EV02: A Phase I trial to compare the safety and immunogenicity of HIV DNA-C prime-nyvac-c boost to NYVAC-C alone

Transcription

1 Vaccine (2008) 26, available at journal homepage: EV02: A Phase I trial to compare the safety and immunogenicity of HIV DNA-C prime-nyvac-c boost to NYVAC-C alone Sheena McCormack a,,1, Wolfgang Stöhr a,1, Tristan Barber b,1, Pierre-Alexandre Bart c,1, Alexandre Harari c, Christiane Moog d, Donatella Ciuffreda c, Cristina Cellerai c, Miranda Cowen b, Romilda Gamboni c,séverine Burnet c, Ken Legg b, Elizabeth Brodnicki a, Hans Wolf e, Ralf Wagner e, Jonathan Heeney f,g, Marie-Joëlle Frachette h, Jim Tartaglia i, Abdel Babiker a, Giuseppe Pantaleo c, Jonathan Weber b a MRC Clinical Trials Unit, 222 Euston Road, London NW1 2DA, UK b Imperial College, St Mary s Hospital, London, UK c Division of Immunology and Allergy, Department of Medicine, Centre Hospitalier Universitaire Vaudois, University of Lausanne, Lausanne, Switzerland d Institut de Virologie, Université Louis Pasteur, Strasbourg, France e Institute of Medical Microbiology, University of Regensburg, Regensburg, Germany f Biomedical Primate Research Centre, Rijswijk, The Netherlands g University of Cambridge, Cambridge, UK h Sanofi Pasteur, Lyon, France i Sanofi Pasteur, Toronto, Canada Received 17 October 2007; received in revised form 30 January 2008; accepted 4 February 2008 Available online 6 May 2008 KEYWORDS HIV preventative vaccine; Phase I; DNA-C Summary The aim of this randomised controlled trial was to see if the addition of 4 mg/ml DNA-C priming given by the intramuscular route at weeks 0 and 4 to NYVAC-C at weeks 20 and 24, safely increased the proportion of participants with HIV-specific T-cell responses measured by the interferon (IFN)- ELISpot assay at weeks 26 and/or 28 compared to NYVAC-C alone. Although 2 individuals discontinued after the first DNA-C due to adverse events (1 vaso-vagal; 1 transient, asymptomatic elevation in alanine transaminase), the vaccines were well tolerated. Three others failed to complete the regimen (1 changed her mind; 2 lost to follow-up). Of the 35 that completed the regimen 90% (18/20) in the DNA-C group had ELISpot responses compared to 33% (5/15) that received NYVAC-C alone (p = 0.001). Responses were to envelope in the majority Corresponding author. Tel.: ; fax: address: smc@ctu.mrc.ac.uk (S. McCormack). 1 The contribution of these authors to the trial was equal X/$ see front matter 2008 Elsevier Ltd. All rights reserved. doi: /j.vaccine

2 Comparison of DNA-C prime-nyvac-c boost to NYVAC-C alone 3163 (21/23). Of the 9 individuals with responses to envelope and other peptides, 8 were in the DNA-C group. These promising results suggest that DNA-C was an effective priming agent, that merits further investigation Elsevier Ltd. All rights reserved. Introduction In spite of twenty-five randomised controlled trials of biomedical prevention interventions against HIV other than against mother to child transmission there is only consistent evidence to support circumcision [1 3]. A safe HIV vaccine that can prevent infection remains a global health priority. After disappointing results from efficacy trials of recombinant protein, efforts shifted towards regimens that elicit cellular immunity. The chance of preventing infection with these constructs remains small, but viral control delaying the progression to AIDS and onward infectivity could still be a public health gain for countries with high prevalence [4,5]. However, in September 2007 this strategy suffered a setback when two Phase IIb efficacy trials exploring the Merck adenovirus type 5 vector HIV clade B vaccine were stopped after the interim analysis because there was no evidence of effect. A possible weakness of the adenovirus type 5 products was the lack of env in the insert, and a weakness generic to all candidates thus far is the failure to elicit heterologous neutralising antibodies. This paper describes the second trial in the EuroVac Programme, and the first to be sponsored by the EuroVacc Foundation, an international not for profit organisation created in 2002 with the primary objective of developing safe and effective preventive vaccines against HIV/AIDS. The Foundation involves academic and pharmaceutical partners in eight European countries. The first immunogen in the portfolio to enter clinical trial was a recombinant poxvirus vector, NYVAC-C, containing an insert derived from the Chinese HIV-1 clade C R5 CRF 07 strain (97CN54). This was well tolerated [6], and more immunogenic than anticipated with demonstrable interferon (IFN)- ELISpot responses in 50% of vaccinees assessed. At the time the trial described in this paper (EV02) started, there was scepticism about the ability of DNA to prime cellular responses, but increasing confidence in the safety of DNA-C constructs. The DNA-C construct investigated in EV02 carries the same insert as the NYVAC-C recombinant virus and was explored in humans for the first time. This trial set out to determine whether or not the highest feasible DNA-C prime (4 mg) could safely boost cellular responses to NYVAC-C compared to NYVAC-C alone. Methods Vaccine products The recombinant NYVAC-C and the plasmid DNA vector were manufactured according to Good Manufacturing Practice. Both express the same HIV genes. The vaccine NYVAC-C (vp2010), lot S3914, was produced by sanofi pasteur (formerly Aventis Pasteur, Lyon, France) and formulated with a potency of CCID 50 /ml. It was presented in a liquid form, in a volume of 1 ml in single dose 3 ml vials, which have been stored at 20 C. Prior to use each vial was thawed at room temperature and the vaccine was administered by intramuscular injection into the deltoid muscle of the nondominant arm by a member of the study team. The DNA-C vaccine was produced by Cobra Biomanufacturing (Keele, UK) and prepared for the trial in vials by ProPharma, all according to Good Manufacturing Practice. The presentation was in liquid form, with an extractable volume of 2 ml in 5 ml vials stored at 20 C. The appearance was clear and the composition per millilitre as follows: 1.05 mg DNA-C, 1.57 mg Tris HCl buffer, mg EDTA, 9 mg NaCl, and water to 1 ml. Prior to use the vials were thawed at room temperature and the DNA-C vaccine was administered by intramuscular injection of 2 ml into each vastus lateralis muscle (total of 4 ml per vaccination) by a member of the study team. Study design EV02 was a randomised Phase I trial without a placebo control conducted in two clinical centres: Centre Hospitalier Universitaire Vaudois (CHUV), Lausanne and St Mary s Hospital, Imperial College of Science, Technology and Medicine, London. Participants were randomised to receive DNA-C (DNA + NYVAC group; n = 20) or nothing (NYVAC alone group; n = 20) at weeks 0 (randomisation) and 4, followed in both groups by NYVAC-C at weeks 20 and 24. Allocation was open for participants and clinical investigators but blinded for laboratory personnel undertaking and interpreting the assays. Participants attended clinic at screening (within 6 weeks of randomisation), at immunisation time points, and, for examining possible solicited adverse events, once within the first 3 days and again 1 week after each immunisation time point. Of note, participants in the NYVAC alone group followed the same visit schedule as participants of the DNA + NYVAC group. For collecting information on unsolicited events and/or immunological analyses all participants further attended clinic at weeks 8, 18, 26, 28, and 48. The primary and secondary safety endpoints in EV02 were the same as in EV01 [6]. Study population Recruitment of healthy volunteers for EV02, enrolment and allocation to vaccination schedule was conducted in the same way as in EV01 [6]. All volunteers provided written informed consent to participate in the trial at screening and enrolment. Safety evaluations Local and systemic events that are recognized to be associated with licensed vaccines were solicited systematically at clinical centres and with a diary card given to the participant in the same way as in EV01. Data on other clinical events and laboratory events were collected with an open

3 3164 S. McCormack et al. question at each visit and through routine scheduled laboratory investigations, respectively. Routine haematology and chemical pathology were performed at screening and twice after each vaccination in accredited laboratories attached to each centre. In addition, DNA and antinuclear antibodies were examined at screening and weeks 4, 18, 24, 28, and 48. All events were graded in the same way as in EV01 and those reported in clinic assessed for relationship to study product (unrelated, unlikely, possibly or probably). Immunogenicity evaluations Cellular assays The immunogenicity of DNA-C and NYVAC-C was assessed by the quantification of T-cell responses (spot forming units (SFUs)) evaluated using the IFN- ELISpot assay with a panel of 8 pools of peptides (15-mers overlapping by 11 amino acids), namely ENV1/2, GAG1/2, NEF, GAG/POL, GAG/POL/NEF1/2. The assays were performed in the same laboratory and in the same way as in EV01 [6] on cryopreserved peripheral blood mononuclear cells at weeks 0, 5, 20, 24, 26, 28, and 48. The distribution of T-cell responses in CD4 and CD8 T-cell populations was assessed using polychromatic flow cytometry in responders with IFN- ELISpot responses in the range of 100 SFU/10 6 cells or above. Humoral assays The induction of HIV-specific antibodies was assessed using an ELISA assay at the same time points as the T-cell assays. Briefly, recombinant gp140 of CN54 (1 g/ml) (Simon Jeffs, Imperial) in 100 mm Na 2 HCO3 (ph 9.6) was coated on plates (Maxisorp, Nunc), one night at 4 C. After elimination of the solution and blocking step with PBS, 5% non-fat dry milk, serum dilutions made in PBS, 5% milk, 3% Tween 20 (beginning at 1/20) were added for 1 h at 37 C. Antibodies bound to the coated gp140 were revealed by an anti-human IgG-HRP conjugate (1/25 000, A1070, Sigma). Optical density (OD) of samples was measured in duplicates and human sera from HIV-1 seronegative and seropositive subjects were used as controls in all ELISAs. All analyses were performed at the Institut de Virologie, University of Strasbourg, using standardised equipment and operating procedures. The inhibitory activity of antibodies was assessed with 3 different assays. First, a multiple round neutralisation assay on PBMC was carried out with the homologous primary isolate CN54 using experimental conditions previously described [7]. Second, antibody effects in a single cycle infection of primary isolate Bx08 in the engineered cell line (TZMbl) was measured as reported [8]. Third, the inhibition, by antibodies, of HIV-1 Bal multiplication in macrophages was determined according to the method developed previously [9]. Statistical methods All clinical event and routine laboratory data were included in the safety analyses. These considered the number (%) of participants experiencing an event, and the duration of events, for (1) local or systemic events solicited within 7 days of vaccination, (2) all unsolicited events occurring within 28 days of an immunisation or during total followup (48 weeks), (3) unsolicited events considered possibly or probably related to study product, and (4) laboratory adverse events. When the same event was reported on the clinic form and diary card and the grades differed, the maximum grade was reported; any event existing prior to immunisation that persisted after the vaccination was only considered to be an adverse event if the severity or the grade of the event increased post immunisation. To evaluate the safety of DNA-C, two comparisons were made based on the maximum grade per participant: (1) solicited local events were analysed in the DNA + NYVAC arm, comparing events after the DNA-C immunisations with events after the NYVAC-C immunisations in the same participants, using the symmetry test. (2) Solicited systemic events were compared between the two randomisation groups after the first 5 weeks, when the DNA + NYVAC group had received both DNA-C immunisations and the NYVAC alone group had not received any vaccination, using Fisher s exact test. An ELISpot result was defined as positive if the number of SFUs was 55 SFU/10 6 cells and 4-fold the negative control (as in EV01). The primary immunogenicity endpoint was a positive ELISpot result at weeks 26 and 28: each participant was classified as a responder if there was at least one positive response against any of the HIV peptides at weeks 26 or 28, and as a non-responder if responses at these weeks were all negative. The magnitude of a ELISpot response was described as the sum of SFUs of positive responses for each of the different HIV proteins, i.e. env, gag, pol and nef, without subtraction of background. An antibody response was classified as present if the OD measured with sera collected after immunisation was at least three times greater than the OD obtained with the corresponding pre-immune serum. Comparisons of categorical variables were made using Fisher s exact test. The magnitude of a response was compared between groups using nonparametric statistical tests (magnitude over time: paired Wilcoxon or Friedman test; comparison between randomisation groups: Mann Whitney test). The magnitude of ELISpot and ELISA responses was compared using non-parametric correlation. Trial management The study protocol was approved by St Mary s LREC (Local Research Ethics Committee) and GTAC (Gene Therapy Advisory Committee) for the UK, and Commission d éthique de la recherche clinique (Université de Lausanne) for Switzerland. Regulatory approval was also obtained from the competent authorities in both countries, MHRA (Medicines and Healthcare Products Regulatory Agency) for the UK and Swissmedic for Switzerland. The trial was overseen by a Trial Coordinating Committee (TCC) which included the Principal Investigators from each centre, a representative from sanofi pasteur, and two independent members including the Chair. A Data and Safety Monitoring Committee (DSMC) was also appointed to review the design and protocol and met three times during the trial. The data management, site monitoring, analysis and report writing were coordinated by the UK Medical Research Council Clinical Trials Unit (MRC CTU) in collaboration with the Clinical Centres, sanofi pasteur and the EuroVac scientific programme coordination.

4 Comparison of DNA-C prime-nyvac-c boost to NYVAC-C alone 3165 Results Trial population Of 50 healthy volunteers at low risk of HIV infection screened, 10 were excluded because of meeting one or more exclusion criteria (n = 8) or withdrawal of consent (n =2) (Fig. 1). The remaining 40 were enrolled and randomised between 21st February and 22nd August Due to recruiting difficulties in London and reopening of the Lausanne site, 25 participants were enrolled in Lausanne, and 15 in London. As a result of this differential enrolment, a chance imbalance between the two groups emerged with 23 participants allocated to receive DNA-C and NYVAC-C, and 17 allocated to NYVAC-C alone. Ninety percent were white, 50% female, and the median age was 32 years (Table 1). A few volunteers had minor laboratory abnormalities before enrolment: raised serum creatinine (grade 1: n = 2), raised ALT (grade 1: n = 1, grade 2: n = 1), raised serum glucose (grade 1: n = 1), or decreased lymphocytes (grade 1: n = 2). All abnormalities were considered clinically insignificant or were normal at repeat testing, and all but one (grade 2 ALT, see below) did not deteriorate during the trial. After the first DNA-C vaccination, immunisations were discontinued in two participants due to adverse events. The first participant developed a vaso-vagal reaction after administration of the first 2 ml DNA-C with loss of consciousness for about 10 s, followed by nausea and vomiting. The episode resolved spontaneously without any medication. Whilst this was classified as a grade 2 event, the clinical team decided not to proceed with the second 2 ml dose. Furthermore, it was agreed that even though this event did not meet the protocol recommendations for discontinuing immunisations, it was appropriate to do so in the context of a Phase I study, and an episode that could plausibly recur with subsequent immunisations. The second participant was discontinued because of a severe but transient elevation in transaminase (see below). Both participants attended all visits after discontinuation. A further three participants also received no NYVAC-C: one participant who had received two DNA-C immunisations (worst event: grade 2 myalgia) decided not to receive further vaccinations but she attended visits at weeks 28 and 48, and two NYVAC-C alone participants were lost to follow-up at weeks 5 and 8, respectively. In summary, 35 participants completed the immunisations and all scheduled visits. For analysis, all safety Figure 1 Participant disposition. Multiple reasons possible; * exclusion before start of vaccination; ** self-withdrawal at week 18, after having received complete DNA vaccinations; *** participants last seen at weeks 5 and 8, respectively.

5 3166 S. McCormack et al. Table 1 Baseline characteristics and follow-up Characteristic DNA + NYVAC (n = 23) NYVAC (n = 17) Total (n = 40) Demographic data Centre Lausanne 14 (61%) 11 (65%) 25 (63%) London 9 (39%) 6 (35%) 15 (38%) Sex Female 11 (48%) 9 (53%) 20 (50%) Male 12 (52%) 8 (47%) 20 (50%) Ethnicity White 21 (91%) 15 (88%) 36 (90%) Black 1 (4%) 1 (6%) 2 (5%) Other 1 (4%) 1 (6%) 2 (5%) Age (years) 34 (23 46) 23 (23 41) 32 (23 43) BMI (kg/m 2 ) 24 (20 26) 23 (21 25) 23 (20 25) Routine laboratory parameters Haemoglobin (g/dl) 14 (13 15) 14 (13 15) 14 (13 15) White cell count (10 9 cells/l) 5.7 ( ) 6.2 ( ) 6.2 ( ) Neutrophils (10 9 cells/l) 3.3 ( ) 3.4 ( ) 3.4 ( ) Lymphocytes (10 9 cells/l) 1.9 ( ) 2.0 ( ) 2.0 ( ) Platelets (10 9 cells/l) 241 ( ) 284 ( ) 272 ( ) CD4 cell count (10 6 cells/l) 858 ( ) 1070 ( ) 971 ( ) ALT (U/l) 20 (15 31) 19 (16 24) 20 (15 29) AST (U/l) 27 (22 30) 28 (21 37) 27 (21 33) Alkaline phosphatase (U/l) 60 (49 70) 64 (53 74) 60 (51 71) Bilirubin ( mol/l) 10 (10 12) 11 (10 13) 10 (10 13) Creatinine ( mol/l) 87 (83 100) 90 (83 96) 88 (83 97) Glucose (mmol/l) 4.9 ( ) 5.2 ( ) 5.1 ( ) Follow-up (weeks) 48 (47 48) 48 (46 49) 48 (47 49) Note: Values are number (%) or median (interquartile range). examinations after discontinuation of the vaccination scheme were ignored for local and systemic solicited events. Scheduled safety visits were incomplete in 6/40 participants (4 DNA + NYVAC, 2 NYVAC alone group) who each missed one of the two visits due after immunisation (nobody missed both): three participants each missed one visit, two participants missed two visits, and one participant missed three visits. In 5 participants (2 DNA + NYVAC, 3 NYVAC alone group), diary cards were missing for one immunisation time point. Both a missing diary card and a missing safety visit occurred in one participant (after 1st DNA). Safety grade 3 or 4 adverse events There were four grade 3 or 4 adverse events reported (two in each randomisation arm), only one of which was considered clinically significant. One event was probably a recording error (see below, case 4). Case 1 (DNA + NYVAC group): This male aged 55 had an ALT of 402 IU/ml (grade 4, Upper Limit of Normal (ULN) 40 IU/ml) 1 day after the first DNA-C immunisation. In repeated tests, ALT had decreased to 65 IU/ml (grade 1, ULN 50 IU/ml) 7 days later, and to a normal value of 35 IU/ml 13 days after the vaccination. Of note, at screening this volunteer s ALT was raised at 114 (grade 2), but the repeat was 72 IU/ml (grade 1) and the participant was recovering from a viral illness. ALT was normal in the third specimen at 36 IU/ml and so he was enrolled. There was no history of hepatitis, travel, excessive alcohol or recreational drugs to explain this rise in ALT and the participant remained clinically well throughout. Additional investigations (liver ultrasound, virology) were normal, and the investigators concluded that this event was possibly related to vaccine. Following the advice of an independent hepatologist, immunisations were discontinued in this individual, and it was also agreed not to enrol any individuals that had abnormal transaminases at any time during the screening process. Case 2 (NYVAC alone group): Grade 3 induration with erythema (grade 1) and swelling (grade 2) following the 1st NYVAC-C immunisation. The participant had experienced a similar reaction following a licensed immunisation in the past. Symptoms (itching and pain) were only mild and it was decided to proceed with the 2nd NYVAC-C immunisation which did not elicit remarkable local events. Case 3 (DNA + NYVAC group): Grade 3 pain at injection site and myalgia (muscular pain in left arm which restricted movement) 3 days after his 1st NYVAC-C immunisation with regular intake of painkillers. The 2nd NYVAC-C was given in the opposite arm without remarkable events.

6 Comparison of DNA-C prime-nyvac-c boost to NYVAC-C alone 3167 Case 4 (NYVAC alone group): A grade 3 fever was noted during a database generated safety review, recorded at week 4 visit but without immunisation by then. When the participant was questioned later she had no recollection and thought that she had probably recorded the wrong temperature. The investigators concluded that this was most likely to be a recording error. Local events Solicited local events within 1 week after each vaccination are summarised in Table 2. After both DNA vaccinations, all 23 participants in the DNA + NYVAC group had experienced at least one local adverse event, mostly pain and erythema. Apart from moderate pain in two participants all events were mild. The frequency of events and the event pattern were very similar across the two DNA-C vaccinations. After the vaccinations with NYVAC-C, 18/20 participants in the DNA + NYVAC group and 12/15 in the NYVAC alone group had at least one adverse event, again mostly pain and erythema. There was one severe event in each randomisation arm (pain and induration, respectively; see above), and another four participants experienced moderate events (pain, swelling and/or itching) but most events were mild. Solicited local events in the DNA + NYVAC arm were similar after the DNA-C and after the NYVAC-C immunisations (Fig. 2) and pairwise comparisons of the maximum grade per participant did not reveal statistically significant differences, neither overall (p = 0.38) nor broken down by event (data not shown). There was also no significant difference between the two randomisation groups after the vaccinations with NYVAC-C. The median duration of local events was 1 day (iqr 1 2, range 1 35). Five events lasted for more than 7 days, 3 (pain, erythema, swelling) in the DNA + NYVAC and 2 (erythema, induration) in the NYVAC alone group. Systemic events Solicited systemic events within 1 week after each vaccination are summarised in Table 3. The most common events were malaise and headache. After both DNA-C vaccinations, 13/23 (57%) participants in the DNA + NYVAC group had experienced at least one systemic adverse event, and, during the same period but without vaccination, 6/17 (35%) in the NYVAC alone group (comparison of maximum grade: p = 0.13 and ignoring the dubious fever in NYVAC alone arm: p = 0.11). There was no statistically significant difference between the groups for any event. Events were mostly mild, apart from a few events of moderate malaise, myalgia and/or headache (all in DNA + NYVAC arm; Fig. 3). The frequency of events and the event pattern were very similar across the two DNA-C vaccinations. After the vaccinations with NYVAC-C, 11/20 participants in the DNA + NYVAC group and 11/15 in the NYVAC alone group had at least one adverse event. There was one severe myalgia in the DNA + NYVAC arm (see above), and another four participants (3 NYVAC alone, 1 DNA + NYVAC) experienced moderate events (headache, malaise, myalgia and/or chills/rigors). Again, the majority of events were mild. Overall, about 75% in each randomisation arm experienced solicited systemic events. The median duration of systemic events was 1 day (iqr 1 2, range 1 26). Two events (malaise) lasted for more than 7 days, both occurring in one participant in the DNA + NYVAC group. Laboratory abnormalities In addition to the grade 4 raised ALT (see above), there were five grade 2 abnormalities (1 hyperbilirubinaemia (DNA + NYVAC), 1 hyperglycaemia (DNA + NYVAC), and 3 decreased lymphocytes (1 DNA + NYVAC, 2 NYVAC alone)), which were not considered clinically relevant. Furthermore, 50 grade 1 abnormalities occurred similarly across the randomisation arms, mostly decreased lymphocytes (n = 16), hyperglycaemia (n = 14), hyperbilirubinaemia (n = 6), raised AST (n = 5), and raised ALT (n = 4). Whereas no participant showed DNA antibodies after randomisation, seven developed antinuclear antibodies, firstly found at week 4 in one participant (NYVAC alone), at week 18 in 5 (3 DNA + NYVAC, 2 NYVAC alone), and at week 30 in 1 (DNA + NYVAC). Of note, the three participants in the NYVAC alone group had detectable antibodies before the first immunisation. In 5/7 participants the antibodies were found on more than one occasion, and in all but one participant (NYVAC alone group) antibodies disappeared by week 48. Non-solicited events Overall, 35 participants (21 DNA + NYVAC, 14 NYVAC alone group) reported at least one non-solicited adverse event within 28 days following vaccination (total number of events: n = 86). Of these, 9 participants (8 (35%) DNA + NYVAC, 1 (6%) NYVAC alone group) reported events considered to be possibly or probably related to study products: 6 local events not solicited or after 7 days, and 9 systemic events. Ten of these events were grade 1, and 5 were grade 2: erythema and induration occurring in one NYVAC alone participant 9 days after 1st NYVAC-C, chest infection and sore throat in one participant 3 days after the 2nd DNA-C, and one event of heartburn on the day of 1st DNA-C. In addition, 39 grade 1 and 8 grade 2 events were reported beyond 28 days from vaccination, however, none of these were considered possibly or probably related to the study products. Immunogenicity: T-cell responses Based on an intention-to-treat analysis of all 40 participants (assuming missing equals non-response), 83% (19/23) in the DNA + NYVAC group responded at weeks 26/28 in the ELISpot assay compared to 35% (6/17) in the NYVAC alone group (difference 47% (95% CI 20 75%; p = ). Of those that completed the immunisation schedule (Fig. 4A), the proportion of responders was 90% (18/20) in the DNA + NYVAC group compared to 40% (6/15) in the NYVAC alone group (difference 50% (95% CI 22 78%; p = 0.003). Of note, one of the 6 responders in the NYVAC alone group had a weak response against pol at all time points including prior to vaccination when it was just below the cut-off for classification as positive. Therefore, this T-cell response was clearly nonspecific, and, after re-classification of this participant, the

7 Table 2 Number (%) of participants with solicited local events 1st DNA 2nd DNA Overall DNA 1st NYVAC 2nd NYVAC Overall NYVAC Overall (n =23) (n =21) (n =23) (n = 20) N (n =15) (n = 20) N (n =15) (n = 20) N (n =15) (n = 23) N (n = 17) Pain Grade 1 17 (74%) 18 (86%) 20 (87%) 13 (65%) 8 (53%) 13 (65%) 10 (67%) 14 (70%) 11 (73%) 17 (74%) 11 (65%) Grade 2 1 (4%) 1 (5%) 2 (9%) 0 1 (7%) 2 (10%) 1 (7%) 2 (10%) 1 (7%) 4 (17%) 1 (6%) Grade (5%) (5%) 0 1 (4%) 0 Itching Grade 1 1 (4%) 1 (5%) 2 (9%) 5 (25%) 3 (20%) 4 (20%) 3 (20%) 6 (30%) 4 (27%) 7 (30%) 4 (24%) Grade (5%) 0 1 (5%) 0 1 (4%) 0 Erythema Grade 1 13 (57%) 12 (57%) 17 (74%) 10 (50%) 5 (33%) 10 (50%) 4 (27%) 13 (65%) 7 (47%) 20 (87%) 7 (41%) Swelling Grade 1 2 (9%) 1 (5%) 3 (13%) 2 (10%) 2 (13%) 3 (15%) 5 (33%) 3 (15%) 6 (40%) 5 (22%) 6 (35%) Grade (5%) 1 (7%) 2 (10%) 0 2 (10%) 1 (7%) 2 (9%) 1 (6%) Induration Grade (5%) 1 (7%) 1 (5%) 1 (7%) 2 (10%) 2 (13%) 2 (9%) 2 (12%) Grade Grade (7%) (7%) 0 1 (6%) Clear blisters Grade 1 1 (4%) 0 1 (4%) (4%) 0 Blood blisters Grade (5%) 1 (4%) (4%) 0 All events Grade 1 20 (87%) 20 (95%) 21 (91%) 13 (65%) 8 (53%) 14 (70%) 11 (73%) 14 (70%) 10 (67%) 17 (74%) 10 (59%) Grade 2 1 (4%) 1 (5%) 2 (9%) 1 (5%) 1 (7%) 3 (15%) 1 (7%) 3 (15%) 1 (7%) 5 (22%) 1 (6%) Grade (5%) 1 (7%) (5%) 1 (7%) 1 (4%) 1 (6%) Total 21 (91%) 21 (100%) 23 (100%) 15 (75%) 10 (67%) 17 (85%) 12 (80%) 18 (90%) 12 (80%) 23 (100%) 12 (71%) Note: For each event, the maximum grade is shown for each participant per time point. All events summarises the maximum grade across all events. D + N: DNA-C + NYVAC-C arm; N: NYVAC-C alone arm S. McCormack et al.

8 Table 3 Number (%) of participants with solicited systemic events 1st DNA or nothing 2nd DNA or nothing Overall DNA or nothing 1st NYVAC 2nd NYVAC Overall NYVAC Overall (n = 23) N(n =17) (n = 21) N(n =17) (n = 23) N(n =17) (n = 20) N(n =15) (n = 20) N(n =15) (n = 20) (n = 15) (n = 23) N(n = 17) Fever Grade (5%) 0 1 (4%) (13%) 0 1 (7%) 0 3 (20%) 1 (4%) 2 (12%) Grade Grade (6%) a 0 1 (6%) (6%) Chills/rigors Grade 1 1 (4%) 0 2 (10%) 1 (6%) 3 (13%) 1 (6%) 2 (10%) 2 (13%) 3 (15%) 1 (7%) 3 (15%) 3 (20%) 5 (22%) 3 (18%) Grade (5%) (5%) 0 1 (4%) 0 Malaise Grade 1 4 (17%) 2 (12%) 2 (10%) 3 (18%) 5 (22%) 4 (24%) 5 (25%) 5 (33%) 5 (25%) 4 (27%) 6 (30%) 6 (40%) 8 (35%) 9 (53%) Grade 2 1 (4%) 0 1 (5%) 0 2 (9%) 0 1 (5%) 1 (7%) 0 1 (7%) 1 (5%) 2 (13%) 2 (9%) 2 (12%) Myalgia Grade 1 1 (4%) 0 4 (19%) 1 (6%) 3 (13%) 1 (6%) 2 (10%) 2 (13%) 1 (5%) 4 (27%) 3 (15%) 4 (27%) 5 (22%) 4 (24%) Grade 2 1 (4%) (4%) (7%) 0 1 (7%) 1 (4%) 1 (6%) Grade (5%) (5%) 0 1 (4%) 0 Headache Grade 1 3 (13%) 3 (18%) 3 (14%) 1 (6%) 4 (17%) 4 (24%) 3 (15%) 1 (7%) 5 (25%) 5 (33%) 7 (35%) 4 (27%) 8 (35%) 6 (35%) Grade 2 1 (4%) (4%) 0 1 (5%) 2 (13%) 1 (5%) 0 1 (5%) 2 (13%) 2 (9%) 2 (12%) Nausea Grade 1 2 (9%) 1 (6%) 2 (10%) 2 (12%) 4 (17%) 2 (12%) (20%) 0 3 (20%) 4 (17%) 5 (29%) Vomiting Grade (5%) 0 1 (4%) (7%) 0 1 (7%) 1 (4%) 1 (6%) All events Grade 1 6 (26%) 3 (18%) 7 (33%) 4 (24%) 9 (39%) 5 (29%) 6 (30%) 5 (33%) 8 (40%) 7 (47%) 9 (45%) 8 (53%) 13 (57%) 10 (59%) Grade 2 3 (13%) 0 1 (5%) 0 4 (17%) 0 1 (5%) 2 (13%) 1 (5%) 2 (13%) 1 (5%) 3 (20%) 3 (13%) 2 (12%) Grade (6%) 0 1 (6%) 1 (5%) (5%) 0 1 (4%) 1 (6%) Total 9 (39%) 3 (18%) 8 (38%) 5 (29%) 13 (57%) 6 (35%) 8 (40%) 7 (47%) 9 (45%) 9 (60%) 11 (55%) 11 (73%) 17 (74%) 13 (76%) Note: For each event, the maximum grade is shown for each participant per time point. All events summarises the maximum grade across all events. D + N: DNA-C + NYVAC-C arm; N: NYVAC-C alone arm. a Likely to be a recording error. Comparison of DNA-C prime-nyvac-c boost to NYVAC-C alone 3169

9 3170 S. McCormack et al. Figure 2 Local adverse events in participants allocated to DNA-C + NYVAC-C after DNA-C and after NYVAC-C. difference between the two randomisation arms was 57% (95% CI 29 84%; p = 0.001). T-cell responses following DNA-C + NYVAC-C were significantly higher than those in the NYVAC alone group. At the primary endpoint weeks 26 and 28, the median IFN- secreting T-cells was 299 and 246 SFU/10 6 cells in those receiving the combination regimen compared to 131 and 76 SFU/10 6 cells in those immunised with NYVAC-C alone (p = and 0.005, respectively, Fig. 4B). Duration was linked to magnitude of response and, in the DNA + NYVAC group, there were still 80% responders 6 months after the completion of the vaccination, compared to 13% in the NYVAC alone group. Of note, in the DNA + NYVAC group the number of responders at week 24, i.e. after the first NYVAC-C immunisation, was 85%, similar to weeks 26 or 28, and there was no statistically significant difference in the magnitude. Figure 3 Systemic adverse events during the first 5 weeks in participants allocated to DNA-C + NYVAC-C (after two vaccinations with DNA-C) and in those allocated to NYVAC-C alone (no vaccination during this time period).

10 Comparison of DNA-C prime-nyvac-c boost to NYVAC-C alone 3171 Figure 4 ELISpot responses in participants with complete vaccination (n = 35): (A) number of responders by allocation; (B) cumulative spot forming units of positive responses (individual responses and group box plot, showing minimum, 25% percentile, median, 75% percentile, and maximum). Env-specific responses were observed in 21 responders and responses to other peptides were found in 11 participants (gag: n =6, pol: n =5, nef: n = 1; 9 participants (8 DNA + NYVAC, 1 NYVAC alone) had responses to env and at least one other peptide). CD4 and CD8 T-cell responses were polyfunctional as indicated by the ability to secrete a number of cytokines such as IL-2, IFN- and TNF-, by the granulation activity (for CD8 T-cells) and by the proliferation capacity [10]. These responses were mediated by CD4 T-cells in all responders, and by CD8 T-cells in 47% of responders. Epitope mapping was performed for the env-specific responses and revealed an average of 4 env-derived peptides being recognized per responder. The immunogenicity data have been written up in detail elsewhere [11]. Immunogenicity: T-cell responses influence of gender and of previous smallpox vaccination The number of responders to HIV antigens in the NYVAC alone and in the DNA + NYVAC group, respectively, was 5 of 8 and 9 of 9 in women compared to 0 of 7 and 9 of 11 in men (exact p = 0.01). Considering positive responses only in the DNA + NYVAC group, there was a trend for women to have higher mean SFUs compared to men, with 588 versus 407 at week 26 (p = 0.24) and 379 versus 263 at week 28 (p = 0.34), respectively.

11 3172 S. McCormack et al. Figure 5 Magnitude (mean + standard deviation) of ELISpot responses by gender and previous smallpox immunisation in participants allocated to DNA-C + NYVAC-C. Previous smallpox vaccination had no effect on the immunogenicity of the NYVAC alone group, since 2 and 1 subjects had been immunised for smallpox out of 5 responders and 10 non-responders, respectively. In the DNA + NYVAC group, the 2 non-responders (n =2)and9of 18 (50%) responders had been previously vaccinated for smallpox. There were sufficient responders in the DNA + NYVAC group to separate subjects according to gender and previous smallpox vaccination. The 2 non-responders were both men Figure 6 Antibodies against gp140 CN54 over time by allocation: (A) proportion of responders; (B) magnitude of responses (individual responses and group mean).

12 Comparison of DNA-C prime-nyvac-c boost to NYVAC-C alone 3173 that had been previously vaccinated. Furthermore, the magnitude of the peak T-cell responses in men with a previous smallpox vaccination was significantly lower as compared to men who had not been immunised for smallpox at weeks 24, 26, 28 and 48 (p = 0.01, 0.03, 0.03, and 0.02). There also was a numerical difference to the group of women with a previous smallpox vaccination although this did not reach statistical significance (p = 0.30, 0.05, 0.33, and 0.48) (Fig. 5). Immunogenicity: humoral responses IgG antibody responses against gp140 CN54 at week 26 were found in 4/15 (27%) participants of NYVAC alone and in 15/20 (75%) participants of DNA + NYVAC group, respectively (Fig. 6A). In the DNA + NYVAC group, no antibodies were detected after DNA-C, and only 15% responded after the first NYVAC-C immunisation. The magnitude of antibody levels was also greater in the DNA + NYVAC group (Fig. 6B). However, antibody responses were transient and none of the NYVAC alone group and only 1 (5%) participant of the DNA + NYVAC group had a measurable antibody response at week 48. The vaccine-induced antibodies also failed to show neutralising activity in the 3 assays performed. There was a significant association between presence of an antibody response and presence of a T-cell response at weeks 26/28 (of 35 participants, 15 were responders in ELISA and ELISpot assay, and 9 were non-responders in both; p = 0.035), However, within those responding in both assays there was no correlation of the response magnitude. Discussion There was only one adverse event of note in the trial, and this was the transiently raised ALT following the first DNA-C in a single individual with a pre-existing unexplained elevation, who was asymptomatic throughout with complete recovery by the second repeat test 13 days later. Although the participant was willing to continue, the independent medical expert consulted recommended that immunisations be discontinued since this was a healthy volunteer Phase I trial. Laboratory abnormalities, particularly in liver transaminases, are common and it is standard clinical practice to repeat the investigation before deciding to exclude a volunteer. As a result of this event, which occurred 9 weeks after enrolment opened, future volunteers were excluded on the basis of a single abnormal liver transaminase. Interestingly, this resulted in the St Mary s site excluding a higher proportion of screened volunteers compared to the first 9 weeks and CHUV overall, such that St Mary s was not able to achieve the original target of 20 enrolled participants. Of the 50 grade 1 abnormalities, 9 were raised transaminases, 5 in DNA-NYVAC and 4 in NYVAC alone. There were no grade 2 or 3 elevations. DNA-C and NYVAC-C were otherwise well tolerated, and there was no evidence to suggest that the DNA-C prime exacerbated local or systemic reactions following NYVAC-C. Antinuclear antibodies developed in seven participants during the trial, but only transiently at low titre. Furthermore, antibodies were detected before the first immunisation in the NYVAC alone group, so could not have been induced by the study products. The DNA-C prime significantly boosted the cellular immune responses to NYVAC-C in an intention-to-treat analysis counting missing as non-response, with an even higher proportion of responders when the analysis was limited to those individuals that completed the immunisation schedule. Responses were strong with high levels of SFU per million cells, and more importantly, durable with 80% of participants that completed the DNA-C + NYVAC-C regimen still positive in the ELISpot assay at the final scheduled visit. The majority of responses were to env peptides, with only 30% of those that completed immunisations responding to other peptides. These have been further characterised using flow cytometry, phenotypic profiling and epitope mapping, and 9 new peptides/epitopes have been identified that were not previously listed on the Los Alamos database [11]. These results indicate that the DNA-C prime-nyvac-c boost strategy is as effective in eliciting vaccine-induced T-cell responses as the trivalent Ad5 gag-pol-nef from Merck or the multiclade DNA/Ad5 candidate vaccine of the Vaccine Research Center (VRC) (R. Koup et al., unpublished data). Compared to these other immunogens, the DNA/NYVAC regimen is env dominant, and in light of the recent Phase IIb clinical trial results of the MRK-Ad5, this could be an advantage. The multiclade DNA/Ad5 candidate vaccine induced more pol-specific responses and a higher proportion of CD8 T- cell responses, with the majority of vaccinees demonstrating CD4 and CD8 responses (R. Koup et al., unpublished data). This could be due to differences inherent in the products, but it may also be explained by the difference in schedule with three DNA prime in the VRC design compared to the two in the EuroVacc trial. On the basis of these preliminary data, the next step in the EuroVacc programme will be to see whether a three DNA-C prime + single NYVAC-C boost regimen can increase either the proportion of participants with responses to env plus another peptide pool, or the percentage of CD8 T-cell responses compared to the prime boost regimen described in this paper. The decision to boost three DNA with only one NYVAC-C is based on the observation that in EV02 the proportion of T-cell responders and the response magnitude after the first NYVAC-C was similar to the response after the second NYVAC-C. There is a broad scientific consensus that a successful vaccine to prevent HIV-1 transmission must be able to elicit both HIV-specific T-cells and neutralising antibodies. Unfortunately, despite strong T-cell responses in the DNA-C + NYVAC-C regimen, antibody responses were only transient and no neutralising activity was found. This is similar to other products investigated so far and may have contributed to the recent failure of the Ad5 vaccine to prevent HIV infections. However, vaccine regimes eliciting strong T-cell responses such as DNA-C + NYVAC-C are likely to provide a priming platform for recombinant protein boosting to achieve sufficient antibody responses. By chance, the two groups were balanced for gender and vaccinia status. Combining the data for NYVAC-C alone from EV02 with those from EV01 (where 4 of 5 women but 1 of 5 men had ELISpot responses) gives an absolute difference of 61% more responders among women (p = 0.004). This influence of gender was not seen in the proportions of responders in the DNA + NYVAC group, but there was a trend towards higher magnitude T-cell responses in women compared to men. This difference was not statistically significant and

13 3174 S. McCormack et al. the observation could be chance, or could be due to the small numbers. Interestingly, previous exposure to vaccinia appeared to make no difference to females whereas in men, in spite of small numbers, the magnitude of the peak T-cell responses was significantly lower as compared to those who had not been immunised for smallpox at weeks Furthermore, the 2 single non-responders were both men with a previous smallpox vaccination. Taken together these results indicate that the DNA- C/NYVAC-C has comparable cellular immunogenicity to other vaccine strategies that have entered Phase IIB/III clinical trials. The env-dominant response may represent an attractive alternative to Ad5-based vaccination strategies, and an ideal priming platform for recombinant proteins. Acknowledgements We are indebted to the volunteers participating in the trial, who generously gave up their time. We also wish to thank Song Ding from the EuroVacc Foundation for the support in the management of the trial, all EuroVacc program investigators, and the following members of trial staff at the coordination and clinical centres: Patrick Kelleher, Mary Rauchenberger, and Mona Khonkarly. Trial Steering Committee: Geoffrey Schild (Chair), Frank de Wolf, Guiseppe Pantaleo, Jonathan Weber, Abdel Babiker (observer), Marie Miossec (observer), Ralf Wagner (observer), Hans Wolf (observer). Data and Safety Monitoring Committee: Anne McLaren (Chair), Ferdinand Wit, Jürg Schifferli. Funding: Funds for the trial came from the European Commission through two scientific awards (QLK2-CT and QLK2-CT ) which was distributed to the partners, and coordinated by Professor Peter Liljeström based at the Karolinska Institute, Stockholm, Sweden. In addition each institutional partner provided financial support through the time of core senior staff and access to clinical facilities and technical assistance. References [1] Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, Puren A. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 trial. PLoS Med 2005;2(11):e298. [2] Bailey RC, Moses S, Parker CB, Agot K, Maclean I, Krieger JN, et al. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet 2007;369(9562): [3] Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, Nalugoda F, et al. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet 2007;369(9562): [4] Johnston MI, Fauci AS. An HIV vaccine evolving concepts. N Engl J Med 2007;356(20): [5] Gupta SB, Jacobson LP, Margolick JB, Rinaldo CR, Phair JP, Jamieson BD, et al. Estimating the benefit of an HIV- 1 vaccine that reduces viral load set point. J Infect Dis 2007;195(4): [6] Bart PA, Goodall R, Barber T, Harari A, Guimaraes-Walker A, Khonkarly M, et al. EV01: a Phase I trial in healthy HIV negative volunteers to evaluate a clade C HIV vaccine, NYVAC-C undertaken by the EuroVacc Consortium. Vaccine 2008;26(25): [7] Burrer R, Salmon-Ceron D, Richert S, Pancino G, Spiridon G, Haessig S, et al. Immunoglobulin G (IgG) and IgA, but also nonantibody factors, account for in vitro neutralization of human immunodeficiency virus (HIV) type 1 primary isolates by serum and plasma of HIV-infected patients. J Virol 2001;75(11): [8] Li M, Gao F, Mascola JR, Stamatatos L, Polonis VR, Koutsoukos M, et al. Human immunodeficiency virus type 1 env clones from acute and early subtype B infections for standardized assessments of vaccine-elicited neutralizing antibodies. J Virol 2005;79(16): [9] Holl V, Hemmerter S, Burrer R, Schmidt S, Bohbot A, Aubertin AM, et al. Involvement of Fc gamma RI (CD64) in the mechanism of HIV-1 inhibition by polyclonal IgG purified from infected patients in cultured monocytederived macrophages. J Immunol 2004;173(10): [10] Precopio ML, Betts MR, Parrino J, Price DA, Gostick E, Ambrozak DR, et al. Immunization with vaccinia virus induces polyfunctional and phenotypically distinctive CD8(+) T cell responses. J Exp Med 2007;204(6): [11] Harari A, Bart P, Stohr W, Tapia G, Garcia M, Medjitna- Rais E, et al. An HIV-1 Clade C DNA Prime, NYVAC Boost Vaccine Regimen Induces Vigorous, Broad, Polyfunctional and Long-Lasting T-Cell Responses. J Exp Med 2008;205(1):