Deficient Mismatch Repair as a Predictive Marker for Lack of Benefit from 5-FU based Chemotherapy in Adjuvant Colon Cancer

1
Deficient Mismatch Repair as a Predictive Marker
for Lack of Benefit from 5-FU based Chemotherapy
in Adjuvant Colon Cancer

• Authors DJ Sargent et al
• Date posted June 26 2008

2
Background/Rationale

• Defective mismatch repair (MMR) is characterized
  by presence of MSI and loss of MLH1, MSH2, MSH6
  or PMS2 expression
• 15 of sporadic colon cancer has evidence of
  dMMR
• Clinical correlations right sided, female,
  early stage, better outcome
• Ribic/Gallinger NEJM 2003 demonstrated that
  patients with MSI-H have improved survival
  compared to MSS and MSI-L
• Their data also suggested that patients with
  defective MMR tumours do not benefit from 5-FU
  based chemotherapy

3
Methods

• Sargent et al used data and tumour specimens
  from RCTs of 5-FU based treatment vs control that
  were not used in Ribic analysis
• MMR determined by MSI testing if possible,
  otherwise IHC
• Primary EP was DFS comparing 5-FU based
  treatment to control by MMR status secondary EP
  was OS primary focus was stage II patients

Results

• 5 RCT with n491 patients
• 49 stage II
• 15 dMMR (as expected)
• Findings validated the 2003 Ribic study
• 5-FU based therapy of no value in dMMR patients
• 5-FU based therapy of benefit in pMMR patients

4
Results (contd)

• Consistency of findings justified pooling the
  Sargent and Ribic datasets
• Pooled data, n1027
• Pooled data validated dMMR as a prognostic
  marker in untreated patients and showed no
  suggestion of benefit from 5-FU based treatment
  in dMMR patients
• Significant OS decrement to 5-FU based treatment
  in stage II patients
• Stage II dMMR 5 yr OS Untreated93,
  Treated75, p0.03
• p0.014 for treatment by MMR status interaction
• Sargent et al take home message was In a
  patient being considered for 5-FU based therapy
  (stage II), mismatch repair status by MSI or IHC,
  should be tested to determine whom not to treat.

5
STUDY COMMENTARY

• Due to limited tissue resources and low
  prevalence of dMMR this study included only 165
  dMMR patients
• This study lends supports the assertion that MMR
  is a potentially useful prognostic and predictive
  biomarker in colon cancer
• These data are only for 5-FU/LV and therefore
  would generally apply to low risk stage II and
  not to any stage II (ie hi risk stage II) patient
  being considered for oxaliplatin based adjuvant
  therapy

6
BOTTOM LINE FOR CANADIAN MEDICAL ONCOLOGISTS

• Given the small numbers and retrospective nature
  of this study, it would be difficult to make
  definitive recommendations based on the results
  however given the lack of prospective evidence
  and the significant detriment seen in treating
  selected patients, oncologists with access to MMR
  testing could use this information in a
  discussion regarding adjuvant therapy with their
  patients
• HOWEVER
• since most clinicians do not yet have access to
  reliable and centralized MMR testing, and
• many Canadian oncologists do not routinely offer
  adjuvant chemotherapy to low risk stage II colon
  cancer patients
• The results of this study are unlikely to have
  much of an impact on Canadian oncology practice.
• Ongoing E5202 (NCIC CRC.3) will contribute
  greatly to our understanding of this issue in the
  FOLFOX era