MS clusters: chance or design?

MS Clusters; Chance or Design?

Whenever a story of a "cluster" breaks in the media, the telephones at the Society ring off the wall.

Why? We know perfectly well that no virus, no bacterium, no chemical or other environmental agent has ever been flagged as the culprit. Yet, when we hear news of an inordinate number of cases in one geographic area, suddenly the possibilities of infection or contagion come alive. The public reacts. And epidemiologists, who study the spread of the disease, are alerted.

Several steps are necessary for determining if a multiple sclerosis "cluster" is legitimate. Epidemiologists note that often some patients in a reported cluster do not have multiple sclerosis at all. Experts therefore recommend that reported patients be carefully examined by neurologists. Once a diagnosis is confirmed in a given patient, it must then be determined whether multiple sclerosis occurred after the patient became a resident of the place in question. It is also important to ascertain where the patient lived during his childhood, because multiple sclerosis appears to be linked to the location in which people spend the first 15 years of their lives, even though clinical manifestations may surface much later.

One of the earliest and most famous clusters known to MS investigators was an "epidemic" that occurred on the Faroe Islands, a Danish possession lying in the north Atlantic between Norway and Iceland. Though the inhabitants are Nordic and considered a high-risk group for the disease, there are no known reports of MS among native-born residents before 1943.

In the early 1960s a Washington, D.C. neurologist, Dr. John Kurtzke, became intrigued with a report by a Danish investigator, K. Hyllested, about 29 cases of MS in the Faroes that had occurred starting in 1943. "This looked like a real epidemic," he said. "Obviously, the disease had to have been brought into the Faroes since it hadn't been there before."

The only thing brought into the Faroes in the 1940s was a battalion of British troops who occupied the islands as a protective measure during World War II. Assuming an incubation period of a few years, this would tally with an epidemic onset in 1943. The epidemic occurred only in people past puberty at the time of the British occupation. There was a second epidemic among those exposed as children, who developed multiple sclerosis only several years after puberty. In fact, a third, smaller wave of cases surfaced among those born after the British left. Many of the occupation soldiers were from the Scottish High-lands, where the MS prevalence is quite high: 90 cases per 100,000, comparable to rates in the northern U.S. In Dr. Kurtzke's view, if MS is somehow triggered by a virus, the disease may have been brought to the Faroese by the occupying forces.

A grantee of the Society since 1977 and author of the famous Kurtzke scale that delineates levels of disability, the neurologist is continuing a broad surveillance of new cases in the Faroes, scrutinizing areas in which no MS has been reported thus far, and analyzing results of 5,000 questionnaires sent out locally.

Clusters are usually first noticed by residents of a community, and Mansfield, Massachusetts was no exception. In 1971 a resident suggested that an unusually large number of people with multiple sclerosis had lived in the town from birth. This led to an investigation by a team working then at the University of Virginia School of Medicine.

Drs. Richard Eastman and the late David Poskanzer examined 17 patients and found 14 of them with probable or possible multiple sclerosis. This meant an MS prevalence rate of 141 per 100,000. The doctors carefully examined the histories of the patients but found no increase in childhood diseases or infectious illnesses among them. Eight patients had lived within several blocks of one another in the 1930s, and there was speculation that water contamination during that time might have been a factor in the MS cluster. But, the authors concluded it was not.

The next cluster to hit the headlines was in Saskatoon, in Saskatchewan, Canada, where a patient with multiple sclerosis, whose sister also had the disease and later died, insisted she had found a strikingly high prevalence of MS in and around the small community of Henribourg, 50 miles north of Saskatoon. She had identified 27 people who either had MS or had died affected with it. Based on the population, this was one person out of 11. Very high if, indeed, these were multiple sclerosis cases. Some could have been misdiagnosed.

Scientists at the University of Saskatchewan tried to review conditions as they had existed in Henribourg during and before World War II, when almost all the afflicted people were living in the village. Toxicologists examined local soil, water and base materials comparing them with samples from other areas. Their findings proved inconclusive.

An MS cluster that drew scare headlines was reported in Key West, Florida in 1985. A University of Miami physician, Dr. William Sheremata, announced he had found 23 cases of multiple sclerosis among the 26,000 residents of the island resort. Within a few months another five cases were uncovered, and Dr. Sheremata called in an expert to confirm that the patients really had MS. It appeared that most of the patients were natives of Key West. They hadn't migrated there, taking with them a "northern prevalence rate."

What endowed this cluster with unusual interest was that Key West has a tropical climate, yet had an apparent multiple sclerosis prevalence rate of 84 per 100,000. In other parts of the South, MS occurs in about five out of every 100,000. However, as Dr. Sheremata has pointed out, MS is not a communicable disease, doctors aren't required to report it, and MS specialists are in short supply in the South. So the accepted low prevalence of the disease in Florida might be due to incomplete case ascertainment.

Complicating the picture was the finding by Dr. Hilary Koprowski and colleagues at the Wistar Institute in Philadelphia of HTLV-1 (human T lymphotropic) virus in the blood and spinal fluid of some of the Key West patients. Since HTLV-1 causes tropical spastic paraparesis, a disease which is common in the Carribean and clinically similar to mild progressive MS, many experts think some of Dr. Sheremata's patients may well have had this disease.

Galion, Ohio, with a population of 12,391, found itself with 25 cases of MS in 1986. Epidemiologists investigating the cases came up with some intriguing facts. Over 90 percent of the patients had lived in Galion since childhood, with seven out of ten living on the north end of town. Three patients grew up on the same block, attended the same high school, and were almost the same age. Nothing in the history of Galion pointed to any common agent except that in 1960 a patch of land on part of an old cemetery was dug up so that a new high school gymnasium could be built. The loose earth was offered to anyone who would take it away. Many did. About five years ago a reporter from the Galion Inquirer did a search of old death certificates and found that about half of Galion's deaths in 1987 were attributed to neurological illness.

Recently a team from the Cleveland Clinic took a hard look at the Galion cluster, which was getting so much local publicity it was affecting the social and economic life of the town. Reporting at the American Academy of Neurology last April, the scientists said that 27 percent of the alleged cases were not multiple sclerosis at all. The remaining 73 percent had definite or probable MS; however, of these, 27 percent had experienced their first symptoms before moving to Galion. The prevalence was still high, though not greater than expected by chance alone. Furthermore, the apparent excess could be explained, the team said, by the fact that industry has shifted from Galion over the years, and healthy people have moved away with it, leaving a disproportionate number of disabled behind. The prevalence rate, the team concluded, was "not outstandingly high."

MS clusters may sometimes crop up in a factory setting. In 1987 the October issue of Neurology included a report of 20 cases of multiple sclerosis at a manufacturing plant in Rochester, New York. When a team of scientists from the University of Rochester under Dr. Randolph Schiffer checked workers' records, they found that 10 had developed MS during one 10-year period, 1970-79, when two or three cases would have been the norm.

Since the plant used zinc as a principal raw material and workers might have inhaled or swallowed zinc dust, the scientists measured zinc levels in the blood of employees with multiple sclerosis, a group of MS patients who were not plant employees, and a group of healthy people in the plant. There were no significant differences in zinc levels among the groups studied, though all people working in the plant (MS and non-MS) showed higher levels of the metal.

Zinc seemed of possible interest to the Rochester investigators because it is known to affect the body's immune regulation. Obviously, some of the plant workers are genetically predisposed to MS. Dr. Byron Waksman, the Society's vice president for research, points out that people with genetic susceptibility do not all get the disease. "It may be," he says, "that zinc 'resets their thermostat' so that a larger proportion gets MS."

However, if zinc had that effect on the workers at the plant, presumably other immunologically-based disorders would also have increased. Since they did not, the Society's Medical Advisory Board decided the study showed no particular risk factor associated with the development of MS.

To get a general perspective on cluster research, INSIDE MS contacted Dr. Leonard Kurland, senior consultant and former chairman of medical statistics and epidemiology at the Mayo Clinic in Rochester, Minnesota, who has published widely on the subject.

"There are two things you have to be sure of before you start talking clusters," he said. "First you must be certain that people in the cluster really have the disease. As we all know, many other illnesses mimic multiple sclerosis. Secondly, it's essential to find out where these people were residing at the time of onset of their MS and during the years before the onset of symptoms.

"The cluster may simply mean that these people assembled in the community at some later date. This happened with an apparent cluster of eight cases of MS in Duxbury, Massachusetts in 1959. When we examined the patients, indeed they all had multiple sclerosis but only one had developed it locally; the other seven had had MS before their arrival in the town. They all came to Duxbury because it was known to be a charming retirement community. So this washed out that cluster."

Other factors, too, may dim the luster of a cluster. Says Dr. Carmel Armon, research fellow in neuroepidemiology working with Dr. Kurland at Mayo, "Remember that in the past many mild cases of multiple sclerosis used to go undiagnosed. Now we are using more sensitive diagnostic techniques. It is therefore easier to identify MS cases, including milder ones. So the use of improved technology helps create an apparent cluster.

"Moreover," he adds, "you might have a community which has never had a neurologist in residence before. Then a neurologist comes in to set up his practice. More cases of multiple sclerosis will be identified in that community because there is someone there now to spot them."

Dr. Armon is trying to apply a mathematical formula to calculate the probability of the chance occurrence of apparent clusters, taking into consideration the cases that are not reported. "You don't hear any reports about a town that has too few cases of multiple sclerosis," he observes.

Dr. Armon is factoring into his model the "multiple affiliations" each of us has which increase the likelihood that "abnormal" MS prevalence rates may occur with one or another affiliation purely by chance.

It is often the case, Dr. Armon says, that for a cluster to have epidemiologic significance, it must be larger than what is being reported. People instinctively react when they become aware of six or eight cases of multiple sclerosis in a small town.

"But we need to counterbalance that instinct," Dr. Armon says. "Perhaps 16 or 18 cases might be needed to make a significant cluster. For example, there are many thousand towns of 10,000 throughout the country. We can show almost with certainty that at least one town of that size could come up with six or eight cases just by chance.

"I want to develop some rules of thumb for determining when a collection of MS cases reported by a community is really of epidemiological significance." Dr. Armon plans to present a preliminary report on his analysis at the American Neurological Association meeting in September.

Meanwhile, multiple sclerosis clusters continue to emerge. The latest one was reported in the April 10 Duluth News-Tribune. An article described 14 people with multiple sclerosis who lived in or near a town called Carlton, which has a population of 860. A Duluth neurologist, Dr. Joanna Woyciechowska, had noticed that a number of her MS patients were from the Carlton area and notified the Society. Since most Carlton residents have spent most of their lives there, it may be easier to find some common link among them, if one exists.

"It's almost impossible to trace such links in a large, transient city population," says Dr. Mathilde Solowey, special consultant for a national study of clusters being conducted by the Society. Recently, she and Margaret Calvano, Society director of information, began targeting certain clusters for investigation. Carlton is one of them. The 14 patients in Carlton are being sent questionnaires on their symptoms, diagnostic tests, hospitalizations, and residential history. When these are completed they will be sent, along with the patients' medical records, to neuroepidemiologist Gustavo C. Roman of Lubbock, Texas, who will sort out possible common traits and characteristics.

As Dr. Solowey says, "The answer may not be found for years, if ever. Cluster research seldom yields conclusive results. But if there is something people with MS patients have in common, we'll find it. Just because we haven't found it yet doesn't mean it's not there."