A Medical Revolution Too Late for the Man Who Started It

Illustration of Sam Gambhir with an image of cancerous cells superimposed over half his suit
Society of Nuclear Medicine and Molecular Imaging / Getty / Katie Martin / The Atlantic

When I finally met Sanjiv “Sam” Gambhir in person—last November, after months of planning—I knew he was dying.

Gambhir knew it, too. Seated in his small, bland office at the end of a warren of hospital hallways in Palo Alto, he was visibly depleted from the cocktail of treatments, some highly experimental, that were being deployed to save him from cancer. But Gambhir, a titan in the field of diagnostic medicine, wasn’t interested in dwelling on his own fate. Instead, he wanted to reimagine reality.

Well known among his colleagues for his “What if?” thought experiments, Gambhir took the occasion of my writing about the end of his life to reflect on the limits of writing itself. “I see writing as a primitive form of communication,” he declared in his commanding nasal voice. If Homo sapiens had evolved from the beginning with modern-day video and audio technologies, he surmised, there would have been little need to perfect written communication—and we’d be all the better for it today.

The moment was classic Gambhir. His riff was admittedly absurd, but it was meant as provocative speculation. He had made a career out of spinning outlandish visions into realities. Working at the nexus of medicine, physics, and engineering, he had spent nearly 40 years masterminding ways to see and study the infinitesimal biochemical and cellular events in living people—something that had never been done before.

Recommended Reading

The goal of his research: cancer detection. Over some 700 scientific papers and 40 patents, Gambhir, a physician-scientist at the Stanford University School of Medicine and a principal investigator in the massive health-monitoring research endeavor Project Baseline, had pioneered steps toward finding the earliest, currently unfindable, cancer cells before they had a chance to spread.

Then, in the spring of 2019, his pursuit took on unanticipated urgency. His own bone marrow was seeded with a metastasis that had migrated from an unknown site somewhere in his body. Gambhir convened a scientific all-star team from around the globe, but even they couldn’t figure out the source. Medicine calls this rare malignancy “carcinoma of unknown primary”—a sad, darkly poetic, and, in Gambhir’s case, tragically apt name.

The median survival time after diagnosis is three to four months. Gambhir made it 16 months. On July 18, he died at the age of 57.

It was a staggering blow to the field. Gambhir’s “research and innovations have, with no uncertainty, founded modern medicine’s approach to early disease diagnostics,” said Lloyd Minor, the dean of the Stanford University School of Medicine, in a statement after Gambhir’s death. Last month, the Journal of Nuclear Medicine, a highly influential publication that has always featured on its covers spectral images of human scans or incomprehensible graphs from scientific papers, ran a tender, full-page photo of Gambhir, his dark eyes full of curiosity, his expression serene—the journal’s first cover with a flesh-and-blood human being.  

Over the last year of his life, I had four probing interviews with Gambhir about his work and his life; I also spoke with his colleagues and with his wife, Aruna Gambhir, a biotech executive. These were the last formal interviews that Gambhir granted. What he gave me was an intimate portrait of a vast mind coming to terms with the limits of his own place in his grand perspective. Gambhir’s discoveries opened radically new directions in cancer medicine. But, like so many visionaries, Gambhir carried the burden of knowing that he would not live to see his full aspirations become reality. Right up to the end, his cancer remained a medical mystery.


In cancer treatment, timing is everything. With many tumors, there is a lag of 20 years or more between the first cancer cell and the onset of end-stage disease. Most cancer therapies ultimately fail for a simple reason: They aren’t applied until after a cancer is discoverable by conventional diagnostics—which is usually too late.   

When cancer cells emerge, they are confined to a single small site and are homogeneous, making them good targets for drug treatments. But as the cells proliferate and spread, they mutate. Standard chemotherapy kills the most vulnerable cells, leaving those that have evolved genetic defenses to survive and multiply. Cutting-edge treatments such as immunotherapies, which have been celebrated in mainstream-news reports for miraculously melting away advanced cancers revealed by conventional diagnostics, are undeniably valuable when they succeed, but they vary widely in their outcomes and can cause life-threatening side effects.      

Gambhir’s revolutionary ambition was to erase the long, silent, lethal interval between the first deviant cell and a grim lab report. Rather than curing cancer once it had already spread, he wanted to stop the first abnormal cells from ever gaining ground. To achieve this almost inconceivable goal, he proposed developing ways to monitor people while they are still healthy, in order to catch the earliest, subtle transition to disease. If surgeons could pinpoint a tiny mass of aberrant cells, smaller than the tip of a ballpoint pen, he explained, they could easily remove it before it had a chance to spread and cause irreversible harm.

In our first conversation, in June 2019, Gambhir likened cancer progression to a movie that we glimpse in only a few scattered frames—or “snapshots,” as he called them—near the end. “We need the whole cancer movie,” he said, including footage of the very first rogue cells. In his lab, he searched for telltale biomarkers from cells that have started to careen out of control—faint biological signals deep in the body such as proteins, DNA or RNA, circulating tumor cells, or immune cells. He called this signal capturing “remote sensing,” invoking astrophysicists’ use of the term when trying to understand what’s going on in a star millions of light years away.

    As recently as the 1990s, when less was known about cancer biology and when radiological techniques were less sensitive, the notion of intercepting nascent tumors would have struck most scientists as outrageous. Gambhir was repeatedly rejected for research grants early in his career. But at UCLA, where he started out, and then more extensively at Stanford, Gambhir and his team built instruments to record an individual’s cancer movie. Among his lab’s fantastical creations: a slender magnetic wire that can sample the body’s entire blood supply and snag stray cells sloughed off by otherwise elusive tumors, and a “smart toilet” that analyzes urine and stool to detect biomarkers of colorectal cancer, a potential alternative to colonoscopies. In the works: a photo-acoustic smart bra that enables doctors to “hear light” reflected by newly forming blood vessels around a microscopic breast tumor.

These devices were just the beginning. Gambhir foresaw a time when a patient would swallow a pill, the contents of which would enter all of the body’s cells and, should they encounter a cancer cell, send up the equivalent of a biological flare. He predicted that within 40 or 50 years, the annual physical will be a quaint relic, because we will instead be surrounded by passive health trackers in our homes and cars and wearable devices, including bedsheets that gauge heart and lung function, mirrors that measure vital signs using radar, toothbrushes that biochemically probe saliva, and even sensors that can register “smellprints” of lung and ovarian cancer.   

Gambhir’s ultimate vision was a bioinformatician’s dream—though perhaps a dystopia to anyone troubled by the prospect of being surveilled by their bedsheets. He conjured up always-on technology that monitors the entire body, just as jet engines’ hundreds of sensors continually generate real-time data during flight. “Imagine a cockpit where … you can see the status of each and every cell of our trillions of cells,” he told me. “To me, that’s not impossible.”


Prevention can be a hard sell. Unlike cures, which are dramatic and definable—and, therefore, attractive investments—staving off disease in the first place is uneventful and vague, because nothing happens. Even for much of the scientific community, preempting diseases like cancer through continuous passive surveillance is a risky bet. How can the results—essentially, nonevents—be measured? How much evidence would it take to prove the concept? As scientists scramble to compete for federal grants doled out at ever-shrinking pay lines, who would bet their careers on such an untested path? In fiscal year 2018, the most recent period for which information is available, only 5.7 percent of the National Cancer Institute budget was allotted to prevention and control, even though cancer is poised to be the leading cause of death in the United States (surpassing heart disease) and, later in this century, around the world.

That dilemma has led many scientists to seek deep-pocketed private investment, and Gambhir was no exception. As a leader of Project Baseline, he helmed what is likely to be the most expensive collaborative study (estimated to be upward of $100 million) ever undertaken by the academy and industry, with participating institutions that include Stanford University and Duke University, as well as Verily, the life-science division of Google’s parent company, Alphabet. Launched in 2017, Project Baseline will eventually follow 10,000 volunteers across the U.S. for four years, monitoring scores of biological functions through an extensive battery of tests, home sensors, and wearables. The study’s aim: to describe in minute detail each individual’s healthy “baseline” or threshold, and how it imperceptibly shifts to disease.

Privacy experts have raised issues around the use of medical data for commercial purposes. Earlier this year, for instance, Verily sparked concerns about data collection when it used the Project Baseline platform to roll out a much-discussed coronavirus-screening website. Gambhir was acutely aware of the danger. According to his colleagues, he stipulated that Project Baseline data be stored under the strongest possible security. He also recognized that standards and regulations for data sharing, confidentiality, and privacy would take years to come to fruition.

This was a common tension throughout Gambhir’s career: a crystal-clear sense of where he was going, with an agonizingly complete understanding of the obstacles that would need to be overcome to get there. Cyriac Roeding, the CEO of the cancer-detection biotech company Earli Inc., told the hundreds of mourners at Gambhir’s memorial service that Gambhir once warned him, “The world of biology will always find a way to screw you over. It will be harder than you want, and it will take longer than you want.” Yet Gambhir always seemed content to wait. Naive or not, he had faith that society would catch up with technology. One of his mantras was “We’re in the field of research, not search. The prefix ‘re-’ is extremely important.” As he told me, “The secret is not to give up, to take those decades and not look for short-term gratification, because most of the time, you won’t get lucky.”     


In the last months of his life, Gambhir acknowledged that his own luck had run out. “I tell my wife, ‘If I just had more time, I know I could solve this,’” he told me last fall. I asked him how much time he would need to solve his own cancer case. His answer, fittingly, was “decades.”

This was not the first time for Gambhir that the gap between his futuristic vision and the present came with tragic personal consequences. When he and I met in November, he wore a thin, red, braided cloth bracelet: the kalava, a symbol of protection in Hinduism. “I’m not a religious person at all,” he said, gazing down at his wrist. “It’s a connection to my son.” In 2015, Sam and Aruna Gambhir lost their only child, 16-year-old Milan, to an aggressive and highly lethal brain tumor called glioblastoma multiforme, one of the many cancers that Gambhir had been studying at the time in his lab.

As the Gambhir family learned afterward, Milan’s cancer was likely tied to Li-Fraumeni syndrome, a rare, inherited condition caused by a genetic mutation that dramatically raises the risk of many types of cancer. The mutation was passed down from Aruna, who has twice battled breast cancer as a result of the DNA defect. “If [Milan] had been born 100 years from now,” Gambhir said in a 2018 talk, “the tools of precision health would possibly have allowed him to live much, much longer.”

In the final months of Milan’s life, Gambhir confided to family members and close friends that he felt as if his mind had “opened.” His scientific imagination, already expansive, seemed to stretch even further, with an invigorated sense of purpose. Unlike Milan, Gambhir carried no family history of cancer, so he had no reason to expect that the many productive years he had left in his career would be cut short. “I thought we had had enough random bullets with our son,” he told me. Still, he said, it had been easier to come to terms with his own untimely diagnosis than with his child’s. “When your son’s life is at stake, it’s even more than your own life.”

Just before Thanksgiving last year, Gambhir learned that the cancer that had been confined to his bone marrow had spread. He was in severe pain. “I’m back to square one,” he told me by phone in December. “Just have to bear through it.” By January, it was obvious that his cancer had metastasized.

As the coronavirus pandemic seeded itself around the world, Gambhir entered his own medical lockdown, sequestered at home as he cycled over the subsequent months through different chemo regimens, suffering toxic reactions to almost all of them. “With Sanjiv, the problem was that he knew too much … But he still had hope,” Aruna told me in August. “He was fighting until the very end.”

Early in our conversations, Gambhir admitted that living in the present moment—that earnest cultural cliché—was “a foreign concept” to him. Later in our talks, I mentioned that the idea of time, in all of its dimensions, seemed to course through his work. He agreed. “The part that humans can’t comprehend—and I myself have a difficult time comprehending it—is time. We are a tiny speck in this massive universal clock.” That enormity, he added, worked against him. “It’s frustrating, because from the pure scientific-discovery part, you would like to be able to … [go] back to that full movie.”

“Do you mean the cancer movie or another movie?” I asked.

“The cancer movie, the life movie, the universal-clock movie,” he said. He was skipping across temporal scales, as if to say that great science requires a kind of God’s-eye view of existence, intimate and infinite—though, empiricist that he was, he didn’t believe in God. “Boy, what I’d do to have a peek into what things’ll be like hundreds, if not thousands, of years out,” he said.

Gambhir called the medical revolution that he inspired “inevitable.” And though the revolution came too late to save his own life, it will, by the end of this century, save countless others from premature death—the time horizon on which he had always set his sights.