How NGS is Making Personalized Cancer Vaccines a Reality

How NGS is Making Personalized Cancer Vaccines a Reality

Next-generation sequencing has revolutionized the detection of tumor neoantigens

Miriam Bergeret, MSc

According to The American Cancer Society, approximately 1.8 million Americans will be diagnosed with cancer in 2020. Most patients will undergo surgery and chemotherapy, but they may soon be able to receive a new kind of treatment in the form of a custom-made cancer vaccine thanks to rapid next-generation sequencing (NGS). 

Cancer vaccines, which generate a strong immune response against a patient’s cancer cells based on their unique molecular profile, are the latest advancement in the field of immunotherapy. Initial clinical trials have shown promising results.1-5 

“It really is an evolutionary war between the tumor and the immune system… our goal is to get the T cells to win every time.”

Although most patients have an immune response to their cancer, cancerous cells often have ways to evade immune activity. Existing immunotherapies aim to reactivate that immune response, but they only work for about 30 percent of patients.6,7 Thus, researchers have been looking for other ways to reactivate immune cells and boost the antitumor response—and using a targeted vaccine is one way to do that. 

“What we want to do is give a person the ideal medicine for the disease, so it does the most good and the least damage,” said Dr. Nir Hacohen, the director of the Massachusetts General Hospital’s Center for Cancer Immunology, while discussing his work with cancer vaccines in 2017 during an Advances at Mass General Cancer Center podcast.8 “It really is an evolutionary war between the tumor and the immune system… our goal is to get the T cells to win every time.”

Cancer vaccines before NGS

How Cancer Vaccines Work

While traditional vaccines are used to prevent infections, cancer vaccines are used to treat disease that is already in the body. To accomplish this, the vaccine trains a patient’s T cells to find and kill cancer cells based on the presence of abnormal proteins, called neoantigens, on their cell surface. Neoantigens arise from the unique DNA mutations found in cancer cells, which means neoantigens are tumor-specific and do not occur anywhere else in a patient’s body. This makes neoantigens the ideal target for personalized cancer therapy as they only generate an immune response against the cancer cells and there is no risk of harming normal tissues.

The idea of treating cancer using vaccines has been around for decades with the first vaccines developed from patient tumor cells in the 1980s.9 During the 1990s, researchers identified the first tumor-associated antigens, such as melanoma-associated antigen 1.10 However, the first tumor-associated antigen vaccine was not approved by the FDA until 2010.11 Unfortunately, tumor-associated antigen vaccines have not produced the results clinicians were hoping for, in part because these antigens are also expressed by healthy tissues and the immune system normally does not mount an immune response against cells it recognizes as self. Thus, the key to creating a robust antitumor response is identifying the differences that make each patient’s tumor unique, which was not feasible before the advent of NGS technology as the sequencing process was costly and cumbersome. 

NGS has fundamentally changed our understanding of cancer

Sequencing of thousands of human cancer samples through projects such as The Cancer Genome Atlas ( and the International Cancer Genome Consortium ( has improved scientists’ understanding of how cancers develop. Researchers have discovered great genetic differences not only between patients but also between individual cancer cells from the same tumor, leading to genetically diverse subpopulations with different susceptibilities to treatments.12-15 These genetic differences emphasize why the one-sizefits-all approach of tumor-associated antigen vaccines does not work well for cancer therapy and why we need personalized cancer treatments. 

In addition to the genetic differences, large-scale sequencing of cancer samples has revealed that tumor cells that express more neoantigens are associated with a stronger immune response and better patient survival.16-18 Thus, researchers believe that harnessing the immune system’s natural ability to find and kill cancer cells is the best way to treat cancer in the long term.7,9 

Creating personalized neoantigen vaccines

Free eBook: Genetic Testing for Cancer  

“What’s been super exciting over the last decade with the availability of next-generation sequencing is that suddenly we have a way to find tumor neoantigens,” said Dr. Catherine Wu, an oncologist and professor of medicine at Harvard Medical School, during a live talk for the Dana-Farber Cancer Institute this past April.But the process is still far from simple. 

To create a personalized neoantigen vaccine, researchers first compare a patient’s tumor cells to their normal cells using whole-exome sequencing to identify DNA mutations unique to the tumor cells.9 Neoantigens are made from the mutated proteins that arise from these DNA mutations, but not all DNA mutations are translated into mutated proteins, so researchers use RNAsequencing to determine which of the DNA mutations are actually expressed in a patient’s tumor cells, further narrowing the list of potential neoantigen candidates.9

Adding a further layer of complexity, only some of the mutated proteins can bind to and be processed by HLA class I molecules into neoantigens. Therefore, researchers use a neural network-based algorithm to predict which of the mutated proteins are most likely to undergo this transformation and be presented on the surface of tumor cells as neoantigens.16,19,20 These neoantigens are the most likely to be detected by T cells to produce a strong tumor-specific immune response. The number of neoantigens included in a vaccine varies by patient, but, so far, up to 20 different neoantigens have been included in a single personalized vaccine.1

Neoantigen vaccines stimulate tumor-specific T cells

Source: Hu, Zhuting, et al. “Towards personalized, tumourspecific, therapeutic vaccines for cancer.” Nature Reviews Immunology 18.3 (2018): 168–182

Clinical trials of personalized cancer vaccines in solid tumors have shown that neoantigen vaccines are safe and can generate tumor-specific T cells that only recognize the tumor and nothing else without serious side effects.1-5 For example, a recent vaccine trial for glioblastoma, a type of brain tumor that usually has very low levels of immune cells, found an increase in tumor-infiltrating T cells: “We have direct evidence that what we're stimulating peripherally on the limbs of a vaccine is generating T cells that have neoantigen specificity that are actually able to track inside of the tumor,” said Dr. Wu, who led the phase I/Ib glioblastoma vaccine trial.3 

But scientists have only just begun to investigate neoantigen vaccines. While therapeutic efforts directed at targeting neoantigens have shown promising results, immune resistance remains an issue. Ongoing clinical trials have combined personalized cancer vaccines with other immunotherapies to try to overcome immune resistance and increase the effectiveness of tumor-specific T cells generated from the vaccines. For example, Roche is currently leading an open-label phase 1a/1b trial combining a personalized neoantigen vaccine with anti-PD-L1 therapy to treat melanoma and non-small cell lung cancer, among other types of cancer ( identifier: NCT03289962). 

Developing better prediction algorithms Since the first clinical trials, researchers have improved the neoantigen prediction algorithms using mass spectrometry data of HLA-bound neoantigens from different patient-derived tumor cell lines.21 In particular, the new high-throughput tool HLAthena ( can better predict HLA-binding preferences for a wider variety of patients. While research has mostly focused on HLA class I molecules to generate T cells that kill tumor cells, they have also begun working on tools to identify neoantigens capable of binding to HLA class II molecules,7 which generate memory T cells for long-term protection against tumor cells. 

New call-to-action

At present, not all neoantigens in a vaccine produce a T cell response. This may be because, in addition to binding to HLA class I molecules on the surface on tumor cells, neoantigens must bind to T cell receptors to produce an immune response, and whether candidate neoantigens can bind to T cell receptors is not yet part of the selection process.21 Therefore, there are still many opportunities to improve neoantigen selection in order to produce the best antitumor response. 

The availability of NGS has brought personalized cancer vaccines to the forefront of cancer research by allowing clinicians to identify tumor-specific neoantigens on a patient-by-patient basis, with promising results from early clinical trials. As researchers work to improve high-throughput technologies and prediction algorithms, it is clear that personalized cancer vaccines will continue to make an impact in the field of immunotherapy. 


1. Ott, Patrick A., et al. “An immunogenic personal neoantigen vaccine for patients with melanoma.” Nature 547 (2017): 217–221. 

2. Sonntag, Katja, et al. “Immune monitoring and TCR sequencing of CD4 T cells in a long term responsive patient with metastasized pancreatic ductal carcinoma treated with individualized, neoepitope-derived multipeptide vaccines: a case report.” Journal of Translational Medicine 16 (2018): 23. 

3. Keskin, Derin B., et al. “Neoantigen vaccine generates intratumoral T cell responses in phase Ib glioblastoma trial.” Nature 565 (2019): 234–239. 

4. Hilf, Norbert, et al. “Actively personalized vaccination trial for newly diagnosed glioblastoma.” Nature 565 (2019): 240–245. 

5. Johanns, Tanner M., et al. “Detection of neoantigen-specific T cells following a personalized vaccine in a patient with glioblastoma.” Oncoimmunology 8.4 (2019): e1561106. 

6. Wei, Spencer C., et al. “Fundamental mechanisms of immune checkpoint blockade therapy.” Cancer Discovery 8.9 (2018): 1069–1086. 

7. Wu, Catherine J. “Addressing Cancer Clonal Heterogeneity: Developing Personalized Cancer Vaccines.” DF/HCC Connect:Science, 2 April 2020,

8. “The Future of Immunology in Cancer Care with Nir Hacohen, PhD.” Antitumor Immunity and Its Evasion by Tumors by Advances at Mass General Cancer Center, 1 July 2017, 

9. Hu, Zhuting, et al. “Towards personalized, tumour-specific, therapeutic vaccines for cancer.” Nature Reviews Immunology 18.3 (2018): 168–182. 

10. van der Bruggen, P., et al. “A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma.” Science 13.5038 (1991): 1643–1647. 

11. Kantoff, Philip W., et al. “Sipuleucel-T immunotherapy for castration-resistant prostate cancer.” The New England Journal of Medicine 363.5 (2010): 411–422. 

12. Campbell, Peter J., et al. “Subclonal phylogenetic structures in cancer revealed by ultra-deep sequencing.” PNAS 105.35 (2008): 13081–13086. 

13. Alexandrov, Ludmil B., et al. “Signatures of mutational processes in human cancers.” Nature 500 (2013): 415–421. 

14. Vogelstein, Bert, et al. “Cancer genome landscapes.” Science 339.6127 (2013): 1546–1558. 

15. Lawrence, Michael S., et al. “Mutational heterogeneity in cancer and the search for new cancer-associated genes.” Nature 499 (2013): 214–218. 

16. Rooney, Michael S., et al. “Molecular and genetic properties of tumors associated with local immune cytolytic activity.” Cell 160.1-2 (2015): 48–61. 

17. Brown, Scott D., et al. “Neo-antigens predicted by tumor genome meta-analysis correlate with increased patient survival.” Genome Research 24.5 (2014): 743–750. 

18. Giannakis, Marios, et al. “Genomic correlates of immune-cell infiltrates in colorectal carcinoma.” Cell Reports 15.4 (2016): 857–865. 

19. Fritsch, Edward F., et al. “HLA-binding properties of tumor neoepitopes in humans.” Cancer Immunology Research 2.6 (2014): 522–529. 

20. Rajasagi, Mohini, et al. “Systematic identification of personal tumor-specific neoantigens in chronic lymphocytic leukemia.” Blood 124.3 (2014): 453–462. 

21. Sarkizova, Siranush, et al. “A large peptidome dataset improves HLA class I epitope prediction across most of the human population.” Nature Biotechnology 38 (2020): 199–209.