During the 10 years 1968–78, eight fatal cases of primary amoebic meningoencephalitis occurred in the Waikato area. All had been swimming at places along the primeval course of the Waikato River between Taupo and Matamata.

This paper reports a ninth fatal case of primary amoebic meningoencephalitis, involving a 10-year-old girl who swam at Okauia Springs in one or more of the ‘Opal Springs’ pools on the west bank of the Waihou River (just opposite the ‘Crystal Springs’ pool area) over Easter (21–24 April 2000), as well as in natural warm pools in the Rotorua area during the same period.

In late April 2000, a 10-year-old from Waipa District was notified to the Medical Officer of Health of Waikato District Health Board, with primary amoebic meningoencephalitis. In the week prior to the onset of her symptoms, the child had swum in two geothermal pools, one a natural, undeveloped spring near Lake Rotoma, and one part of a commercial pool complex in Matamata. She was known to have put her head underwater in both locations, as did many others present at the same times. Both exposures occurred during Easter. There had been no other exposures to thermal or other waters during the incubation time or during the two weeks of school holidays, which had immediately preceded Easter. The presenting symptom was headache, which began three and four days after the two swimming exposures. The headache was severe and persistent, with fever and vomiting developing over 48 hours. There was no rash. Clinically the child presented as a case of bacterial meningitis with pleocytosis and an elevated protein. Clinically she was thought to have pneumococcal meningitis and was treated with ceftriaxone and vancomycin. Her continued clinical deterioration and reduction in level of consciousness prompted a review of the diagnosis and consideration of less common causes of meningitis. Motile amoebae were seen in cerebrospinal fluid (CSF), and have since been confirmed by culture and polymerase chain reaction (PCR) as Naegleria fowleri. The child died three days after admission despite intensive therapy with intravenous and intrathecal amphotericin B, and intravenous rifampicin and co-trimoxazole.

At the commercial complex she had spent several hours in a chlorinated Olympic pool and used a playground-style slide into the pool repeatedly. She was thought by her family to have spent some time in a second pool at the complex, which has a warning posted alongside it advising swimmers not to put their heads underwater. The water in this pool received no treatment other than sand filtration. The family spent about an hour in the natural soda springs pool the following day. It was raining, and about 20–30 others were using the pools at the same time. Other members of the family reported putting their heads underwater, as did other pool users. After confirmation of the diagnosis, the public health response involved immediate public notification of the incident and the known risk factors via the media, and review of both pool sites. Follow-up actions included liaison with territorial local authorities, which are responsible for safe management of all swimming pools in their areas.

The pool complex closed voluntarily while investigations were undertaken. The Olympic pool at the commercial complex had, before the child’s illness was notified, been drained and the surface stripped for repainting. It is filled in series after the two smaller pools are filled. Water from the Olympic pool could not therefore be tested for amoebae. The Olympic pool was, at the time of the case’s exposure, being tested for free available chorine (FAC) once daily before bathers arrived; at which time it was described as always being at 2 parts per million (ppm). It now has a different inflow, an automatic chlorine-dosing system, and a separate sand-filtration system serving this pool only. It was reopened after repainting and an intensive phase of monitoring of FAC, to ensure that levels remain around the minimum acceptable level of 2 ppm throughout the day. More regular monitoring of FAC is now in place as part of a detailed management plan. The two smaller untreated pools at the complex are maintained at a higher temperature than the chlorinated Olympic pool, and are not chlorinated. The water passes through a sand filter and the pools are filled in series, with the Olympic pool filling last. Both filters are backwashed (twice) daily, with backwash water running to waste. Both smaller pools had extensive cracks in the concrete floor. In the deeper of the two pools, a multiply fissured crack of approximately 10–20 cm depth ran approximately five metres across much of the floor. It contained paint flakes and other loose debris. The water in the Olympic pool was exposed to these cracks during filling. Four samples were taken for testing for amoebae, from the bore water, the water after sand filtration, the debris from the crack and from a separate bore feeding a private pool not used by the case. Only the bore water for the main pool complex was free of amoebae, but none of the isolates were Naegleria fowleri. The pool owner reported that the cracking of the two smaller pools in the main complex was thought to be due to intrusion of groundwater from below, as the cracks are known to have been repaired many times. The pools are thought to have been built in the early 1900s. Repairs since undertaken involve the construction of a new floor above the existing one, with space and drainage to prevent groundwater build up.

A wet mount of the CSF was examined using phase contrast microscopy at 400X magnification. Three hundred micro litres of the CSF was cultured on a 0.25% NaCl agar plate onto which a lawn of Escherichia coli had been spread. The plate was incubated in a humidified atmosphere at 37 °C for 24 hours; a further wet mount was made from the plate and examined for the presence of Naegleria-like amoebae. The plate was then flooded with 5 ml of sterile water and incubated at 37 °C for 90 min. Following incubation the plate was examined for the presence of the flagellate stage using a phase contrast inverted microscope at 200X magnification. Amoebic isolates were subjected to DNA amplification. Two sets of primers p3f (5-gctatcgaatggattcaagc) p3r (5-cactactcgtggaaggctta)1 and primers B1f (5-atgcagtagtttgggcg) and B2r (5-actgtgatatttcatcattg)2 were used for DNA amplification. Briefly, DNA was extracted from the CSF specimen and amoebic trophozoites from the saline agar plate, as previously described.3 The PCR solution contained 1X PCR amplification buffer (10 mM Tris-HCL pH 8.3, 50 mM KCL, 2.5 mM MgCl2), 200 μM deoxynucleoside triphosphates, primers at 0.4 μM, 1.25 units of amplitaq Gold, and 5 μl of extracted DNA. Template DNA was denatured at 95 °C for 10 minutes. A total of 40 cycles was performed, during which the DNA was denatured at 94 °C for 30 sec and primers annealed at 55 °C for 30 sec and extended at 72 °C for 1.5 min. The PCR B1/B2 products from the CSF and culture plates were cycle sequenced in both directions using Big-Dye chemistry with the respective forward and reverse primers. The PCR negative controls included sterile water as a mock template control for the amplification reaction mix and N. gruberi DNA as a control for primer specificity.

One litre of water from the two pools (ie, Okauia Springs natural pool and commercial pool complex) was filtered through a 5 micron filter and the filter then inverted onto a saline agar plate containing a lawn of E. coli. The agar plates were incubated at 37 °C for up to 10 days and any amoebic isolates tentatively identified via morphology and flagellation. Speciation of Naegleria sp. was performed via amplification of the amoebic ribosomal internal transcribed spacers (ITS) using the primer set ITSf 5-gaacctgcgtagggatcattt and ITSr 5-tttcttttcctccccttatta.4 DNA amplification and sequencing conditions were as reported above.

The CSF looked purulent (WBC 1590 x 106/l, 90% neutrophils, protein 2.4 g/l, glucose 4.3 mmol/l). Examination of the wet mount from the CSF revealed a small number of amoeboid trophozoites that exhibited characteristic eruptive flowing of the protoplasm typical of Naegleria spp. Culture of the CSF on saline agar plates also resulted in the growth of typical limax amoebae. The flagellation test supported the provisional diagnosis of Naegleria spp. Amplification of the DNA extracted from amoebae isolated from culture produced two different sized amplicons: 1.5-kbp using primer set p3f/r and 678-bp using primer pair B1f/B2r. DNA extracted from the CSF and amplified with primer set B1f/B2r also produced a 678-bp amplification product. A BLAST5 search of the sequence results for both the CSF and the culture B1f/B2r amplification products identified N. fowleri with a 100% nucleotide sequence match. Examination of the thermal pool waters revealed no N. fowleri isolates either by culture or DNA amplification. However, a thermophilic Naegleria sp. was isolated from the Okauia Springs which was later identified as Naegleria lovaniensis via sequencing of the ribosomal internal transcribed spacer amplification product.

Amoebic meningoencephalitis has an incubation time of 3–10 days, and presents with headache, fever, vomiting and later reducing consciousness. In the early stages it cannot be clinically distinguished from bacterial meningitis, and the advice to give parenteral penicillin or other antibiotic effective against meningococci still applies to any presentation suggestive of meningitis.

The disease had not been reported in New Zealand since 1978.6 Eight cases have been notified previously in New Zealand. All cases followed swimming in geothermal water between Lake Taupo and Matamata; all have been fatal.

The first four of these cases swam in 1968 in the warm Okauia Springs pool on the east bank of the Waihou River, known as ‘Crystal Springs’. Amoebae pathogenic to mice were identified from the third of these cases.7 Because the microbiologist identified an amoeboid slime mould in the CSF culture of this case, he considered the slime mould to have been the pathogen.8 More likely, an amoeba of the genus Naegleria was the true culprit, with the slime mould as a contaminant.9 A fifth case involved a 12-year-old boy who swam in the same pool in 1971. His CSF revealed amoebae on microscopy and culture. Immunoperoxidase staining of brain tissue at Massey University, Palmerston North, of both the third and fifth cases demonstrated the amoebae to be Naegleria fowleri.10 In addition, specific immunofluorescence of brain sections of these two cases at the Amoebiasis Diagnostic and Research Unit, London, identified the amoebae as N. fowleri (personal communication between WP Stamm, Amoebiasis Diagnostic and Research Unit, St Giles Hospital, London, UK, and CM David, Director of Laboratory Services, Waikato Hospital, Hamilton, NZ, 1976). The sixth case, a 21-year-old male, visited the ‘Golden Springs’, near the Mihi bridge of the Waikato River in 1972. The amoeba was identified at the Institute of Medical and Veterinary Science, Adelaide, serologically and by mouse pathogenicity tests, as N. fowleri.11 The seventh case, a 15-year-old boy, bathed in the warm Otumuheke stream at its entry into the Waikato River above the Huka Falls. Amoebae pathogenic to mice were identified at Massey University, Palmerston North, as N. fowleri.12 The eighth case, an adult male, related again in 1978 to the Okauia Springs pool near Matamata, called the ‘Crystal Springs’; which has now been closed.6

The natural pool may be the more likely source of this infection, because of direct contact between the soil and the water. Warm water and soil are the natural habitat of Naegleria fowleri. Public health management of this incident has focused on promoting safe behaviour in natural geothermal pools (ie, users should not dive or jump in, and should not immerse their faces in any way) and safe management of commercial pools. Assiduous exclusion of soil from the water source and the pools, filtration, adequate and sustained chlorination and/or high water turnover are the mainstays of safe commercial pool operation.

Ensuring that swimming pools do not present a risk to health is the responsibility of city and district councils. Pools run according to the guidelines for geothermal pools (which are part of the NZ Standard for swimming pool operation) are at very low risk of harbouring N. fowleri. There are at least 14 species of the free-living Naegleria identified, two of which, N. fowleri and N. australiensis, are regarded as pathogenic.4 Although N. australiensis is virulent for mice there have been no recorded infections of man with this species. N. lovaniensis on the other hand resembles N. fowleri for growth at temperatures up to 45 °C, cytopathogenicity for tissue culture, and antigenicity. Historically mouse pathogenicity was used to differentiate between these two species, but the recognition that N. australiensis was pathogenic to mice, albeit less so than N. fowleri rendered the test non-specific.1 Consequently genotypic methods have been developed for the reliable speciation of this genus of free-living amoebae.4 Using sequence variation in the ribosomal internal transcribed spacer region, N. fowleri, N. lovaniensis, N. australiensis and N. gruberi have all been identified in thermal waters in New Zealand (unpublished observations).

This case illustrates the continuing threat of amoebic meningoencephalitis in New Zealand, and describes the use of modern molecular methods in its diagnosis.

Author information: Ray T Cursons, Lecturer, Department of Biological Sciences, University of Waikato; Jamie W Sleigh, Professor of Anaesthesia and Intensive Care, Intensive Care Unit; Dell Hood, Consultant, Department of Community Medicine; David Pullon, Retired Paediatrician, Waikato Hospital, Hamilton

Acknowledgments: The laboratory work was in part funded by the Waikato Medical Research Foundation.

Correspondence: Dr Ray T Cursons, c/o Department of Biological Sciences, University of Waikato, Hillcrest, Hamilton. Fax: (07) 838 4324; email: r.cursons@waikato.ac.nz