The first time someone pitched Genentech’s senior leadership on a personalized cancer vaccine, it did not go well. “I thought there was going to be a riot,” Ira Mellman, then Genentech’s head of research oncology, recalls.
From across the table, he watched the scientific review committee grimly shaking their heads as his team member and longtime collaborator Lélia Delamarre made her case. Then he overheard the head of clinical development turn to the person sitting next to him and mutter, “Over my dead body. A vaccine will never work.”
That was in 2012. Cancer immunotherapy, which uses a person’s own immune system to attack tumors, is now one of medicine’s most promising fields, and one of the greatest breakthroughs in oncology in decades. But it took a long time to get there. Until the recent advent of a new class of blockbuster immunology drugs, the field was notorious for questionable science, hype, and spectacular disappointments.
And what Mellman and his team were proposing that day went further than turbocharging immune cells to make them better able to attack cancers. They were talking about a vaccine precisely tailored to stimulate the immune system to react to specific tumors. If it worked, the approach could, in some cases, be even more potent than other types of immunotherapy. But it faced a series of daunting hurdles. If Genentech, a San Francisco–based biotech company owned by the Swiss pharma giant Roche, were to attempt to develop a vaccine that could attack individual tumors, it wouldn’t just have to accept new scientific advances; it would also have to embrace an entirely new and untested business model. That’s because the vaccine Mellman and Delamarre envisioned could not be manufactured the traditional way, in large batches that could be packaged in bulk, warehoused, and dispensed off the shelf at your local pharmacy.
When Mellman and Delamarre said “personalized,” they really meant it. The composition of each vaccine would be based on the characteristics of each patient’s tumor DNA. The company would have to, in essence, make a separate treatment for every single patient.
Nor would this be the kind of drug you could order up with a prescription in hand and get in a few days, like Genentech’s highly successful cancer drugs Herceptin and Avastin. To create this drug, the company would have to orchestrate a multi-step process for each patient, performed at multiple sites. Each patient would need a biopsy, the tumor tissue would have to undergo full genome sequencing, the results would require complex computational analysis, and the individual vaccines would then need to be designed and queued up for manufacture. Theoretically, if the vaccines were to be produced on a large scale, this would have to happen hundreds of times a week. And it would have to happen fast.
If any single step in the process went awry, if a shipping mistake occurred or a batch was contaminated, it could prove deadly—because cancer doesn’t wait.
No wonder the Genentech leadership was so skeptical.
After that calamitous first pitch meeting, Mellman and Delamarre retreated to their laboratories. They returned a few months later with more exciting data: they had identified specific targets on cancer cells, targets that would readily be attacked by immune cells. They also had fresh, convincing research from a growing number of other academic groups on the feasibility of their approach. And, critically, they had a preliminary plan for how Genentech itself might take the first tentative steps toward making tailor-made treatments an economically viable product.
This time the reception was different. The committee signed off on an exploration that would culminate in 2016 with a $310 million deal with BioNTech, a German company that has a technique for producing personalized vaccines to target tumors. Last December, the partners launched a massive round of human testing, targeting at least 10 cancers and enrolling upwards of 560 patients at sites around the globe.
At Genentech headquarters, Mellman and Delamarre’s small team has grown by now into an army of hundreds, consisting not just of lonely lab workers but supply-chain specialists, regulatory experts, diagnosticians, and a whole host of consultants, all focused on the laborious task of figuring out how the production of their promising new product—should it continue to demonstrate the powerful effects seen so far—might be scaled up in a way that won’t bankrupt the company.
“It’s never been done, so we are learning as we go,” says Sean Kelley, the project team leader overseeing the effort.
Nor are Genentech and BioNTech the only companies now pushing into this new territory. In late 2017, Moderna, a biotech based in Cambridge, Massachusetts, announced that, in partnership with pharmaceutical giant Merck, it intended to start human trials with a vaccine targeting solid tumors. Another company, Neon Therapeutics, founded by researchers at Dana Farber Cancer Institute and Washington University, treated its first patient in phase 1 trials in May with a similar vaccine derived using a different method. It raised $100 million in an IPO this summer, driven largely by optimism over its approach.
The technology for the first truly personalized cancer vaccine is not yet proven. And these therapies are all likely to be expensive, Mellman acknowledged recently, sitting in a spacious conference room outside his office at Genentech’s headquarters in South San Francisco. But he insists that if it’s all done right, the extra costs and thinner margins will be more than offset by the sheer number of people who would use the treatment.
“You can imagine a scenario where every single cancer patient would benefit from this vaccine,” he says. “That’s unheard of.”
Fighting against yourself
Scientists have been intrigued for decades by the possibility that cancer’s greatest strength—its ability to mutate and evolve—might also be one of its greatest vulnerabilities.
Mutations in cellular DNA are, after all, what cause cancer in the first place, by prompting the cells carrying them to grow and proliferate uncontrollably. As far back as the 1940s, some researchers were arguing that it might be possible to put the immune system’s cellular bloodhounds onto the scent of a specific tumor by somehow priming them with a vaccine that helped it recognize the tumor’s mutations. A number of researchers have experimented and continue to experiment with techniques that involve removing immune cells from the body, genetically engineering them, and then reinfusing them in the hopes of triggering a robust response. Other cancer immunologists have focused on developing drugs to turn off molecular switches on the immune system’s T cells that can interfere with their ability to attack.
But until recently, the scientific tools simply didn’t exist to take the sophisticated personalized approach Genentech is now pursuing—an approach that requires scientists to fully characterize an individual cancer tumor, identify the most attackable mutations, and then design a personalized vaccine that would provoke the immune system to target them.
The problem was identifying the right target molecules on the tumor cell, or—as researchers thought of them—the antigens that would catch the attention of the immune cells. “It was so much work to identify antigens in the past,” says Robert D. Schreiber, director of immunotherapy at Washington University. “You could do all this work, and then you end up with one antigen from one individual that is not necessarily ever seen again in any other individual.”
That all changed with the advent of cheap genetic sequencing. In 2008, five years after the Human Genome Project published the sequence of the first human genome, scientists published the first genome sequence of a cancerous cell. Soon after, scientists began to compare the DNA in tumor cells and healthy cells to characterize the myriad ways that they differed. These studies confirmed that all cancer cells contain hundreds—if not thousands—of mutations, most of which are unique to each tumor.
In 2012, a team of German researchers, led by scientists at BioNTech, sequenced a widely used mouse tumor cell line designed to mimic human melanoma cells. They identified 962 mutations and used RNA sequencing to identify 563 that were expressed in genes. The group then created vaccines made of protein fragments that contained 50 of the mutations and injected them into mice to see if this would prime the immune system to respond. About one third—16 of the mutations—were detected by the immune system, and five of those generated an immune response designed specifically to attack any cell found to harbor such mutations.
It was concrete evidence suggesting that genome sequencing could be used to design an effective cancer vaccine capable of putting the immune system on the trail of multiple mutations at the same time—and that such a vaccine might indeed provoke the immune system to attack a tumor. The race was on to answer the next logical questions: Why is it that the human immune system can be stimulated to attack some mutations and not others? And how can we figure out which mutations are most likely to be vulnerable?
At the urging of Mellman, Delamarre took Genentech’s own lab mice and sequenced their tumor cells, identifying 1,200 individual mutations not present in normal tissue. Then she measured how T cells naturally responded to them. Of those 1,200 mutations, she found, the mice’s immune system had begun to mount attacks against only two.
To answer why only those two mutations appeared to attract an immune response, Delamarre took a closer look at the interaction between the cancer DNA and a key component of the mouse immune system known as major histocompatibility complex, which in humans is called the human leukocyte antigen system (HLA). The HLA complex comprises 200 different proteins that protrude from cellular surfaces like microscopic thumbtacks on a poster board. When passing immune cells detect the presence of a protein fragment that doesn’t belong—a piece of an unwanted virus or bacterium, or a mutation—they sound the alarm and cause the body to attack it.
Delamarre had determined that roughly seven of the 1,200 tumor mutations she’d identified were displayed on the cellular surface by HLA. When she examined the structure of these seven protein fragments, something got her attention: in the two that the immune system had recognized, the mutations were prominent on the cellular surface, facing up toward passing immune cells. Those the immune system had ignored faced down and were hidden in grooves in the cellular surface or obscured on the edges of the HLA. The immune system attacked those two mutations because they were the easiest to detect. By injecting mice with a vaccine designed to target those two mutations, she could enhance their bodies’ ability to fight the tumors.
Together, these findings were what helped her and Mellman convince Genentech’s review committee that a cancer vaccine was worth pursuing.
Facing the music
Genentech’s headquarters, in an industrial park just off California’s Highway 101, is a sprawling campus of glass buildings, hulking warehouses, and grassy courtyards. On a sunny morning this past August, cheerful groups of men and women in shirtsleeves and T-shirts strolled casually through a courtyard outside the company cafeteria. A band was setting up, getting ready to regale the lunchtime crowd with some blues, while nearby some kitchen workers prepared outdoor grills to cook food for employees.
Much of this is paid for by cancer drugs. Genentech won approval for its first cancer treatments in 1997, and since then the company has fielded no fewer than 15 of them.
But a cancer vaccine is unknown territory. The initial human trials that Genentech and BioNTech launched last year are shaping up as a test not just of the vaccine’s efficacy but of the two partners’ ability to scale up the new technology. By design, the geographic scope and the number of conditions targeted in the trial are broad—so far Genentech and BioNTech have opened sites in the US, the UK, Belgium, Canada, and Germany, and they are likely to expand to other nations around the globe.
Producing the vaccines even for the small number of patients in early trials “was an extremely challenging process,” says BioNTech CEO Ugur Sahin, a veteran cancer researcher who cofounded the company in 2008. “Everything was driven by pipetting and by people on the bench producing the vaccines,” he says. “So we had a very small capacity.”
BioNTech has been able to automate some functions and reduce the time it takes to manufacture each vaccine from three months to about six weeks. It is shooting to get that down to four weeks by the end of the year.
The company can now produce hundreds of vaccines in a year—it aims to reach 1,500 over the next year. But if Genentech and BioNTech are ever to bring the product to market, they will need to be able to produce between 10,000 and 20,000 a year, Sahin says.
In San Francisco, teams from Genentech and BioNTech track progress in a designated space, consisting of a suite of rooms. On the walls, there are huge charts spelling out the patient status, the manufacturing and supply chain, the duration and schedule for each activity. “The key thing is that on paper it can look like a very coordinated process, but if any of those steps break down, then you can be in a situation where you have to start over,” Genentech’s Sean Kelley notes.
A number of unanticipated challenges have arisen. Early on, the team was surprised to discover that workers at BioNTech were contractually prohibited from working on weekends—so there was no one to receive patient tissue samples arriving then.
Gregg Fine, a senior medical director who is overseeing the trials, says he has been surprised by how variable the turnaround time has been at clinics and labs where patient biopsies themselves are collected and analyzed—a problem, since individual vaccines can’t be manufactured until the samples are received.
The issue, Fine believes, is that patients with metastatic cancer may have problems getting to the doctor in a timely manner because they are too sick. And many collection sites don’t yet have a procedure for flagging their samples as urgent, which means they can get lost in the stack with other biopsies.
Getting the vaccines back to the patients themselves has also proved problematic. At least one vaccine has been held up at customs in New York City.
For now, the problems are manageable and informative because the number of patients is relatively small. But all these problems will have to be solved if the vaccines are ever to go mainstream. “You’re not going to be able to wait six months for a vaccine if you have a patient with fast-progressing pancreatic cancer,” says Kelley.
Genentech officials declined to speculate about the eventual price of the vaccine, insisting it was too early to know. “It’s going to be more expensive,” says Kelley. “This will cost us much more to make per person.”
The cost of sequencing might come down, building out a manufacturing network would increase efficiencies, and new assays might be developed, or new technologies that allow the cheaper manufacture of the vaccines themselves. “We’ve done estimates, and we feel that right now it is viable, but we would like it to become, obviously, more and more viable,” he says.
For now, though, one of the most promising advances in cancer research remains an experimental treatment. It might be a medical breakthrough, but it is facing a familiar logistical challenge: how to get the product cheaply and quickly where it needs to go.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Everything you need to know about artificial wombs
Artificial wombs are nearing human trials. But the goal is to save the littlest preemies, not replace the uterus.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.