Specific humoral and cellular immune responses in cancer patients undergoing chronic immunization with a VEGF-based therapeutic vaccine

Vaccine 35 (2017) 3582–3590

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Specific humoral and cellular immune responses in cancer patients undergoing chronic immunization with a VEGF-based therapeutic vaccine Yanelys Morera a,⇑, Javier Sánchez a, Mónica Bequet-Romero a, Katty-Hind Selman-Housein b, Ana de la Torre c, Francisco Hernández-Bernal a, Yenima Martín c, Acralys Garabito c, Jesús Piñero b, Cimara Bermúdez a, Josué de la Torre b, Marta Ayala a, Jorge V. Gavilondo a a

Center for Genetic Engineering and Biotechnology, 31 Ave. between 158 and 190 Streets, Cubanacán, Playa, Havana, Cuba Center of Medical and Surgical Research, 216 and 11B Streets, Siboney, Playa, Havana, Cuba c Celestino Hernández Hospital, # 564 Cuba Street, Santa Clara City, Cuba b

a r t i c l e

i n f o

Article history: Received 22 November 2016 Received in revised form 2 May 2017 Accepted 7 May 2017 Available online 20 May 2017 Keywords: Cancer Active immunotherapy VEGF Angiogenesis Clinical trial Therapeutic vaccine

a b s t r a c t CIGB-247 is a cancer therapeutic vaccine, based on recombinant modified human vascular endothelial growth factor (VEGF) as antigen, in combination with the adjuvant VSSP, a bacterially-derived adjuvant. The vaccine have demonstrated efficacy in several murine malignancy models. These studies supported the rationale for a phase I clinical trial where safety, tolerance, and immunogenicity of CIGB-247 was studied in patients with advanced solid tumors at three antigen dose level. Surviving individuals of this clinical trial were eligible to receive off-trial voluntary re-immunizations. The present work is focus in the immunological follow up of these patients after approximately three years of immunizations, without additional oncological treatments. Long term vaccination was feasible and safe. Our results indicated that after sustained vaccination most of the patients conserved their seroconversion status. The specific antiVEGF IgG titer diminished, but in all the cases keeps values up from the pre-vaccination levels. Continued vaccination was also important to produce a gradual shift in the anti-VEGF IgG response from IgG1 to Ig4. Outstanding, our results indicated that long-term off-trial vaccination could be associated with the maintaining of one reserve of antibodies able to interfere with the VEGF/Receptor interaction and the production of IFNc secretion in CD8+ cells. The results derived from the study of this series of patients suggest that long term therapeutic vaccination is a feasible strategy, and highlight the importance of continuing the clinical development program of this novel cancer therapeutic vaccine candidate. We also highlight the future clinical applications of CIGB-247 in cancer and explain knowledge gaps that future studies may address. Registration number and name of trial registry: RPCEC00000102. Cuban Public Clinical Trial Registry (WHO accepted Primary Registry). Available from: http://registroclinico.sld.cu/. Ó 2017 Published by Elsevier Ltd.

1. Introduction The concept of blocking angiogenesis as a new therapeutic option in cancer [1], was first reduced to clinical practice in 2004 with the FDA approval of Bevacizumab (AvastinÒ), a humanized monoclonal antibody that neutralizes the pro-angiogenic effects of human Vascular Endothelial Growth Factor (VEGF), and has shown clinical benefits in colorectal, kidney, and non-small cell lung cancers [2]. ⇑ Corresponding author at: Pharmaceuticals Department, Biomedical Direction, Center for Genetic Engineering & Biotechnology, P.O. Box 6162, Havana, Cuba. E-mail address: [email protected] (Y. Morera). http://dx.doi.org/10.1016/j.vaccine.2017.05.020 0264-410X/Ó 2017 Published by Elsevier Ltd.

These results, and others [3] have underscored the relevance of anti-angiogenic strategies for cancer therapy and stimulated research in this field. One of such research scenarios involves the development of cancer therapeutic vaccines that employ as antigens different molecules involved in tumor angiogenesis, or expressed in the tumor vasculature [4]. Active immunotherapy approaches are additionally attractive because of its lesser adverse side effects, compared to other anti-tumor drugs, and their potential to stimulate other anti-tumor immunological mechanisms, apart from those specific to the vaccine [5,6]. During the last decade our team has been developing CIGB-247, a cancer therapeutic vaccine candidate that uses as antigen a

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

recombinant version of human Vascular Endothelial Growth Factor (VEGF), in combination with a the powerful VSSP bacteriallyderived adjuvant [7,8]. In mouse experimental models, the vaccine is immunogenic and inhibits tumor growth and metastasis. CIGB247 elicits specific antibodies that block the interaction VEGF and VEGF receptors, and stimulates direct cell cytotoxicity against tumor cells that secrete VEGF [7,9,10]. After preclinical tests done in rats, rabbits, and non-human primates [11,12], the Cuban regulatory authority (CECMED) approved in 2011 the development of a multi-center phase I clinical trial with CIGB-247 in cancer patients. The study, code named CENTAURO, involved thirty patients with advanced solid tumors, most of which had received all available onco-specific therapies without response. The individuals were distributed in three cohorts of ten patients that received either 50, 100 or 400 mg of the antigen, combined with 200 mg of the VSSP. Subcutaneous immunization was done once a week, for up to 8 weeks, followed by a reimmunization on week 12. The detailed results of the CENTAURO trial were first reported by us in 2014 [13]. CIGB-247 was found to be safe and tolerable in the final trial evaluation. Immunized patients had serum specific anti-VEGF IgG antibodies, the ability to block VEGF-VEGF Receptor 2 (KDR) interaction in an in vitro competitive ELISA assay, and positive IFN-c ELISPOT tests after in vitro peripheral blood mononuclear cell (PBMC) stimulation with a mutated VEGF molecule. The highest antigen dose vaccine combination produced the best results in the three specific immune response assays done, with higher frequencies of positive patients and individual test values. Patients surviving the trial were eligible to start off-trial voluntary supervised CIGB-247 re-immunizations every 4 weeks with the best antigen dose. In the same article we also presented evidences of clinical benefit, and positive results in the specific immune response tests, for patients that had been off-trial re-immunized for close to two years. The present publication reports the evolution of the specific immune response tests in eight of the original CENTAURO trial patients that, by late 2015, had received between 52 and 56 immunizations with CIGB-247, and survived at least for three years subsequent to their inclusion in the trial, with no other treatment than vaccination.

2. Materials and methods 2.1. The CIGB-247 vaccine CIGB-247 combines human VEGF and the VSSP bacterially derived adjuvant. The antigen is a recombinant fusion protein representative of human VEGF isoform 121 [7], produced as a lyophilized material under GMP conditions in 400 mg vials by the Development Unit of Center for Genetic Engineering and Biotechnology (CIGB, Havana, Cuba). VSSP are very small sized particles obtained from Neisseria meningitides outer membrane, mixed with N-acetyl GM3 [8]. VSSP is produced in liquid form by the Center for Molecular Immunology of Havana, Cuba. Lots used in the patients reported in this paper were: VED11401/0, VED11403/0, VED11101/0, VED12402/0, VED12403/0, VED13401/0, VED14401/0, D421VE/0, and D521VE/0 for CIGB-247, and 711002, 711001, 711201, 711301, 711401, 711403, 711501, and 711601 for VSSP. Antigen and VSSP were stored at 4 °C. At the moment of vaccination, one antigen vial was re-hydrated with a pre-calculated amount of USP injection water, and mixed with the established quantity of VSSP, up to a final volume never exceeding 1 mL per injection dose.

3583

2.2. Patients The CENTAURO Phase I clinical trial (RPCEC00000102, Cuban Public Clinical Trial Registry) [13], was developed between 2011 and 2012, and involved 30 patients with advanced solid tumors that were submitted to subcutaneous weekly vaccinations with CIGB-247, for 8-weeks, at three different antigen dose levels (50, 100, and 400 mg), and 200 mg of VSSP, followed by a reimmunization on week 12. A final safety, immunological, and clinical evaluation was done for all patients on week 16 after trial inclusion. Surviving individuals were eligible for supervised offtrial re-immunizations with 400 mg of antigen and 200 mg of VSSP, based on consent and their physician’s criteria. Re-immunizations were done approximately every four weeks, and carried on until death, intolerance, marked disease progression or patient’s withdrawal of consent. Eight off-trial re-immunized individuals that had survived for approximately 3 years after trial inclusion (with no other specific oncological treatment) voluntarily provided blood and cell samples for the study reported in this paper. Table 1 depicts the patient’s original trial code name, trial entry diagnosis, total number of immunizations received, accumulated survival since trial inclusion, as well as initial and 2015 clinical status (RECIST criteria) [14]. 2.3. Serum and plasma samples Stored serum and plasma samples representative of prevaccination status, of weeks 13 and 16 after inclusion in the CENTAURO trial, and of those taken at different time points during the off-trial vaccination phase (see details in each of the following sections, or in the Results chapter), were used in this study. Venous blood was collected using a blood collection set with a preattached holder (Beckton Dickinson 367355), and placed into EDTA or serum separator tubes for plasma and serum analyses, respectively. Samples were immediately stored at
70 °C, until used. 2.4. Enzyme linked immunosorbent assays (ELISAs) reagents Human VEGF isoform 121 (rhVEGF) was produced in CHO cells [15]. Skimmed milk powder (A0830) and Tween 20 (A1389) were supplied by AppliChem. HRP-conjugated goat anti-human IgG antibodies (Fc c fragment specific; Jackson Immunoresearch Laboratories, 109-035-098; 80 ng/mL) were used for detecting human serum total IgG. Biotinylated mouse monoclonal antibodies specific for human IgG1 (ab9975), IgG2 (ab99785), IgG3 (ab99830) and IgG4 (ab99824) were purchased from Abcam. Recombinant human KDR receptor 2/Fcc chimera (Sigma, V6758) was used in competitive ELISAs as described below. Bevacizumab, a commercially available monoclonal antibody specific to human VEGF (Roche, Switzerland) was used as a positive control for VEGF/KDR blockade. Biotinylated goat antibody specific for human KDR (BAF357) was supplied by R&D Systems and used at 0.1 mg/mL to detecting VEGF/KDR interaction in the competitive ELISA. Streptavidin-peroxidase conjugate (Sigma, S5512), was used at 1:30,000–1:35,000 dilutions. 2.5. ELISA for specific anti-human VEGF IgG antibodies The IgG anti-VEGF ELISA used in this study has been previously described in detail [16]. Briefly, wells were coated overnight with 2.5 mg/mL rhVEGF at 4 °C (100 mL/well), and later blocked with 2.5% goat serum, 2% skim milk, and 0.05% Tween20. Following a blocking step with 2.5% goat serum, 2% skim milk, and 0.05% Tween20 (250 mL/well), serum samples were diluted with blocking

3584

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

Table 1 Patient basic information. Patient code

Diagnosis at trial onset

Total number of immunizations (*)

AS (**) (months)

CH-11 CH-19 CH-18 CH-25 CH-28 CH-07 CH-15 CQ-17

Peritoneal metastases from an ovarian adenocarcinoma Peritoneal metastases from an uterus-ovary adenocarcinoma NSCLC with metastases in both lungs Duodenum adenocarcinoma with pancreas infiltration Lung and bone metastases from an alveolar soft-part sarcoma Liver, lymph nodes, and ovarian metastases from a small intestine carcinoid tumor Pancreatic neuroendocrine carcinoma with adrenal, lymph node, and spleen metastases Pure mediastinum seminoma

56 54 54 54 52 56 52 54

48 46 46 47 45 49 47 47

RECIST status (***) Initial

2015

SD PD PD SD PD SD PD SD

CR CR CR CR SD SD SD PR

*

The number includes the 9 vaccinations made during the 12 weeks of the CENTAURO Phase I clinical trial. Accumulated survival (2015) since the beginning of immunizations with CIGB-247. *** Response Evaluation Criteria in Solid Tumors (RECIST) classification: CR – Complete Response, PR – Partial Response, SD – Stable Disease, PD – Progressive Disease, R – Relapse, y patient dies of disease progression. **

buffer and added to the wells (100 mL/well) for 1 h at 37 °C. Specific IgG antibodies were detected with a HRP-conjugated goat antihuman IgG antibody, diluted with 2% skim milk, and 0.05% Tween20. Plates were developed using H2O2 as substrate and OPD. After 15 min, the reaction was stopped by the addition of 2.0 N H2SO4 (50 ll/well), and the absorbance was measured at 492 nm. To declare a given serum sample taken after vaccination to be positive for specific IgG anti-VEGF antibodies, the patient’s prevaccination sample titer was first subtracted from that of the given sample, and the resulting value denominated ‘‘specific antibody titer”. This value should be 2 times that of the pre-vaccination sample titer, and also 1:100, for the specific sample to be considered positive. 2.6. IgG subclasses assays Antigen-specific IgG1, IgG2, IgG3, and IgG4 antibodies were determined by indirect ELISA using biotinylated mouse monoclonal anti-human subclass-specific antibodies (IgG1, IgG2, IgG3 and IgG4). Briefly, microtiter plates were coated with rhVEGF (2.5 mg/mL) and blocked with 4% BSA. Sera were diluted with 0.4% BSA and incubated during 1 h at 37 °C. The subsequent steps of the reaction were developed as described above. Samples were considered positive in the test when the titer values produced were at least two times higher than pre-vaccination values. A given serum sample was classified as having ‘‘detectable” anti-VEGF IgG1, IgG2, IgG3, or IgG4 antibodies if the value resulting from the subtraction of the pre-vaccination sample titer from that of a given time point sample was 1:10. Samples with values below 1:10 were classified as with ‘‘non-detectable” specific IgG1, IgG2, IgG3, or IgG4 antibodies. For each patient, the IgG subclass classified as ‘‘detectable” with the highest specific antibody titer was considered as ‘‘predominant”. 2.7. Competitive ELISAs for the detection of VEGF/KDR interactions blockade Competitive ELISAs for detection of VEGF/KDR interaction blockade by serum have been previously described in detail [16]. Briefly, plates were coated overnight at 4 °C with rhVEGF. After three washes with 0.1% Tween 20 in PBS, the plates were blocked with 4% BSA for 1 h at 37 °C, followed by new washes. Serial dilutions of test sera (1/50, 1/100, 1/200, 1/400), Bevacizumab (Roche; 1 mg/mL) or dilution buffer were added (100 mL/well) and incubated for 1 h at 37 °C. Then, 100 mL of 25 ng/mL of KDR-Fc were added to the wells (12.5 and 62.5 ng/mL final concentration,

respectively) and additionally incubated for 45 min at 37 °C. After washing, wells were incubated with biotinylated anti-human KDR antibody, the latter followed by streptavidin-peroxidase conjugate. After washings, the subsequent steps of the reaction were developed as described in previous sub-sections. Maximum binding values of KDR to the plate coating were obtained from wells incubated with dilution buffer and VEGF receptor/Fc c chimera. The inhibition caused by a given serum sample to VEGF/KDR interaction was expressed in percentage, according to the following formula:

% inhibition ¼ 100%
½ðabsorbance of test sample =maximum binding absorbanceÞ  100 A given serum sample was classified with positive blocking activity when the value resulting from the division its % inhibition by that of the pre-vaccination sample was 2. Individuals with at least one given sample obtained during vaccinations that showed positive blocking activity were considered to be positive for KDR blocking. 2.8. ELISPOT assays The IFN-c ELISPOT assay was performed similarly to previously described [13]. Briefly, PBMC isolated from venous puncture blood and combined with EDTA and cryopreserved, were thawed and in vitro activated for 72 h with mutated VEGF (7.5 lg/mL) in serum-free medium (OpTmizer, GIBCO-Invitrogen). Cells were seeded in anti-IFN-c antibody pre-coated plates (MABTECH 3420-2APT). VEGF-induced IFN-c secretion by CD8+ cells was assessed using previously activated cells, submitted to additional re-stimulation for 72 h in the presence of 3.75 lg/mL of mutated VEGF, 1 lg/mL of anti-CD28 antibody (MABTECH) and recombinant IL-2 (R&D). CD8+ cells were selected by seeding onto anti-CD8/anti-IFN-c pre-coated plates (EL3094, R&D) that were cultured for an additional 72 h. IFN-c secreting clones were detected 18–20 h later, as recommended by the manufacturer (R&D). Spots were counted using ELIExpert and EliAnaliyze software (version XX, Germany) and controlled by human audit. Quadruplicates of VEGF-activated cells were compared to those of nonactivated PBMC for every time point. A positive response was considered to be vaccine induced if a significant increase in the number of spots was detected for activated cells, as compared to non-activated samples, using a web tool (www.sharp.org/zoe/ runDFR/). Induction of IFN-c in pre-vaccination cell samples was also taken into consideration as control, whenever these cells were available.

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

Patient blood samples used for ELISPOT tests shown in this article corresponded to week 48 after trial inclusion, and at different time points during weeks 132 to 144, depending on individual extraction time. For comparison purposes, week 13 IFN-c ELISPOT results used in this article were taken from a previous publication [13]. 2.9. Platelet VEGF Platelet VEGF concentrations were measured in triplicate using a commercially available sandwich ELISA kit (R&D Systems; catalogue Nos. SVE00). All supplied standard reagents and solutions were used in accordance with the manufacturer0 s instructions. VEGF released per platelet was calculated using the following formula [17]:

Platelet VEGF ¼ ½ðserum VEGF-plasma VEGFÞ  ð1
hematocritÞ=platelet count Platelet VEGF was expressed in picograms of VEGF per million platelets. For each individual, platelet VEGF values were determined 7–10 days after a given re-immunization, depending on patient availability and normalized using its corresponding baseline value. The variation of VEGF (denominated DVEGF) was expressed in percentage and was calculated using the following formula:

3585

immunization time, followed by those obtained for samples taken one week later. It can be seen that sampling time had a minor impact on the detected antibody titer levels (four samples in three patients slightly increased titer values). Conversely, VEGF/KDR blocking activity values went over the positive cut-off in the second samples for patients CH18, CH25, and CH15, and increased in one or more second sampling times for patients CH19, or CH11, CH17, CH07, and CH28. In 19 samples from different time points and individuals in which positive antibody titers were documented, no associated positive blocking activity was found. Conversely, 23 samples from seven patients that showed positive blocking activity had below cut-off anti-VEGF antibody titers. 3.2. Evolution of anti-VEGF IgG subclass with re-immunization time Changes in specific anti-VEGF IgG subclass with immunization time were evaluated in samples taken from seven of the patients one week after the last CENTAURO trial vaccination (week 13) or off-trial re-immunization (weeks differ between patients). Table 2 shows that IgG1 was predominant at week 13 in six individuals; IgG3 was also detected in four of the patients. After long-term CIGB-247 vaccination (weeks 129–145) IgG4 was the predominant subclass, exception made of patient CH-19, where IgG3 was the most important specific IgG detected.

DVEGF ¼ ½ðlevels after re-immunization=pre-vaccination levelsÞ  100
100% Based on criteria established by other authors [18], a decrease in D was defined by values 
30%, and an increase when values were 30%. Other values indicate no change. 3. Results 3.1. IgG antibody titers and VEGF/KDR blocking activity Anti-VEGF IgG antibody titers and VEGF/KDR blocking activity were studied in serum samples taken at different off-trial reimmunization times, and compared to week 16 values (one month after the last immunization of the trial’s vaccination scheme). Results are depicted per patient in Fig. 1. All individuals were positive for anti-VEGF IgG antibodies at week 16, except patient CH-07. With time, titer values showed a tendency to drop in most individuals, becoming negative for patient CH-18 after week 48, and after weeks 100, 112, or 108 for patients CH-25, CQ-17, and CH-28, respectively. Antibody titers also dropped with time for patients CH-19 and CH-15, but remained positive or borderline positive. The same Figure shows that all individuals were positive for VEGF/KDR interaction blocking activity at week 16, exception made of patient CH-11. This patient started to show positive results in the blocking activity test only after re-immunization. Blocking activity values in samples from patients CH-19, CH-18, CH-28, and CH-07 dropped with time, without becoming negative, exception made of some inter-spaced negative results for CH-18. CH-25, CH-15 had some negative samples in the test after weeks 96 or 104, respectively. In the case of CQ-17, blocking activity values substantially increased with time. Because referred samples were routinely taken at the moment of immunization (approximately one month after the previous vaccination time), additional blood samples taken approximately one week after a given re-immunization to assess its immediate effect. Anti-VEGF specific IgG titers, as well as VEGF/KDR blockade results are depicted in Fig. 2. The first values correspond to re-

3.3. IFN-c and CD8+ IFN-c ELISPOTs Table 3 depicts the results of the cellular immune response studies, using samples taken at week 48, and thereafter (a time point between weeks 132 and 144), after trial inclusion. Positive IFN-c secretion after short term stimulation of cells sampled on week 48 was documented in patients CH-07, CH-11, CH-15, CH-18, CH-19, CH-25, and CH-28, similar to what had reported previously at week 13. Patient CQ-17 was classified for the first time as positive to the test on week 48. Four out of the six patients with available samples for the CD8+ IFN-c ELISPOT test were also classified positive in this assay, at this time point. At the later sampling times studied, six of the eight individuals retained their positive classification, taking into consideration one or both ELISPOT tests. CQ-17 returned to a negative status, and CH25 became negative. 3.4. Platelet VEGF levels in samples taken after long term immunization with CIGB-247 As shown in Fig. 3, patients CH-15, CQ-17, CH-18, and CH-19 showed no significant change in platelet VEGF levels, with respect to pre-vaccination values, at sampling times close to three years of CIGB-247 immunization. In the case of patients CH-07, CH-11, and CH-25, a substantial increase in their pre-vaccination platelet VEGF levels was documented. 4. Discussion Because of the immune depressed state of cancer patients and the ‘‘self” nature of many tumor-associated antigens used in cancer therapeutic vaccines, it is accepted that these anti-tumor active immunotherapy procedures should involve both powerful vaccine adjuvants and chronic immunization of patients, in order to break tolerance, sustain a specific immune response, and eventually show clinical benefits for the patients [19,20]. Concerning chronic immunization of patients, adverse side effects on one side, and the presence of objective clinical improve-

3586

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

Fig. 1. Specific IgG antibodies and VEGF/KDR blocking activity in serum samples from cancer patients re-immunized off-trial with CIGB-247. Anti-VEGF specific IgG antibodies (inverse of titer) are shown as open black squares, and % of VEGF/KDR blocking activity as grey bars. Samples shown were taken always before receiving the corresponding monthly off-trial re-immunization. Week 0 (pre-vaccination) is shown as reference. Black hyphened, and grey dot plus hyphen lines, are the positive cut-off values for specific antibody titer and blocking activity, respectively. See Material and Methods for more details.

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

3587

Fig. 2. Immediate effect of monthly re-immunization on anti-VEGF IgG antibody titers and VEGF/KDR interaction blockade values. Anti-VEGF specific IgG antibodies (inverse of titer) are shown as black open squares and % VEGF/KDR blocking activity as grey bars. The consecutive weeks in the x axis, after week 16, and linked by the downward pointing horizontal square bracket symbols, refer to the moment when blood samples were taken, the first coinciding with the time the patient attended consultation for reimmunization, and the second one week later. Weeks 0 (pre-vaccination) and week 16 are shown as references. The discontinued black hyphened, and the grey hyphen plus dot lines are the positive antibody titer and VEGF/KDR blocking activity cut-off values, respectively, and are the same shown in Fig. 1.

3588

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

Table 2 Anti-VEGF specific IgG subclasses. Patient code

Week

IgG1

IgG2

IgG3

IgG4

CH-11

13 141 13 133 13 133 13 132 13 145 13 141 13 129

+++ – +++ + na – +++ – +++ – +++ – +++ –

– – – – na – – – – – – – – –

+ – + +++ na – – – – – + – + +

– +++ – + na +++ – +++ – +++ – +++ – +++

CH-19 CH-18 CH-25 CH-07 CH-15 CH-28

Notes: + + +: Predominant subclass; + positive;
negative; na: sample not available for this testing; see Material and Methods for more details.

Table 3

c-IFNand CD8+ c-IFN ELISPOTs results. Patient code

CH-07 CH-11 CH-15 CQ-17 CH-18 CH-19 CH-25 CH-28

Week 13 (*)

Week 48

Weeks 132–144 (**)

c–IFN

c–IFN

CD8+ c–IFN

c–IFN

CD8+ c–IFN

+ + + – + + + +

+ + + + + + + +

+
+ nd nd +
+

+ + +

+
+

+ +
nd + +
+

Notes: c–IFN = gamma Interferon ELISPOT test; c–IFN CD8+ = specific gamma Interferon CD8+ cells ELISPOT test; nd = not done due to lack of adequate sample or availability; + and
symbols represent samples classified as positive or negative, respectively. See Materials and Methods for more details. * Week 13 results were taken from a previous publication [13] for comparison purposes. ** Weeks are expressed as a range, because not all samples were taken at the same week for different patients.

Fig. 3. Platelet VEGF in cancer patients undergoing chronic vaccination with CIGB247. Y-axis are patient codes and weeks when the sample was taken for comparison with pre-vaccination values. X-axis shows the DVEGF values, expressed as %. Broken lines are cut-off values. A given sample was classified as showing a reduction in platelet VEGF when % values are 
30. Percent values 30 indicate an increase in platelet VEGF. Values within the
29 to +29% ranges indicate no significant change. See Materials and Methods for more details.

ments on the other are, by far, the most important guiding elements in defining how much time a given vaccine preparation should be administered to the patients. However, monitoring the specific immune response to the vaccine is also relevant to this matter. In fact, long-term specific immune response tests not only

contribute to our understanding of the potential anti-tumor immunological mechanisms elicited by immunization and their time course, and lead to vaccine dose adjustments, but can evolve into being part of each patient’s stratification and/or prognostic considerations [20]. In the present article we focus in the immunological follow up of eight of the long-term surviving CENTAURO patients, which have been chronically immunized with CIGB-247 for at least 3 years. In a previous 2014 publication [13] we had shown that surviving patients from CENTAURO phase I clinical trial that had been off-trial re-immunized with CIGB-247 on a monthly basis, maintained or improved their previous status, with respect to three specific immune response tests, i.e. anti-VEGF antibodies in serum, the ability of serum to block VEGF/KDR interaction, and a ELISPOT assay that measured gamma IFN secretion by antigen stimulated blood cells. In this new study we have found that anti-VEGF IgG antibody titers tend to decrease with time in most of the eight long-term vaccinated patients, with respect to week 16 end-of-trial values. While in some patients this behavior leads to below cut-off negative values, we found that titers never permanently drop to or below pre-vaccination status. Because blood for the humoral response tests done with these patients was routinely drawn from patients at the moment of re-immunization, i.e. one month after the previous vaccination, we also compared the antibody titers in samples taken at CIGB-247 vaccination time, and a week after being re-immunized. To our surprise, we found no important differences between the two samples. Antibody titers have been reported to drop or stabilize after long-term vaccination with other vaccines that use tumorassociated antigens. In breast cancer patients vaccinated with MUC1-KLH [21], and in melanoma patients immunized with GD3-lactone-KLH [22], both conjugated with QS-21, antibody titers values decreased after sustained immunization, albeit values never reached pre-vaccination levels [21,22]. Conversely, for a cancer vaccine that uses Epidermal Growth Factor (EGF) as antigen, antibody titers reached a plateau in many patients, and remained at high levels as long as vaccination proceeded [19]. Despite the aforementioned, an additional explanation to our findings could be related to the goat anti-human IgG polyclonal antibody preparation used as detecting reagent in our antibody ELISA test [16]. Because sensitivity of whole anti-IgG conjugates has been reported to be higher for IgG1 subclass antibodies [23,24], and we found a marked skewing towards specific IgG4 with CIGB-247 vaccination time, the apparent drop in specific antibody titers we have detected after long-term chronic immunization could be related to a change in response quality, that remains undetected because of test limitations. IgG4 switch has been reported to occur after repeated antigen exposure [25], probably associated with the terminal position of the Cc4 cassette, and results similar to ours have been shown in colorectal carcinoma patients immunized with a recombinant carcinoembryonic antigen (CEA) vaccine for three years [26]. These authors documented a progressive shift from IgG1, present during the first weeks of immunization, to IgG4, arising only after months of treatment. Some authors claim that this switch may also happen concomitant with an increase in antigen affinity [27] a matter we have not yet explored in the case of CIGB-247. This aspect related to our ELISA tests characteristics could also play a role in the VEGF receptors, and particularly KDR (VEGF receptor 2) [16] play a transcendental role in translating the effects of VEGF binding by endothelial cells and their precursors, into tumor neo-angiogenesis [28]. This makes the measurement of VEGF/KDR blocking activity a very relevant element in the understanding of the potential capacity of the CIGB-247 vaccine of

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

eliciting, and maintaining an effective anti-VEGF neutralizing antibody response. In our study we also found a decrease in blocking activity with time in six of the studied patients. But different from what was seen in the antibody tests, blocking values remained mostly positive. Moreover, one patient gained VEGF/KDR blocking activity only after long term re-immunization, and in another individual blocking activity values increased with re-immunization. Also different from what was shown in the case of anti-VEGF antibodies, studies related to possible immediate effects of re-immunization showed that blocking values returned to previously ‘‘lost” positive cut-off levels or increase, a week after re-immunization. Apparent disparities were found when comparing the two aforementioned humoral immune response tests in seven of the eight long-term immunized patients. For samples with positive antibody titers but no associated blocking activity the most plausible explanation could be that, at some specific sampling time points circulating anti-VEGF antibodies lacked functional VEGF/KDR blocking ability. In the case where blocking activity was detected without apparent positive anti-VEGF IgG antibody results, we can put forward at this point to non-excluding reasons to this phenomenon. The first one is based on the already described possible subclass detection limitations of the antibody ELISA assay, which would exclude the detection of IgG4 neutralizing antibodies. A second explanation takes into account that VEGF/KDR interaction can be blocked in our corresponding test by anti-VEGF antibody classes different from IgG. In our previous CENTAURO trial paper [13] we had already reported that anti-VEGF IgM could be detected in serum samples during the trial time. We have very recently expanded these results, and in a submitted article we now show that anti-VEGF specific IgM can be also detected after long-term re-immunization with CIGB-247. The generation of anti-tumor cytotoxic T lymphocytes (CTL) is a desired hallmark of many cancer therapeutic vaccine strategies [29], but because of test complexities, antigen-specific T cells have been evaluated only at one or two time points after vaccination in the majority of studies, and little is known about the length of this response [30]. Despite the limitations that the use of peripheral blood cells imposes to the functional interpretation of ELISPOT IFN-c and other cytokine secretion assays, these tests are the most frequently used tools employed to measure an antigen-specific T cell response. In these, stimulation of peripheral blood lymphocytes by peptides of different length is used to measure both CTL frequency and the existence of a ‘‘helper” or ‘‘cytotoxic” response in the vaccinated patients [31]. In our previous CENTAURO trial article [13] we had shown that most patients surviving at up to one year of vaccination were positive in an ELISPOT IFN-c secretion test where patients’ PBMC were activated for a short time with a mutated variant of human VEGF. These results, while indicative of the vaccine’s capacity of specifically stimulating cellular immunity, were nevertheless not informative about the CD4+ or CD8+ nature of the response. This limitation has been overcome with the use of single and double activation test variants described in the present paper. We have now been able to show that long-term off-trial vaccination associates to a positive status in one or both assays for a majority of the studied patients, and that CIGB-247 can specifically stimulate the production of IFN-c secretion in CD8+ cells, the latter a relevant aspect to take into account when discussing the vaccine’s potential anti-tumor mechanisms. The use of tumor markers [32] that can help in characterizing the evolution of disease and the effects of vaccination will be an important part in the design of future Phase II/III trials for the CIGB-247 vaccine. Within this context, and because of the nature of our vaccine, angiogenesis-related biomarkers could also be potentially useful.

3589

The measurement of VEGF blood/platelet levels has been evaluated in studies done with other anti-angiogenic products [33], albeit with inconclusive results. We first studied this biomarker in the CENTAURO phase I trial and found that by week 13, with respect to pre-vaccination time, platelet VEGF levels had decreased in patients vaccinated with the higher antigen dose [13]. Platelet VEGF levels were re-examined now in the eight long-term vaccinated patients, all of which have shown objective clinical benefits after participating in the CENTAURO study and off-trial [13,34]. Our present results showed either no significant change with respect to pre-vaccination values, or increased platelet VEGF levels. Because of the patient sample size, and the original design of our phase I trial, it is not possible to reach definitive conclusions about the potential prognostic value of the test, but the inclusion of platelet VEGF studies as a potential biomarker in CIGB-247 vaccination will surely be a matter of debate in the design future clinical trials. Overall, the results discussed herein on long-term CIGB-247 offtrial vaccinated patients, add to recent evidences included in a detailed series case report of our clinical findings [34], highlight the importance of continuing the clinical development program of this novel cancer therapeutic vaccine candidate. The clinical follow up of CENTAURO patients has shown an accumulated survival time ranging between 4.5 to 4.9 years after trial inclusion. Of the eight patients, two ovary cancers and one NSCLC have achieved complete response status. A patient with a duodenum adenocarcinoma with pancreas infiltration has recently relapsed after two years free of disease. An individual with a pure seminoma has a documented partial response. The other three patients (a metastatic alveolar soft-part sarcoma, a small intestine carcinoid tumor, and a metastatic pancreas neuroendocrine carcinoma) have maintained a stable disease status for several years [34]. Taking into consideration that the CENTAURO clinical trial was not designed to prove CIGB-247 efficacy, it is required to be cautious when advancing the implications of the vaccine in the clinical findings and their relationship with the induced immune response. Nevertheless, is encouraging that all patients included in this follow up showed some evidence of positive response to vaccination, in terms of the specific immune tests designed by us to this effect. The low toxicity profile of CIGB-247 opens interesting possibilities in the future combination of our vaccine with, or after, chemotherapy, radiotherapy, or even other anti-angiogenic treatments, and in the chronic application of the vaccine for long-term cancer control purposes. The vaccine candidate is now being considered for Phase II studies in specific tumors as ovarian and metastatic colorectal cancer. We will continue monitoring the specific immune response to the vaccine in future trials, as part of our efforts to characterize the potential anti-tumor mechanisms of CIGB-247. Conflict of interest disclosure Authors YMD, JS, MBR, FHB, CMV, MAA, and JVG, are employees of the Center for Genetic Engineering and Biotechnology, where the CIGB-247 vaccine was developed. The remaining authors have no conflict of interests. No honoraria, consulting fees or payments for seminar presentations, speeches or appearances have been received by any of the authors. The study was sup-ported by Heber Biotec (http://www.heberbiotec.com), Chemo (http://www.chemogroup.com), and the Ministry of Public Health of Cuba. References [1] Folkman J. Angiogenesis: an organizing principle for drug discovery? Nature Rev Drug Discov 2007;6:273–86.

3590

Y. Morera et al. / Vaccine 35 (2017) 3582–3590

[2] Keating GM. Bevacizumab: a review of its use in advanced cancer. Drugs 2014;74:1891–925. [3] Petrovic N. Targeting angiogenesis in cancer treatments: where do we stand? J Pharm Pharm Sci: Publ Canad Soc Pharm Sci, Soc Canad des Sci Pharm 2016;19:226–38. [4] Wentink MQ, Huijbers EJ, de Gruijl TD, Verheul HM, Olsson AK, Griffioen AW. Vaccination approach to anti-angiogenic treatment of cancer. Biochimica et Biophysica Acta. 2015;1855:155–71. [5] Guo C, Manjili MH, Subjeck JR, Sarkar D, Fisher PB, Wang XY. Therapeutic cancer vaccines: past, present, and future. Adv Can Res 2013;119:421–75. [6] van der Burg SH, Arens R, Ossendorp F, van Hall T, Melief CJM. Vaccines for established cancer: overcoming the challenges posed by immune evasion. Nature Rev Cancer 2016;16:219–33. [7] Morera Y, Bequet M, Ayala M, Lamdán H, Agger EM, Andersen P, et al. Antitumoral effect of active immunotherapy in C57BL/6 mice using a recombinant human VEGF protein as antigen and three chemically unrelated adjuvants. Angiogenesis 2008;11:381–93. [8] Estévez F, Carr A, Solorzano L, Valiente O, Mesa C, Barroso O, et al. Enhancement of the immune response to poorly immunogenic gangliosides after incorporation into very small size proteoliposomes (VSSP). Vaccine 1999;18:190–7. [9] Bequet M, Morera Y, Ayala M, Ancízar J, Soria Y, Blanco A, et al. CIGB-247: a VEGF-based therapeutic vaccine that reduces experimental and spontaneous lung metastasis of C57Bl/6 and BALB/c mouse tumors. Vaccine 2012;30:1790–9. [10] Pérez Sánchez L, Morera Diaz Y, Bequet-Romero M, Ramses Hernández G, Rodríguez Y, Castro Velazco J, et al. Experimental studies of a vaccine formulation of recombinant human VEGF antigen with aluminum phosphate. Human Vacc Immunother 2015;11:2030–7. [11] Morera Y, Bequet M, Ayala M, Velazco JC, Pérez PP, Alba JS, et al. Immunogenicity and some safety features of a VEGF-based cancer therapeutic vaccine in rats, rabbits and non-human primates. Vaccine 2010;28:3453–61. [12] Morera Y, Bequet M, Ayala M, Pérez PP, Castro J, Sánchez J, et al. Antigen dose escalation study of a VEGF-based therapeutic cancer vaccine in non human primates. Vaccine 2012;30:368–77. [13] Gavilondo JV, Hernández F, Ayala M, de la Torre AV, de la Torre J, Morera Y, et al. Specific active immunotherapy with a VEGF vaccine in patients with advanced solid tumors. Results of the CENTAURO antigen dose escalation phase I clinical trial. Vaccine 2014;32:2241–50. [14] Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Europ J Cancer (Oxford, England: 1990) 2009;45:228–47. [15] Sánchez Ramírez J, Morera Diaz Y, Musacchio Lasa A, Bequet-Romero M, Muñoz Pozo Y, Pérez Sánchez L, et al. Indirect and competitive enzyme-linked immunosorbent assays for monitoring the humoral response against human VEGF. J Immun Immunochem 2016. [16] Sánchez Ramírez J, Morera Díaz Y, Musacchio Lasa A, Bequet-Romero M, Muñoz Pozo Y, Pérez Sánchez L, et al. Indirect and competitive enzyme-linked immunosorbent assays for monitoring the humoral response against human VEGF. J Immun Immunochem 2016;37:636–58. [17] Adams J, Carder PJ, Downey S, Forbes MA, MacLennan K, Allgar V, et al. Vascular endothelial growth factor (VEGF) in breast cancer: comparison of

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25] [26]

[27] [28]

[29]

[30]

[31] [32] [33]

[34]

plasma, serum, and tissue VEGF and microvessel density and effects of tamoxifen. Cancer Res 2000;60:2898–905. Recchia F, Candeloro G, Necozione S, Bisegna R, Bratta M, Rea S. Immunotherapy in patients with less than complete response to chemotherapy. Antican Res 2009;29:567–72. González G, Crombet T, Lage A. Chronic vaccination with a therapeutic EGFbased cancer vaccine: a review of patients receiving long lasting treatment. Curr Can Drug Targets 2011;11:103–10. Vatsan RS, Bross PF, Liu K, Theoret M, De Claro AR, Lu J, et al. Regulation of immunotherapeutic products for cancer and FDA’s role in product development and clinical evaluation. J ImmunoTherapy Cancer 2013;1:1–16. Gilewski T, Adluri S, Ragupathi G, Zhang S, Yao TJ, Panageas K, et al. Vaccination of high-risk breast cancer patients with mucin-1 (MUC1) keyhole limpet hemocyanin conjugate plus QS-21. Clin Can Res: Off J Am Assoc Cancer Res 2000;6:1693–701. Ragupathi G, Meyers M, Adluri S, Howard L, Musselli C, Livingston PO. Induction of antibodies against GD3 ganglioside in melanoma patients by vaccination with GD3-lactone-KLH conjugate plus immunological adjuvant QS-21. Int J Cancer 2000;85:659–66. Hart MH, de Vrieze H, Wouters D, Wolbink GJ, Killestein J, de Groot ER, et al. Differential effect of drug interference in immunogenicity assays. J Immunol Methods 2011;372:196–203. Bendtzen K. Immunogenicity of Anti-TNF-a Biotherapies: II. Clinical relevance of methods used for anti-drug antibody detection. Front Immunol 2015;6:109. van Zelm MC. B cells take their time: sequential IgG class switching over the course of an immune response? Immunol Cell Biol 2014;92:645–6. Ullenhag GJ, Frodin JE, Strigard K, Mellstedt H, Magnusson CG. Induction of IgG subclass responses in colorectal carcinoma patients vaccinated with recombinant carcinoembryonic antigen. Cancer Res 2002;62:1364–9. Collins AM, Jackson KJ. A temporal model of human IgE and IgG antibody function. Front Immunol 2013;4:235. Cébe-Suarez S, Zehnder-Fjällman A, Ballmer-Hofer K. The role of VEGF receptors in angiogenesis; complex partnerships. Cell Molecul Life Sci 2006;63:601–15. Ribas A, Butterfield LH, Glaspy JA, Economou JS. Current developments in cancer vaccines and cellular immunotherapy. J Clin Oncol: Off J Am Soc Clin Oncol 2003;21:2415–32. Raez L, Walker G, Baldie P, Fisher E, Gomez J, Tolba K, et al. CD8 T cell response in a phase I study of therapeutic vaccination of advanced NSCLC with allogeneic tumor cells secreting endoplasmic reticulum-chaperone gp96-Igpeptide complexes*. Adv Lung Cancer 2013;2:9–18. Slota M, Lim J-B, Dang Y, Disis ML. ELISpot for measuring human immune responses to vaccines. Expert Rev Vacc 2011;10:299–306. Duffy MJ. Tumor markers in clinical practice: a review focusing on common solid cancers. Med Princ Pract 2013;22:4–11. Alidzanovic L, Starlinger P, Schauer D, Maier T, Feldman A, Buchberger E, et al. The VEGF rise in blood of bevacizumab patients is not based on tumor escape but a host-blockade of VEGF clearance. Oncotarget 2016. Selman-Housein KH, de la Torre A, Hernández-Bernal F, Martin Y, Garabito A, Piñero J, et al. Clinical benefits in patients with advanced solid tumors after long-term immunization with a VEGF therapeutic vaccine. Open J Clin Med Case Rep 2016.