patients 22, 26, and 35 had > 30,000 cpm) after 1 vaccination, suggesting they may have had T cells primed to either HPV-6 L2 or E7.
PBMC were stimulated in vitro with L2E7, the supernatant was collected, and IFN-γ and IL-5 production was determined in capture ELISAs. Figure 2 shows that IL-5 was the dominant cytokine produced by cells after vaccination. IL-5 levels increased throughout the study, peaking at week 8 for all groups. IFN-γ levels were greatest in patients with new warts who received both vaccine and therapy and peaked at week 4, reaching levels similar to IL-5. IFN-γ levels were low in the other 2 groups.
The log10 mean titers of IgG in serum are shown in table 2. No subjects had detectable serum IgG responses to HPV-6 L2E7 before vaccination (all OD values at serum dilutions of 1/30 and 1/120 were < 0.32 and < 0.12, respectively). At 8 weeks after the final vaccination, all subjects had produced anti-L2E7 antibodies (OD values, 0.55–3.04 at 1/30 serum dilution). Only 2 subjects with persistent warts (nos. 25 and 30) made weak IgG responses (OD values, 0.55 and 0.79 at 1/30 serum dilution; both designated as < 1.477 in table 2). All subjects but patient 33 (data not shown) made IgG1 responses. Both IgG2 and IgG3 positive responses were low (30% and 50% of subjects, respectively; data not shown). Only subjects 26 and 29 were positive for IgG4 anti-L2E7 antibodies (data not shown). There were no marked differences in serum titers of IgG or IgG isotype profiles between patients with persistent or newly presenting warts.
For men with persistent or relapsing warts who were treated with vaccine alone and seen for the 8 week follow-up period, 2 of 15 completely responded, 3 partially responded (1 had 95% clearance), and 10 showed no change. In men with new warts who opted for vaccine and podophyllotoxin therapy and who were followed for 8 weeks, 3 of 6 were complete responders, 1 partially responded, and 2 showed no change. In the 3 men with new warts who received vaccine alone, 1 partially responded and 2 showed no change. Four of 5 complete clearances occurred between weeks 4 and 8. HPV typing was only successful in 1 person with vaccine-associated complete clearance; he had HPV-11. During wart regression, no other macroscopic signs were observed. Although the study was not designed prospectively to yield data beyond 8 weeks of follow-up, we attempted to collect clinical data from patient records beyond that time. Such follow-up schedules were not necessarily consistent after completion of mandated trial visits. Most subjects who were treated for genital warts after the end of the trial follow-up at 8 weeks received cryotherapy. Five patients who achieved complete clearance within the 8 weeks of the trial were followed for 10–24 weeks (mean, 17) and received no further therapy and developed no recurrences. Eight patients who did not achieve wart clearance within the trial period cleared lesions after further conventional therapy (range, 1–8 treatments; mean, 3.5). These subjects were followed for an additional 1–20 weeks (mean, 8) after clearance, and no recurrences were reported during this time.
Immunologic mechanisms clearly play a major role in the natural history of genital warts. First, immunodeficiency is known to be associated with up-regulation of HPV infection. Women infected with HIV have a relative risk of recurrent genital warts of 15.9 (95% confidence interval, 6.6–38.6) compared with matched HIV-negative women; HIV-positive women have a 2-fold increased risk of recurrent genital warts across each of five descending CD4 cell strata of 1500 to < 100 × 106/L [15]. Second, spontaneous regression of both cutaneous warts and genital warts can be observed, and the central role of cell-mediated immunity in this phenomenon has been demonstrated [16, 17]. Regressing lesions are characterized by significantly more T lymphocytes (predominantly CD4- and CD45RO-positive with enhanced expression of activation markers) and macrophages than are nonregressing lesions. Keratinocytes also display induction of HLA-DR and intercellular adhesion molecule 1, and the changes are consistent with a delayed-type hypersensitivity response [17]. Why some healthy persons exhibit such spontaneous regression while others do not remains unclear, but in the CRPV model such regression is linked to class II major histocompatibility complex (MHC) haplotype [18].
Therapeutic immunization against papillomavirus infection has been explored in the BPV, CRPV, and mouse models. Rejection of BPV-2-induced cutaneous fibropapillomata was elicited by an L2 vaccine [9], whereas BPV-4-induced mucosal papillomata regressed after immunization with E7 [10]. This regression of mucosal papillomas was immunohistologically similar to that seen in genital warts with a predominance of CD4-positive lymphocytes [19]. Also in this BPV-4 model, an L2 vaccine prevented infection after subsequent viral challenge [10]. In CRPV infection, both E1 and E2 vaccines elicited regression of lesions [20, 21], although regression in this model was associated with a predominantly CD8-positive infiltrate [22]. In mice bearing experimental HPV-16 tumors, both E6 and E7 vaccines caused tumor rejection [23, 24]. We considered that the observations from the BPV-4 model were particularly relevant in choosing an immunogen appropriate to HPV-6/11 disease, and therefore decided that using both E7 and L2 in the form of a fusion protein would offer a reasonable chance of reproducing the results from the animal model systems. Clearly, immunizing with such a protein would be more likely to stimulate MHC class II-restricted CD4 T cell responses, although we made no attempt to measure the phenotype of the responding cells in vivo. However, the immunohistologic data in regressing warts suggested to us that eliciting such a CD4 cell-driven response would be a desirable outcome.
To our knowledge, this is the first clinical trial of a genital wart recombinant vaccine. While it represents an early stage of the development of anti-HPV immunotherapeutic agents, there were a number of positive findings. First, we showed that the TA-GW vaccine is safe and well-tolerated in subjects with genital warts. Patient compliance with the study schedule was high, no significant adverse events were reported, and local reactions were transient and self-limiting. There was no increase in the frequency or severity of adverse reactions in this phase IIa trial compared with the phase I investigation. Second, TA-GW was immunogenic in patients with genital warts. All subjects made L2E7-specific serum antibodies, and 23 of 27 subjects made antigen-specific T cell proliferative responses to L2E7. Third, although this was a nonrandomized study, we observed encouraging clinical responses. In subjects with persistent and relapsing genital warts, the median duration was 12 months, yet in 2 of 15 subjects (1 of whom had warts for 6 years), lesions resolved with vaccine alone in 8 weeks. Also, in the 13 subjects in whom we observed clearance either with vaccine alone or with vaccine followed by conventional therapy, no recurrences were observed during a mean of 11 weeks of follow-up.
We offer the following reservations about our findings. First, although PBMC when cultured with L2E7 in vitro showed stimulation of IL-5 production after vaccination, IFN-γ only increased above baseline in subjects with new warts who received vaccine and topical podophyllotoxin therapy. In addition, less IFN-γ production was seen in subjects with genital warts vaccinated with TA-GW than in the vaccinated healthy volunteers [13]. This suggests a skewing of vaccine-induced immune responses toward a Th2 type response in the HPV-6/11 disease-bearing subjects. Such responses could be analogous to those in mice, where a predominance of a Th2 type immune response over that of Th1 facilitates the persistence of chronic hepatitis B virus infection, which can be reversed with stimulation of Th1 type responses [25, 26]. Second, we did not delineate any specific association between the immunologic responses and the observed clinical outcomes. Whether it may be necessary to measure cellular responses within genital wart lesional tissue to observe such a correlation remains to be determined [27]. Third, the nonrandomized nature of the study without a control group mandates caution in interpretation of the clinical responses, as similar remission rates have been observed in the placebo arms of some double-blind randomized trials [5].
Nevertheless, we believe this study is a valuable addition to knowledge in the field of immunotherapy against HPV disease [28]. Further studies are underway using the HPV-6 L2E7 immunogen with newer adjuvant formulations known to efficiently induce Th1 type responses [12, 29]. Randomized double-blind controlled trials to assess efficacy and immune correlates of regression will be central in such development.
We thank Stephen Inglis and Saveria Campo for assistance and encouragement during the vaccine development program.
Informed consent was obtained from all study subjects. The study was reviewed and approved by the research ethics committee, General Infirmary, Leeds, UK.
Financial support: Cantab Pharmaceuticals Research Ltd.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy