Background
Methods
Searches
Study selection
Inclusion criteria
Data extraction and quality appraisal (risk of bias assessment)
Analysis and synthesis
Results
Study and participant characteristics
Study | Participant selection and disease prevalence a
| Participant characteristics | Reference standard | Microdissection | MSI Panel | MSI thresholds | ||
---|---|---|---|---|---|---|---|---|
Population-based and age-limited single-gate studies
| ||||||||
Barnetson, 2006 | Diagnosed <55yrs of age, consecutive recruitment Disease prevalence 8.5% (95% CI 5.8 to 11.9) | Number: | For all participants: Germ-line DNA obtained from blood leukocytes analysed for MLH1, MSH2, and MSH6 mutations. dHPLC analysis was used for MSH2 and MLH1. Variants noted on chromatography were then sequenced. Mutations were confirmed by re-amplification of an independent sample of DNA and resequencing in both directions. MLH1 and MSH2 were assessed for deletions by MLPA, with products separated on a genetic analyser. | 10μm tumour sections; microdissection performed on purified tumour DNA, and control DNA from blood or normal tissue in the section | Bethesda/NCI panel | MSI-H: >1 marker MSI-L: 1 marker MSS: 0 markers | ||
Recruited | 1259 | |||||||
Receiving RS | 870 | |||||||
Receiving MSI | 352 | |||||||
Gender: | ||||||||
Male | 53.1% | |||||||
Female | 46.9% | |||||||
Age (in years): | ||||||||
Non-carrier | 48.2 (±6.0) | |||||||
Carrier | 42.7 (±7.7) | |||||||
MLH1 | 38.5 (±8.4) | |||||||
MSH2 | 43.8 (±6.1) | |||||||
MSH6 | 49.0 (±3.9) | |||||||
Clinical criteria: | ||||||||
AMS II | 4.0% | |||||||
RBG | 64.0% | |||||||
Poynter, 2008b
| Recruitment through population-based cancer registries (population-based sample), selection process unclearc
Disease prevalence 9.2% (95% CI 7.1 to 11.8) | Number: | For all MSI-H or MSI-L probands and in a random sample of 300 MSS population-based probands: Mutations in MSH2 and MLH1 were detected using a combined approach of dHPLC/direct sequencing and MLPA. Direct sequencing was used to detect MSH6 mutations in cases with absent IHC staining of MSH6. | NR | BAT25, BAT26, D5S346, D17S250, BAT40, MYCL, ACTC, Dl 8S55, D1OS197, BAT34C4 | MSI-H: ≥30% MSI-L: >0% and <30% MSS: 0% | ||
Recruited | 1061 | |||||||
Receiving RS | 726 | |||||||
Receiving MSI | 1061 | |||||||
Gender: NCR
| ||||||||
Age (in years): NCR | ||||||||
Clinical criteria: NCR | ||||||||
Southey, 2005 | Diagnosed <45yrs of age, random recruitment Disease prevalence 30.5% (95% CI 19.2 to 43.9) | Number: | For all MSI-H or MSI-L probands, those that lacked expression of at least one MMR protein and a random sample of 23 patients selected from those who had tumours that were MS stable and did not lack expression of any MMR protein: MLH1, MSH2, MSH6, and PMS2 genes were screened for germline mutations using sequencing approaches or dHPLC. Confirmation of putative mutations was sought using an independent polymerase chain reaction for direct automated sequencing. MLPA was used to detect large genomic alterations in MLH1 and MSH2 on samples from 10 patients who had tumours lacking at least one MMR protein expression and for which no previous mutation had been identified by sequencing. | 5μm tumour sections; microdissection performed on invasive tumour cells from paraffin-embedded archival tumour tissue stained with 1% methyl-green, and normal cells from colonic or lymph node tissue/DNA extracted from peripheral blood lymphocytes | BAT25, BAT26, D2S123, D5S346, D17S250, BAT40, MYB, TGFRII, IGFIIR, BAX | MSI-H: >5 markers MSI-L: 2-5 markers MSS: <2 markers | ||
Recruited | 131 | |||||||
Receiving RS | 59 | |||||||
Receiving MSI | 105 | |||||||
Gender:
| ||||||||
Male | 62.7% | |||||||
Female | 37.3% | |||||||
Age (in years): | ||||||||
37.1 (range 24 to 42) | ||||||||
Clinical criteria: | ||||||||
AMS II | 9.2% | |||||||
RGB | NR | |||||||
High-risk, single-gate studies
| ||||||||
Caldes, 2004 | HNPCC families selected through a clinic for familial cancer, selection process unclear Disease prevalence 58.6% (95% CI 44.9 to 71.4) | Number: | For all participants: Genomic DNA was isolated from peripheral blood lymphocytes was analysed for MLH1, MSH2 and MSH6. DNA was amplified using PCR and all amplicons were subjected to DGGE or cycle sequencing. The MSI-H cases that were negative for mutations were analysed for genomic deletions in MLH1 and MSH2 by Southern Blotting. | 10μm tumour sections; microdissection performed on H&E stained slides with demarked areas containing cancer cells, and corresponding areas on unmarked slides | Bethesda/NCI panel | MSI-H: >1 marker, or 1 marker if BAT26 MSI-L: Not used MSS: 0 markers | ||
Recruited | 58 | |||||||
Receiving RS | 58 | |||||||
Receiving MSI | 58 | |||||||
Gender: NR
| ||||||||
Age (in years): NR | ||||||||
Clinical criteria: NR | ||||||||
Mueller, 2009 | ‘Suspected Lynch syndrome’ participants who met Amsterdam criteria, modified Amsterdam criteria, were ‘HNPCC-like’ or met Bethesda criteria, selection process unclear Disease prevalence 58.3% (95% CI 43.2 to 72.4) | Number: | For all participants: Sequencing and MLPA. Limited details provided. | NR | 5 and 10 panel markers, no further details provided | NR | ||
Recruited | 48 | |||||||
Receiving RS | 48 | |||||||
Receiving MSI | 48 | |||||||
Gender: NR
| ||||||||
Age (in years): NR | ||||||||
Clinical criteria: NR | ||||||||
Overbeek, 2007 | Families history that fulfilled one of the following criteria: 1) Amsterdam II criteria 2) Bethesda guidelines 3) a history very close to the Bethesda guidelines, selection process unclear Disease prevalence NC | Number: | For all participants: Mutation analysis of MLH1, PMS2, MSH2, and MSH6 was performed in DNA from peripheral blood lymphocytes by a combination of either single-strand conformation polymorphism analysis or DGGE and direct sequence analysis. For the detection of large deletions and duplications in MLH1, MSH2, MSH6, and PMS2, MLPA was used. All deletions and duplications were confirmed by Southern blot analysis or with a specific PCR. | NR | BAT25, BAT26, D2S123, D5S346, D17S250 (BAT40 was also added to the standard set of markers but it is unclear for which participants) | Tumours categorised as positive (>2 Bethesda markers) or negative | ||
Recruited | NR | |||||||
Receiving RS | NR | |||||||
Receiving MSI | NR | |||||||
Gender: NR
| ||||||||
Age (in years): 40.7 (range 29 to 51) | ||||||||
Clinical criteria: NR | ||||||||
Poynter, 2008b
| Recruitment through high-risk clinics (clinic-based sample), selection process unclearc
Disease prevalence 30.9% (95% CI 23.7 to 38.9) | Number: | For all participants: Mutations in MSH2 and MLH1 were detected using a combined approach of dHPLC/direct sequencing and MLPA. Direct sequencing was used to detect MSH6 mutations in cases with absent IHC staining of MSH6. | NR | BAT25, BAT26, D5S346, D17S250, BAT40, MYCL, ACTC, Dl 8S55, D1OS197, BAT34C4 | MSI-H: ≥30% MSI-L: >0% and <30% MSS: 0% | ||
Recruited | 172 | |||||||
Receiving RS | 152 | |||||||
Receiving MSI | 172 | |||||||
Gender: NR
| ||||||||
Age (in years): NR | ||||||||
Clinical criteria: NR | ||||||||
Shia, 2005 | Family history that fulfilled one of the following criteria: I ) Amsterdam I or II criteria 2) a set of relaxed AC three or more colorectal cancers among the first and second-degree relatives of a family that we referred to as ''HNPCC-like,'' and 3) Bethesda criteria, selection process unclear Disease prevalence 49.2 (95% CI 36.1 to 62.3) | Number: | For all participants: Each of the exons of MLH1, MSH2 and MSH6 was amplified by PCR, and heteroduplex analyses were performed using dHPLC. DNA fragments that displayed an abnormal chromatogram were sequenced directly. Cases with tumours that exhibited MSI but in which a point mutation was not detected were analysed for large deletions in MLH1 and MSH2 using a procedure based on the multiplex PCR of short fluorescent fragments. | Microdissection performed on DNA from paraffin-embedded tissue blocks. No further details reported | BAT25, BAT26, D2S123, D17S250, BAT40, PAX6, MYCL1 | Tumours categorised as positive (≥30%) or negative | ||
Recruited | 83 | |||||||
Receiving RS | 83 | |||||||
Receiving MSI | Unclear d
| |||||||
Gender:
| ||||||||
Male | 43.6% | |||||||
Female | 56.3% | |||||||
Age (in years): | ||||||||
50 (range 23 to 78) | ||||||||
Clinical criteria: | ||||||||
AMS II | 38.2% | |||||||
RBG | 8.2% | |||||||
Reference standard positive study
| ||||||||
Hendriks, 2003 | Germline mutation in MLH1, MSH2 or MSH6, selection process unclear Disease prevalence 84.8% (95% CI 68.1 to 94.9)e
| Number: | For all participants: DGGE or Southern blotting. Limited details provided | Microdissection not specifically reported, paired tumour and normal tissue DNA samples were used | BAT25, BAT26, D2S123, D5S346, D17S250, BAT40, MSH3 and MSH6 | MSI-H: >1 Bethesda markers MSI-L: 1 Bethesda marker MSS: 0 Bethesda markers | ||
Recruited | 45 | |||||||
Receiving RS | 45 | |||||||
Receiving MSI | 33 | |||||||
Gender:
| ||||||||
Male | 35.6% | |||||||
Female | 40.0% | |||||||
Age (in years): | ||||||||
MLH1 48 (range 29 to 90) | ||||||||
MSH2 40 (range 23 to 61) | ||||||||
MSH6 62 (range 26 to 84) | ||||||||
Clinical criteria: NR |
Risk of bias in individual studies
Domain | Item | Population-based, single-gate | High-risk,single-gate | Other | ||||||
---|---|---|---|---|---|---|---|---|---|---|
Barnetson 2006 [16] | Southey 2005 [23] | Poynter 2008a [21] | Caldes 2004 [17] | Mueller 2009 [19] | Overbeek 2007 [20] | Poynter 2008 a [21] | Shia 2005 [22] | Hendriks 2003 [18] | ||
Patient selection | Was a consecutive or random sample of patients enrolled? | Y | Y | U | U | U | U | U | U | U |
Was a case-control design avoided? | Y | Y | Y | Y | Y | Y | Y | Y | Nb
| |
Did the study avoid inappropriate exclusions? | Y | Y | U | Y | U | U | U | U | Y | |
Could the selection of patients have introduced bias? | L | L | U | U | U | U | U | U | Uc
| |
Is there concern that the included patients do not match the review question? | L | L | L | L | L | L | L | L | L | |
Index test | Were the index test results interpreted without knowledge of the results of the reference standard? | U | U | U | U | U | U | U | U | U |
If a threshold was used, was it pre-specified? | U | U | U | U | U | U | U | U | U | |
Could the conduct or interpretation of the index test have introduced bias? | U | U | U | U | U | U | U | U | U | |
Is there concern that the index test, its conduct, or interpretation differ from the review question? | L | L | L | L | L | L | L | L | L | |
Reference standard | Is the reference standard likely to correctly classify the target condition? | Y | Y | Y | Y | Y | Y | Y | Y | Y |
Were the reference standard results interpreted without knowledge of the results of the index test? | U | U | U | U | U | U | U | U | Y | |
Could the reference standard, its conduct, or its interpretation have introduced bias? | U | U | U | U | U | U | U | U | L | |
Is there concern that the target condition as defined by the reference standard does not match the review question? | L | L | L | L | L | L | L | L | L | |
Flow and timing | Was there an appropriate interval between index test(s) and reference standard? | U | U | U | U | U | U | U | U | U |
Did all patients receive a reference standard? | Y | N | N | Y | Y | Y | Y | Y | Y | |
Did patients receive the same reference standard? | Y | N | N | N | U | U | N | N | U | |
Were all patients included in the analysis? | N | N | N | N | Y | U | N | U | N | |
Could the patient flow have introduced bias? | U | U | U | U | U | U | U | U | U |
Sensitivity and specificity
Author, year | Sensitivity (%) | Specificity (%) |
---|---|---|
Single-gate, population-based samples
| ||
Poynter, 2008a [21] | 100.0 (93.9, 100.0) | 61.1 (57.0, 65.1) |
Barnetson, 2006 [16] | 66.7 (47.2, 82.7) | 92.5 (89.1, 95.2) |
Southey, 2005 [23] | 72.2 (46.5, 90.3) | 87.8 (73.8, 95.9) |
Single-gate, high-risk samples
| ||
Caldes, 2004b [17] | 79.4 (62.1, 91.3) | – |
Mueller, 2009 [19] | 91.3 (72.0, 98.9) | – |
Overbeek, 2007b [20] | 90.0 (59.6, 98.2) | – |
Poynter, 2008c [21] | 86.8 (71.9, 95.6) | – |
Shia, 2005b [22] | 100.0 (85.8, 100.0) | – |
Reference standard positive study
| ||
Hendriks, 2003 [18] | 88.0 (68.8, 97.5) | – |
Secondary outcomes
Author, year | LR+ | LR− | PPV (%) | NPV (%) | Diagnostic yield | Concordance |
---|---|---|---|---|---|---|
Single-gate, population-based samples
| ||||||
Poynter, 2008a [21] | 2.57 (2.32, 2.85) | 0.00 (NE)b
| 20.8 (16.2, 26.0) | 100.0 (99.0,100.0) | 44.5 (40.6 to 48.5) | 64.7 (60.9 to 68.4) |
Barnetson, 2006 [16] | 8.94 (5.54, 14.20) | 0.36 (0.21, 0.60) | 45.5 (30.4, 61.2) | 96.8 (94.1, 98.4) | 12.5 (9.2 to 16.4) | 90.3 (86.8 to 93.2) |
Southey, 2005 [23] | 5.92 (2.48, 14.10) | 0.32 (0.15, 0.67) | 72.2 (46.5, 90.3) | 87.8 (73.8, 95.9) | 30.5 (19.2 to 43.9) | 83.0 (71.0 to 91.6) |