Guam’s “Skeleton Key” Enigma

A Tiny South Pacific Island’s Rare Neurodegenerative Disease Sent Researchers on a Decades-long Hunt for Answers that Could Unlock Alzheimer’s and More.  

By Steve Graff

UMATAC, GUAM—On a hillside along a steel blue ocean cove here at the southern tip of the South Pacific island sits a cemetery. It’s small but hard to miss from the road that winds around the water’s edge and through this sleepy, centuries-old village. The 50 or so white tombstones—many adorned with colored Mary and Jesus statues and tall crosses, all facing the water—shine bright on the half-manicured, lush green grounds. Palm trees and flowered shrubs sway in the wind while sounds of the sea fill the air. Fake flowers and wreaths lean up against some of the chipped graves, deteriorated by harsh rains and harsher sun.

It's a serene place to mourn locals who have passed, but also a solemn reminder of the town’s plight. Many buried at Umatac Cemetery in the decades following World War II died of a debilitating, rare neurological disease that has yet to be fully explained, and that has since nearly vanished from existence as mysteriously as it appeared. Sharing many of the characteristics of Parkinson’s and Alzheimer’s disease, and amyotrophic lateral sclerosis (ALS), lytico-bodig is a complex medical mystery that has captivated but befuddled scientists for more than half a century.

Though folklore suggests the disease has been around much longer—two centuries even— the first official reports come from 1904. International attention came to lytico-bodig after the United States took back Guam from the Japanese more than 40 years later, when an American neurologist stationed here picked up on the disease’s emergence among the Chamorro people who are native to Guam. By the mid-1950s, it was a full-blown epidemic on an island of 60,000, afflicting people at a rate 100 times greater than similar neurodegenerative diseases did in other parts of the world. And Umatac bore the brunt of it. 

The search to understand lytico-bodig since then has been a continual source of frustration, with no shortage of etiological hypotheses: genetics, the cycad tree, metals, prions, algae, the water, bats, a virus, a parasite, some combination of these. None has been proven as a definitive cause.

Instead, investigators found only “decades of dead ends,” said Gerard D. Schellenberg, PhD, a professor of Pathology and Laboratory Medicine in the Perelman School of Medicine (PSOM) at the University of Pennsylvania, who studied the disease and the potential mutations behind it.

Even as the number of cases dwindled, researchers have flocked to the island to try to crack the case—not just to help the locals, but because they believe it’s a skeleton key, a neurological Rosetta stone.  

“If we could solve Guam disease at some point, it would help with these other neurodegenerative disorders,” Schellenberg said, though with skepticism. “Or, it may be the other way around.”

A Treasured Island

Decades before John Q. Trojanowski, MD, PhD, a professor of Geriatric Medicine and Gerontology in the PSOM and director of Penn’s Center for Neurodegenerative Research, studied lytico-bodig or even heard of it, he found himself on Guam, a tropical, coral reef island in the Pacific Ocean—a speck on the world map 8,000 miles west of Philadelphia and no bigger than Harrisburg, Pa. He was just four years old.

His father, then a captain in the U.S. Air Force, was stationed here as America worked to build up its military installations after the war. Some of Trojanowski’s earliest recollections are of the 30-day boat ride he and his parents took to Guam, with small snapshot memories he has managed to hold onto: fish jumping out of the water, his mother’s hysteria when she could not find him after he locked himself in a bathroom while playing.

By the time Trojanowski and his parents arrived in 1950, the disease was on the rise. The Chamorro people call it lytico-bodig (lytico from the Spanish word paralytico meaning weakness and bodig from the Chamorro word for listlessness), but in 1961 researchers assigned a formal scientific name which reflects its connections to other vexing neurological diseases, amyotrophic lateral sclerosis-parkinsonism-dementia, or ALS/PDC.

Virginia M.-Y. Lee, PhD, and John Q. Trojanowski, MD, PhD, began studying Guam's lytico-bodig disease in 1996. It has led
them to new approaches to Alzheimer's disease. Photo by Peggy Peterson

Trojanowski did not learn about the disease until he reached medical school at Tufts University in Boston in the mid-1970s. He worried then that he had been exposed to something during his short time here. He hadn’t, he decided, since it was primarily found in Chamorro people. The disease left his mind until it intersected decades later with his research on Alzheimer’s—which he would find shared molecular similarities, in addition to some symptomatic ones, with lytico-bodig.

The symptoms plaguing Guam residents in the second half of the 20th century included tremors, paralysis of the arms and legs, missed memories, and bouts of dementia. The afflicted were typically over 50, though reports of patients as young as 15 exist. Technically, lytico-bodig represents two different diseases: Patients with lytico present with ALS-like symptoms that usually appear first, while bodig patients have symptoms resembling Parkinson’s disease with occasional dementia. Some patients suffer from all those maladies.

Severe deterioration of the body and mind can start within just a year of diagnosis, doctors and families have said. Patients eventually become bedridden, relying on families for care, and succumb from the inability to swallow and breathe on their own. Being a full-time caregiver for sick relatives is part of the culture on Guam, where it’s not uncommon that extended families already live together.

“When she was first diagnosed, it was more her repeating herself, along with dementia,” said Lucy Cruz, an Umatac resident who left her job as a pediatric nurse in 1998 to care for her mother, who lived four years with the disease. “She only remembered things from the past and what was happening right [then]. Twitching of the hands; she was unstable,” she said. “Eventually, she lost her speech and swallowing mechanisms. That was right before she died.”

Cruz also has uncles, aunts, and a grandfather who died from lytico-bodig, all buried in the Umatac Cemetery, a few hundred yards from her house and theirs. Most everyone on her street had someone in their home suffering from the disease at one time, she lamented.

During the 1950s, local Navy doctors and off-island researchers, including Leonard T. Kurland, MD, a young neurologist from the National Institute of Mental Health and Mayo Clinic, took note of the disproportionate numbers of sufferers in Umatac, one of the island’s poorest and most isolated communities. The disease struck people all over the island, but nearly half of those living in the village of about 600 perished from the disease at its peak in the 1950s. Overall, the incidence rate among the Chamorro was 50 to 100 times that of the global average for ALS during this time, Kurland and colleagues reported in 1954 in a Neurology paper.

That same year, Kurland stressed the importance of the village and the island in the U.S. Armed Forces Medical Journal. It’s a geographic isolate, making it, theoretically, easier to home in on what’s causing neurological disease, here and elsewhere. The National Institutes of Health (NIH) was convinced. A year later, it established a research center on the island that ignited the great search.

Lytico-bodig was now the number one cause of death among Chamorro people.

Batty Ideas

In the decades before Trojanowski developed a research interest in lytico-bodig as a faculty member at Penn in the 1990s, several theories had already surfaced. A familial link was clear, meaning a genetic factor was possible. But Filipino immigrants who moved to the island also developed the disease, so many researchers surmised the cause was more of a shared environmental exposure rather than a genetic disposition.

Scientists turned to the Chamorro diet in 1960s for perhaps the most well-known of theories: the seeds of the cycad (Cycas micronesica), a prehistoric, palm-like tree that’s nearly extinct on the island. Chamorros harvested these baseball-sized seeds to make flour—known as fadang locally—for tortillas or as a thickener for soups, but only after soaking them in water for days to rid them of toxins. Locals knew they were poisonous to animals and people if not properly handled, with reports dating back centuries of people becoming violently ill after ingestion, some with brain damage. Kurland, working under the NIH center, and a visiting anthropologist named Marjorie Whiting marked the seeds as the culprit for lytico-bodig.

However, no one could reproduce a connection between cycad seeds and lytico-bodig in the lab. Consumption never drove animals to acquire the actual disease, and their brains never showed the unique set of neurofibrillary tangles that Kurland discovered a few years before. (Those tangles, a surprising finding at the time that further linked lytico-bodig to Alzheimer’s and ALS, were later key to Trojanowski’s involvement with the disease.) 

By the late 1970s, after six NIH conferences on the cycad theory, it mostly faded away, only to be resurrected in the mid-1980s with a fresh clue. Peter Spencer, MD, then director of the Institute of Neurotoxicology at the Albert Einstein College of Medicine, reported in Science that feeding cynomolgus monkeys large amounts of β-N-methylamino-L-alanine (BMAA), one of the main toxins released by the cycad seed, affected their motor neurons. Scientists eyed the finding with criticism because a person would have to eat hundreds of pounds of fadang in a week to show the effects seen in the monkey. Also, because Chamorros washed the seeds, only small amounts of BMAA would have been left behind.

“It seems unlikely that these low levels could cause the delayed and widespread neurofibrillary degeneration of nerve cells observed in amyotrophic lateral sclerosis and the parkinsonism-dementia complex of Guam,” Mark W. Duncan, PhD, a neurologist at the National Institute of Neurological Disorders and Stroke, and colleagues concluded in a 1990 Neurology study.

Other theories materialized amid the early cycad suggestions.

In 1976, Carleton Gajdusek, MD, received the Nobel Prize in medicine for his work on kuru, an infectious neurodegenerative disease (and happened to share the prize that year with Baruch Samuel Blumberg, MD, a Penn undergraduate alumnus who later became a professor of Medicine and Anthropology at PSOM). Gajdusek showed that kuru was transmitted through cannibalism in Papua New Guinea. That research prompted him to suggest that lytico-bodig may be another form of transmissible spongiform encephalopathy. Perhaps Chamorros were consuming infected tissue.

However, studies couldn’t support Gajdusek’s idea. Monkeys fed infected tissue never acquired lytico-bodig, so he abandoned the theory in 1982. He and others also suggested that low levels of calcium and magnesium concentrations in the soil and water caused an excess absorption of aluminum. Aluminum seemed a reasonable culprit because, in the 1970s, it had emerged as a possible cause of Alzheimer’s. However, a closer look at the soil and water in Umatac and other parts of Guam didn’t reveal any abnormal levels, studies showed. That theory, too, fell to the wayside.

A new cycad hypothesis came up in the early 2000s, this one from Paul Allan Cox, PhD, the director of the Institute for Ethnomedicine in Wyoming, and the late Oliver Sacks, MD, a neurologist and author of The Island of the Colorblind, an oft-cited travelogue of his island-hopping experiences in the Pacific. They blamed the fruit bat, a delicacy on Guam that locals call fanihi, near extinction today in large part because of its consumption. The bat, which fed on the seeds, “bioaccumulated” BMAA in its fat tissue to levels that could cause motor weakness and other neurocognitive deficiencies if consumed over long periods of time, the researchers reported in monkey studies published in Neurology. The decline of the bat population also paralleled the decline of the disease, they said.

The hypothesis grabbed people’s attention but hit a nerve in the scientific community. Cox’s assay techniques, cycad sampling and use of museum bats, were questioned by many scientists, and follow-up studies never backed up the findings.  

“I didn’t ever really believe the cycad story,” Trojanowski said. “It was never compelling to me. It’s good entertainment, but not science.”

John Steele, MD, a neurologist who studied lytico-bodig for nearly 30 years on Guam, shared that doubt. “They invested years and years of very intense study, yet nothing came of that,” said Steele, who was a co-author on the 1990 Neurology paper questioning the neurotoxicity in monkeys, and a self-described friend of Cox. “I don’t believe it.”

Steele, a Canadian who made a name for himself studying progressive supranuclear palsy, another rare neurological disease that shares many of the traits of lytico-bodig, arrived on Guam in 1983, after the NIH had closed the doors on its research center here following a significant decline in the incidence of cases and a lack of any breakthroughs. Steele split his time as a neurologist at the Veterans Affairs Medical Center and the Navy Hospital. The decline of lytico-bodig cases complicated the search for its cause, but helped firm up the belief that the disease wasn’t genetic—or at least not solely genetic, Steele said. Purely genetic diseases don’t just stick around for a single generation.

Ambitious and passionate, Steele stayed with it, studying and treating patients, teaming up with researchers from different fields along the way, and pursuing his own investigations of a novel detail: markings on the eye seemingly left by a parasite (See sidebar: “Keeping an Eye on the Disease”). He lived in Umatac for 10 of his years on Guam, where he would often find himself sitting on the graves at the cemetery to reflect, looking out over the water, hoping something would click or reveal itself.

“You’d think it would have been very easy to have come up with the reason, the cause for such a localized disease confined to Guam, and pretty much confined to Chamorros,” Steele said. “But that wasn’t the case.”

Over all this time, there was no consensus among the Chamorro people, either. Verena Keck, PhD, of the University of Heidelberg in Germany, noted in her book, The Search for A Cause: An Anthropological Perspective on a Neurological Disease in Guam, Western Pacific, that many locals were re-interpreting theories such as overall poor eating habits, not one seed, or “bad blood” from interfamilial marriages. Others pointed to military waste or a locally infamous and vengeful Spanish Catholic priest. Legend has it, in the 1700s, a family stole mangoes from a tree outside of his Umatac church, and as a consequence, he put a curse of fatal paralysis on that family and its descendants.

“If we could solve Guam disease at some point, it would help with these other neurodegenerative disorders,” said Gerard Schellenberg, PhD. “Or, it
may be the other way around.” Photo by Peggy Peterson

Untangling a Disease

Halfway across the world at Penn in the 1990s, Trojanowski joined the search for the “skeleton key” that might unlock the secrets of both lytico-bodig and other, more common neurodegenerative diseases, with his life and lab partner, Virginia M.-Y. Lee, PhD, a professor of Pathology and Laboratory Medicine at PSOM and director at its Center for Neurodegenerative Research. As Alzheimer’s researchers, their focus was on neurofibrillary tangles formed by abnormal versions of proteins called tau that clump together within nerve cells and their processes, a discovery the pair made in 1991.

Tau tangles are hallmarks of Alzheimer’s disease and an underlying feature in both lytico and bodig that Trojanowski and others believed could reveal clues to its cause. One of lytico-bodig’s distinctions rests in the topography of its tangles: they are more prominent in the spinal cord compared to other diseases. Otherwise, the brain of a lytico-bodig patient is remarkably similar to the brain of an Alzheimer’s patient.

“It was intriguing because it was dominantly a tauopathy, a neurodegenerative disease in which tau pathology is the main underlying brain abnormality,” Lee said. And the correlation between the tau tangles and cognitive deficiencies closely mimicked Alzheimer’s disease. “So, this was an opportunity to study a pure tau disease and its impact on the nervous system.”

A key collaboration in that study began in 1996 when Trojanowski and Lee met Schellenberg for the first time at a medical conference at another researcher’s poster presentation on lytico-bodig. “John walked up, and we started talking about genetics,” recalled Schellenberg, who was on faculty at the University of Washington at the time. They all decided to team up to study lytico-bodig shortly thereafter and have worked together in collaborative studies of neurodegenerative tauopathies continuously ever since. It’s how Schellenberg ended up coming to Penn in 2008.

The three were named co-investigators on a grant awarded to researchers across the country in 1997 by the National Institute on Aging (NIA) of the NIH. After a lull in NIH support for lytico-bodig following the closure of the Guam research center, the grant represented a renewed investment in the disease. It would last 10 years and add up to nearly $20 million. Collaborators on the project also included the principal investigator W.C. Wiederholt, MD, and Douglas Galasko, MD, from the University of California, San Diego, and Ulla-Katrina Craig, DrPH, from the University of Guam.

Some interaction between aging, genetics, and the environment served as the driving hypothesis for this grant. By this time in the late ’90s, scientists could depend on more advanced genetic technology to help answer these questions. While Trojanowski and Lee worked in Philadelphia to investigate the pathogenesis and development of neurofibrillary tangles, Schellenberg, in Washington, focused on the genetics of tau proteins and the interaction of genetic and environmental risk factors in disease.

Trojanowski and Lee led studies showing that many of the abnormal proteins in other neurological diseases existed in lytico-bodig patients in addition to tau, including Aβ in Alzheimer-like amyloid plaques, alpha-synuclein that Lee and Trojanowski showed were the building blocks of Parkinson’s disease Lewy bodies, and TDP-43 inclusions that Lee and Trojanowski discovered are the pathological brain signatures of frontotemporal degeneration and ALS. 

Finding the misfolded protein TDP-43 made the disease unique and important, in part because it helped better distinguish the disease. Some asymptomatic Chamorros in their 30s and 40s and many over the age of 50 had some variation of tangles in their brains. This group, however, had little or no TDP-43 pathology prior to age 70, the Penn team found in studies of control subjects, while patients with lytico-bodig did. Pathological TDP-43 might therefore reflect a more specific disease process, the authors concluded in a 2007 Acta Neuropathologica paper.

Trojanowski and Lee further spurred what seemed to be a tantalizing approach to therapies. They first engineered a mouse with spinal cord tangles and motor weakness to model lytico-bodig, a feat in itself published in the journal Neuron, as the first authentic model of the disease. Then they successfully offset the spinal cord tau pathology for the first time by using paclitaxel, a known cancer compound that functions as a “microtubule stabilizing” drug. That proof-of-principle led to other animal models and investigations with the drug in studies of Alzheimer’s and other diseases involving misfolded tau.

Trojanowski and Lee then discovered a more brain penetrant and therapeutically effective microtubule stabilizing drug for the treatment of Alzheimer’s, lytico-bodig, and related tauopathies, called epothilone D. Remarkably, the drug slowed tau buildup in the brain, helped improve cognition due to the build-up of tau tangles, and reduced nerve dysfunction in mice, several of the team’s studies showed.

But this, too, became another dead end in the search for answers to lytico-bodig: Despite epothilone D’s promise, the drug was shelved in 2013 when Bristol-Myers Squibb, the drug’s maker, shifted its focus away from Alzheimer’s. Trojanowski considered that act an abandonment of one of the most promising drugs that mitigates tau pathology in mouse models of Alzheimer-like and lytico-bodig-like tau pathology and neurodegeneration.  

Meanwhile, Schellenberg’s risk-factor subproject on the NIA grant led him down a breadcrumb trail from genes to environmental exposures and back to genetics.

He and his colleagues found that variants in the tau gene known as MAPT contributed to a higher risk of developing lytico-bodig, but they weren’t dominant mutations. That meant if it the variants were to blame, they needed help to trigger the disease, like an environmental factor.  

Schellenberg was surprised to find that consumption of fadang, the flour of the cycad seed, did show up as a risk factor in epidemiological studies. But that didn’t necessarily mean the consumption contributed to the disease, he noted. It might just correlate with a more rural lifestyle.

“The problem is that when you say environmental risk factor, the first thing that comes to mind is that people are eating a toxin,” Schellenberg said. “It could also mean that you are consuming something protective. A [different] diet perhaps protects you against the lytico-bodig on Guam. You may be looking for something rare that disappeared, or something that has now become common that protects.”

The concurrent demographic shifts on Guam and changes in the disease rate made it challenging to find answers based on environmental exposures. Cases of lytico-bodig peaked in the 1950s, right around the time the island started to become more westernized, and slowly declined thereafter. Save for a peak in bodig cases in the 1990s, by the mid-2000s, the disease pretty much disappeared altogether. By then, Guam looked more like mainland America with amenities, restaurants, grocery stores, construction, and more. People changed the way they ate and lived, researchers from the NIA grant noted in a 2003 paper in the American Journal of Epidemiology. And that made the epidemiological studies a virtual dead end.

Into Obscurity

When the NIA grant ended in 2007, Daniel Perl, MD, who studied the disease for decades and is now at the Uniformed Services University of the Health Sciences in Bethesda, Md., said the NIH urged him to keep lytico-bodig specimens intact so they could be made available to investigators in the future. Today, he said, some of the rare and limited samples he's held onto are with four different research groups, and three of these projects have full or partial NIH support.

Schellenberg’s last paper on lytico-bodig was published in JAMA Neurology in 2013, five years after the grant ended. While lytico-bodig afflicted mainly Chamorro people, cases of a very similar disease on the Kii Peninsula in Japan and in Western New Guinea had scientists linking them together, though several studies have muddied that water, including Schellenberg’s.

A mutation known as C9orf72, linked to ALS in Western populations, was found to be the driver of disease in the Japanese cases; however, Schellenberg and colleagues found no such mutation in Chamorro patients. They concluded that the mutation wasn’t the cause of lytico-bodig.

“So that was the last thing,” Schellenberg said. “I have samples, all sorts of information. If something new comes up, I’ll find a way to pursue it.”

As lytico-bodig fades into relative obscurity on Guam, the onetime epidemic lingers in the memory of the generations old enough to remember and mourn those buried at the Umatac Cemetery and others like it across the island. And it endures in the form of its continuing influence on research into similar diseases.

Trojanowski and Lee revived their work with microtubule stabilizing compounds after the grant ended but with another drug, dictyostatin, and a focus on Alzheimer’s and other tauopathies, like progressive supranuclear palsy. A 2016 study in Acta Neuropathologica Communications showed that dictyostatin mirrored many of the promising results found with the related, abandoned drug epothilone D in the mouse model of lytico-bodig. That finding reinforced the power and potential of microtubule stabilizing drugs to treat tau diseases, the authors said.

“Guam contributed to new ideas for therapies to treat Alzheimer’s,” Trojanowski said. “But I don’t see a path forward for lytico-bodig. Its time has come and gone. Time to move onto the next challenges.”

Sign up here to subscribe to the email edition of Penn Medicine magazine.