Although RT-qPCR is the recommended method for monitoring responses to TKI
therapy, it might not be the best assay for patients with deep responses considered for
treatment discontinuation. The current consensus is that therapy discontinuation is
followed by relapse in at least 50% of patients and that persisting LSCs with low BCRABL1 expression or quiescent LSC with no BCR-ABL1 expression are the source of the
relapse. Therefore, identifying a sensitive method for detecting and accurately
quantifying residual disease would help identify patients with the lowest likelihood of
relapse, and predict those who are at a higher risk of disease recurrence while off
therapy.
In the era of rapid technological advances and their deployment in molecular diagnostics
labs, we sought to investigate the potential of some of these technologies in improving
the current clinical management of CML. Digital PCR is a recently developed method
of absolute molecular quantification that has the potential for complementing RT-qPCR
not only in monitoring MRD while in deep molecular remission or during therapy
discontinuation, but also in simplifying the standardization efforts. NGS, on the other
hand, has the potential for increasing the sensitivity of TKD mutation detection. Unlike
the case in other diseases such as virology and solid tumours, the clinical value of these
technologies is not yet known in CML. Therefore, the focus of this Ph.D. thesis was to
develop protocols and proof-of-principle data facilitating the implementation of these
methodologies in routine testing and consequently investigating their clinical
significance in the routine management of CML patients.
In chapter two, we investigated an NGS-assisted DNA-based dPCR approached for
detecting and quantifying low levels of BCR-ABL1 positive disease. When applied to
samples with undetectable disease by RT-qPCR, DNA-based dPCR provided a marked
improvement in sensitivity, not only over RT-qPCR, but also compared to real-time
qPCR and to RT-dPCR. Although more sensitive, this method is not yet ready for the
immediate implementation in routine testing. The impact of residual disease level as
assessed by DNA-based dPCR at the time of treatment withdrawal on outcome is
currently being investigated within the UK based DESTINY clinical trial. If validated
in clinical trials of stopping TKI, the technique will permit a more personalised approach
to recommendations for dose reduction or drug cessation in individual patients, ensuring
that therapy is withdrawn only from patients with the highest chance of long-term
remission. In addition, it will allow timely therapeutic intervention to prevent the
occurrence of overt relapse after therapy discontinuation.
RT-qPCR remains the gold-standard for monitoring residual disease in CML despite of
its various limitations including the compromised precision at the lower end of the
calibration curve, in addition to the laborious requirement of continuous assay
validation. Therefore, efforts to improve on the current gold-standard are appreciated.
In chapters three and four, we sought to investigate the performance of the E.A.C. assay
on different RT-dPCR platforms and found that false positive signals detected in the
negative controls limit the accurate quantification of residual disease in samples
classified below MMR. We also showed that the performance of the RainDrop® RTdPCR platform had excelled compared to the other two RT-dPCR platforms allowing a
sensitivity of at least 5-logs on the IS. The false positivity detected on the three RTdPCR platforms using the E.A.C. assay, albeit at different levels, indicated that the noise
is most likely platform and/or assay design related. Therefore, further work is required
to eliminate false positivity before RT-dPCR could be adopted for the routine
monitoring of BCR-ABL1 transcript levels in response to TKI therapy.
Mutations in the BCR-ABL1 TKD are the most studied cause of resistance to different
TKI therapies and play a crucial role in planning patients’ management. In chapter five,
we validated an amplicon deep sequencing approach on the Ion Torrent PGM next
generation sequencing platform followed by the application of the method for the
prospective testing of referral sample over a period of one year. The aim was to evaluate
the performance of the platform and assess the practical need for replacing the current
gold-standard with NGS, notwithstanding its potential technical superiority. From
technical point of view, we demonstrated that the platform has an LoQ and LoD of 5%
and 1%, respectively. Low-level contamination occurring during the runs came as a
persisting problem, dictating a cautious interpretation of mutations below 5%. Although
NGS demonstrated superior sensitivity compared to the current gold-standard, we didn’t
find enough evidence that supports an immediate need for replacing the gold-standard
in routine clinical testing, except when performed after TKI resistance with the aim of
guiding therapeutic intervention.