Communicable Diseases Intelligence Volume 33 No 2 - June 2009
Surveillance of Creutzfeldt-Jakob disease in Australia: 2009 update
This report provides a snapshot of Creutzfeldt-Jakob disease (CJD) in Australia for the period 1 April 2008 to 31 March 2009. In this short report, the Australian National CJD Registry provides undated Australian human transmissible spongiform encephalopathies (TSE) figures and discuss a recently published investigation of geographical TSE clustering in regional New South Wales.
Communicable Diseases Surveillance
Genevieve M Klug, Alison Boyd, Victoria Lewis, Amelia McGlade, Christiane Stehmann, Colin L Masters, Steven J Collins Abstract
In Australia, the occurrence of all human transmissible spongiform encephalopathies (TSEs) is surveyed by the Australian National Creutzfeldt-Jakob Disease Registry (ANCJDR). While prospective surveillance commenced in October 1993, the ANCJDR also retrospectively ascertained cases that occurred between 1970 and 1993. During the surveillance period of 1 April 2008 to 31 March 2009, the ANCJDR received 90 suspect TSE case notifications, which is slightly increased from previous annual surveillance periods. Based on the total number of probable and definite CJD cases, ascertained between 1993 and 2009, the Australian age-adjusted mortality rate is 1.18 deaths per million per year. In this short report, we provide updated Australian human TSE figures and discuss a recently published investigation of geographical TSE clustering in regional New South Wales. Commun Dis Intell 2009;33:188–191.
Australian National Creutzfeldt-Jakob Disease Registry surveillance update Since 1 October 1993, the ANCJDR has been notified of 1,345 suspected cases of CJD, arising in both the prospective and retrospective ascertainment periods. For the prospective period, the average annual rate of suspect cases notified to the ANCJDR was 3.1 per million per year. Fluctuations in these annual notification rates have been observed (Figure 1) and the reasons for these have been discussed previously.3 More recently, the rate of notifications from 2006 to 2008 has been sustained at a higher level in comparison to the longer term average observed for 1993–2008. The increased number of notifications is most likely underpinned by increased referrals to the ANCJDR of cerebrospinal fluid for 14-3-3 protein testing, particularly for the 2007–2008 period where CSF referrals have increased by 43% in comparison with the previous 6 year average. This sustained increased level of CSF referrals has given the ANCJDR confidence that the introduction of 'fee-for-service' from 1 January 2007 has not detrimentally affected CSF referrals and consequently rates of suspect case notification, as initially speculated.
Figure 1: Annual age-standardised Creutzfeldt-Jakob disease mortality rates and suspect case* notification rates, 1993 to 2008
During the 2008–2009 reporting period to 31 March 2009, 90 suspect cases were added to the Register, which is a 17% increase from the previous surveillance year. Of these 90 suspect case notifications, 20 have been reclassified as definite TSE cases. A further 15 cases that were notified prior to 1 April 2008 have also been confirmed as definite (11) or probable (4) TSE since the last update. The Register has a large group of cases still under investigation (184), with the majority of these still alive as at 31 March 2009. Despite active investigation of all suspect cases, the number of incomplete cases continues to expand, although final outcomes for the large majority of all suspect cases have been obtained through detailed investigation. The number of incomplete cases is of concern to the ANCJDR as for many of these cases, a conclusive outcome may not be achievable. Follow-up can be difficult for cases that have been notified several years previously. In 541 cases, CJD has been excluded while 608 cases have been classified as definite (395) or probable (213) CJD and a further 12 cases fulfil the possible case definition (Table 1). A sustained elevation of the annual CJD incidence and the proportions of autopsy-confirmed cases for the surveillance years of 2005–2008 has been observed (Figure 2). As previously discussed, this relates to the growing number of notifications and pro-active approach of the ANCJDR of seeking autopsy in all clinically suspect cases.3
Table 1: Classification of cases by the Australian National Creutzfeldt-Jakob Disease Registry, 1 January 1970 to 31 March 2009
Age-standardised mortality rates were calculated using the Australian Bureau of Statistics 2000 estimated resident population for Australia.
The aetiologic proportions of all Australian TSE cases are consistent with previous observations.3 Cases classified as sporadic CJD comprise 90.5% of all Australian cases, while 8.2% of cases are genetic and the remaining 1.3% are iatrogenic. During the 2008–2009 surveillance period, 4 new cases of familial TSE were classified, while no further iatrogenic CJD and no cases of vCJD were identified in Australia.
Based on the 608 definite and probable cases, TSE incidence peaks at 4.9 cases per million per year in the 65–69 year age group, an incidence almost 5 times the reported global incidence. As sporadic cases comprise the majority of cases, the peak incidence in this group closely aligns with overall TSE rates. Since the last surveillance period, the median age at death for sporadic cases has remained unchanged; 66 years (males, 65 years; females, 67 years), the proportion of female sporadic cases has remained consistent at 53% and their median duration of disease similar at 4 months. A slightly shorter disease duration is observed in males (median, 3 months). Genetically determined TSEs have a younger age at death (medians, overall 59 years; males, 51 years; females, 62 years) and longer illness duration (medians, overall 6 months; males, 4 months; females, 7.5 months) when compared to sporadic cases. The sex ratio for familial cases is slightly biased towards females with 58% of the cases being female.
Based on all definite and probable TSE cases, the average, age-adjusted mortality rate in Australia for the 1970–2009 period is 0.88 deaths per million per year. A restriction of the timeframe to the prospective surveillance period of 1993–2009 provides a more reliable representation of the true national figures with a mortality rate of 1.18 deaths per million per year. By individual state and territory, the TSE mortality rates (Table 2) in some jurisdictions have altered since the previous update.3 Notably, the average age-standardised mortality rates during 1993–2009 have increased in Western Australia, the Australian Capital Territory and the Northern Territory, while in Tasmania, the rate has continued to decline and is currently around half the rate observed in Victoria and Western Australia (Table 2). This strongly suggests case under-ascertainment in Tasmania for this specific period. In the remaining states and territories, no significant changes in the longer-term mortality rate average (1993–2009) have occurred since last reported.
Analysis of incidence rates by state and territory over the last decade highlights the strengths and weaknesses of surveillance in the various regions (Table 2). In Victoria, case numbers have remained constant over this period and have resulted in the highest mortality rate in Australia. In contrast, a lower than expected mortality rate in Queensland, South Australia and Tasmania was observed. This decline was concerning as the 10-year timeframe provided us with a recent snapshot of confirmed cases, excluding the diluting influence of the earlier prospective surveillance years. Broadened surveillance and diagnostic responsibilities, changes to privacy legislation around 2000 and less accessible autopsy services in various regions may have contributed to the lower mortality in the specific states.
Suspect case notification between this and the previous reporting period have remained stable in the Australian Capital Territory, the Northern Territory, South Australia and Western Australia (Figure 3). In contrast, a 70% increase in notifications has been observed in Victoria. While the total number of CSF referrals arising in Victoria has remained unchanged, there has been an increase in the number of clinically suspect TSE cases added to the Register, derived from CSF referrals. Marked declines in notifications in the large populations of New South Wales and Queensland, and to a lesser degree in Tasmania, were observed in 2008. The impact of these lower notifications may be reflected in a reduced number of confirmed CJD cases for this period. A contractual agreement between the ANCJDR and Queensland Health Department was established in May 2008 to evaluate all cases of suspect TSE. The impact of this agreement on TSE incidence in Queensland will be of particular interest.
Analysis of a potential Creutzfeldt-Jakob disease cluster During 2008, the ANCJDR published findings from an investigation conducted assessing an increased number of sporadic CJD cases within a coastal region of New South Wales during the period 1993–2006.4 Statistical analysis identified a spatially significant cluster in 3 contiguous statistical local areas, consisting of 13 definite and 1 probable CJD case. An epidemiological review of ANCJDR case data for the 14 cases did not reveal a plausible cross-over or point source transmission event to explain the cluster of cases.
One potential hypothesis for the significant finding related to the region's clinicians and their management of potential CJD cases. To investigate this theory, further evaluation was undertaken comparing the regional area with the entire state, emphasising rates of referrals for 14-3-3 CSF testing, rates of case notification to the ANCJDR and suspect CJD post-mortem rates. These observations were chosen to objectively quantify an intensity of surveillance and how this relates to incidence rates.
Our analyses demonstrated that the cluster area maintained a higher level of surveillance and clinical awareness compared with the entire State of New South Wales. The population-based rate of notification of all suspect cases to the ANCJDR was 68% greater in the cluster area than for New South Wales (age-adjusted RRMH: 1.68, 95% CI=1.36–2.10) and similarly, the population-based rate of request for CSF testing was 59% greater than the state referral rate (age-adjusted RRMH: 1.59, 95% CI=1.25–2.02). No difference between the likelihood of a suspect case being confirmed as probable or definite CJD (all types or sporadic only) was observed, suggesting that once CJD was questioned as a diagnosis in a clinical setting, the likelihood of a case being assessed for CJD classification was no different in the circumscribed area to the entire state. In contrast, a difference did exist in the proportion of cases that were assessed by neuropathological examination (biopsy or autopsy), with the cluster area having an almost two and half times greater neuropathological examination rate in suspect cases compared with New South Wales (age-adjusted RRMH, 2.34, 95% CI=1.56–3.51). Simply stated, approximately double the intensity of surveillance translated to a doubling of the CJD incidence rate.
One of the distinguishing features of the 14 cluster cases provided another key piece of supporting evidence for enhanced surveillance. The cohort displayed a significantly older age at death when compared with sporadic CJD cases from New South Wales and Australia overall. Analysis of autopsy data in Austria, where autopsy of all suspect CJD cases is mandatory, suggests global under-ascertainment of older age CJD cases.5 Hence, the ability to detect older and less typical cases in this cluster region suggests clinicians manifested a greater than usual suspicion of CJD and atypical presentations.
These findings have provided us with a hypothesis that intensity of surveillance for rare disorders can be quantified and this can positively correlate with higher incidence. It further suggests that the true incidence of CJD in Australia may be almost twice the currently observed average rate of 1.18 cases per million per year. A further exploration of this hypothesis is needed within and between individual nations and may give us an improved understanding of methodologies for optimal surveillance for rare conditions such as CJD.
Surveillance of Creutzfeldt-Jakob disease in Australia: 2008
Do Reports of sCJD Clusters Matter? January 13, 2004 JP Morgan North American Equity Research
There have been seven cases of human sCJD clusters identified in the US in the last 15 years, in which people in a specific location were diagnosed with sCJD, resulting in rates between 2.1 and 8.4 deaths per million people for that specific location compared with the national average of one in 1 million. · There is no proven link between sCJD and BSE, and hence it is considered a different disease from vCJD (which has been linked to BSE). However, the existence of clusters raises the question of "contamination" or "infection", and also raises the hypothesis that rather than cases of sCJD these might have been cases of vCJD. · Clusters are not spontaneous, they normally have a source. Moreover, some cases of sCJD may have been improperly diagnosed as Alzheimer's.· We continue to believe that as long as no further cases of BSE-positive cows are found in North America and the industry has respected the 1997 ban on animal feed for live cattle, beef consumption in the US will not suffer. · Moreover, due to political pressure we expect key overseas markets (Japan, South Korea, and Mexico) to open up to US beef in the next six months - the recent 20% drop in cattle prices can be attributed mainly to these import bans. · However, two concerns linger and should be kept in mind by investors, 1) Has the 1997 ban on animal feed for live cattle been honored by the beef industry? 2) Can clusters of cases of sporadic CJD (or sCJD) really be a variant of CJD and indeed be linked to BSE? In this note we focus on the issue of sCJD clusters, and the potential impact that the growing debate on clusters could have on beef consumption in the US. United States Foods
Pablo E. Zuanic(1-212) email@example.comChristopher
M. Bledsoe(1-212) firstname.lastname@example.orgDaniel
State of Our Views Regarding BSE in the US We continue to believe that as long as no further cases of BSE-positive cows are found in North America and the industry has respected the 1997 ban on animal feed for live cattle, beef consumption in the US will not suffer. Moreover, due to political pressure we expect key overseas markets (Japan, South Korea, and Mexico) to open up to US beef in the next six months - the recent 20% drop in cattle prices can be attributed mainly to these import bans.
· However, two concerns linger and should be kept in mind by investors, 1) Has the 1997 ban on animal feed for live cattle been honored by the beef industry? The government's General Accounting Office says it has not; 2) Can clusters of cases of sporadic CJD (or sCJD as it is commonly known) really be a variant of CJD and indeed be linked to BSE (vCJD is the scientific term for the disease linked to mad cow)?
· In this note we focus on the issue of sCJD clusters.
· Do sCJD Clusters Matter?
· The media focus (and as a result, the consumer at large) since December 23, thus far, has been on the potential of new BSE-positive cows being found, and on the various initiatives the authorities are taking to prevent an outbreak of BSE. However, the apparent existence of sCJD clusters in the US has received little publicity. If sporadic (or spontaneous) CJD is really spontaneous, it should not be found in population clusters. The fact that it indeed has been found in clusters raises concerns.
· Understanding the "Difference" Between sCJD and vCJD
· Prior to 1996 there was only one known type of CJD, and it was called "sporadic" or "spontaneous" because it was unclear where it came from, or how it was generated. In 1996 scientists in England "discovered" a "variant" of CJD (vCJD), which they indicated could be linked to the animal form of the disease (BSE or Mad Cow). Experts kept vCJD separate from sCJD because unlike the new vCJD the original sCJD could not be directly linked to BSE. However, not enough is known to be fully certain that sporadic CJD is truly spontaneous and has no external catalyst. The other notable difference between vCJD and sCJD is the incubation period. Whereas sCJD has an average incubation period of 40 years and is exceedingly rare in young people, vCJD can affect people of all ages and has a much shorter incubation period of just two to five years. An even more relevant difference is that sCJD is found in 1 out of 1 million people per annum, or 5,000 cases per year on a global basis, while only 180 human cases of vCJD (the type of CJD linked to BSE) have ever been reported.
· Existence of Clusters of sCJD May Imply They Are Really Cases of vCJD There have been seven sCJD clusters identified in the US in the last 15 years, in which people in a specific location were diagnosed with sCJD, resulting in rates between 1.2 and 8.4 deaths per million people for that specific location compared with the national average of one in 1 million. The existence of clusters raises the question of "contamination" or "infection", and also raises the hypothesis that rather than cases of sCJD these might have been cases of vCJD. Clusters are not spontaneous, they normally have a source.
· A cluster consists of two statistical improbabilities: 1) multiple cases occurring in a relatively limited geographic area, and 2) multiple cases occurring within the same time period. The most recent cluster was found in Cherry Hill, New Jersey. The others have been found in Lehigh, Pennsylvania (1986-90), Allentown, Pennsylvania (1989-92), Tampa, Florida (1996-97), Oregon (2001-02), and Nassau County, New York (1999-2000). Given that sCJD occurs randomly in one out of one million cases, it is a statistical rarity to find an sCJD cluster - let alone six. The following tables highlight known clusters in the US.
· Table 1: Clustered sCJD Deaths
· Local sCJD Deaths
· Time Span State Local Area Pop. (MM) Period (mo.) Total Ann'lized 1986-1990 PA Lehigh Valley 0.5 48 18 4.5 1989-1992 PA Allentown 2.5 36 15 5.0 1996-1997 FL Tampa 2.2 18 13 8.7 1996-1999 TX Denton .01 38 4 1.3 1999-2000 NY Nassau County 1.3 12 7 7.0 2001-2002 OR Entire State 3.4 24 14 7.0 2000-2003* NJ Cherry Hill Area 1.7 36 12 4.0 Source: JPMorgan.
The second table, below, shows what portion of the state's total expected sCJD cases (as based on a one per million occurrence) were found in the local cluster, comparing the local cluster's portion of cases with the local area's portion of the state's total population. The greater the factor between the former and the latter suggests a higher statistical improbability that the cluster is spontaneous (sCJD).
· Table 2: Clustered sCJD Deaths vs. Expected State Cases
· Annual Statewide Local Area (% of:)
· Time Span State Local Area sCJD Deaths* exp. state cases state pop. 1986-1990 PA Lehigh Valley 11.9 37.8% 4.5% 1989-1992 PA Allentown 12.0 41.7% 20.8% 1996-1997 FL Tampa 14.1 61.5% 15.7% 1996-1999 TX Denton 20.9 6.1% .02% 1999-2000 NY Nassau County 18.1 38.7% 7.4% 2001-2002 OR Entire State 3.4 205.9% 100.0% 2000-2003* NJ Cherry Hill Area 8.0 50.0% 21.6% * *State cases are extrapolated based on state population and the 1 per million national average. Source: JPMorgan.
The CDC Is Currently Investigating the New Jersey Cluster The US Center for Disease Control (CDC) has opened an investigation into a cluster of deaths in an area surrounding Cherry Hill, New Jersey. Specifically, after dismissing the case when it was first brought to their attention earlier in 2003, the agency has since reversed course and on December 31, 2003 sought out information from Janet Skarbek. Skarbek, a Cinnaminson, New Jersey CPA believes she has uncovered a common link between seven deaths in the local area and a restaurant at the now-closed Garden State Race Track. All of the deaths had first been identified as the randomly occurring (one out of one million) cases of sporadic Creutzfeldt-Jakob Disease (sCJD), and six of the deaths occurred between 2000 and 2003.
Science Expanding its Knowledge of CJDs
Have Cases of sCJD Been Overlooked?
Dr. Omar Bagasra believes that a 29 year old that died of presumably sCJD in the New Jersey cluster may have died from a new, mutated form of CJD since sCJD has a typical incubation period of 40 years and is limited to elderly patients in almost all cases. Moreover, he suspects that the link between the seven local deaths (clustered geographically and chronologically) indicates that the new form of the disease is caused by some external catalyst, unlike the randomly occurring sporadic CJD (sCJD). He adds, though, that there may actually have been other unreported CJD-related deaths in the area since the disease is often misdiagnosed as Alzheimer's.
Diagnoses of Alzheimer's Might Have Been Cases of CJD
Lawrence Schonberger, the CDC epidemiologist who contacted Janet Skarbek on December 31, is quoted separately as saying that sCJD is underreported on death certificates, and that about 14 percent of cases are missed. In fact, due to similarities between sCJD and Alzheimer's disease, a 1998 Yale study found that as many as 13 percent of Alzheimer's deaths are actually sCJD, but conservative estimates place this number closer to 1 percent. If we extrapolate this finding to the 50,000 Alzheimer's deaths each year in the US, the number of actual sCJD deaths per year is somewhere between 500 and 6500. But, for us this raises additional questions, since at a rate of one per million, the US should not experience much more than 300 sCJD deaths in a single year. Furthermore, Alzheimer cases have grown 50-fold in the last 25 years from 857 cases in 1979 to 50,000 cases in today (albeit part of the increase could very well be attributed to improvements in reporting).
Can sCJD Be Caused by External Agents?
A recent study out of Imperial College in London has led some to believe that the same prions that cause the BSE-related vCJD may also cause a disease that manifests itself in a way that more closely resembles sporadic CJD. John Collinge, the scientist that conducted the experiment, is basing this assertion on findings in the study's mice, which were injected with BSE prions. As expected, some of the mice developed symptoms from vCJD, but unexpectedly, others suffered from symptoms that more closely resembled sCJD. If true, the implications are significant, as it will force scientists to consider whether cases of sporadic CJD may actually have been caused by consumption of contaminated beef.
Does All This Matter?
For now we await the results of the CDC investigation of the New Jersey cluster. Previous investigations have found clusters to be just coincidences. While this may be the case, we believe that the media may start focusing more on the issue of clusters, and that the debate could raise consumers' concerns about beef. But even in the worst case scenario that these clusters were indeed linked to BSE, one could still make the argument that these cases were generated before the 1997 ban on animal feed for cattle was imposed, and that hence chances of contamination since then are unlikely. Still, the question lingers, has the 1997 ban been respected? Will consumers concerns increase as the discussion of CJD clusters hits the national media? Bottom Line: If no new cases of BSE-positive cows are found and the issue of CJD clusters is disregarded by consumers, then the effect on beef consumption will be negligible. On the other hand, new cases of infected cows and/or a wider debate of CJD clusters could indeed have an effect on beef sales. ...END
Subject: J.P. MORGAN INVESTORS WARNED ABOUT POTENTIAL SPORADIC CJD AND BSE RELATION From: "Terry S. Singeltary Sr."
BSE (Mad Cow) Update: Do Reports of sCJD Clusters Matter?
snip... see full text ;
Thursday, November 05, 2009
Incidence and spectrum of sporadic Creutzfeldt-Jakob disease variants with mixed phenotype and co-occurrence of PrPSc types: an updated classification
Friday, October 23, 2009
Creutzfeldt-Jakob Disease Surveillance Texas Data for Reporting Years 2000-2008