Manna from hell

Manna from hell Photographs by Mirko Beović/ 1 "This so-called Chinese-herb nephropathy is characterized by a pattern of interstitial fibrosis similar to that of Balkan endemic nephropathy," the authors wrote. Pavo Lukšić (above) and his horse (below) in front of his uncle's house. Grollman, a pharmacologist who has pushed for stricter regulation of herbal supplements, says he immediately phoned a Croatian colleague. Grollman was soon on a plane

Julia C. Mead
Oct 31, 2007

Manna from hell

Photographs by Mirko Beović/ 1 "This so-called Chinese-herb nephropathy is characterized by a pattern of interstitial fibrosis similar to that of Balkan endemic nephropathy," the authors wrote.

Pavo Lukšić (above) and his horse (below) in front of his uncle's house.

Grollman, a pharmacologist who has pushed for stricter regulation of herbal supplements, says he immediately phoned a Croatian colleague. Grollman was soon on a plane, expecting to learn that patients in Croatia had ingested Aristolochia-based decoctions. Instead, he learned something entirely different, and spent the next few years pulling together a binational, multidisciplinary research team that not only tracked the pathway of exposure, but also eyeballed the compounds that form when Aristolochia couples with DNA, thus enabling the identification of the mutational fingerprint of endemic nephropathy.

In the process, his team developed a hypothesis for ongoing work: Black houses, those desolate symbols of families dispersed by what was once a mysterious toxin, may not be confined to an isolated smidgen of central Europe.

The first laboratory findings of what is now known as endemic nephropathy are uremia, low-molecular proteinuria, anemia that is more severe than is typical of early-stage chronic kidney disease, and hypertension, rarely present in other kidney diseases. Those signs generally appear between the ages of 40 and 60. The kidneys turn fibrous, shrink from the cortex inward, and are eventually reduced to a third their normal size, smaller than with other forms of nephropathy. Oddly, the glomerulus is spared.

The disease is unyieldingly fatal, although kidney transplants could offer a cure. While such transplants are rare in Croatia, healthier diets and better medications for hypertension are helping forestall the inevitable. Outcomes have improved as dialysis clinics opened across the region in the 1970s.

Bojan Jelaković, Croatia's top expert on endemic nephropathy.

One of them was at the Josip Benćević Hospital in Slovanski Brod, a hub for 14 endemic villages strung along the Sava River. At the city's modest hospital, whose façade still bears the blast marks of hand grenades lobbed more than a decade ago, one-third of the dialysis patients (31, as of August 2007) have endemic nephropathy, and 50 more are being tracked through earlier stages of disease, says the hospital's senior nephrologist, Ninoslav Leko. Six to eight new diagnoses are made each year.

Significantly fewer patients today reach end-stage renal failure before age 65, says Leko. Between 1957 and 1960, the average age at death was 45; now, it's 69.2

An unanticipated problem presented with that increased lifespan, however. Patients lived longer, but half of them were developing malignant tumors in the renal pelvis and ureter, often bilaterally. "I talk all the time about the need to remove those things," says Jelaković, adding that endemic nephropathy patients have the highest known prevalence of transitional-cell cancer of the upper urinary tract.

Researchers were keen to find the cause of the disease in the 25 years after the disease was first identified. No fewer than 10 symposia aired an evolving list of suspected risk factors. Heavy metals and viruses were ruled out early, and a consensus gelled around an unknown environmental toxin. In the 1990s, scientists proposed, and then disproved, the hypothesis that underground coal deposits were contaminating the drinking water. War and poverty later diverted attention, says Jelaković, who has become his country's emissary of endemic nephropathy. "I gave a lecture four years ago that I called 'Endemic Nephropathy: The Forgotten Disease,'" he says.

Pockets of endemic villages are scattered among five Balkan nations, along a 500-kilometer stretch of the Danube River basin

In October 2006, a University of Zagreb symposium devoted two days to the two remaining suspects: ochratoxin A, a fungus found in stored grain, coffee, and dried fruit, and Aristolochia clematitis, a common weed with heart-shaped leaves and a pale yellow flower that resembles a down-turned pitcher. Both are nephrotoxins. Ochratoxin A is found across North and South America and Europe, and A. clematitis, originally a Mediterranean cousin to some 450 known Aristolochia species, has wandered across Europe. But why was the disease confined to a small patch of the Balkan Peninsula?

In a literature search at the Zagreb University library during the summer of 2003, Grollman says he found the first report of a correlation between nephropathy and A. clematitis in a 1957 paper.3 A veterinarian identified chronic renal failure and proteinuria among horses that ate hay contaminated with the weed. Later, tissue samples from rabbits and rats that were fed the plant showed signs of chronic interstitial nephritis, which is the fibrosis and atrophy that are now established pathologic features of the disease in humans.4

The pathway of exposure to humans remained a mystery. So Grollman went with Jelaković to Slavonski Brod to talk with dialysis patients. "I was sure they had been taking herbal medicines," Grollman says. "And then we interviewed them through a translator and that made clear they hadn't." The patients, shown pictures of the plant, each identified it as "vućja stopa": Wolf's Paw, a common weed. Grollman says he and Jelaković later walked around the farms outside the city and found the plant on the outer fringes of the fields. When they tramped the plowed rows, they found it entwined among the wheat stalks.

That discovery caused Grollman, whose lab previously worked on the mechanisms of DNA damage and cancer, to suspect that endemic nephropathy is not all that endemic. Rather, he reckons the disease might exist anywhere people ingest Aristolochia. In other words, endemic nephropathy and aristolochic acid nephropathy, the disease that struck the Belgians and just a few patients elsewhere, were the same.

After Grollman returned to Stony Brook in 2003, Andrea Fernandes, then a senior research assistant in his lab, began documenting the history of Aristolochia use and quantifying the dosages in ancient and modern remedies. Fernandes says Aristolochia species grow in every climate range and on every continent, except perhaps Australia, and many have been prized for centuries for their supposed medicinal value (see 5 "We felt maybe we could use the same approach on DNA adduct in vivo," says a coauthor, Robert J. Turesky, who is with the health department's Division of Environmental Disease Prevention in Albany.

The team not only identified the presence of two analytes - slightly differing molecules from the dA-AL adduct, labeled dA-AL I and dA-AL II - in the Rhode Island woman's renal cortex, but also looked for them in tissue samples from a patient who had never been exposed to Aristolochia. The adducts were absent.

Turesky says that was the first time this method was used to characterize the identity of a human DNA adduct derived from an environmental, causal agent. More important, it convinced Grollman that adducts could be identified in endemic nephropathy patients. Pathologists in Slavonski Brod had been preserving kidneys and tumors in jars of formalin and later in paraffin for four decades, hoping the technology to solve the endemic nephropathy riddle would someday come to Croatia.

Instead, some 850 Croatian samples traveled to the technology. Many, however, were too degraded by time or end-stage disease to extract the DNA, or for three renal pathologists to see clear signs of endemic nephropathy. Grollman asked for newer, frozen samples. From those, the pathologists were able to select samples from four patients who met the World Health Organization's clinical diagnostic criteria for endemic nephropathy. They also had control samples from five patients with upper urinary tract cancers who lived in nonendemic parts of Croatia, and from five others with more common forms of renal disease.

Grollman says it soon became clear that extracting the DNA in Croatia would be more efficient, so he enlisted Branko Brdar. Now retired from the Rudjer Bošković Institute in Zagreb, Brdar had also worked earlier at Stony Brook, investigating a genetic treatment of the HIV 1 strain. Brdar did the extractions, but could not do p32-postlabeling analysis in Zagreb. Shibutani could, and he found both the dA-AL and dG-AL adducts in the renal cortexes of all four patients with the disease but none in the ten control subjects. Then he also found both adducts in tumor samples from three residents of endemic villages.

Finally, more than five years after Grollman first read the NEJM article linking the two types of nephropathy, he had convincing proof that they were the same disease with the same causal agent, but separate pathways of exposure.

Ever cautious, however, Grollman still wasn't ready to publish. Instead, he handed Brdar's DNA extractions to Neda Slade, a molecular biologist who had been one of Brdar's students in Zagreb, but who was then working under Ute Moll in Stony Brook's pathology department. They stained the material for possible p53 mutations and, where more than 10% of the cells held the stain, it went to Roche's lab, she says. "At first, we were thinking about sequencing the p53 exon by exon," Slade says. "But that would have been painful." They sent Roche tumor samples, all transitional-cell carcinomas, from seven women and four men who had lived in endemic villages for at least 15 years. Five samples had been fixed earlier in formalin and embedded in paraffin, the remainder frozen in liquid nitrogen. Lin Wu says her team at Roche was able to sequence exons 2 to 11 on nine samples and, where a crosslink between DNA and protein from the fixative made that difficult, exons 5 to 8 on the rest.

Altogether, 19 base mutations were found. Of those, mutations at A:T base pairs accounted for 17, and 15 resulted when A:T flipped to T:A. Roche had identified the mutational fingerprint of aristolochic acid exposure using an assay its scientists had been trying to develop since 1998, Wu says. When the team, including members of Grollman's lab, Turesky, and Roche researchers, published their results in PNAS in July,6 it was the first public release of their technique, she says.

That same month, Grollman won a five-year, $8.17 million grant renewal from the National Institute of Environmental Health Sciences to broaden his study of the disease that now appears destined to become known solely as aristolochic acid nephropathy. That work may involve further collaboration with Ute Moll, who coauthored a paper, published in August, showing that the p73 cancer-suppressor gene takes over when p53 is lost.7 Her finding could help test whether genetic predispositions account for only half of endemic nephropathy patients having cancer. The ongoing work, on human tissue and on mice, is also likely to investigate the fibrogenic pathway through the kidneys, track the biotransformation of aristolochic acid, and validate the adducts as biomarkers.

The work could send Grollman, or at least his investigation, to China. China saw its first class-action lawsuit filed in 2002; 100 plaintiffs, all with renal failure, sued the nation's oldest supplier of traditional medicines, claiming Aristolochia mixtures were to blame. Grollman says that colleagues in China have confided to him their suspicions that aristolochic acid nephropathy accounts for a large portion of the renal failure there.

Croatia may be in line for membership in the European Union, but research funding remains scarce there. Despite the recent discoveries, Jelaković still funds his own medical diplomacy, shuffling visitors to the dialysis clinic in Slavonski Brod, the endemic villages, and the wheat fields, and to meet his colleagues. Croatia is only a microcosm of the problem, he says. Considerable anecdotal information but a dearth of data on prevalence in Bosnia, Bulgaria, and Rumania leave unconfirmed the widely held suspicion that the disease cuts a broader, longer swathe across the region than has been documented. The only recent data from Serbia involve just one small endemic pocket.

Poverty and political antipathy mean endemic nephropathy has gone virtually ignored in those countries. Just this autumn, Jelaković has begun visiting physicians in Serbia and Bosnia. Romania, which under socialistic control has refused to even acknowledge that the disease existed there, has done little research. Bulgaria, in an attempt to obliterate the evidence, once picked up an entire village and moved it to a nonendemic location, the sum total of its attention to the disease, he says.

The disease could develop in a proportional number of refugees from what is now Bosnia, just as it did in Ukrainians who fled the Russian Revolution some 80 years earlier. "It's unbelievable," says Leko. "Whatever we do, even with the war, even with improved communications and treatment, we always have the same numbers."

"It's unbelievable," says Leko. "Whatever we do, even with the war, even with improved communications and treatment, we always have the same numbers."
Standing next to Pavo Lukšić's cart, Jelaković says that later this year, health workers and agronomists will try to convince the farmers to avoid harvesting A. clematitis with their wheat. He says this softly and in English, as if not wanting to offend Lukšić, who had insisted the water was to blame. The farmer waved a weathered hand toward his uncle's abandoned, crumbling house, then abruptly urged his horse forward. As his wooden cart rounded a bend, Jelaković translated Lukšić's parting words: "He said, 'You can look, but don't forget.'"

1. J.L. Nortier et al., "Urothelial carcinoma associated with the use of a Chinese herb (Aristolochia fanghi)," N Engl J Med, 342:1686-92, 2000.
2. M. Miletić-Medved et al., "Recent data on endemic nephropathy and related urothelial tumors in Croatia," Wiener Klinische Wochenschrift, 117:604-9, 2005.
3. M. Martinčić, [in German] Veterinarski Arhi, 27:51-9, 1957.
4. M. Ivić, "Etiology of endemic nephropathy," Lijec Vjesn, 91:1273-81, 1969.
5. A. Goodenough et al., "Novel LC-ESI/MS/MS(n) method for the characterization and quantification of 2'-deoxyguanosine adducts of the dietary carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine by 2-D linear quadrupole ion trap mass spectrometry," Chem Res Toxicol, 19:263-76, 2007.
6. A.P. Grollman et al., "Aristolochic acid and the etiology of endemic (Balkan) nephropathy," Proc Natl Acad Sci, 104:12129-34, 2007.
7. F. Talos et al., "p73 suppresses polyploidy and aneuploidy in the absence of functional p53," Mol Cell, 27:647-59, 2007.