Molecular testing for Lynch syndrome in people with colorectal cancer: Systematic reviews and economic evaluation

colorectal cancer
cost-effectiveness analysis
diagnostic testing
discrete event simulation
genetic testing
Lynch syndrome
systematic review
Authors

Snowsill, T.

Coelho, H.

Huxley, N.

Jones-Hughes, T.

Briscoe, S.

Frayling, I. M.

Hyde, C.

Published

Sep 2017

Abstract

Background: Inherited mutations in deoxyribonucleic acid (DNA) mismatch repair (MMR) genes lead to an increased risk of colorectal cancer (CRC), gynaecological cancers and other cancers, known as Lynch syndrome (LS). Risk-reducing interventions can be offered to individuals with known LS-causing mutations. The mutations can be identified by comprehensive testing of the MMR genes, but this would be prohibitively expensive in the general population. Tumour-based tests - microsatellite instability (MSI) and MMR immunohistochemistry (IHC) - are used in CRC patients to identify individuals at high risk of LS for genetic testing. MLH1 (MutL homologue 1) promoter methylation and BRAF V600E testing can be conducted on tumour material to rule out certain sporadic cancers. Objectives: To investigate whether testing for LS in CRC patients using MSI or IHC (with or without MLH1 promoter methylation testing and BRAF V600E testing) is clinically effective (in terms of identifying Lynch syndrome and improving outcomes for patients) and represents a cost-effective use of NHS resources. Review methods: Systematic reviews were conducted of the published literature on diagnostic test accuracy studies of MSI and/or IHC testing for LS, end-to-end studies of screening for LS in CRC patients and economic evaluations of screening for LS in CRC patients. A model-based economic evaluation was conducted to extrapolate long-term outcomes from the results of the diagnostic test accuracy review. The model was extended from a model previously developed by the authors. Results: Ten studies were identified that evaluated the diagnostic test accuracy of MSI and/or IHC testing for identifying LS in CRC patients. For MSI testing, sensitivity ranged from 66.7% to 100.0% and specificity ranged from 61.1% to 92.5%. For IHC, sensitivity ranged from 80.8% to 100.0% and specificity ranged from 80.5% to 91.9%. When tumours showing low levels of MSI were treated as a positive result, the sensitivity of MSI testing increased but specificity fell. No end-to-end studies of screening for LS in CRC patients were identified. Nine economic evaluations of screening for LS in CRC were identified. None of the included studies fully matched the decision problem and hence a new economic evaluation was required. The base-case results in the economic evaluation suggest that screening for LS in CRC patients using IHC, BRAF V600E and MLH1 promoter methylation testing would be cost-effective at a threshold of £20,000 per quality-adjusted life-year (QALY). The incremental cost-effectiveness ratio for this strategy was £11,008 per QALY compared with no screening. Screening without tumour tests is not predicted to be cost-effective. Limitations: Most of the diagnostic test accuracy studies identified were rated as having a risk of bias or were conducted in unrepresentative samples. There was no direct evidence that screening improves long-term outcomes. No probabilistic sensitivity analysis was conducted. Conclusions: Systematic review evidence suggests that MSI- and IHC-based testing can be used to identify LS in CRC patients, although there was heterogeneity in the methods used in the studies identified and the results of the studies. There was no high-quality empirical evidence that screening improves long-term outcomes and so an evidence linkage approach using modelling was necessary. Key determinants of whether or not screening is cost-effective are the accuracy of tumour-based tests, CRC risk without surveillance, the number of relatives identified for cascade testing, colonoscopic surveillance effectiveness and the acceptance of genetic testing. Future work should investigate screening for more causes of hereditary CRC and screening for LS in endometrial cancer patients.

Citation

BibTeX citation:
@article{t.2017,
  author = {Snowsill, T. and Coelho, H. and Huxley, N. and Jones-Hughes,
    T. and Briscoe, S. and Frayling, I. M. and Hyde, C.},
  title = {Molecular Testing for {Lynch} Syndrome in People with
    Colorectal Cancer: {Systematic} Reviews and Economic Evaluation},
  journal = {Health Technology Assessment},
  volume = {21},
  number = {51},
  date = {2017-09-01},
  url = {https://tristansnowsill.co.uk/molecular-testing-for-lynch-syndrome-in-people-with.html},
  doi = {10.3310/hta21510},
  langid = {en},
  abstract = {Background: Inherited mutations in deoxyribonucleic acid
    (DNA) mismatch repair (MMR) genes lead to an increased risk of
    colorectal cancer (CRC), gynaecological cancers and other cancers,
    known as Lynch syndrome (LS). Risk-reducing interventions can be
    offered to individuals with known LS-causing mutations. The
    mutations can be identified by comprehensive testing of the MMR
    genes, but this would be prohibitively expensive in the general
    population. Tumour-based tests - microsatellite instability (MSI)
    and MMR immunohistochemistry (IHC) - are used in CRC patients to
    identify individuals at high risk of LS for genetic testing. MLH1
    (MutL homologue 1) promoter methylation and BRAF V600E testing can
    be conducted on tumour material to rule out certain sporadic
    cancers. Objectives: To investigate whether testing for LS in CRC
    patients using MSI or IHC (with or without MLH1 promoter methylation
    testing and BRAF V600E testing) is clinically effective (in terms of
    identifying Lynch syndrome and improving outcomes for patients) and
    represents a cost-effective use of NHS resources. Review methods:
    Systematic reviews were conducted of the published literature on
    diagnostic test accuracy studies of MSI and/or IHC testing for LS,
    end-to-end studies of screening for LS in CRC patients and economic
    evaluations of screening for LS in CRC patients. A model-based
    economic evaluation was conducted to extrapolate long-term outcomes
    from the results of the diagnostic test accuracy review. The model
    was extended from a model previously developed by the authors.
    Results: Ten studies were identified that evaluated the diagnostic
    test accuracy of MSI and/or IHC testing for identifying LS in CRC
    patients. For MSI testing, sensitivity ranged from 66.7\% to 100.0\%
    and specificity ranged from 61.1\% to 92.5\%. For IHC, sensitivity
    ranged from 80.8\% to 100.0\% and specificity ranged from 80.5\% to
    91.9\%. When tumours showing low levels of MSI were treated as a
    positive result, the sensitivity of MSI testing increased but
    specificity fell. No end-to-end studies of screening for LS in CRC
    patients were identified. Nine economic evaluations of screening for
    LS in CRC were identified. None of the included studies fully
    matched the decision problem and hence a new economic evaluation was
    required. The base-case results in the economic evaluation suggest
    that screening for LS in CRC patients using IHC, BRAF V600E and MLH1
    promoter methylation testing would be cost-effective at a threshold
    of £20,000 per quality-adjusted life-year (QALY). The incremental
    cost-effectiveness ratio for this strategy was £11,008 per QALY
    compared with no screening. Screening without tumour tests is not
    predicted to be cost-effective. Limitations: Most of the diagnostic
    test accuracy studies identified were rated as having a risk of bias
    or were conducted in unrepresentative samples. There was no direct
    evidence that screening improves long-term outcomes. No
    probabilistic sensitivity analysis was conducted. Conclusions:
    Systematic review evidence suggests that MSI- and IHC-based testing
    can be used to identify LS in CRC patients, although there was
    heterogeneity in the methods used in the studies identified and the
    results of the studies. There was no high-quality empirical evidence
    that screening improves long-term outcomes and so an evidence
    linkage approach using modelling was necessary. Key determinants of
    whether or not screening is cost-effective are the accuracy of
    tumour-based tests, CRC risk without surveillance, the number of
    relatives identified for cascade testing, colonoscopic surveillance
    effectiveness and the acceptance of genetic testing. Future work
    should investigate screening for more causes of hereditary CRC and
    screening for LS in endometrial cancer patients.}
}
For attribution, please cite this work as:
Snowsill, T., Coelho, H., Huxley, N., Jones-Hughes, T., Briscoe, S., Frayling, I. M., and Hyde, C. 2017. “Molecular Testing for Lynch Syndrome in People with Colorectal Cancer: Systematic Reviews and Economic Evaluation.” Health Technology Assessment 21 (51). https://doi.org/10.3310/hta21510.