Utility of cell-based vaccines as cancer therapy: Systematic review and meta-analysis

ABSTRACT

Cell-based therapeutic cancer vaccines use autologous patient-derived tumor cells, allogeneic cancer cell lines or autologous antigen presenting cells to mimic the natural immune process and stimulate an adaptive immune response against tumor antigens. The primary objective of this study is to perform a systematic literature review with an embedded meta-analysis of all published Phase 2 and 3 clinical trials of cell-based cancer vaccines in human subjects. The secondary objective of this study is to review trials demonstrating biological activity of cell-based cancer vaccines that could uncover additional hypotheses, which could be used in the design of future studies. We performed the systematic review and meta-analysis according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The final review included 36 studies − 16 single-arm studies, and 20 controlled trials. Our systematic review of the existing literature revealed largely negative trials and our meta-analysis did not show evidence of clinical benefit from cell-based cancer-vaccines. However, as we looked beyond the stringent inclusion criteria of our systematic review, we identified significant examples of biological activity of cell-based cancer vaccines that are worth highlighting. In conclusion, the existing literature on cell-based cancer vaccines is highly variable in terms of cancer type, vaccine therapies and the clinical setting with no overall statistically significant clinical benefit, but there are individual successes that represent the promise of this approach. As cell-based vaccine technology continues to evolve, future studies can perhaps fulfill the potential that this exciting field of anti-cancer therapy holds.

Introduction

The anti-cancer clinical responses seen with immune checkpoint inhibitors (ICIs) and chimeric antigen receptor (CAR)-T cell therapy have firmly established the potential of the immune system to control, or even eradicate cancer. Their success, however, has been restricted to a subset of patients. ICIs rely largely on the presence of preexisting tumor infiltrating cytotoxic T lymphocytes to produce anti-tumor effect, and these are absent in a significant proportion of patients.^1,2 Therapeutic cancer vaccination is an active immunization strategy, which aims to stimulate an adaptive immune response against tumor antigens and generate tumor-specific functional immune effectors, including cytotoxic T lymphocytes.

Therapeutic cancer vaccines have been researched for decades, and different approaches have been used to develop vaccine constructs for anti-cancer therapy. Broadly, these can be classified as cell-based vaccines, viral vector vaccines and molecular vaccines consisting of either peptides, deoxyribonucleic acid (DNA) or ribonucleic acid (RNA).³ Cell-based cancer vaccines use either replication incompetent cancer cells (autologous patient-derived tumor cells or allogeneic cancer cell lines) or autologous antigen presenting cells (APCs) loaded with cancer antigens. Tumor cell-based vaccines have the advantage of involving a broad range of neoantigens, eliminating the need to prospectively identify target antigens, though effective antigen presentation can be a limiting factor.⁴ Dendritic cells (DCs) are professional APCs optimal for priming of T cell responses and DC-based therapeutic cancer vaccines have been extensively studied.⁵ Sipuleucel-T (Provenge ®), a vaccine consisting of ex vivo activated peripheral blood mononuclear cells, including DCs, became the first FDA approved therapeutic cancer vaccine after the phase 3 IMPACT trial showed improved overall survival in patients with metastatic castration-resistant prostate cancer.⁶

The increasing role of ICIs in multiple cancers has led to renewed interest in personalized cancer vaccines as a means to induce or augment the anti-tumor immune response.⁷ Importantly, since cancer vaccines and ICIs target different stages of the cancer immune cycle, vaccine-ICI combinations have the potential to synergistically overcome each individual therapy’s resistance mechanisms.⁸ The cell-based cancer vaccines are backed by a strong biological rationale since they mimic the natural immune process.⁹ They have had encouraging pre-clinical results, and trials of cell-based cancer vaccines have demonstrated their ability to mount tumor-specific T-cell responses, but are yet to realize their full potential in terms of durable clinical benefit.³ There is little consensus on their role, which cancer types are likely to benefit and which vaccine approach provides maximum efficacy. The current evidence is inconclusive due to limited number of studies, with relatively smaller sample sizes. This provides the rationale for undertaking this systematic review and meta-analysis, with an attempt to combine existing evidence and analyze the overall effect of cellular cancer vaccines. Hence, we sought to assess the clinical utility of cellular cancer vaccines, as measured by Response Evaluation Criteria in Solid Tumors (RECIST) response or survival benefit, by systematically reviewing and comprehensively summarizing available data on cell-based cancer vaccines from published clinical studies. The primary objective of this study is to perform a systematic literature review with an embedded meta-analysis of all published Phase 2 and 3 clinical trials of cell-based cancer vaccines in human subjects. The secondary objective of this study is to review trials demonstrating biological activity of cell-based cancer vaccines that could uncover additional hypotheses, which could be used in the design of future studies. Therefore, a more granular review of individual studies independent of the meta-analysis was done, focused on detecting biological activity as opposed to RECIST responses, as defined below.

Materials and methods

The systematic review and meta-analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.¹⁰

Types of studies

In the systematic review, we included all published Phase 2 and 3 clinical trials of cell-based cancer vaccines in human subjects. We considered cell-based cancer vaccines to include autologous tumor cell, allogeneic tumor cell and dendritic cell vaccines. We included both single arm and multiple arm/controlled studies. Single-arm studies were included if they reported objective responses using RECIST. Multiple arm studies were included if they reported hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) for the association between cancer vaccines and overall survival (OS). The studies were characterized based on the clinical setting being evaluated (adjuvant vs metastatic vs recurrent) and the controlled studies were also classified based on the duration of follow-up being reported (<5-year vs ≥5 year). We excluded studies in which patients were treated simultaneously with chemotherapy, immunotherapy, and/or radiation therapy along with a cell-based cancer vaccine.

Types of outcome measures

All included single-arm studies utilized the RECIST guidelines, and the meta-analyzed outcome measures reported were Complete Response (CR), Partial Response (PR), Stable Disease (SD), Progressive Disease (PD) and Disease Control Rate (DCR; [DCR=CR+PR+SD]). For multiple arm studies, the meta-analyzed outcome measure was Overall survival (OS), reported as pooled hazard ratios (HRs) with 95% confidence interval. This was reported separately for studies in the adjuvant setting, metastatic setting and with ≥5-year survival follow-up.

Search strategy

A database search was performed using the PubMed database and ClinicalTrials.gov to identify published results of all Phase 2 and 3 clinical trials of cell-based cancer vaccines from January 1, 1985, to December 31, 2021. The https://pubmed.ncbi.nlm.nih.gov/search terms were “cancer vaccine” (limited to clinical trials between 1985 and 2021). The ClinicalTrials.gov search terms were Condition or disease = “Cancer,” Search Term = “Vaccine,” Study Type = “Interventional Studies (Clinical Trials),” Study Results = “Studies with results,” Phase = “Phase 2 and Phase 3.” No limitations were placed on cancer type or disease stage. Only full-text original clinical research articles published in English were included. The search also excluded conference abstracts, editorials, commentaries, case reports, and study protocols because of incompleteness of data or small sample sizes.

Data collection process

Three independent reviewers (AT, EC, JK) screened the titles and abstracts that were identified from the database search for eligibility. The selected studies were then reviewed as full text by two reviewers independently (AT, KT). Any conflicts were resolved by discussion with a third reviewer (GC). All noted reviewers are included in this manuscript’s author list. Relevant data was then extracted from the studies selected after full-text review and conflict resolution. The data included study design, type of cell-based cancer vaccine, cancer type, clinical setting, sample size and the outcome measures reported.

Quality assessment

The single-arm studies were assessed for quality using the Methodological Index for Non-Randomized Studies (MINORS).¹¹ The quality of controlled trials was evaluated using the Cochrane Risk-of-Bias Tool.¹² Quality assessment was done by two independent reviewers (AT and KA), with conflict resolution done via discussion with a third reviewer (GC).

Statistical analysis

In this study, Stata Statistical Software version 17 was used to conduct all statistical analyses. Depending on the nature of the variable, each clinical outcome was analyzed separately using the meta-analysis of proportions (metaprop command) or the meta-analysis of effect sizes (metan command). The meta-analysis summary estimate of the proportions represents the estimated overall rate of the outcome, while the summary estimate of the effect sizes represents the estimated overall association between the vaccine and the outcome. The summary estimates were predicted and presented using random effects estimators. To stabilize the variances of certain proportions that were equal to (or close to) 0 or 1, the Freeman-Tukey (FT) double arcsine transformation was employed. The pooled estimate of the rate was then back transformed and presented with their Wald 95% CI estimated using the Score method. The meta-analysis of effect sizes employed the DerSimonian-Laird (DL) estimator for the variance estimation in the random-effects model. The I² statistic and its P-value were used to estimate heterogeneity between studies.

Results

Studies selected

A total of 2,941 studies were identified in the initial database search. After screening of the titles and abstracts, 101 publications were selected for full-text review. After reviewing the full-text articles, 49 studies were included, and no additional articles were identified via review of references. The final review after data extraction included 36 studies, 16 single-arm studies,^13–27 and 20 controlled trials.^6,28–46 Figure 1 shows the PRISMA flow diagram for steps in the study selection process.

Figure 1.

PRISMA flow diagram.

Quality assessment

Table 1 shows the quality assessment of the single-arm studies using MINORS. There was one study of good quality (score 15–16), 15 studies of moderate quality (score 8–15) and 0 studies of poor quality (score <8). Figure 2 shows the Risk of Bias assessment for the randomized controlled trials along with the sources of funding for the study, when reported. While most of the newer studies were well designed with low risk of bias, lack of blinding was a common drawback seen in the older studies leading to some concerns of bias.

Table 1.

Quality assessment of the single-arm studies.

Study	Item
	1	2	3	4	5	6	7	8	Total score
Lee et al.13	2	0	2	2	2	2	2	0	12
NemunaitisNemunaitis et al.14	2	0	2	2	2	2	2	0	12
Powell et al.15	2	0	2	2	2	2	2	0	12
Avigan et al.16	2	0	2	2	2	2	2	0	12
Fishman et al.17	0	2	2	2	2	2	2	0	12
Ribas et al.18	2	2	2	2	2	2	2	2	16
Oshita et al.19	0	0	2	2	2	2	2	0	10
Mackiewicz et al.20	2	0	2	2	2	2	2	2	14
Mackiewicz et al.20	2	0	2	2	2	2	2	2	14
Chia et al.21	2	2	2	2	2	2	2	0	14
Creelan et al.22	2	0	2	2	2	2	2	2	14
Florcken et al.23	2	0	2	2	2	2	2	0	12
Bapsy et al.24	2	0	2	2	2	2	2	0	12
Xi et al.25	0	2	2	2	2	2	2	0	12
Mackiewicz et al.26	2	0	2	2	2	2	2	2	14
Herbert et al.27	0	2	2	2	2	2	2	0	12
Percentage of total included studies that are adequate	75	31	100	100	100	100	100	31	6.25

Score: 0- not reported, 1- reported but inadequate, 2- reported and adequate.

Items: 1- A clearly stated aim, 2- Inclusion of consecutive patients, 3- Prospective collection of data, 4- Endpoints appropriate to the aim of the study, 5- Unbiased assessment of the study endpoint, 6- Follow-up period appropriate to the aim of the study, 7- Loss to follow up less than 5%, 8- Prospective calculation of the study size.

Figure 2.

Quality assessment of the controlled studies.

Study characteristics

Table 2 summarizes the major characteristics of all included studies. Among the 16 single-arm studies included, 12 (75%) were conducted in the advanced/metastatic setting, 3 (19%) looked at recurrent tumors and 1 study, Powell et al.¹⁵, looked at primary malignant pleural mesothelioma irrespective of stage, since curative surgical resection was not standard of care (SOC).¹⁵ Of the 20 controlled studies included, 16 (80%) were conducted in the adjuvant setting and 4 (20%) were conducted among patients with advanced/metastatic disease. Most of the studies (53%) looked at dendritic cell vaccines, while 8 (22%) and 9 (25%) of the studies tested autologous and allogeneic tumor cell vaccines, respectively. Melanoma was the most common cancer type, in 10 (28%) of the studies. For the controlled studies, nine studies (45%) reported OS data with length of follow-up less than 5 years and 11 studies (55%) reported OS for 5 years or longer.

Table 2.

Characteristics of all studies included in the systematic review.

Characteristic	Type of study		Total; N = 36
	Single arm; N = 16	Controlled; N = 20
Sample Size
Median	29.5	122
Range	8–99	15–689
Age
Range	27–85	19–91
Sex (% Female)
Range	13.6% − 62.3%	10.3% − 52.9%
Clinical setting
Adjuvant	0	16 (80%)	16 (44%)
Advanced/Metastatic	12 (75%)	4 (20%)	16 (44%)
Recurrent	3 (19%)	0	3 (8%)
NA	1 (6%)	0	1 (3%)
Type of cancer
Colon	0	3 (15%)	3 (8%)
Colorectal	0	2 (10%)	2 (6%)
Glioblastoma	0	3 (15%)	3 (8%)
Hepatocellular carcinoma	1 (6%)	2 (10%)	3 (8%)
Malignant glioma	0	1 (5%)	1 (3%)
Melanoma	5 (31%)	5 (25%)	10 (28%)
NSCLC	2 (13%)	1 (5%)	3 (8%)
Prostate Cancer	1 (6%)	2 (10%)	3 (8%)
Renal cell carcinoma	3 (19%)	1 (5%)	4 (11%)
Nasopharyngeal carcinoma	1 (6%)	0	1 (3%)
Malignant mesothelioma	1 (6%)	0	1 (3%)
Multiple	2 (13%)	0	2 (6%)
Type of cellular vaccine
Dendritic cell vaccine	9 (56%)	10 (50%)	19 (53%)
Autologous tumor cell vaccine	2 (13%)	6 (30%)	8 (22%)
Allogeneic tumor cell vaccine	5 (31%)	4 (20%)	9 (25%)
Publication Year
1989	0	1 (5%)	1 (3%)
1993	0	1 (5%)	1 (3%)
1996	0	1 (5%)	1 (3%)
1998	0	1 (5%)	1 (3%)
2000	0	1 (5%)	1 (3%)
2004	0	1 (5%)	1 (3%)
2005	1 (6%)	2 (5%)	3 (8%)
2006	2 (13%)	1 (5%)	3 (8%)
2007	1 (6%)	0	1 (3%)
2008	1 (6%)	0	1 (3%)
2009	0	1 (5%)	1 (3%)
2010	1 (6%)	1 (5%)	2 (6%)
2012	4 (25%)	2 (10%)	6 (17%)
2013	2 (13%)	0	2 (6%)
2014	1 (6%)	2 (10%)	3 (8%)
2015	2 (13%)	1 (5%)	3 (8%)
2016	0	1 (5%)	1 (3%)
2017	0	1 (5%)	1 (3%)
2018	1 (6%)	1 (5%)	2 (6%)
2021	0	1 (5%)	1 (3%)
Length of Follow-up (for Overall Survival)
Less than 5 years	-	9 (45%)	-
Five years or longer	-	11 (55%)	-

Outcomes

Response based on RECIST criteria from single-arm studies

Figure 3 depicts the findings of the meta-analysis evaluating objective response to cell-based cancer vaccines in single-arm trials, using the RECIST criteria. The overall disease control rate (DCR), calculated as the proportion of complete responses (CR), partial responses (PR) and stable disease (SD) was 0.44 (0.36–0.53) using the random effects model (Figure 3a). Looking at the breakdown of the different responses, the proportion of CR was 0.02 (0.00–0.06) (Figure 3b); PR proportion was 0.08 (0.06–0.11) (Figure 3c); and the proportion of SD was 0.31 (0.24–0.40) (Figure 3d). Correspondingly, the rate of progressive disease (PD) was 0.56 (0.47–0.64) (Figure 3e).

Figure 3.

Meta-analysis of objective response to cell-based cancer vaccine therapy in single-arm trials, using the response evaluation criteria in solid tumors (RECIST). Figure 3a depicts the findings of the meta-analysis evaluating the disease control rate (DCR) with cell-based cancer vaccines. Figures 3b–e depict the findings of the meta-analysis evaluating the rates of complete response (CR), partial response (PR), stable disease (SD) and progressive disease (PD), respectively.

Overall survival (OS) from controlled studies

Figure 4 depicts the findings of the meta-analysis evaluating the association between cell-based cancer vaccine therapy and clinical outcome (OS) in controlled trials. Overall, there was no statistically significant improvement in OS across different settings. In the adjuvant setting, a pooled HR of 0.89 (0.79–1.00) was seen (Figure 4a) while in the advanced/metastatic setting, a pooled HR of 1.08 (0.76–1.52) was seen (Figure 4b). Looking specifically at studies that reported survival data with ≥5-year follow-up, a pooled HR of 0.91 (0.80–1.03) with a random effects model was observed (Figure 4c).

Figure 4.

Meta-analysis of cell-based cancer vaccine therapy and overall survival (OS). Figure 4a depicts the findings of the meta-analysis evaluating the association between cell-based cancer vaccine therapy and clinical outcome (OS) in the adjuvant setting. Figure 4b depicts the findings of the meta-analysis evaluating the association between cell-based cancer vaccine therapy and clinical outcome (OS) in the advanced/metastatic disease setting. Figure 4c depicts the findings of the meta-analysis evaluating the association between cell-based cancer vaccine therapy and clinical outcome (OS) with ≥ 5-year survival follow-up.

Literature review to evaluate biologic activity

While our rigorous systematic review and meta-analysis revealed several negative studies and an overall lack of clinical effect, there are important positive studies that highlight the promise of cell-based cancer vaccines. As we looked beyond the stringent inclusion criteria of our systematic review, we identified significant examples of biological activity of cell-based cancer vaccines that are worth highlighting with the aim of providing a separate and useful summary of approaches involving cellular vaccines that have the potential to be of clinical benefit.

The studies involving Sipuleucel-T, which consists of autologous dendritic cells pulsed with prostatic acid phosphatase (PAP) linked to granulocyte macrophage-colony stimulating factor (GM-CSF), stand out as successful trials that have led to change in clinical practice. Four different trials looking at its effect in metastatic castrate-resistant prostate cancer (CRPC) have demonstrated biological activity that has translated to improved clinical outcomes and has led to Sipuleucel-T becoming FDA approved for metastatic CRPC.^6,45,47,48 Several other studies show evidence of biological activity, though they did not meet their primary endpoints. In 2013, Skachkova et al. conducted a study of DCs pulsed with mechanically activated microparticles of lyophilized autologous tumor cells in Stage IIB – IIIA Non-Small Cell Lung Cancer (NSCLC) following surgery.⁴⁹ They observed distant recurrences in 25% (15/60) of patients who received only surgery (disease-free period – 9.55 ± 1.37 months) versus only 10% (6/60) of patients treated with DC-immunotherapy (disease-free period – 24.92 ± 2.39 months).⁴⁹ Liau et al. studied autologous tumor lysate-pulsed dendritic cells in newly diagnosed glioblastoma after surgery. Though the cross-over nature of the study made interpretation of the result difficult, the median overall survival (mOS) of 23.1 months from surgery compares favorably with the mOS of 15–17 months from surgery typically achieved with standard of care (SOC).⁵⁰

Phase III studies of autologous cancer immunotherapies, all in the adjuvant setting, have also had positive data. This included irradiated autologous cancer cells with mixed with Bacillus Calmette-Guerin (BCG) in stage I and III colorectal cancer (CRC) after surgery, which failed to meet the primary endpoint, but did see a significant improvement in survival and disease-free survival (DFS) in eligible colon cancer patients who received autologous cell therapy.²⁹ In another study of autologous tumor cells mixed with BCG in stage II and stage III colon cancer patients, the intention-to-treat (ITT) recurrence free interval was better for active treatment compared with observation (OR 0.75, 95% CI, 0.56–1.00).⁵¹ In a second study of the same therapy a beneficial effect of autologous cell immunotherapy was statistically significant for all endpoints including recurrence-free interval (RFI), OS, and recurrence-free survival (RFS) in Stage II colon cancer patients, while there was no beneficial effect in stage III patients.³⁴ A study of autologous tumor cell-lysate in M0 renal cell carcinoma (RCC) post-radical nephrectomy initially showed no significant differences between immunotherapy versus no treatment, but a later analysis of the ITT population showed 5-year and 70-month hazard ratios (HRs) for tumor progression were 1.58 (95% CI 1.05–2.37) and 1.59 (95% CI 1.07–2.36), respectively, in favor of the immunotherapy group and the 5-year and 70-month PFS rates were 77.4% and 72.0%, respectively, in the immunotherapy group versus 67.8% and 59.3%, respectively, in the control group.^52,53

For allogeneic cell therapies in the therapeutic setting, a study of Stage III/IV NSCLC patients who did not progress after platinum-based chemotherapy compared treatment with four transforming growth factor (TGF)-β2-antisense gene-modified, irradiated, allogeneic NSCLC cell lines (Belagenpumatucel-L) with placebo.⁴⁰ Although there were statistically insignificant increases in OS and PFS (median OS 20.3 versus 17.8 months; and, PFS 4.3 months versus 4.0 for immunotherapy vs placebo, both favoring the immunotherapy group, but not statistically significant), in a pre-specified Cox regression analysis of 161 patients who received chemoradiation therapy prior to randomization, median survival was 28.4 months with immunotherapy versus 16.0 months with placebo; HR 0.61, p = .032. In the adjuvant setting, a study of allogeneic melanoma cell lysate plus the detoxified Freund’s adjuvant compared with observation in resected, clinically node negative melanoma.³⁸ While the overall endpoint of OS and DFS was not met (statistically insignificant increases in 5-year and 10-year survival), if the analysis was focused on patients who were HLA-A2+ and/or HLA-Cw3+, there were significant differences in both and OS at 5 years 10 years. Other studies of immunotherapy or a similar live vaccinia virus-augmented allogeneic polyvalent melanoma cell lysate did not show significant improvement in survival. But, it is noteworthy that these studies used cell lysates and not live cells, which may make a difference in mechanism of action and activity.^31,54,55

Discussion

Therapeutic cancer vaccination, particularly cell-based cancer vaccines, has been an attractive concept for a long time that is yet to fulfill its promise in the clinical setting. Despite extensive pre-clinical and clinical work, success has remained elusive with only Sipuleucel-T gaining FDA approval, while most attempts have resulted in negative clinical trials. The utility of meta-analysis lies in teasing out effect when individual studies are limited by smaller sample sizes. This provides an important rationale for undertaking this study, as we notice individual trials trending toward survival benefit but failing to reach statistical significance. Another reason to systematically review and meta-analyze the trials on cell-based cancer vaccines is to provide a snapshot and graphically represent the current literature on the topic. This study, to the best of our knowledge, is the first to systematically review the entire spectrum of cell-based cancer vaccines for all cancer types and analyze efficacy. We sought to evaluate measures of objective response from single-arm studies as well as OS outcomes from controlled studies.

Among the included single-arm studies, the highest DCR was observed in the trial by Xi et al., looking at a dendritic cell vaccine for prostate cancer.²⁵ They also had the highest partial response rate of all the included studies, confirming the utility of cellular vaccines in prostate cancer. The maximum proportion of complete responders were observed in the trials by Mackiewicz et al. using an allogeneic whole cell vaccine in melanoma, another cancer known to be responsive to immunotherapy.^20,26 On the other hand, high rates of progressive disease were seen in the studies by Chia et al. looking at advanced metastatic nasopharyngeal carcinoma, and Florcken et al. looking at metastatic renal cell cancer.^21,23 These are both, of course, aggressive diseases that are challenging to control.

Among the controlled studies that were included, the majority were negative trials and did not demonstrate survival benefit. One of the early studies by Hoover et al. in 1993 was a successful trial, showing survival benefit with an autologous tumor cell vaccine in colon cancer.²⁹ Though limited by sample size, this early trial highlights the promise of cellular cancer vaccines. Since that time, systemic chemotherapy and radiation therapy has improved significantly, possibly contributing to the lack of positive trials with cellular cancer vaccines. Except for the studies with Sipuleucel-T in prostate cancer that have been discussed earlier, the only trials demonstrating survival benefit were Yamanaka et al., Cho et al. and Yao et al.^35,36,42 All of these looked at dendritic cell vaccines in patients with glioblastoma, which has poor prognosis with conventional treatments.

Prior meta-analyses on this subject have mostly focused on a particular cancer type. Some of them primarily analyzed the safety of the vaccines while some of them looked at cell-based and peptide vaccines. Botrel et al. in 2012 pooled together the already strong evidence of Sipuleucel-T in metastatic CRPC to further confirm its survival benefit.⁵⁶ Wang et al. in 2015, and Zhou et al. in 2016 showed survival benefit with therapeutic vaccines in NSCLC though their meta-analyses included cellular and peptide vaccines.^57,58 Chen et al. in 2018 meta-analyzed studies on DC-based vaccines in HCC and showed enhanced anti-tumor immunity as well as improved survival.⁵⁹ More recently, Han et al. in 2023 looked at therapeutic vaccines for HCC, though they included peptide and cell-based vaccines, and focused on safety.⁶⁰ Lv et al. in 2020 looked at phase II RCTs of DC vaccines in GBM and observed improved survival, though they included studies that used other concurrent therapies in the vaccine arm comparing best supportive care as control.⁶¹ Shamshiripour et al. also meta-analyzed studies of DC-based therapy for gliomas, concluding it was safe with minimal grade III/IV adverse events.⁶² Dafni et al. in 2021 meta-analyzed cancer vaccines (peptide and cell-based vaccines) in breast and ovarian cancer, looking primarily at ORR and observed a modest response rate.⁶³ These studies have provided important evidence for potential benefits of therapeutic cancer vaccines in specific cancer types.

Our systematic review of the existing literature revealed largely negative trials and our meta-analysis did not show evidence of clinical benefit from cell-based cancer-vaccines. A major limitation of this analysis is the heterogeneity of the included studies. Since our focus was solely on cell-based cancer vaccines, irrespective of other factors, a diverse range of studies have been included in our analysis. We have attempted to minimize the impact of heterogeneity by using a random-effects model, as well as classifying the studies based on clinical setting and length of follow-up, yet significant variability still exists with respect to cancer type and vaccine platform used, as reflected in the I² statistics. While this limits the interpretation of the results, this was expected by virtue of the design of our study and is reflective of the variability of existing literature on the topic of therapeutic cancer vaccination. Broadly speaking, the review suggests a low level of activity for this therapeutic modality as it has been implemented in prior trials.

However, as evidenced by the success of Sipuleucel-T among men with metastatic castration-resistant prostate cancer, there have been successful studies of cellular cancer vaccines. In order to evaluate the potential differentiating features of the successful study, a more granular approach was employed. This revealed that several phase III studies had convincing evidence of biological activity, although some failed to meet their primary endpoints. Interestingly, and contrary to the convential wisdom, phase III studies in the therapeutic setting were as successful or more successful than those in the adjuvant or neoadjuvant settings. This is counter-intuitive as an earlier stage patient might be expected to have a more robust immune system and thus be able to mount a more potent immune response to the vaccination. This may suggest that even in late-stage cancer patients, the immune system may still be robust enough to develop significant anti-cancer activity.

As mentioned earlier, the combination of ICI and therapeutic cancer vaccines has a strong biological rationale to produce synergistic anti-tumor effect. In our systematic review and meta-analysis, we excluded any published trials that were evaluating these combinations. This was done to try and gauge the effect of vaccination alone, without the confounding effect of concurrent treatments. Already there exists early evidence of potential efficacy of the ICI-vaccine combination approach, and we recognize this as a limitation of our study.^64–66

Despite these limitations, our study provides an important and timely overview of what has been observed thus far in the field of therapeutic cell-based cancer vaccines and should help serve as a starting place for future trials evaluating the biological activity and clinical impact of cell-based cancer vaccines. Our analysis has identified individual successes that underscore the promise of this therapeutic cell-based cancer vaccine approach. The summary of trials included in this study, along with the pooled meta-analysis, aims to provide an initial step for future, more targeted research.

Conclusions

The existing literature on cell-based cancer vaccines is highly variable in terms of cancer type, vaccine therapies and the clinical setting. Overall, with our applied systematic review and meta-analysis, we found no statistically significant benefit with cell-based cancer vaccines in the adjuvant setting or the metastatic setting. However, individual success, especially with dendritic cell-based vaccines and in the advanced cancer setting, have been well documented. As our understanding of cancer immunology keeps improving and cell-based vaccine technology continues to evolve, future studies can perhaps fulfill the potential that this exciting field of anti-cancer therapy holds.

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

AT, KA, AS, WW, and GC prepared and drafted the manuscript. GC, AN, WW, TV, ML, GDP, and GP planned the concept and scope of the review. AT, KA, EC, KT, and JK collected the data. AN, SVD, TS, and FV contributed data or analysis tools. KA and AT performed the analysis. AT, ML, GDP, WW, AS, GP, and GC all substantially edited the review. All authors contributed to the article and approved the submitted version.