Kinetics of tumor-specific T-cell response development after active immunization in patients with HER-2/neu overexpressing cancers

Clinical Immunology (2007) 125, 275 280 available at www.sciencedirect.com www.elsevier.com/locate/yclim Kinetics of tumor-specific T-cell response development after active immunization in patients with HER-2/neu overexpressing cancers Lupe G. Salazar a,, Andrew L. Coveler a, Ron E. Swensen a, Theodore A. Gooley b, Vivian Goodell a, Kathy Schiffman a, Mary L. Disis a a Tumor Vaccine Group, Division of Oncology, University of Washington, Box 358050, 815 Mercer Street, Seattle, WA 98109, USA b Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, Seattle, WA 98109, USA Received 20 February 2007; accepted with revision 12 August 2007 Available online 29 October 2007 KEYWORDS HER-2/neu; Vaccination; Immunity; Kinetics Abstract The ability of a cancer vaccine to elicit a specific measurable T-cell response is increasingly being used to prioritize immunization strategies for therapeutic development. Knowing the optimal time during a vaccine regimen to measure the development of tumorspecific immunity would greatly facilitate the assessment of T-cell responses. The purpose of this study was to overview the kinetics of HER-2/neu-specific T-cell immunity evolution during and after the administration of HER-2/neu peptide-based vaccination in the adjuvant setting. Furthermore, we questioned whether the presence of preexistent HER-2/neu T-cell immunity or the timing of immunity development over the course of active immunization influenced the intensity of the elicited HER-2/neu-specific T-cell immunity. Our findings demonstrate that maximal tumor-specific immune responses may occur toward the end of the vaccination regimen or even after the scheduled vaccines have been completed. Additionally, the presence of tumor antigen-specific immunity prior to vaccination is associated with greater magnitude immune responses. 2007 Elsevier Inc. All rights reserved. Introduction Corresponding author. Fax: +1 206 685 3128. E-mail address: lsalazar@u.washington.edu (L.G. Salazar). The assessment of antigen-specific immune responses after active immunization with cancer vaccines is one method to prioritize vaccines for further development. As recently recommended by The Cancer Vaccine Clinical Trial Working Group [1], therapeutic cancer vaccines should be investigated in two types of clinical studies: proof-of-principle vaccine trials and efficacy trials. Proof-of-principle vaccine trials such as the one described here should be conducted in the adjuvant setting in patients who do not have rapidly progressing disease to allow for sufficient time for biologic (i.e., generation of immune response) and potential clinical activity to develop. Additionally, immune responses should be evaluated sequentially with a minimum of 3 time points (baseline and at least 2 follow-up time points) and the frequency and magnitude of an immune response should be defined for the population under study. Un- 1521-6616/$ see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.clim.2007.08.006

276 L.G. Salazar et al. fortunately, few vaccine studies to date have been conducted in such a manner and as such the optimal timing for evaluation of the immune response is not well defined. Furthermore, very little data exist on the frequency and magnitude of an immune response in the post-vaccination setting once active immunization has ceased. A more detailed understanding of T cell activation and antigen processing has lead to a wealth of novel vaccination strategies that may significantly potentiate tumorspecific immunity. Thus, the identification of factors associated with the development of T cell responsiveness after vaccination remains critical. Murine studies would suggest that the full potential of repeated vaccinations may not be reached until late in the vaccine course or at a distant time point after completing vaccination [2]. Moreover, the intensity or strength of a vaccine-induced response appears to be vital to its efficacy in tumor rejection [3]. Lastly, evidence indicates that serial tumor peptide-based vaccination can successfully boost preexistent tumor-driven T cell responses [4]. Based on these limited findings and recognizing that few studies have explored the kinetics of the development of T cell responses to vaccination with cancer antigens, we retrospectively evaluated the evolution of HER-2/neu-specific T cell immunity over the course of active immunization and at distant time points post-immunization in a large number of cancer patients. We questioned whether there was an optimal time to measure the development of T cell immunity after a cancer vaccine as well as what role preexistent immunity had in the either augmenting or inhibiting the vaccinated response. Methods and materials Patient population Thirty-eight subjects with HER-2/neu overexpressing cancer were enrolled in a Phase I HER-2/neu peptide vaccine study, completed 6 vaccinations as previously described [5], and were available for evaluation. Briefly, all subjects had completed standard conventional therapy for their disease so that they had no detectable disease or were stable on hormonal or bisphosphonate therapy only. Additionally, all subjects had been off and were able to remain off chemotherapy or other immune modulatory therapies for a minimum of 30 days before enrollment and for the 7-month duration of the study (6 monthly vaccines plus 1 month post- 6th vaccine follow-up). Subjects were vaccinated with one of three vaccine formulations each containing three different peptides derived from the HER-2/neu protein. All peptides were potential helper epitopes of HER-2/neu predicted by computer modeling, and empiric testing found them to be immunogenic [6]. The HER-2/neu peptide-based vaccines were admixed with GM-CSF and administered intradermally once monthly for 6 months to the same regional draining lymph node site [5]. Detection of peripheral blood T cell responses HER-2/neu-specific T cell responses were measured at baseline, 30 days after each vaccination, and then every 2 months for a period of 1 year from start of immunization. To evaluate the effectiveness of this peptide-based vaccine strategy in generating T-cell immunity to HER-2/neu protein, [7] both HER-2/neu peptide-and protein-specific responses were evaluated. In addition, T cell responses were continually measured every 1 2 months in several subjects for a total period of 18 months from start of immunization. T cell proliferation was assessed at different time points using a modified limiting dilution assay as previously described [7]. While patient samples were assayed at different time points, the modified limiting dilution assay is designed for detecting low frequency lymphocyte precursors based on Poisson distribution [8], and has recently been shown to have adequate sensitivity and reproducibility in the ability to measure a broad range of T cell responses; and was determined to be a better discriminator of low-level immune responses when compared to ELISPOT [9]. Results were evaluated as a standard stimulation index (SI), defined as the mean of all 24 experimental wells divided by the mean of the control wells (no antigen). An SIN2 was considered evidence of an immunized response based on analysis of a reference population [5]. If subjects had an SI N 2 at baseline, i.e. preexistent immunity to HER-2/neu [10], a post-vaccination response was defined as positive if it was a minimum of 2 times baseline. Subjects were considered to have developed an immune response if they had a positive response to at least one peptide in the immunizing mixture. Statistical methods Subjects who developed immunity by the 6th vaccine were separated into 2 groups: those who achieved a positive immune response early or late in the vaccine course. An early response was defined as a positive response developing by 30 days after the third vaccine, i.e. during the first half of the vaccine regimen. A late response was defined as a positive response developing after the fourth vaccine, i.e. during the second half of the vaccine regimen. The probability of developing an immune response to vaccination was summarized using cumulative incidence estimates, where death without an immune response was regarded as a competing risk [11]. The magnitude of both peptide and protein immune response was evaluated as follows: the maximum immune response (SI) for each patient was selected regardless of whether it occurred before or after the 3rd vaccine, and the mean maximum immune response was calculated for each group. Two-sample t-tests were used to test the difference between the maximum immune response of the 2 groups. All reported p values are twosided. Subjects who had a positive immune response at baseline were compared to those who did not have immunity at baseline. The magnitude of the peptide and protein immune response was evaluated as follows: the maximum immune response (SI) after vaccination for each patient was selected regardless of whether it occurred before or after the 3rd vaccine. The Mann Whitney test was used to test the difference between the protein responses. An unpaired t-test with Welch s correction was used to test the difference between immune responses for the peptide responses.

Kinetics of tumor specific immunity 277 Results The majority of advanced stage cancer patients develop HER-2/neu peptide-specific T cell immunity early in the course of vaccination Twenty subjects had complete evaluation of T cell responses (baseline, then every 2 months for a period of 1 year from start of immunization), 14 subjects had T cell evaluation at baseline, then every 2 months until 30 days after the last vaccine. Two subjects had T cell evaluation done at only 2 time points after baseline due to early study termination. No patients received other interim treatments during the 1-year testing period. Fig. 1A shows the maximum SI to Figure 2 Early immunity predicts the magnitude of HER-2/neu peptide-specific T cell response. Shown are the mean maximum immune responses for peptide T cell responses in subjects who developed early immunity (defined as a positive response developing by 30 days after the third vaccine, i.e. during the first half of the vaccine regimen) or late immunity (defined as a positive response developing after the fourth vaccine, i.e. during the second half of the vaccine regimen) during the course of vaccination. immunizing HER-2/neu peptides at each time point for each individual. Data suggest that peptide-specific immune responses occur early in the course of vaccination. Fig. 1B depicts the probability of developing a HER-2/neu peptidespecific Tcell response by a particular number of vaccinations in the subjects who completed the entire course of immunizations. The probability of developing Tcell immunity by the 3rd vaccine, to at least one peptide in an individual s immunizing mix, was 82%. The probability of developing Tcell immunity to at least one peptide in an individual s mix after the 6th vaccine was 95%. Thus, the majority of subjects with advanced stage cancer were able to generate an immune response to at least one of the HER-2/neu peptides in their vaccine early, after only 3 vaccinations. In addition, the magnitude of HER-2/neu peptide-specific T cell responses was greater in subjects who demonstrated detectable immunity by the 3rd vaccine. The average SI was 13.5 (range, 2.0 59) for subjects achieving a positive immune response during the first half of the vaccine regimen and 4.6 (range, 1.2 15.5) for those achieving a positive immune response during the second half of the vaccine regimen (Fig. 2). The difference in mean maximum SI between the two groups was significant (p = 0.004). Figure 1 The majority of advanced stage cancer patients develop HER-2/neu peptide-specific Tcell immunity early in the course of vaccination. (A) The maximum SI to HER-2/neu peptides for each of the 38 subjects is shown at baseline, after each vaccination, and during follow-up. Vaccination time points correspond with evaluation of HER-2/neu-specific T-cell responses which were measured 30 days after each vaccination, before the next immunization as follows: time point 0 =1st vaccine; time point 1=30 days after 1st vaccine; time point 2=60 days after 1st vaccine; time point 3=90 days after 1st vaccine; time point 4 =120 days after 1st vaccine; time point 5=150 days after 1st vaccine; time point 6=180 days after 1st vaccine. Dotted line depicts SI of 2.0. (B) The estimated probability of developing a HER-2/neu peptide-specific T cell immune response before and after each vaccination is shown. The majority of advanced stage cancer patients develop HER-2/neu protein-specific T cell immunity later in the course of vaccination Fig. 3A shows the SI to HER-2/neu protein at each time point for each of the 38 subjects. As compared to the peptidespecific responses, more subjects developed protein-specific immunity later in the course of vaccination, even after active immunizations ended. Fig. 3B depicts the probability of developing a HER-2/neu protein-specific T cell response by a particular time point where the vaccination occurred during months 1 through 6. The probability of developing a proteinspecific Tcell response was 42% by the 3rd vaccine compared to 62% by the 6th vaccine. An additional 13% of subjects

278 L.G. Salazar et al. developed a protein-specific T cells response at a distant time point after completion of all 6 vaccinations. All subjects that developed a protein-specific T cell response did so during vaccination or within 6 months of completing all 6 vaccinations. Preexisting immunity predicts magnitude of the HER-2/neu protein-specific T cell response Five of the 38 subjects had preexistent protein immunity. The magnitude of protein-specific T cell response was greater in subjects who did demonstrate a positive immune response at Figure 4 Preexisting immunity predicts the magnitude of HER-2/neu protein-specific T cell response. The median maximum protein immune response (SI) during vaccination in subjects with (Yes) or without (No) pre-existent immunity is shown. Dotted line at 2.0 represents baseline. baseline as shown in Fig. 4. The median protein-specific Tcell response after vaccination was significantly different (p =0.0089) between those with pre-existing immunity (median SI 5.1 (range 3.8 32.1)) and those without preexisting immunity (median SI 2.2 (range 0.7 13.6)). Discussion Figure 3 The majority of advanced stage cancer patients develop HER-2/neu protein specific T cell immunity later in the course of vaccination. (A) The maximum SI to the HER-2/neu protein for each of the 38 subjects is shown at baseline, after each vaccination and during follow-up. Vaccination time points correspond with evaluation of HER-2/neu-specific T-cell responses which were measured 30 days after each vaccination, before the next immunization as follows: time point 0= 1st vaccine; time point 1=30 days after 1st vaccine; time point 2=60 days after 1st vaccine; time point 3=90 days after 1st vaccine; time point 4= 120 days after 1st vaccine; time point 5=150 days after 1st vaccine; time point 6=180 days after 1st vaccine. The asterisk ( ) indicates the value of 32 achieved on one subject after the second vaccination. (B) The estimated probability of developing a HER-2/neu protein specific Tcell immune response before and after each vaccination and during follow-up is shown. There are many new vaccine strategies being developed to target human self-tumor antigens [12]. Novel vaccine approaches are designed to circumvent mechanisms by which tumors escape immune surveillance. When vaccines approach the transition from pre-clinical models to human clinical trials much emphasis has been placed on the quantitative measurement of the immune response, particularly a T cell response, as an initial assessment of vaccine potency. As the use of cancer vaccines moves into the adjuvant setting where evaluation of clinical response, i.e. prevention of disease relapse, will require larger number of patients, an indicator in Phase I studies of vaccine immunogenicity has increasing importance. The immunogenicity of a novel vaccine may significantly impact whether that approach progresses to later phase studies. There have been few investigations attempting to determine the optimal time during a vaccine regimen to measure the development of tumor-specific immunity. Most studies focus on assessing the development of T cell responses only during active immunization. We overviewed the kinetics of the evolution of T cell immunity during and after the administration of a HER- 2/neu peptide-based vaccine [5]. Data presented here demonstrate that maximal tumor-specific immune responses may occur towards the end of the vaccination regimen or even after the scheduled vaccines have been completed. In addition, the presence of tumor antigen-specific immunity prior to vaccination is associated with greater magnitude immune responses. A major concern for the use of peptide-based vaccines is the possibility of developing peptide-specific immunity with no response to native protein. It is likely that peptide-based vaccine strategies will be effective only if the peptide-specific responses generated allow recognition of native protein. Unlike previous studies which have reported that T cells

Kinetics of tumor specific immunity 279 induced by peptides are unable to recognize the antigen processed and presented naturally [13], the vaccination strategy reported here was highly effective in generating both HER-2/neu peptide- and protein-specific T cell immunity. While generation of peptide-specific immunity after peptide immunization is a reflection of a patient s immune competence and their ability to be immunized, detection of protein-specific response implies that the native protein was taken up by APCs and processed so that the natural peptide epitope presented by the APC class II MHC molecules was in configuration similar to that of the immunizing peptides. Additionally, the peptide was presented in a concentration high enough to be recognized by immune T cells. Findings in our study demonstrate that the majority of subjects developed HER-2/neu peptide-specific T cell responses early or by the 3rd vaccine. Furthermore, those subjects who had a detectable peptide-specific immune response by the third vaccine had very robust peptidespecific Tcell responses which were significantly greater than those in subjects who demonstrated a detectable peptidespecific immune response later in the vaccine regimen (after the 3rd vaccine). One may speculate that these early and more robust peptide-specific immune responses were in part due to effective boosting of preexistent immunity by peptide vaccination resulting in rapid activation and expansion of T cells which had been spontaneously primed by endogenous tumor antigen. When is the optimal time to measure the development of an immune response in relation to vaccination? Most studies assess immunity during active immunization. Data presented here would suggest that monitoring response at least a year after vaccinations have ended would allow the identification of all the subjects capable of responding. Unlike the early HER-2/neu peptide-specific responses, the observed HER-2/ neu protein-specific T cell responses in this study primarily occurred after completion of 6 vaccines with thirteen per cent of these responses occurring up to 12 months after starting vaccination. This may be due to the more involved and complex process that measurable protein-specific T-cell responses imply. Indeed, the generation of HER-2/neu protein-specific responses was associated with epitope or determinant spreading as previously reported [5] suggesting that the immune repertoire to HER-2/neu evolves during the course of vaccination. Thus, the detection of protein-specific Tcells after peptide immunization implies that the immunizing peptides represent natural epitopes of the HER-2/neu protein, and this naturally expressed HER-2/neu protein is being processed and presented in an augmented fashion. As such, it is likely that protein responses to vaccination may take longer to evolve and only by monitoring immunity long term will detection of such responses be possible. While the optimal time for measurement of the development of tumor-specific immunity has not been well defined, recent studies have indirectly provided information that would argue for extended immunologic monitoring. Indeed, recent findings by Monsurro and colleagues showed that protracted exposure to antigen stimulation broadened and intensified the extent of the immune response. However, the full potential of vaccinations may not be reached until late in the vaccine course or until a distant time point after completing vaccination [2]. These findings as well as data presented here would suggest that ongoing assessment of immune responses after active immunization has stopped is not only reasonable but necessary. Identification of factors associated with T cell responsiveness to vaccination remains critical. Data presented here suggest that those subjects with preexistent immunity potentially achieved the highest level of HER-2/neu-specific T cell immunity after vaccination. Indeed, Speiser and colleagues investigated whether pre-vaccine Tcell activation differed between patients who responded versus those that did not respond to a Melan-A peptide-based vaccine [14]. Peptide antigen-specific T cells were quantified and characterized ex vivo before and after vaccination and interestingly, several patients were found to have a relatively high percentage of pre-activated Melan-A-specific T cells prior to the initiation of vaccination. More importantly, positive immune responses to vaccination occurred primarily in those patients with preexistent immunity to Melan-A-specific T cells prior to vaccination. These findings suggested that vaccine responder patients might primarily be those with spontaneous (i.e. tumor-driven) immune pre activation and as such, significant activation and expansion of antigenspecific T cells may be the result of combined or sequential stimulation by antigen provided by both the tumor and the vaccine. Thus, developing vaccines that target antigens that already demonstrate an increased incidence of endogenous immune responses in cancer patients may improve the vaccine potency and potentially, therapeutic efficacy. In the clinical development paradigm for cancer vaccines, the ability of a cancer vaccine to elicit a specific measurable T cell response is increasingly being used to prioritize immunization strategies for therapeutic development. Data presented here would suggest that immunologic monitoring should be performed for a more extended period of time than just during active immunization. Assessing immune responses for up to a year after the end of active immunization would be a reasonable time frame to capture all patients capable of responding to a particular vaccine. In addition, data demonstrate that patients who have a preexistent antigenspecific immune response, detectable at the time of initiating vaccination, achieve higher levels of tumor-specific T cell immunity overall most likely due to the boosting of the memory response. Acknowledgments This work was supported for MLD by the Cancer Research Treatment Foundation and NIH grants R01CA7516 and K24CA85218 and for LGS by NIH grant K23CA100691. Patient care was performed in the University of Washington General Clinical Research Center, NIH grant M01-RR-00037. We would like to thank Sally Zebrick for assistance in manuscript preparation. References [1] A. Hoos, et al., A clinical development paradigm for cancer vaccines and related biologics, J. Immunother. 30 (1) (2007) 1 15. [2] V. Monsurro, et al., Kinetics of TCR use in response to repeated epitope-specific immunization, J. Immunol. 166 (9) (2001) 5817 5825.

280 L.G. Salazar et al. [3] A. Perez-Diez, et al., Intensity of the vaccine-elicited immune response determines tumor clearance, J. Immunol. 168 (1) (2002) 338 347. [4] D.E. Speiser, et al., A novel approach to characterize clonality and differentiation of human melanoma-specific T cell responses: spontaneous priming and efficient boosting by vaccination, J. Immunol. 177 (2) (2006) 1338 1348. [5] M.L. Disis, et al., Generation of T-cell immunity to the HER-2/neu protein after active immunization with HER-2/ neu peptide-based vaccines, J. Clin. Oncol. 20 (11) (2002) 2624 2632. [6] M.L. Disis, M.A. Cheever, HER-2/neu oncogenic protein: issues in vaccine development, Crit. Rev. Immunol. 18 (1 2) (1998) 37 45. [7] M.L. Disis, et al., Generation of immunity to the HER-2/neu oncogenic protein in patients with breast and ovarian cancer using a peptide-based vaccine, Clin. Cancer Res. 5 (6) (1999) 1289 1297. [8] J.C. Reece, H.M. Geysen, S.J. Rodda, Mapping the major human T helper epitopes of tetanus toxin. The emerging picture, J. Immunol. 151 (11) (1993) 6175 6184. [9] V. Goodell, C. dela Rosa, M. Slota, B. MacLeod, M.L. Disis, Sensitivity and specificity of tritiated thymidine incorporation and ELISPOTassays in identifying antigen specific Tcell immune responses, BMC Immunology 8 (21) (2007). [10] M.L. Disis, et al., Pre-existent immunity to the HER-2/neu oncogenic protein in patients with HER-2/neu overexpressing breast and ovarian cancer, Breast Cancer Res. Treat. 62 (3) (2000) 245 252. [11] T.A. Gooley, et al., Estimation of failure probabilities in the presence of competing risks: new representations of old estimators, Stat. Med. 18 (6) (1999) 695 706. [12] G. Curigliano, et al., Breast cancer vaccines: a clinical reality or fairy tale? Ann. Oncol. 17 (5) (2006) 750 762. [13] T.Z. Zaks, S.A. Rosenberg, Immunization with a peptide epitope (p369 377) from HER-2/neu leads to peptide-specific cytotoxic T lymphocytes that fail to recognize HER-2/neu + tumors, Cancer Res. 58 (21) (1998) 4902 4908. [14] D.E. Speiser, et al., Disease-driven T cell activation predicts immune responses to vaccination against melanoma, Cancer Immun. 3 (2003) 12.