Early warning for Lyme disease: Lessons learned from Canada Nick Hume Ogden, National Microbiology Laboratory @ Saint-Hyacinthe Talk outline The biology of Lyme disease emergence in the context of climate change Early warning by model-based assessments of risk Early warning by active field surveillance Early warning by passive tick surveillance 2 1
THE BIOLOGY OF LYME DISEASE AND CLIMATE CHANGE 3 4 Lyme disease in North America Bacterial infection caused by Borrelia burgdorferi s.s. Transmitted amongst wild animals hosts (rodents, squirrels, birds etc) by ticks (Ixodes scapularis and I. pacificus), which also transmit infection to humans Causes mild disease initially: flu-like with classic skin lesion (erythema migrans) Then develops into more severe disease Lyme arthritis, neuroborreliosis and heart block 2
Activity proportion Days to oviposition Days to eclosion 25 2 15 1 5 18 16 14 12 1 8 6 4 2 R 2 =,6582 5 1 15 2 25 3 Temperature R 2 =,8283 5 1 15 2 25 3 Temperature Days to moult Days to moult Diversity of vectors and vector-borne diseases Insects: Mosquitoes: Malaria, Dengue, Chikungunya, West Nile virus Fleas: Bartonellosis, Plague Bugs: Chagas Deer flies: Tularaemia Blackflies: Onchercerciasis Ticks Hard-bodied (Ixodid) ticks : Lyme, Babesiosis, Anaplasmosis, Ehrlichiosis, Powassan, Deer-tick virus, Rocky Mountain Spotted Fever Soft-bodied (Argasid) ticks: Relapsing fever 5 Impact of climate on vector and vector-borne diseases 25 y = 13,1x -1,4278 2 15 y = 11179x -2,5468 R 2 =,8833 1 5 y = 34234x -2,279 5 1 15 2 25 3 1 Temperature 9 8 y = 1595,8x -1,282 7 R 2 =,8268 6 5 4 3 2 1 5 1 15 2 25 3 Temperature Randolph & Rogers Nature Rev Micro 23 Ogden et al. J. Med. Entomol. 24 1.9.8.7.6.5.4.3.2.1 Immatures Adults 5 1 15 2 25 3 Temperature ( o C) Ogden et al. Int. J. Parasitol. 25 Extrinsic incubation period 6 3
Ticks versus mosquitoes LC = 4 weeks TC = 3 weeks LC = 2.5 years TC = 1 year 7 Ticks versus mosquitoes LC = 4 weeks TC = 3 weeks LC = 2.5 years TC = 1 year 8 4
Ticks versus mosquitoes LC = 4 weeks TC = 3 weeks LC = 2.5 years TC = 1 year 9 Ticks versus mosquitoes LC = 4 weeks TC = 3 weeks LC = 2.5 years TC = 1 year 1 5
Number of reported cases Vector lifecycle characteristics determine disease patterns Lyme disease Slow inexorable spread then more or less constant risk West Nile virus Rapid spread then epidemic behaviour Year 11 Climate change Warming (especially for us!) Long term change in rainfall patterns Climate variability and extreme weather events 12 6
Emerging/re-emerging vector-borne disease risks for Canada Warming & climate variability: Epidemics/re-emergence of endemic VBD WNV, CSV, Warming in Canada Climate change abroad: Introduction of exotic VBDs Chikungunya, Zika, Dengue, RVF, JE, Malaria Warming in North America: Northward spread of Lyme, EEE, La Crosse, HME 13 Early warning for tick-borne versus mosquito-borne disease Our needs for early warning for these vector-borne diseases differ Mosquito-borne: Given suitable environmental conditions - rapid spread after invasion Followed by endemic state and Weather-driven epidemics Need to know WHERE and WHEN risk occurs Timescale for early warning is SHORT (next week, next month) Tick-borne: Given suitable environmental conditions slow spread after invasion Followed by endemic state Some weather driven interannual variations in risk Need to know WHERE and WHEN risk occurs Timescale for early warning is LONG (months to decades) 14 7
EARLY WARNING MODEL- BASED RISK ASSESSMENTS 15 Assessment of Lyme disease risk emergence in Canada Lyme disease risk 22 I. pacificus I. scapularis 16 8
Key determinants of Lyme disease risk Suitable habitat for ticks: assessed by field study (Ogden et al. JME 26a) Suitable host densities: assessed previous field studies Dispersal of population-seeding ticks into Canada by migratory birds: assessed by surveillance/field study (Ogden et al. JME 26b, AEM 28) Temperature threshold for tick population persistence: obtained by simulation modelling (Ogden et al. 25) Algorithm using temperature from GCMs and tick dispersion developed and mapped Photo by Bill Hilton Jr (www.hiltonpond.org) qe Eggs No. eggs per adult (e) hl PEP (q) Hardening Larvae POP (x) z ql Questing Larvae Basic HFR: Immature ticks Temperature Basic HFR: Adult ticks Rodent Nos. ql fl Feeding Larvae r L to N (s) Rodent Nos. N to A (v) el Engorged Larvae Questing Nymphs qn Feeding Nymphs qn fn u Egg-laying Adults y Engorged Feeding qa Questing Adults Adults Adults w qa el fa Deer Nos. Engorged Nymphs el 17 Projected distributions of Ixodes scapularis year 2 High risk Moderate risk Low risk Risk of birdborne ticks Ogden et al. Int. J. Health Geogr. 28 18 9
Projected distributions of Ixodes scapularis year 22 High risk Moderate risk Low risk Risk of birdborne ticks Ogden et al. Int. J. Health Geogr. 28 19 Projected distributions of Ixodes scapularis year 25 High risk Moderate risk Low risk Risk of birdborne ticks Ogden et al. Int. J. Health Geogr. 28 2 1
Projected distributions of Ixodes scapularis year 28 High risk Moderate risk Low risk Risk of birdborne ticks Ogden et al. Int. J. Health Geogr. 28 21 Validated by surveillance Risk map validation Ogden et al. IJHG 28 Field surveillance Ogden et al EHP 21 Passive surveillance Leighton et al J App Ecol 212 ON Montreal NY VT NH 1 2 3 4 5 6 Index of certainty for the occurrence of I. scapularis populations in the field Human case surveillance 29-212 Ogden et al CCDR 22 11
Number of CSD with tick populations 24 214 23 Spatiotemporal coincidence of I. scapularis invasion in Canada with warming evidence of VBD emergence event with climate change 6. 5. 4. R using ANUSPLIN data R using CRCM4.2.3 Number of CSDs with tick populations 12 1 8 R 3. 6 2. 4 1. 2. 197 198 199 2 21 22 23 24 25 26 27 Ogden et al. Environ Health Perspect 214 Leighton et al J App Ecol 212 There has been an increasing number of cases of Lyme disease in Canada, and Lyme disease vectors are spreading along climate-determined trajectories (Koffi et al., 212; Leighton et al., 212). UN-IPCC AR-5 12
Prevalence EARLY WARNING BY SURVEILLANCE 25 Surveillance for emerging tick populations provides about 5 years early warning depending on tick seasonality in source populations.24.2 Within Cluster Outside Cluster Model predicted.16.12.8.4 22 24 26 28 21 Year Significant Lyme disease risk (B. burgdorferi infection prevalence > 5%) emerges in new tick populations in about 5 years in eastern Canada But emerges in one or so years in central Canada Due to number of incoming engorged larvae on hosts in spring: MW > NE Ogden et al., J Appl Ecol 213 13
Active surveillance method 1. Field surveillance for ticks Dragging a 1m 2 flannel across the woodland floor using standardised time/distance/pattern +/- rodent capture Ticks and rodent samples tested for tick-borne pathogens by PCR Advantages: Gold standard method Immediate result Identifies risk to the public (what s jumping on the drag = what could jump on people) Detects ticks all through the season (adults, then nymphs, then larvae, then adults) Disadvantages: Expensive need field teams Geographically limited 27 Lyme disease risk in Canada by active field surveillance Ogden et al. CCDR 214 Bouchard et al. Can Vet J 215 Gabriele-Rivet et al. Plos One 215 28 14
Ontario study: Clow et al Zoonoses & Public Health 216 14 sites visited May- Oct 214 I. scapularis present at 2 sites (triangles) 66 ticks total = 1 tick /person-hour B. burgdorferi present at 5 sites I. scapularis- and B. burgdorferi-positive sites spatially clustered Low tick density and low infection prevalence surveillance is early warning 29 The passive tick surveillance program Passive surveillance for I. scapularis has occurred in Canada since 199 Ticks collected from patients at, submitted from, medical and veterinary clinics Ticks collection and species identification by P/Ts (NFL, NB, QC, ON, MB, SK, AB - PCR for Borrelia burgdorferi at PHAC National Microbiology Lab (NML) Passive surveillance data: Provide a long dataset (199 to present) Have a wide geographic coverage Are sensitive but non-specific due to detection of bird-dispersed ticks (particularly by ticks from dogs) 3 15
Prevalence Obtaining early warning signals from passive tick surveillance data Ogden et al Environ Health Perspect 214 Alert maps: Koffi et al J Med Entomol 212.24.2.16 Within Cluster Outside Cluster Model predicted Leighton et al. 212 J Appl Ecol.12.8.4 22 24 26 28 21 Year Cluster analysis: Ogden et al Environ Health Perspect 21 31 Human case surveillance Ogden et al CCDR 215 32 16