Pr

1
Determination of cycle cut off in real time PCR
for the detection of regulated plant pathogens
(Phytophthora ramorum)
Viviane PLANCHON , Robert OGER , Anne
CHANDELIER
Introduction and background
Real time PCR concepts of threshold cycle and
cut-off
Real-time PCR enzymatic reaction for the
amplification of DNA fragment with a fluorescent
detection. The production of fluorescence at each
cycle depend on the concentration of targeted DNA
in the reaction.
The real-time PCR approach has gained increasing
acceptance as a laboratory technique suitable for
diagnostic purposes, notably for quarantine
fungi . Given the high importance of an accurate
detection for quarantine plant pathogens, and the
risk of false positive or false negative results,
a detection method must be correctly validated.
Phytophthora ramorum is a fungus of the subclass
of Oomycetes. Phytophthora ramorum has been
found mainly on Rhododendron and Viburnum, but
the list of plant species affected by the disease
is large (Camelia, Kalmia, Syringa, Vaccinium,
Arbutus, Quercus, Fagus, etc..). The
manifestation of symptoms of Phytophthora ramorum
is very rapid on Viburnum species, Camelia and
Rhododendron species. Due to its
regulated status, the identification of
Phytophthora ramorum infection must be notified
to the national Plant Protection Service
(Commission decision 2002/757/EC). In this
context, it is therefore highly recommended to
have a reliable detection method.

• threshold fixed level of fluorescence in the
  exponential phase
• Ct value PCR cycle number at which the
  fluorescence passes
• the threshold, enable to distinguish a positive
  from a negative sample

Symtoms are leaf spots, shoots wilting and
necrosis of twigs. The petiole is often necrotic.
On Fagaceae (beech, oak), it forms cankers on the
trunk and branches, with the presence of spots
and viscous exudates (reddish brown to black).
Problematic results with late signals Ct value
close to the end of the PCR run
Objectives

• To develop a statistical procedure to determine
• the cycle cut off
• the corresponding limit of detection (LOD)
• of a real-time PCR used as a qualitative method
  (presence/absence) for the detection for a
  quarantine plant pathogen (Phytophtora ramorum).
• This approach takes into account the risks of
  false positive and false negative results, both
  probabilities being fixed at a defined value.

The cut-off is a relative value which depends on
different parameters (method, threshold, PCR
machine, reagents, )
The cycle cut off or cut-off PCR cycles number
above which any sample response is considered as
a false positive
Methodology and results
To fit a distribution to these Ct values, the
method of quantiles was used (PROC NLIN, SAS) to
get initial parameters values. The choice of a
distribution (normal, exponential, etc.), that
gives the best adjustment of the left part of the
negative samples truncated distribution, was
based on root mean square error (RMSE). The
distribution with the lowest RMSE was selected
and its parameters used to evaluate Ct with a
defined risk of false positive results.
First step cut-off determination
Determination of the parameters of the
distribution of the Ct values for a set of 86
negative samples (truncated distribution), for a
producers risk of 1
Data samples from the survey organised by the
Belgian Plant Protection Service (N120) in 2009
Positive and negative status are defined by
reference methods
Producers risk (risk of false positive results)
the risk to reject a negative sample
Negative sample (truncated dist.)
Consumers risk (risk of false negative results)
the risk to accept a positive sample
probability on non detection
Second step LOD determination
What is LOD (limit of detection) ? It corresponds
to the minimum result (number of copies) which,
with a stated probability, can be distinguished
from a suitable blank value
Estimation of - the normal standard variable Z
for a wide range of concentration level - the
corresponding probability
Data genomic DNA from a pure culture of P.
ramorum diluted in DNA from healthy plant
(Rhododendron)
LOD (P 95) for a cut-off of 34.39 ?

• - 6 DNA concentration levels (1 ng, 100 pg, 10
  pg, 1 pg, 100 fg, 10 fg/PCR)
• N30 replicates/concentration

Normality of the distribution for each
concentration level
P 95
P probability corresponding to the normal
standard variable Z
(a)
(b)
(a)
(b)
Relationships between the mean and the log
transformed DNA concentration the standard
deviation and the log transformed DNA
concentration
Conclusions
Third step validation of the method

• The cycle cut off is an important parameter in
  the real time PCR interpretation, especially
  for samples with low pathogen concentration
• It is a relative value which depends on
  numerous parameters its determination should
  be based on validation data, at the laboratory
  level
• A statistical approach has been developed to
  establish the cut-off and the LOD the number of
  observations used to estimate the cut-off is a
  limiting factor for the estimation of the
  parameters of the truncated distribution
• When using real time PCR, there is a risk of
  false positive and false negative (as with any
  detection method). These risks should be fixed
  and known by the customer

• Data
• another data set (2006-2008)
• negative samples N81 32 determined Ct
  values
• The best distribution normal distribution

Data set 2006-2008 Cut-off 34.66 LOD 45
fg/PCR
Ct
Data set 2009 Cut-off 34.39 LOD 53 fg/PCR
Normal quantiles
Similar results for the lab
Best distribution based on RMSE normal
distribution
AGROSTAT 2010 February 23-26, 2010 Benevento,
Italy
Reference Chandelier A, Planchon V, Oger R
(2010). Determination of cycle cut off in
real-time PCR for the detection of regulated
plant pathogens. EPPO Bulletin. In press.