Tamiflu in the environment

1
Tamiflu in the environment

• Caroline Moermond
• Charles Bodar
• Lonneke van Leeuwen
• Mark Montforts
• Bianca van de Ven
• Suzanne Wuijts
• Monique van der Aa
• Ans Versteegh

2
What is Tamiflu?

• Tamiflu oseltamivir
• Antiviral drug which slows the spreading of viral
  cells through the body
• Registered in Europe (EMA) in 2002

3
Why a risk assessment?

• Swine flu (Mexicaanse griep) questions raised
  about environmental impact of flu-related
  medication
• In the original risk assessment for
  authorisation, pandemic use and drinkwater
  quality were not taken into account
• Advice for the ministry of Environment (VROM) at
  sept. 1, 2009

4
Properties of tamiflu

• Oseltamivir ethylester phosphate transforms in
  the body
• ? first into pro-drug oseltamivir ethylester
• ? then into active compound oseltamivir-acid
• In urine ethylesteracid 14
• Log Kow 1.21 for ethylester log Kow 0.006
  for acid
• Very soluble, low sorption to organic matter
• Almost no hydrolysis or photolysis
• DT50 in water-sediment system is 86 days
• No degradation in surface water in the dark
  during 60 days
• ? very persistent!

5
Environmental risk assessment - general

• For human pharmaceuticals, risk assessment is
  performed according to EMA guideline
• Two phases
• Phase 1 estimation of exposure
• If trigger value of 0.01 µg/L in surfacewater is
  met ? phase 2
• If compound is a hormone ? always phase 2
• Phase 2 environmental fate and effects analysis
• Base set of fate and effects data
• Risk assessment for surface water, ground water,
  and STP
• If necessary (sorption) also risk assessment for
  sediment and soil
• Effect characterisation

6
Estimation of exposure

• Predicted Environmental Concentration (PEC)
• EMA guideline gives a basic calculation which can
  be refined using the prevalence of the disease

• Fraction of market penetration is default 0.01
  (1)
• For oseltamivir daily dose is 150 mg/day for
  curative use and 75 mg/day for preventive use
• ? PECsurfacewater using default values 0,70
  µg/L
• ? use by 30 of inhabitants of a region
  PECsurfacewater 20.9 µg/L
• ? 1 µg/L if 1.5 of all inhabitants is treated.
• (Almost) no degradation in sewage treatment
  plant!

7
Estimation of exposure are these values
realistic?
EMA default exposure scenario 0.7 µg/L
EMA exposure scenario with 30 use 20.9 µg/L
Singer et al, 2007 34 µg/L
KWR calculations for river Rhine during pandemic use 1-10 µg/L
UK drinking water inspectorate model calculations (Watts and Crane Associates, 2007) Max 107 µg/L
Industry models (Straub, 2009) Max 98 µg/L
8
Estimation of exposure are these values
realistic?
EMA default exposure scenario 0.7 µg/L
EMA exposure scenario with 30 use 20.9 µg/L
Rivers in Japan during normal flu season (Söderström et al., 2009) Max 60 ng/L
STP effluent in Japan 2008/2009 flu season (Ghosh et al., 2010a) Max. 293 ng/L
Rivers in Japan 2008/2009 flu season (Ghosh et al., 2010a) Max. 190 ng/L
STP influent in Japan 2009/2010 flu season (Ghosh et al., 2010b) Max. 460 ng/L
STP influent Rhine catchment area sept. 09 (Prasse et al.,. 2010) Max. 53 ng/L
River in Germany, sept. 09 (Prasse et al., 2010) Max. 38 ng/L
River Rhine, Germany, sept. 09 (Prasse et al., 2010) Max. 160 ng/L
Measurements agreed very well with modeled
concentrations
Influence from manufacturing plants?
9
Estimation of effects Single species toxicity
tests
10
Estimation of effects criteria for toxic effects

• Chronic ecotoxicity studies required by EMA
  guideline, because exposure is also chronic
• NOEC no effect concentration
• LOEC lowest effect concentration
• LC50 concentration at which 50 of the test
  animals has died (Llethal)
• EC50 concentration at which 50 of the animals
  show an effect (behaviour, growth, reproduction,
  etc.)

11
Criteria for toxic effects
12
Estimation of effects


fish
algae
crustacea
Aquatic
ecosystem
?
13
Estimation of effects

• Predicted No Effect Concentration (PNEC)
• PNEC lowest NOEC / 10
• Use of assessment/extrapolation factor, covering
• intra- and inter-species variation
• short-term to long-term extrapolation
• intra- and inter-laboratory variation
• lab-to-field extrapolation

14
Risk characterisation of surface- and groundwater

• PEC/PNEC lt 1 negligible risk
• PEC/PNEC gt 1 potential risk
• Oseltamivir in surface water
• Worst-case PEC 20.9 µg/L
• PNEC based on lowest NOECs ( 1000 µg/L) 100
  µg/L
• PEC/PNEC 0.21
• Oseltamivir in ground water
• Worst-case PEC 20.9 µg/L / 4 5.2 µg/L
• PNEC based on lowest NOECs ( 1000 µg/L) 100
  µg/L
• PEC/PNEC 0.05 ? no risk for direct
  ecotoxicity

15
Risk characterisation sewage treatment plants

• PEC for sewage treatment plants is factor 10
  higher than PECsurfacewater (no dilution) ? PEC
  209 µg/L
• Respiration-inhibition test (OECD 209) was not
  performed
• Biodegradation tests showed no effect on
  micro-organisms at 200 µg/L ? PNEC 200 / 10
  20 µg/L
• Another study showed an effect at 360 µg/L
• PEC/PNEC 10.5 ? potential risk
• There are indications that tamiflu may have an
  effect on microbial biofilms
• Combined effect of antibiotics and tamiflu
  unclear

16
Risk characterisation drinking water

• No guidance in EMA guideline
• No general drinking water standard for
  pharmaceuticals (Drinkwaterrichtlijnen 98/83/EG
  and 75/440/EEG)
• Because of the enzymatic mode of action, the
  compound belongs to the group of pesticides
• ? Pesticide drinking water standards 0.1 µg/L
• General signal value for alle anthropogenic
  compounds 1 µg/L
• Expected concentrations of tamiflu during
  pandemic use are above these values.

17
Risk characterisation drinking water

• Is this a problem?
• Oseltamivir does not sorb to organic carbon ?
  active carbon filtration will not remove
  oseltamivir
• Other types of filtration may remove oseltamivir,
  but data are scarce
• The safety margin between the calculated
  concentrations and human toxicological effect
  values is large enough not to expect effects.

18
Antiviral resistance formation

• Out of scope of this risk assessment
• But EC50 for influenza virus is 80-230 ng/L
  (Gubareve et al., 2001 Monto et al. 2006).

EMA default exposure scenario 0.7 µg/L
EMA exposure scenario with 30 use 20.9 µg/L
Rivers in Japan during normal flu season (Söderström et al., 2009) Max 60 ng/L
STP effluent in Japan 2008/2009 flu season (Ghosh et al., 2010a) Max. 293 ng/L
Rivers in Japan 2008/2009 flu season (Ghosh et al., 2010a) Max. 190 ng/L
STP influent in Japan 2009/2010 flu season (Ghosh et al., 2010b) Max. 460 ng/L
STP influent Rhine catchment area sept. 09 (Prasse et al.,. 2010) Max. 53 ng/L
River in Germany, sept. 09 (Prasse et al., 2010) Max. 38 ng/L
River Rhine, Germany, sept. 09 (Prasse et al., 2010) Max. 160 ng/L
19
Questions?