Lyodura use and the risk of iatrogenic Creutzfeldt–Jakob disease in Australia

Creutzfeldt–Jakob disease (CJD) is a fatal, transmissible, neurodegenerative disorder belonging to the group known as the transmissible spongiform encephalopathies (TSEs). CJD can occur without explanation (sporadic), secondary to mutations in the prion protein gene (PRNP), or as a complication of medical treatment using contaminated therapeutic agents or equipment (iatrogenic). Although corneal grafts and neurosurgical equipment have been associated with disease transmission, the most common causes of iatrogenic CJD have been treatments involving human-derived cadaveric pituitary hormones or dura mater.1

The first identified case of CJD in a dura mater recipient was reported in the United States in early 1987.2 In response, the US Food and Drug Authority issued a safety alert in April 1987, seeking immediate discontinuation of use of the identified dura mater batch (Lyodura batch #2105).3 A second patient with CJD linked to Lyodura was detected in New Zealand in 1988,4 but the specific batch could not be identified. This has remained a frequent difficulty when tracing contamination sources.

As of January 2003, over 120 CJD cases related to dura mater use had been detected globally, with 97 in Japan.5 These cases were predominantly associated with Lyodura, a commercial product produced since 1969 by B Braun Melsungen AG (based in Germany). Only a few reports suggest the possibility of CJD after use of dura mater from other commercial or non-commercial sources.6,7,8

Lyodura consists of lyophilised, irradiated human dura mater sourced post mortem. Additional processing with immersion in a solution of sodium hydroxide (1 M) was instituted in 1987, with a noticeable reduction in Lyodura-related cases thereafter.9 Lyodura has been used in a number of countries, including Australia, Japan, Canada, the United States and the United Kingdom, mainly in neurosurgery, but also in orthopaedic, otological, dental, urological, gynaecological and cardiac procedures (Box 1) (Dr L Schonberger, Assistant Director, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Ga, USA, personal communication).

Most cases of CJD associated with dura mater have occurred in Japan. In 1996, in response to the growing incidence, a survey was undertaken of Lyodura use in almost 3000 Japanese healthcare institutions. This estimated that up to 100 000 people received Lyodura grafts between 1983 and 1987,9,10 and up to 220 000 between 1979 and 1991 (out of a total of 260 000 who received dura mater grafts). Use of these grafts greatly declined after 1991 but may have continued until 1997.11,12 Assuming that all cases of CJD associated with dura mater were a consequence of Lyodura use, the overall risk of Lyodura-associated CJD in Japan is approximately 0.04%.5

Lyodura was approved by the Australian Therapeutic Goods Administration for importation and use in Australia in 1972. The product licence was withdrawn in early May 1987, shortly after recognition of the first case of CJD linked to Lyodura use.

To date, five cases of CJD have been epidemiologically linked to neurosurgical use of Lyodura in Australia. Their clinical features have already been described13,14,15 and are summarised in Box 2. Patient 1 presented in 1987, about 5 years after implantation of Lyodura. The longest incubation period was in Patient 4, who presented in 1999 after an incubation period of almost 17 years.14 The most recent (fifth) patient died in 2000. Patients 1, 2 and 4 were exposed in 1982, while Patients 3 and 5 were exposed in 1985 and 1986, respectively.

A number of studies have attempted to determine the number of people exposed to Lyodura in Australia. We collated the available information and undertook further enquiries, as a basis for estimating the risk of Lyodura-associated CJD in this country (Box 3).

Quantifying past use of Lyodura in Australia relies on data that cannot be fully confirmed. Initial estimates by the Commonwealth Department of Health and Ageing in discussion with the Therapeutic Goods Administration placed an upper limit of between 5000 and 10 000 individuals potentially exposed to Lyodura. However, from the results reported here, the true number is likely to be much smaller — probably fewer than 2500. If so, the risk of Lyodura-associated CJD is higher in Australia (0.20%–0.23%) than in other countries that have undertaken similar investigations, such as Japan.

Although the risk of Lyodura-associated CJD in Japan has appeared to fluctuate over time, two relatively stable features are a peak in contaminated grafts between 1983 and 1987, and the paucity of Lyodura-associated cases after 1987, when the manufacturer instituted effective decontamination of the tissue with a 1 M solution of sodium hydroxide.1,5,9 Ninety-seven cases associated with dura mater had been recognised to 2003,5 with predictions that further cases are likely until 2020, and that final numbers may reach 160.12 Acknowledging a total of 220 000 people exposed to Lyodura, and assuming all CJD associated with dura mater was associated with Lyodura (which is reasonable based on evidence published to date),11 then the cumulative overall risk of CJD from Lyodura in Japan is currently around 0.04% (95% CI, 0.03%–0.05%) (Box 4). Estimated risk in Australia is much higher — 0.20% to 0.23% (95% CIs, 0.06%–0.47% and 0.07%–0.53%, respectively). With the same assumptions, the risk of CJD in the higher-risk period in Japan (1983–1987) is about 0.08% (95% CI, 0.06%–0.10%), while that in Australia (1982–1986) is 0.43% (95% CI, 0.14%–0.99%). Notwithstanding the need for certain assumptions to facilitate these comparisons, the risk in Australia appears about five times greater than the analogous point estimates of risk in Japan. The reality of the differences is further supported by the lack of overlap between the confidence intervals of the calculated estimates.

Nevertheless, the risk of CJD associated with Lyodura in Australia is well below the risk associated with exposure to human-derived pituitary growth hormone (hGH) in France (Box 4). As of 1999, 55 cases of CJD had been linked to French-derived hGH, from a cohort of 1361 patients, representing an attack rate of about 1 in 25 recipients (4.0%).17 A more recent report suggested the risk could be much higher, at 81 per 1361 (about 6.0%).18

The reason for the higher estimated risk of Lyodura-associated CJD in Australia compared with Japan is not known. Perhaps most likely is chance receipt of a relatively high percentage of contaminated batches of Lyodura and use of multiple pieces of Lyodura per patient in Australia, although the latter is contrary to anecdotal recollections of Australian neurosurgeons. Alternatively, the difference may reflect better case ascertainment in Australia, which, in contrast to Japan, had an established comprehensive, prospective, national surveillance program for CJD, with international comparisons of incidence rates for sporadic CJD attesting to the adequacy of case ascertainment.14

Another potential explanation is genetic difference between the populations of the two countries. For example, homozygosity for methionine at codon 129 of the PRNP gene appears to predispose to iatrogenic CJD.1 However, contrary to expectations, this homozygosity appears more common in the Japanese population (about 92%) than in occidental populations (about 37%).19

Finally, the differences in risk may reflect greater use of Lyodura in Japanese patients with malignancies, whose survival was shorter than the lengthy incubation periods typical of TSEs, or in non-neurosurgical applications, with attendant lower transmission efficiency. Japanese studies suggest that Lyodura-related transmissions have been essentially restricted to neurosurgical procedures,10,11 and our comparative risk analysis was predicated on the assumption that most Lyodura was used in these procedures in both countries. Although some Lyodura was most likely used in non-neurological procedures, the proportion is impossible to quantify in different countries, leaving uncertainty about our risk comparisons.

In addition to neurosurgery and the non-neurological applications identified by Australian surveys, additional uses of Lyodura have been reported in the US (Box 1). Cases of CJD have been linked epidemiologically to Lyodura used in the embolisation of the external carotid artery for treating a nasopharyngeal angiofibroma,20 as well as dura mater used to embolise intercostal arteries before thoracic surgery.7 This reinforces the likelihood that non-neurosurgical use of Lyodura can result in transmission.10 Therefore, risk of Lyodura-associated CJD linked to non-neurosurgical uses appears a genuine possibility, which may be very difficult to identify epidemiologically, given the often deficient state of medical records.

The inability to clearly identify surgical uses of Lyodura, especially non-neurosurgical uses, also raises the possibility of secondary iatrogenic transmissions from Lyodura-treated patients during an extended preclinical or incubation phase. Animal models clearly support the possibility of lengthy presymptomatic periods (which may even exceed the lifespan of hosts) during which transmission is possible.21,22,23,24 Generally, infection control and other guidelines focus on the need to limit transmission risk through identifying patients who received Lyodura during neurosurgery, and do not include guidance on risk associated with other surgical applications. This approach may need to be reconsidered based on studies such as ours.