Head and Neck Squamous Cell Carcinoma: Correlation between molecular cancer subtypes and selected clinical features
Maintained by TCGA GDAC Team (Broad Institute/Dana-Farber Cancer Institute/Harvard Medical School)
Overview
Introduction

This pipeline computes the correlation between cancer subtypes identified by different molecular patterns and selected clinical features.

Summary

Testing the association between subtypes identified by 4 different clustering approaches and 8 clinical features across 259 patients, one significant finding detected with P value < 0.05.

  • CNMF clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that do not correlate to any clinical features.

  • Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'PATHOLOGY.N'.

  • CNMF clustering analysis on sequencing-based miR expression data identified 3 subtypes that do not correlate to any clinical features.

  • Consensus hierarchical clustering analysis on sequencing-based miR expression data identified 3 subtypes that do not correlate to any clinical features.

Results
Overview of the results

Table 1.  Get Full Table Overview of the association between subtypes identified by 4 different clustering approaches and 8 clinical features. Shown in the table are P values from statistical tests. Thresholded by P value < 0.05, one significant finding detected.

Clinical
Features 		Statistical
Tests 		RNAseq
CNMF
subtypes 		RNAseq
cHierClus
subtypes 		MIRseq
CNMF
subtypes 		MIRseq
cHierClus
subtypes
Time to Death logrank test 0.182 0.0678 0.538 0.481
AGE ANOVA 0.483 0.245 0.511 0.247
GENDER Fisher's exact test 0.31 0.124 0.275 0.237
PATHOLOGY T Chi-square test 0.5 0.073 0.603 0.21
PATHOLOGY N Chi-square test 0.136 0.0148 0.652 0.612
PATHOLOGICSPREAD(M) Fisher's exact test 1 0.54 1 1
RADIATIONS RADIATION REGIMENINDICATION Fisher's exact test 0.727 0.209 0.173 0.842
NEOADJUVANT THERAPY Fisher's exact test 0.311 0.0851 0.507 0.631
Clustering Approach #1: 'RNAseq CNMF subtypes'

Table S1.  Get Full Table Description of clustering approach #1: 'RNAseq CNMF subtypes'

Cluster Labels 1 2 3
Number of samples 98 69 85
'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.182 (logrank test)

Table S2.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 250 98 0.1 - 210.9 (13.8)
subtype1 98 35 0.1 - 135.3 (12.4)
subtype2 68 30 0.2 - 142.5 (15.4)
subtype3 84 33 0.1 - 210.9 (16.9)

Figure S1.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.483 (ANOVA)

Table S3.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 252 61.0 (12.1)
subtype1 98 62.2 (11.2)
subtype2 69 60.1 (12.4)
subtype3 85 60.4 (12.7)

Figure S2.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.31 (Fisher's exact test)

Table S4.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients FEMALE MALE
ALL 185 67
subtype1 77 21
subtype2 49 20
subtype3 59 26

Figure S3.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

'RNAseq CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.5 (Chi-square test)

Table S5.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

nPatients T1 T2 T3 T4
ALL 16 70 65 91
subtype1 3 27 30 31
subtype2 6 21 15 26
subtype3 7 22 20 34

Figure S4.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

'RNAseq CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.136 (Chi-square test)

Table S6.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

nPatients N0 N1 N2 N3
ALL 100 40 92 3
subtype1 40 11 36 1
subtype2 20 12 33 1
subtype3 40 17 23 1

Figure S5.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

'RNAseq CNMF subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 1 (Fisher's exact test)

Table S7.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1
ALL 249 2
subtype1 97 1
subtype2 68 0
subtype3 84 1

Figure S6.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

'RNAseq CNMF subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.727 (Fisher's exact test)

Table S8.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 184 68
subtype1 72 26
subtype2 48 21
subtype3 64 21

Figure S7.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'RNAseq CNMF subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.311 (Fisher's exact test)

Table S9.  Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

nPatients NO YES
ALL 42 210
subtype1 18 80
subtype2 14 55
subtype3 10 75

Figure S8.  Get High-res Image Clustering Approach #1: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

Clustering Approach #2: 'RNAseq cHierClus subtypes'

Table S10.  Get Full Table Description of clustering approach #2: 'RNAseq cHierClus subtypes'

Cluster Labels 1 2 3
Number of samples 78 84 90
'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.0678 (logrank test)

Table S11.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 250 98 0.1 - 210.9 (13.8)
subtype1 78 25 0.1 - 135.3 (13.1)
subtype2 83 34 0.1 - 210.9 (17.9)
subtype3 89 39 0.2 - 142.5 (13.3)

Figure S9.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.245 (ANOVA)

Table S12.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 252 61.0 (12.1)
subtype1 78 62.7 (11.2)
subtype2 84 61.1 (11.7)
subtype3 90 59.5 (13.0)

Figure S10.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.124 (Fisher's exact test)

Table S13.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients FEMALE MALE
ALL 185 67
subtype1 63 15
subtype2 62 22
subtype3 60 30

Figure S11.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.073 (Chi-square test)

Table S14.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

nPatients T1 T2 T3 T4
ALL 16 70 65 91
subtype1 2 25 21 24
subtype2 3 20 20 38
subtype3 11 25 24 29

Figure S12.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.0148 (Chi-square test)

Table S15.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

nPatients N0 N1 N2 N3
ALL 100 40 92 3
subtype1 27 8 33 1
subtype2 44 18 18 1
subtype3 29 14 41 1

Figure S13.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

'RNAseq cHierClus subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.54 (Fisher's exact test)

Table S16.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1
ALL 249 2
subtype1 77 1
subtype2 83 1
subtype3 89 0

Figure S14.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

'RNAseq cHierClus subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.209 (Fisher's exact test)

Table S17.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 184 68
subtype1 53 25
subtype2 67 17
subtype3 64 26

Figure S15.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'RNAseq cHierClus subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.0851 (Fisher's exact test)

Table S18.  Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

nPatients NO YES
ALL 42 210
subtype1 15 63
subtype2 8 76
subtype3 19 71

Figure S16.  Get High-res Image Clustering Approach #2: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

Clustering Approach #3: 'MIRseq CNMF subtypes'

Table S19.  Get Full Table Description of clustering approach #3: 'MIRseq CNMF subtypes'

Cluster Labels 1 2 3
Number of samples 32 31 26
'MIRseq CNMF subtypes' versus 'Time to Death'

P value = 0.538 (logrank test)

Table S20.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 89 27 1.5 - 135.3 (12.5)
subtype1 32 8 1.5 - 87.5 (10.7)
subtype2 31 10 4.6 - 135.3 (16.1)
subtype3 26 9 3.2 - 89.8 (13.2)

Figure S17.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq CNMF subtypes' versus 'AGE'

P value = 0.511 (ANOVA)

Table S21.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 89 60.3 (12.6)
subtype1 32 60.8 (14.8)
subtype2 31 61.7 (12.1)
subtype3 26 58.0 (10.1)

Figure S18.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq CNMF subtypes' versus 'GENDER'

P value = 0.275 (Fisher's exact test)

Table S22.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients FEMALE MALE
ALL 66 23
subtype1 23 9
subtype2 26 5
subtype3 17 9

Figure S19.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

'MIRseq CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.603 (Chi-square test)

Table S23.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

nPatients T1 T2 T3 T4
ALL 3 22 25 38
subtype1 2 7 7 16
subtype2 0 9 11 10
subtype3 1 6 7 12

Figure S20.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

'MIRseq CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.652 (Chi-square test)

Table S24.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

nPatients N0 N1 N2 N3
ALL 30 16 39 3
subtype1 13 6 12 1
subtype2 11 3 15 1
subtype3 6 7 12 1

Figure S21.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

'MIRseq CNMF subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 1 (Fisher's exact test)

Table S25.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1
ALL 88 1
subtype1 31 1
subtype2 31 0
subtype3 26 0

Figure S22.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

'MIRseq CNMF subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.173 (Fisher's exact test)

Table S26.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 49 40
subtype1 21 11
subtype2 13 18
subtype3 15 11

Figure S23.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'MIRseq CNMF subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.507 (Fisher's exact test)

Table S27.  Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

nPatients NO YES
ALL 28 61
subtype1 8 24
subtype2 12 19
subtype3 8 18

Figure S24.  Get High-res Image Clustering Approach #3: 'MIRseq CNMF subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

Clustering Approach #4: 'MIRseq cHierClus subtypes'

Table S28.  Get Full Table Description of clustering approach #4: 'MIRseq cHierClus subtypes'

Cluster Labels 1 2 3
Number of samples 35 48 6
'MIRseq cHierClus subtypes' versus 'Time to Death'

P value = 0.481 (logrank test)

Table S29.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 89 27 1.5 - 135.3 (12.5)
subtype1 35 12 3.2 - 89.8 (13.1)
subtype2 48 13 1.5 - 135.3 (12.4)
subtype3 6 2 4.7 - 57.1 (8.6)

Figure S25.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq cHierClus subtypes' versus 'AGE'

P value = 0.247 (ANOVA)

Table S30.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 89 60.3 (12.6)
subtype1 35 58.0 (12.4)
subtype2 48 62.4 (12.8)
subtype3 6 57.2 (10.6)

Figure S26.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq cHierClus subtypes' versus 'GENDER'

P value = 0.237 (Fisher's exact test)

Table S31.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients FEMALE MALE
ALL 66 23
subtype1 25 10
subtype2 38 10
subtype3 3 3

Figure S27.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

'MIRseq cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.21 (Chi-square test)

Table S32.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

nPatients T1 T2 T3 T4
ALL 3 22 25 38
subtype1 2 9 8 16
subtype2 0 13 15 20
subtype3 1 0 2 2

Figure S28.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY.T'

'MIRseq cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.612 (Chi-square test)

Table S33.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

nPatients N0 N1 N2 N3
ALL 30 16 39 3
subtype1 9 7 18 1
subtype2 20 7 19 2
subtype3 1 2 2 0

Figure S29.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY.N'

'MIRseq cHierClus subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 1 (Fisher's exact test)

Table S34.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1
ALL 88 1
subtype1 35 0
subtype2 47 1
subtype3 6 0

Figure S30.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGICSPREAD(M)'

'MIRseq cHierClus subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.842 (Fisher's exact test)

Table S35.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 49 40
subtype1 18 17
subtype2 27 21
subtype3 4 2

Figure S31.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #7: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'MIRseq cHierClus subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.631 (Fisher's exact test)

Table S36.  Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

nPatients NO YES
ALL 28 61
subtype1 13 22
subtype2 13 35
subtype3 2 4

Figure S32.  Get High-res Image Clustering Approach #4: 'MIRseq cHierClus subtypes' versus Clinical Feature #8: 'NEOADJUVANT.THERAPY'

Methods & Data
Input

  • Cluster data file = HNSC.mergedcluster.txt

  • Clinical data file = HNSC.clin.merged.picked.txt

  • Number of patients = 259

  • Number of clustering approaches = 4

  • Number of selected clinical features = 8

  • Exclude small clusters that include fewer than K patients, K = 3

Clustering approaches
CNMF clustering

consensus non-negative matrix factorization clustering approach (Brunet et al. 2004)

Consensus hierarchical clustering

Resampling-based clustering method (Monti et al. 2003)

Survival analysis

For survival clinical features, the Kaplan-Meier survival curves of tumors with and without gene mutations were plotted and the statistical significance P values were estimated by logrank test (Bland and Altman 2004) using the 'survdiff' function in R

ANOVA analysis

For continuous numerical clinical features, one-way analysis of variance (Howell 2002) was applied to compare the clinical values between tumor subtypes using 'anova' function in R

Fisher's exact test

For binary clinical features, two-tailed Fisher's exact tests (Fisher 1922) were used to estimate the P values using the 'fisher.test' function in R

Chi-square test

For multi-class clinical features (nominal or ordinal), Chi-square tests (Greenwood and Nikulin 1996) were used to estimate the P values using the 'chisq.test' function in R

Download Results

This is an experimental feature. Location of data archives could not be determined.

References
[1] Brunet et al., Metagenes and molecular pattern discovery using matrix factorization, PNAS 101(12):4164-9 (2004)
[3] Bland and Altman, Statistics notes: The logrank test, BMJ 328(7447):1073 (2004)
[4] Howell, D, Statistical Methods for Psychology. (5th ed.), Duxbury Press:324-5 (2002)
[5] Fisher, R.A., On the interpretation of chi-square from contingency tables, and the calculation of P, Journal of the Royal Statistical Society 85(1):87-94 (1922)
[6] Greenwood and Nikulin, A guide to chi-squared testing, Wiley, New York. ISBN 047155779X (1996)
Meta

  • Maintainer = TCGA GDAC Team

Copyright © 2011 Broad Institute TCGA GDAC as part of the TCGA Research Network. All rights reserved.
Made with Nozzle