POSTER 32
CML RESPONSE TO THERAPY IN BRITISH COLUMBIA - A TEN YEAR
RETROSPECTIVE ANALYSIS
Sean Young1,2, Darko Curman1, Elaine Law1, Amadea Kristanti1, Sarah Egli1, Tracy Tucker1,2, Ian
Bosdet1,2, Donna Forrest3,4, Stephen Nantel3,4, Aly Karsan1,2 and Stephen Yip1,2
(1)Cancer Genetics and Genomics Laboratory, BC Cancer, Vancouver, Canada
(2)Dept of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
(3)Leukemia/BMT program of BC, BC Cancer, Vancouver, Canada
(4)Dept of Medicine, University of British Columbia, Vancouver, Canada
In the fall of 2017 the Cancer Genetics & Genomics Laboratory (CGL), BC Cancer, updated the
manner by which it tested and reported molecular testing results in patients with chronic
myeloid leukemia to be more in keeping with international standards. Prior to this date results
were reported to the requesting physician simply as the log ratio of BCR-ABL1/BCR. This
transition offered CGL the opportunity to assess approximately 10 years of response data
obtained using a single test method. The patient cohort (n =540) was also uniform in that
standard of care during this period stipulated the initial use of imatinib 400mg daily for
uncomplicated chronic phase patients. Consistent with established norms, changes to initial
therapy were considered only in cases of intolerance or resistance. The present study
evaluates the response to therapy in a 540 patient cohort, overall and at 5 years post initiation
of therapy, based on ELN defined early molecular response milestones.
Data curation resulted in the identification of 540 patients harbouring the p210 transcript
variant (e13a2 or e14a2) with baseline data available. Basic demographic information was
obtained from the laboratory information system. When available, information regarding
therapy at the various touch points was obtained from the patient’s electronic medical record.
Patients were stratified based on their response therapy at 90 days (+/- 30) and 180 days (+/-
30). Average response to therapy for each subgroup over time was then assessed. Only
aggregate data was evaluated and no attempt was made to follow the long term response of
an individual patient to therapy.
This first in-province retrospective population based study clearly demonstrates that early
milestones established in the context of clinical trials can be extrapolated to patients seen in
the community. This notwithstanding, it also highlights a lost opportunity for the very sizeable
minority of patients who do not have testing near the 90/180 day mark. With respect to the
response curves, it should be noted that although these clearly differ between patient groups,
no attempt was made to determine why an individual patient is in a given group. For example,
a patient may find themselves in the worst performing group for any number of reasons
including but not limited to advanced phase disease at presentation, resistance, intolerance,
or poor compliance. The underlying cause will undoubtedly impact their response over the
long term. Another caveat which must be applied to the response curves is that patients who
have had a good response to therapy are underrepresented in the data at later time points as
the testing interval (by local policy or patient/clinician preference) is often longer. This has the
effect of artificially elevating the response curves at later time points. Another point to
consider is that results reported clinically below the quantifiable limit of the assay may have
had three, two, or only one positive replicate. The value recorded in the laboratory
information system was the average of all positive replicates.