Nipah virus infection in humans in Kerala, India: Hypothesis of air-borne transmission

The hypothesis and supporting information

In earlier outbreaks of Nipah in Malaysia, Bangladesh, and Kerala, respiratory transmission was the likely mechanism in most secondary, human-to-human transmission. The virus in oral secretions and other body fluids could be transmitted by inhaling droplets or aerosolised particles. Blood and other body fluids may also form droplets or aerosols that can be inhaled¹⁷. Transmission via fomites could occur through ingestion, inhalation, or perhaps inoculation by fingers into eyes, mouth, or nostrils¹⁸.

In all NiVD outbreaks in Kerala and many outbreaks in Bangladesh, primary cases lacked exposure to bats or their saliva and urine, raising the possibility of another route through which transmission occurred. Inhalation of airborne virus particles as the mechanism of transmission from bats to humans can explain exposure to NiV without contact with bat(s) or bat saliva/urine. Several factors may influence airborne NiV transmission from bats, such as bat distribution and density, viral shedding, aerosol generation, virus survival in the environment, and human exposure to viable virus particles.

Fruit bats (Pteropus species) are widely distributed across the Indian subcontinent and typically roost in large colonies on trees. Following each spillover event in Kerala, bats were sampled from affected areas and tested for viral RNA and NiV antibodies. Viral RNA was detected in bats in three of the five outbreaks, with positivity ranging up to 25 per cent among Pteropus bats. Overall, 20/647 tested bats had the presence of viral RNA (Table)^14,19-21. Despite the proximity of human dwellings to bat roosts, NiV RNA was not consistently detected in bats roosting nearest to the primary case’s residence. In the 2018 outbreak, viral RNA was detected in bats near primary case’s house, within a 5 km radius of the primary case’s residence in the 2019 outbreak, and within a 42-55 km radius of the primary case’s residence in the 2023 outbreak (Table)^14,19-21 . These findings suggest that proximity to bat roosts alone does not directly correlate with NiV infection risk, consistent with patterns observed in several Bangladesh outbreaks¹².

NiV is shed through oral secretions, urine, and faeces. Studies on the related Hendra virus (HeV) in Australia suggest that urine is the primary medium driving bat-to-bat and potentially bat-to-horse transmission. In Bangladesh, NiV RNA was detected in bat urine up to 52 days after the presumed human exposure event, although viral load had declined over time²². Virus shedding from bats is sporadic and influenced by several factors, including age (with juvenile bats shedding more virus), physiological stress (during pregnancy, lactation, and weaning), nutritional deficiencies, habitat disturbances, and co-infections²³. Viral shedding is intermittent as indicated by the seasonality of the spillover pattern of NiV in Bangladesh, Marburg virus in Uganda, and Hendra virus in Australia²³. Viral load may be an important determinant of spillover.

Bats generate aerosols through laryngeal movements during vocalisation and respiration^24,25. Urination from roosting sites may produce aerosols, especially when droplets of bat excreta mix with dust particles²⁶. Evidence from other bat-borne viruses supports the plausibility of airborne transmission. For example, aerosolization of the Marburg fever virus in caves has been associated with infection and outbreaks following visits to caves or mines²⁷. Limited evidence also suggests that airborne rabies transmission may occur in bat caves with dense bat populations and poor ventilation, although other routes cannot be ruled out²⁸. Furthermore, there is speculation about airborne Hendra virus transmission to horses via contaminated aerosols from bat secretions²⁹. This aerosolization process may facilitate the airborne dissemination of NiV.

Virus survival in the environment is essential for airborne transmission. Experimental studies indicated that HeV remained infectious in Pteropus alecto urine (pH ∼7) for over four days at 22°C, with a half-life of approximately 19 h. However, at 37°C, HeV was inactivated within one day, with a reduced half-life of three hours. In contrast, NiV incubated in Pteropus vampyrus urine (pH 2) was rapidly inactivated within thirty minutes at both 22°C and 37°C. These findings suggest that NiV’s environmental stability is limited under most natural conditions²⁷.

Given the above pieces of evidence, airborne transmission of NiV is theoretically possible if humans are exposed to aerosols containing viable virus or free virus particles. However, NiV’s limited environmental stability, particularly in acidic bat urine, may restrict the likelihood of sustained airborne spread. The occurrence of single spillover events across the six NiV outbreaks in Kerala, along with the seemingly random distribution of primary cases, aligns with the hypothesis of rare airborne transmission. Such transmission may occur at an extremely low frequency, but with recurrence. In short, the pattern of NiV spill-over events in Kerala is compatible with rare instances of airborne transmission. It accommodates both recurrence and rarity. No other transmission route has been proposed by anyone.

The way forward

Although the hypothesis of airborne NiV transmission is plausible, laboratory experiments and field investigations are required to confirm or negate this. Studies in the Biosafety Level 4 laboratory may address the virus viability decay rate at different temperatures and humidity. The assumption that stray virus particles may remain viable in the atmosphere must also be tested in the field. Large volumes of air should be sampled by filtration for capturing virus particles. For example, a study conducted in 72 South African High Schools employed high-volume air filtration to detect airborne Mycobacterium tuberculosis and analysed samples using digital droplet PCR³⁰. Similar air sampling methods for NiV around trees with bat roosts may be explored. Chang et al³¹ have described several devices used worldwide to trap airborne viruses from the environment and test them by molecular assays or paper-based strips. Such tools could be valuable in investigating the role of the airborne transmission of NiV coupled with artificial intelligence, to detect fluctuations in virus concentrations around bat roosts due to intermittent shedding, with increased shedding during nutritional stress or birthing²⁹. The experimental Syrian hamster model clearly suggests the possibility of airborne transmission of NiV³². However, laboratory studies are only suggestive in nature and cannot be directly extrapolated to humans. The known animal models for NiV are Syrian hamsters and possibly pigs, as they acted as intermediary hosts in Malaysia. Sentinel surveillance of these animals could also be another way of investigating the possibility of airborne NiV.