Dr. Richard Montali, then the zoo’s head pathologist, got a call from a vet relaying the bad news: Kumari had died, and a truck was on its way with her body. Montali assembled his team to get the necropsy under way as soon as possible.
Dr. Laura Richman was 15 minutes away at the Armed Forces Institute of Pathology at Walter Reed Army Medical Center. She was completing a two-year pathology residency at the National Zoo but was doing some training and course work at Walter Reed. Montali told her to come quickly.
About a dozen doctors worked on Kumari through the night, examining every inch of her inside and out. They came to the table with a list of suspects—poisoning from toxic plants, septic shock from a bacterial infection—but none fit. Like crime-scene investigators, they documented everything, taking notes, pictures, samples.
What they saw was shocking: thousands of tiny hemorrhages in Kumari’s heart, liver, tongue, and intestines. The evidence suggested the presence of a viral or bacterial organism, but they couldn’t be sure until they got the samples under a microscope.
Montali sent the samples to be made into slides. It took two days because the tissues had to undergo a chemical treatment to preserve them. “It was the longest 48 hours of our lives,” Montali says.
When the slides came back, Montali gathered his team around a five-headed light microscope. Richman spotted something in Kumari’s cells: tiny globs called inclusion bodies, a sign of a viral infection. The inclusion bodies were found in cells lining Kumari’s blood vessels. That explained her purple tongue and internal bleeding: The pathogen appeared to have ruptured blood-vessel cells, which caused capillaries to leak and resulted in hemorrhaging.
Richman pored through science journals to see if anyone else had reported mysterious elephant deaths. One case piqued her interest: A circus elephant in Switzerland had died unexpectedly, and scientists found inclusion bodies just like Kumari’s. They had looked at samples under an electron microscope, which offered millions of times greater magnification than the zoo’s microscopes, but they were never able to come up with a diagnosis.
Working alone one night at Walter Reed, Richman fired up an electron microscope and put in one of Kumari’s samples. She noticed dark, round objects with a dense core, which resembled a herpesvirus—a clue. “I was so excited I nearly fell out of my chair,” she says. “I think I called everyone on earth.”
If the virus proved to be herpes, it was unlike any other herpesvirus known to science. In other animals, including humans, herpes resides in nerve and epithelial tissue, such as skin, rather than in blood vessels. And it usually results in blisters and sores, not death.
With the help of John Lehnhardt, who had connections with elephant curators at zoos around the world, Richman turned up more than a dozen undiagnosed elephant deaths and was able to get tissue samples to look at. Nine of those elephants had died from the same virus that killed Kumari.
To confirm the diagnosis, scientists needed to compare the DNA of the virus to that of known herpesviruses. Because efforts to grow the elephant virus proved unsuccessful, Montali and Richman turned to plan B: a polymerase chain reaction (PCR)—a molecular technique in which trace amounts of DNA are copied over and over to yield quantities large enough for analysis.
For a year, Montali and Richman searched for someone with the right tools to run the PCR. “We went around with a little dog-and-pony show to see if we could find other herpes virologists who could help,” says Montali. The problem was, most were busy studying known herpesviruses and had no need—or time—to look for new ones.
Then Montali met Seattle-based virologist Richard Garber, who offered to help. Within a week, Garber called Richman to say that the PCR worked—the viral DNA had been successfully copied.
Garber ran the DNA through a database to search for matches. Sure enough, the herpesvirus was a hit. A comparison showed enough similarities to prove that Kumari’s killer was part of the herpes family but that this virus was a new branch of the tree. Closer analysis revealed there to be two distinct species of the virus in elephants, a version that’s fatal in Asians and another that kills Africans.
Armed with confirmation of her discovery, Richman completed her PhD work on the virus, which became known as the elephant endotheliotropic herpesvirus, or EEHV. In 1999, Richman, Montali, and eight colleagues published a paper in Science describing the virus and its effects.
“We thought we had it all figured out,” Richman says. “Little did we know we’d hardly scraped the iceberg.”
On a recent morning, Erin Latimer stands at a counter in the zoo’s National Elephant Herpesvirus Laboratory—a department the zoo established as part of the pathology lab, anticipating that EEHV would be a major threat to the species. Latimer uses a dropper to transfer elephant blood from one vial to another.
She receives blood samples from zoos and circuses across the country and runs a four-hour test to determine whether or not an elephant is sick. When a test comes back positive, she has a difficult phone call to make.
“I’ve learned a bit of bedside manner, but it never gets easier,” she says. “Telling someone their elephant has herpes feels like a death sentence.”
It turns out EEHV is responsible for about half of the deaths of young Asian elephants in the United States. So far, 37 cases have been identified, including 28 deaths and 9 survivors. In 2006, researchers reported the first known case of EEHV in a wild Asian elephant in Cambodia; more than a dozen cases have been reported since then. The threat to African elephants appears to be far less, which is why most of the research has focused on Asians. Researchers have found only three cases in captive African elephants and none in wild ones.
Without a blood test, EEHV is almost impossible to diagnose. As with Kumari, the early symptoms are vague—lethargy, loss of appetite, swelling of the joints—and most sick elephants are simply described as seeming “off.” To be safe, vets get tests at the first sign of abnormal behavior.
If an elephant is sick, the disease moves rapidly, circulating through the bloodstream and damaging vital organs. Treatment with the human antiviral drug famciclovir has seen some success, but there are no guarantees: Scientists are unsure why the drug works in some cases but not in others.
“There’s not a lot of bigtime science going on with this, so we still have a lot of unanswered questions,” says Dr. Gary Hayward, a human-herpesvirus expert at Johns Hopkins who also studies the elephant virus. He says fewer than a dozen groups worldwide are working on the problem.
Hayward has been involved in EEHV research since Richard Montali and Laura Richman got him interested in the mystery soon after Kumari’s death. Hayward invited Richman to complete her PhD work in his Hopkins lab, and he is one of the authors of the Science paper.
In the 12 years since the paper was published, experts have uncovered 14 distinct herpesviruses in elephants. Five are known to cause the hemorrhagic disease, but one type—EEHV1—is responsible for 90 percent of the deaths.
“Herpesviruses are supposed to be well behaved, quiescent viruses that are well adapted to the host,” says Hayward. “The big question here is why some of these are not.”
Elephants can carry more than one species of the virus, including the deadly ones, and survive. Most deaths have been in calves between ages one and eight. Experts theorize that antibodies in the mother’s milk protect them during infancy, but once a calf is weaned, it’s up to the animal’s immune system to provide protection.
Why do some young elephants die while others live? Hayward thinks the answer might lie in the timing: If an elephant first contracts one of the nonlethal types, the immune system might be stronger and better equipped to combat the more dangerous viruses later. But if a young elephant is initially exposed to a lethal virus, it’s likely its body won’t be able to fight off the disease.
Next: How to fix it—that's the question.