SEARCH

SEARCH BY CITATION

Keywords:

  • head and neck cancer;
  • intratumor genetic heterogeneity;
  • tumor biomarkers;
  • next-generation DNA sequencing;
  • somatic mutations;
  • overall survival

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

BACKGROUND

Although the presence of genetic heterogeneity within the tumors of individual patients is established, it is unclear whether greater heterogeneity predicts a worse outcome. A quantitative measure of genetic heterogeneity based on next-generation sequencing (NGS) data, mutant-allele tumor heterogeneity (MATH), was previously developed and applied to a data set on head and neck squamous cell carcinoma (HNSCC). Whether this measure correlates with clinical outcome was not previously assessed.

METHODS

The authors examined the association between MATH and clinical, pathologic, and overall survival data for 74 patients with HNSCC for whom exome sequencing was completed.

RESULTS

High MATH (a MATH value above the median) was found to be significantly associated with shorter overall survival (hazards ratio, 2.5; 95% confidence interval, 1.3-4.8). MATH was similarly found to be associated with adverse outcomes in clinically high-risk patients with an advanced stage of disease, and in those with tumors classified as high risk on the basis of validated biomarkers including those that were negative for human papillomavirus or having disruptive tumor protein p53 mutations. In patients who received chemotherapy, the hazards ratio for high MATH was 4.1 (95% confidence interval, 1.6-10.2).

CONCLUSIONS

This novel measure of tumor genetic heterogeneity is significantly associated with tumor progression and adverse treatment outcomes, thereby supporting the hypothesis that higher genetic heterogeneity portends a worse clinical outcome in patients with HNSCC. The prognostic value of some known biomarkers may be the result of their association with high genetic heterogeneity. MATH provides a useful measure of that heterogeneity to be prospectively validated as NGS data from homogeneously treated patient cohorts become available Cancer 2013;119:3034—3042. © 2013 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Cancer is believed to arise from the acquisition of multiple mutations that cooperate to transform normal cells.[1] Although all neoplastic cells within a cancer presumably arose from a common ancestor, the progeny of this common ancestor continue to evolve.[2, 3] Hence there may be 1 or multiple dominant progeny subclones, and the evolutionary distance from the progenitor and the other subclones in the cancer is variable.[4] The presence of multiple progeny clones within an individual tumor reflects genetic heterogeneity. Although this concept is now well established,[5-17] to the best of our knowledge biomarkers to quantify this heterogeneity are scant.

It is likely that a greater extent of genetic heterogeneity poses a risk of worse clinical outcome, because a heterogeneous tumor might be more likely to contain subclones of cancer cells that proliferate more rapidly, are prone to metastasis, or are resistant to particular types of therapy.[18-22] Until recently, there had not been a simple, generally applicable measure of genetic heterogeneity to assess this risk that was suitable for use in clinical trials and in future clinical practice.[23]

A genetically heterogeneous tumor is likely to demonstrate wide variability in mutant-allele fractions within next-generation sequencing (NGS) data, with mutations in the ancestral clone at high frequencies and subclone-specific mutations at low frequencies within mixed tumor DNA. Therefore, we recently proposed a simple quantitative measure of genetic heterogeneity based on the variability of mutant-allele fractions,[23] exploiting this consequence of multiple subclones rather than identifying and enumerating subclones directly. This heterogeneity measure, called mutant-allele tumor heterogeneity (MATH),[23] is a percentage ratio of the width to the center of the distribution of mutant-allele fractions among tumor-specific mutated loci. Unlike other measures of genetic heterogeneity, MATH does not depend on preidentifying subclonal markers or on single-cell analysis; rather, it is derived directly from the mixed-population mutant allele frequencies within a tumor. Because NGS of tumor DNA is expected to find clinical application in the near future,[24] MATH could provide a clinically useful way with which to monitor significant, measurable genetic heterogeneity.

Based on publicly available NGS results on 74 patients with head and neck squamous cell carcinoma (HNSCC),[25] we demonstrated that poor-outcome classes of HNSCC possessed high genetic heterogeneity as measured by MATH.[23] Furthermore, MATH values were found to be unrelated to tumor mutation rates, suggesting that genetic heterogeneity involves clinically significant aspects of tumor biology beyond the accumulation of mutations. The possibility remained, however, that MATH was unrelated to clinical outcome per se, but was simply associated with certain pathologic features of HNSCC.

In the current study, we correlated clinical, pathological, and outcome data for these 74 patients and demonstrated that higher MATH was associated with shorter overall survival, especially in those who received chemotherapy. The relation between MATH and outcome in the current study was found to be stronger than that of 2 well-known poor-outcome HNSCC biomarkers: negative human papillomavirus (HPV) status[26, 27] and disruptive mutations in the tumor protein p53 (TP53) tumor suppressor gene.[28, 29] These results support the hypothesis that higher genetic heterogeneity portends a worse clinical outcome in patients with HNSCC, suggest that the prognostic value of some biomarkers may be due in part to their association with high genetic heterogeneity, and demonstrate that MATH provides a useful measure of that heterogeneity to be validated as NGS data from homogeneously treated patient cohorts become available.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Clinical, pathological, and outcome data for the 74 HNSCC patients for whom tumor NGS exome sequencing results had been reported by Stransky et al[25] were imported into the R software environment (R Foundation for Statistical Computing, Vienna, Austria)[30] for analysis. Before surgical removal of tumor tissue, all patients provided informed consent under protocol 99-069, which was approved by the University of Pittsburgh Institutional Review Board. Overall survival was calculated from the date of the surgical procedure from which the tumor sample used for NGS was obtained. Disease staging was based on the seventh edition of the American Joint Committee on Cancer manual,[31] using pathological T and N classifications when available.

Statistical Analysis

Numbers of tumor-specific mutations, MATH values, HPV status, total numbers of mutations, and TP53 mutation status for these tumors had been analyzed previously.[23] The MATH value for each tumor was based on the distribution of mutant-allele fractions among tumor-specific mutated loci, calculated as the percentage ratio of the width (median absolute deviation [MAD] scaled by a constant factor so that the expected MAD of a sample from a normal distribution equals the standard deviation [SD]) to the center (median) of its distribution:

MATH = 100 * MAD/median.

MATH values for these tumors ranged from 19 to 55 dimensionless units,[23] with a mean value of 34, an SD of 10, a median of 32, and first and third quartiles at 26 units and 42 units, respectively.

Bootstrap resampling of individual NGS reads for each tumor previously indicated that each tumor's MATH value had a typical associated SD of 4 units, depending on the number of mutated loci.[23] This SD arises from the sampling of individual DNA fragments among genomic loci and between mutant and reference alleles at each locus during NGS. Therefore MATH values are shown to 2 significant figures.

Relations between MATH and patient and tumor characteristics were assessed using linear models (Student t tests, analysis of variance, or linear regression). Hazards ratios (HRs) with respect to overall survival for MATH and for other patient and tumor characteristics were determined by the Cox proportional hazards analysis (survival package in R). The significance of HRs was based on the Wald test. Differences between survival curves were assessed by the log-rank test. All statistical tests were 2-sided, with significance accepted at P < .05. The receiver operating characteristic curve was obtained with the nearest-neighbor method for survival data developed by Heagerty et al (survivalROC package in R, with a smoothing span of 0.1).[32]

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Relation Between Clinical Characteristics and MATH and Outcomes

As shown in Table 1, the 74 patients ranged in age from 33 years to 76 years, with a median age of 57 years (mean, 58 years; SD, 10 years) at the time of diagnosis. The preponderance of males and of users of tobacco and alcohol is typical of patients with HNSCC.[33] One patient was African American and the other 73 patients were white. At time of last follow-up, 39 patients had died. The median follow-up for surviving patients was 46 months.

Table 1. Relation Between Clinical Variables and MATH and OSa
   Relation to MATHRelation to OS
VariableValueNo. (% of Total)MATH ± SDPHR95% CIP
  1. Abbreviations: 95% CI, 95% confidence interval; HPV, human papillomavirus; HR, hazards ratio; MATH, mutant-allele tumor heterogeneity; OS, overall survival; SD, standard deviation; TP53, tumor protein p53.

  2. a

    Relations between variables and OS were assessed using the Cox proportional hazards analysis, and were restricted to patients for whom values for the variable being considered were available. HRs and 95% CIs are shown for variables with a P value <.10 by the Wald test; only those with P <.05 were considered to be statistically significant.

  3. b

    Age was analyzed as a continuous variable; breakdown of MATH by age groups is provided for illustration.

  4. c

    HPV was assessed via polymerase chain reaction by Stransky et al.[25] Relations between MATH and HPV and TP53 status and the number of tumor-specific mutations in this data set were previously reported by Mroz and Rocco[23] and are presented herein for reference. All tumors were subjected to exome sequencing, and therefore the number of mutations is proportional to tumor mutation rate, conventionally expressed as mutations per megabase of sequenced DNA.[25]

  5. d

    Evidence of nonproportional hazards (P =.014) on chi-square test for trend of coefficient with time. Relation between N classification and survival was found to be statistically significant using the nonparametric log-rank test (P =.00002).

Gender   .053  .87
 Female20 (27%)30 ± 11    
 Male54 (73%)36 ± 10    
Age at diagnosis, yb   .640.964/y(0.934-0.995).024
 33-5017 (23%)36 ± 10    
 51-5618 (24%)32 ± 9    
 57-6519 (26%)34 ± 11    
 66-7620 (27%)35 ± 11    
Tobacco use   .67  .44
 No9 (12%)33 ± 13    
 Yes65 (88%)34 ± 10    
Alcohol use   .43  .20
 No15 (21%)32 ± 10    
 Yes58 (78%)34 ± 10    
 Unknown1 (1%)53    
Family cancer history   .086  .071
 No22 (30%)36 ± 10 1  
 Yes23 (31%)31 ± 9 0.48(0.21-1.07) 
 Unknown29 (39%)36 ± 10    
Tumor site   .29  .95
 Hypopharynx8 (11%)38 ± 10    
 Larynx15 (20%)35 ± 9    
 Oral cavity38 (51%)35 ± 11    
 Oropharynx11 (15%)28 ± 10    
 Sinonasal2 (3%)33 ± 3    
Recurrent tumor   .93  .99
 No67 (91%)34 ± 11    
 Yes7 (9%)35 ± 7    
T classification   .059  .28
 1 or 224 (32%)31 ± 9    
 3 or 446 (62%)36 ± 10    
 Unknown4 (5%)37 ± 9    
Differentiation grade   .34  1
 Well3 (4%)24 ± 2    
 Moderate47 (64%)34 ± 10    
 Poor22 (30%)35 ± 12    
 Unknown2 (3%)39 ± 10    
Perineural invasion   .079  .0092
 No31 (42%)32 ± 9 1  
 Yes36 (49%)36 ± 10 2.60(1.27-5.35) 
 Unknown7 (9%)35 ± 13    
Positive tumor margins   .13  .31
  No57 (77%)35 ± 10    
  Yes11 (15%)30 ± 9    
  Unknown6 (8%)39 ± 12    
Tumor HPV statusc   .013  .21
  Positive11 (15%)27 ± 5    
  Negative62 (84%)35 ± 10    
  Unknown1 (1%)50    
No. of mutationsc   .61  .82
  17-5218 (24%)32 ± 10    
  53-9219 (26%)33 ± 9    
  93-12918 (24%)35 ± 12    
  130-73919 (26%)36 ± 10    
TP53 mutation statusc   .0065  .25
  Wild-type28 (38%)31 ± 8    
  Nondisruptive16 (22%)32 ± 10    
  Disruptive30 (40%)39 ± 11.0029   
N classification    .46  .00009
  0 or 134 (46%)33 ± 10 1  
  2 or 336 (49%)35 ± 11 4.39d(2.09-9.21) 
  Unknown4 (5%)37 ± 9    
Extracapsular spread   .31  .23
  No19 (26%)33 ± 10    
  Yes20 (27%)36 ± 10    
  Lymph nodes not taken9 (12%)31 ± 13    
  Lymph nodes not evaluated26 (35%)34 ± 9    
TNM stage   .46  .066
  II or III18 (24%)33 ± 12 1  
  IV52 (70%)35 ± 10 2.20(0.95-5.08) 
  Unknown4 (5%)37 ± 9    
Additional treatment   .71  .14
  None18 (24%)33 ± 8    
  Radiation alone11 (15%)33 ± 12    
  Chemotherapy alone1 (1%)25    
  Chemoradiation40 (54%)35 ± 11    
  Unknown4 (5%)37 ± 6    

We examined the relations between tumor MATH values and clinical variables. MATH values were not found to be significantly related to any of the variables shown in Table 1, except for the previously reported relations between high MATH and HPV-negative tumors and with tumors having disruptive mutations in the TP53 tumor suppressor gene.[23] It is interesting to note that MATH values were not found to be significantly different between primary and recurrent tumors. Although some true relation between MATH and gender, family history of cancer, T classification, or perineural invasion (PNI) cannot be ruled out in this 74-patient data set, MATH does not simply represent a proxy for some other standard clinical variable.

We also examined the relations between the clinical variables listed in Table 1 and outcome. Of those variables, only age at diagnosis, PNI, and N classification were found to be significantly related to overall survival on univariate Cox proportional hazards analyses. There was no significant survival difference noted between patients who were treated for recurrent disease versus those treated for primary tumors. The well-established high-risk factors of negative HPV status[26, 27] and disruptive TP53 mutations[28, 29] were not found to be significantly related to outcome on univariate analysis, presumably due to the relatively small number of patients or the lack of a uniform treatment regimen. It is interesting to note that the tumor mutation rate itself, as conventionally assessed by the number of mutated loci per megabase of sequenced genomic DNA,[25] was also not related to outcome; as previously reported,[23] genetic heterogeneity of HNSCC assessed by MATH is not significantly related to mutation rate.

Relation Between MATH and Overall Survival

On univariate analysis, higher MATH was found to be strongly associated with shorter overall survival. We began by performing Cox proportional hazard regression of overall survival against MATH taken as a continuous variable, because MATH values ranged from 19 units to 55 units with no obvious subgroups of MATH values.[23] In this analysis, each individual tumor MATH value was related to the corresponding patient's time to death or last follow-up to determine how quickly the hazard of death grew as the MATH values increased. Among all 74 patients, each additional unit of increase in MATH was associated with a 4.7% increased hazard of death (Table 2). This is equivalent to an HR of 5.2 between the tumors with the highest and lowest MATH values.

Table 2. Relation Between MATH and OSa
  Relation to OS
  1. Abbreviations: 95% CI, 95% confidence interval; HPV, human papillomavirus; HR, hazards ratio; MATH, mutant-allele tumor heterogeneity; OS, overall survival; PNI; perineural invasion; TP53, tumor protein p53; U, unit.

  2. a

    Results of a Cox proportional hazards analysis on relation between MATH and OS of patients with tumor exome sequencing results reported by Stransky et al.[25] Each analysis was performed on all patients with values for the variable(s) of interest, and on the subsets involving primary tumors, with the number of patients and of deaths shown. HRs are for MATH unless otherwise noted. MATH and age were analyzed as continuous variables, and therefore the results for those variables are reported as multiplicative change in hazard-per-unit increase in MATH value or per year of age.

  3. b

    Evidence of nonproportional hazards for N classification; P =.048 on chi-square test for trend of coefficient of N classification with time. Relations between the other 3 variables with OS were similar in analysis stratified by N classification to allow for this nonproportionality; in that stratified analysis, the global chi-square test gave a P of .96.

AnalysisDeaths/PatientsHR95% CIP
Univariate39/741.047/U1.017-1.078.002
Primary tumors34/671.044/U1.013-1.075.005
Stratified by recurrence39/741.046/U1.015-1.077.003
Stratified by HPV status39/731.051/U1.018-1.084.002
Primary tumors34/661.046/U1.013-1.081.006
Univariate; HPV-negative subset35/621.050/U1.017-1.083.003
Primary tumors30/551.045/U1.012-1.080.008
Stratified by TP53 mutation status39/741.048/U1.016-1.080.003
Primary tumors34/671.046/U1.014-1.079.005
Univariate; disruptive TP53 subset15/301.088/U1.031-1.15.002
Primary tumors14/291.094/U1.033-1.16.002
Stratified by PNI status36/671.035/U1.002-1.068.035
Primary tumors31/601.030/U0.997-1.064.078
Univariate; subset with PNI25/361.047/U1.006-1.089.023
Primary tumors21/311.041/U0.999-1.084.055
Stratified by T classification (1/2 vs 3/4)36/701.043/U1.012-1.075.006
Primary tumors34/671.042/U1.010-1.075.009
Univariate, subset with T classification >225/461.049/U1.011-1.088.011
Primary tumors24/441.047/U1.008-1.086.016
Stratified by stage (II/III vs IV)36/701.047/U1.015-1.081.004
Primary tumors34/671.047/U1.014-1.082.006
Univariate; subset with stage IV disease29/521.059/U1.020-1.10.003
Primary tumors28/501.057/U1.018-1.098.004
Stratified by N classification (0/1 vs 2/3)36/701.048/U1.016-1.080.003
Primary tumors34/671.049/U1.017-1.083.003
Univariate; subset with N classification >1 (all primary tumors)25/361.056/U1.016-1.096.005
Multivariate (based on variables significantly related to outcome on univariate analyses)33/63  4 × 10−6
MATH 1.043/U1.008-1.080.017
Age 0.946/y0.910-0.982.003
N classification >1 4.92b2.18-11.1.0001
PNI 2.491.15-5.39.021
Multivariate analysis: primary tumors31/60  4 × 10−6
MATH 1.045/U1.008-1.084.016
Age 0.940/y0.904-0.978.002
N classification >1 5.942.45-14.4.0001
PNI 2.481.11-5.53.027
Univariate; patients not receiving chemotherapy13/301.00/U0.945-1.062.96
Primary tumors9/240.989/U0.924-1.059.74
Univariate; patients receiving chemotherapy (all primary tumors)23/411.061/U1.022-1.10.002

To determine whether MATH could be used to classify patients into high-risk and low-risk groups, we then compared patients whose tumors had MATH values above versus those with MATH values below the median value of 32 units. Among all 74 patients, the HR associated with a MATH value above the median was 2.46 (95% confidence interval [95% CI], 1.26-4.79; P = .008, using the Wald test); survival curves are shown in Figure 1a.[25]

image

Figure 1. Relation between mutant-allele tumor heterogeneity (MATH) and outcome is shown in clinically defined subsets of patients with head and neck squamous cell carcinoma (HNSCC). Each panel shows overall survival curves for patients whose tumors had a MATH value above (“High MATH”) versus below (“Low MATH”) the overall median value of 32 units. Survival was calculated from the time of the surgical procedure from which the tumor sample subjected to next-generation sequencing by Stransky et al.[25] was obtained. Crosses represent the last follow-up times for surviving patients. P values are for log-rank tests. Panels represent different subsets of patients. The numbers of patients and numbers of deaths in each subset analysis are shown in Table 2. (a) Comparison of high-MATH and low-MATH groups for all 74 patients is shown. (b) A subset with human papillomavirus (HPV)-negative tumors is shown. (c) A subset in which tumors had disruptive mutations in the tumor protein p53 (TP53) gene is shown; all of these tumors were also negative for HPV. (d) A subset with documented perineural invasion is shown. (e) A subset with American Joint Committee on Cancer stage IV disease is shown. (f) A subset with an N classification of 2 or 3 is shown; all of these patients were determined to have stage IV disease.

Download figure to PowerPoint

The relation between MATH and both HPV status and TP53 mutation status (Table 1) raised the possibility that MATH might not be related to outcomes within groups defined by those variables. Critically, as shown in Table 2, MATH as a continuous variable was still found to be related to outcome when patients were stratified by HPV status or TP53 status. MATH was also found to be significantly related to outcome when patients were stratified by PNI or N classification (Table 2), each of which was also significantly related with outcome (Table 1), or by T classification or TNM stage (Table 2).

Furthermore, MATH was found to be related with outcome within the known or expected worse-outcome subsets of patients defined by each of these variables (HPV-negative, disruptive TP53 mutation, presence of PNI, stage IV disease, N classifications of 2 or 3, and T classifications of 3 or 4). This was true both for MATH as a continuous variable (Table 2) and for categories based on MATH values above versus those below the median (Fig. 1b-1f).[25] These significant relations between MATH and outcome were maintained, except for groups defined by PNI, in stratified or subset analyses restricted to the 67 patients who had primary tumors (Table 2). MATH was also found to be significantly related with outcome on a multivariate analysis that incorporated all 4 variables found to be statistically significant on univariate outcome analysis (Table 2).

Genetic heterogeneity might be expected to have different relations with outcome depending on the type of therapy used. Thus, we evaluated the relation of MATH with outcome within subsets of patients defined by therapy. MATH was not found to be significantly related with outcome in the patients treated with either no adjuvant therapy or with radiation alone as adjuvant therapy (Table 2), although the small number of such patients means that some relation between MATH and outcome in those treatment settings cannot be ruled out.

In contrast, the relation between MATH and outcome was clearly observed in the patients who received systemic chemotherapy, usually combined with radiation (Table 2, last row). In these 41 patients, all having primary tumors, the hazard of death associated with MATH as a continuous variable increased 6.1% per unit, which is equivalent to an HR of 8.4 between the tumors with the highest and the lowest MATH values. In terms of classification, the HR for a MATH value above the median was 4.1 (95% CI, 1.6-10.2) (Fig. 2a). The receiver operating characteristic curve shown in Figure 2b demonstrates the tradeoff between sensitivity and specificity at different points of the MATH classification cutoff for these 41 chemotherapy patients. Thus, the relation between higher MATH with worse outcome was most pronounced for patients who received chemotherapy.

image

Figure 2. Relation of mutant-allele tumor heterogeneity (MATH) with outcome in patients with head and neck squamous cell carcinoma (HNSCC) who were treated with chemotherapy. (a) Survival curves shown are as in Figure 1 for MATH values above versus those below the median in the 41 patients who received chemotherapy (40 also received radiation; all 41 had primary tumors). (b) The receiver operating characteristic curve[32] for MATH estimated from these 41 patients is shown, demonstrating how different MATH cutoff values affect the specificity and sensitivity of outcome classification at survival times of ≥ 24 months. MATH value cutoffs increase from the top right to the bottom left of the solid curve, with the MATH value shown for every fourth tumor. For example, approximately 95% of patients whose tumors have MATH values greater than 40 are predicted to die within 24 months (specificity), but using 40 as a MATH cutoff value only identifies approximately 50% of all patients dying within 24 months (sensitivity). Using the median MATH value as the cutoff, as in panel a, provides greater sensitivity (70%) but lower specificity (80%). The dashed line would have been the sensitivity-specificity relation if there were no relation between MATH and outcome. The area under the curve (AUC) of 0.82 indicates that in 82% of random pairs of patients with 1 patient dying and 1 surviving, the pretreatment MATH value for the surviving patient would be lower.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

These results provide direct evidence, based on novel genomic analysis, that high genetic heterogeneity is related to shorter overall survival. This result is consistent with the long-standing hypothesis that high genetic heterogeneity is a risk factor for a worse outcome in patients with cancer.[18-22] Although the mechanisms linking high genetic heterogeneity with shorter overall survival cannot be determined from these data, the results of the current study are consistent with the hypothesis that genetically heterogeneous tumors are more likely to contain subclones of cancer cells that are resistant to chemoradiation therapy.

A primary role of intratumor genetic heterogeneity in determining clinical outcome may shed some light on the relation between disruptive TP53 mutations and HPV-negative status with worse outcomes in patients with HNSCC.[28, 29] Insofar as TP53 mutations impair both DNA repair processes and the removal of cells that develop additional mutations,[1] early clonal expansion of TP53-mutated cells would be predicted to lead to increased genetic heterogeneity as measured by MATH. The relation of high MATH specifically to disruptive but not to nondisruptive TP53 mutations suggests that disruptive mutations, as defined by the nature and site of the mutation in the p53 protein,[28] are most likely to impair both DNA repair and the removal of cells with newly mutated genomes and thus to promote genetic heterogeneity. Furthermore, high MATH was still associated with shorter overall survival within the subset of patients having disruptive TP53 mutations or when patients were stratified by TP53 status (Table 2). These results suggest that a major influence of disruptive TP53 mutations on outcome may be their tendency to increase genetic heterogeneity. Similarly, HPV-negative tumors have greater genetic heterogeneity compared with HPV-positive tumors, which is consistent with lower genetic heterogeneity as a reason for better outcomes in patients with HPV-positive HNSCC, who are typically treated with concurrent chemoradiation.

These results raise questions about the processes that lead to high genetic heterogeneity within a tumor. The lack of a relation between mutation rate with MATH as well as outcome indicates that mutations alone are not enough. Rather, additional processes must allow for the development and survival of genetically distinct subclones. Disruptive TP53 mutations appear to be involved in some patients, yet processes other than mutations in TP53 can lead to high genetic heterogeneity. Nearly one-third (12 of 37 tumors; 32.4%) of the tumors within the top one-half of MATH values had no TP53 mutation, disruptive or otherwise. Therefore, additional heterogeneity-inducing mechanisms need to be identified. Because high genetic heterogeneity is associated with shorter survival, therapies that target these mechanisms or the resulting heterogeneity itself may represent novel therapeutic approaches.

The results of the current study also raise questions about how therapy might affect intratumor heterogeneity. In particular, if a certain mode of therapy selects for 1 or a few subclones from a tumor, genetic heterogeneity would be decreased initially but might increase later as new subclones arise. Although the average MATH value of the 7 recurrent tumors in the current study did not differ significantly from that of the 67 primary tumors (Table 1), the small number of patients, the variety of prior treatments (1 patient receiving surgery alone, 4 treated with surgery plus radiation, and 2 receiving surgery plus chemoradiation), and the lack of corresponding pretreatment specimens mean that further studies are required to determine both how therapy affects heterogeneity in patients with HNSCC and the clinical implications of heterogeneity in the setting of recurrent disease.

The relation between genetic heterogeneity and outcome was surprisingly strong for this relatively small number of patients, including those having either primary or recurrent disease and without a controlled-treatment study design. This group of patients was evidently too small or too heterogeneous to demonstrate a significant relation between HPV status or disruptive TP53 mutations with outcome, despite the well-established relation between these classifications and outcome reported in studies of HNSCC cohorts that were larger or involved homogeneous treatment regimens.[26-29] In contrast, MATH was found to be significantly related to outcome not only on its own but also within the already high-risk groups defined by those and by other variables (Table 2) (Fig. 1).

MATH values were not found to be significantly related to N classification, the best single prognostic variable in this data set, or to TNM stage. MATH was related to outcome both when patients were stratified by N classification or stage of disease and when analysis was restricted to the subsets of high N classification and high-stage disease. These results thus support the use of MATH as an independent prognostic marker.

As NGS becomes widely used in clinical oncology, calculating MATH from the tumor-specific mutant-allele fractions in NGS results will provide a clinically relevant measure of genetic heterogeneity. MATH is not specific to HNSCC; it can be calculated from NGS results on any type of tumor that has an adequate number of tumor-specific mutations. This method of analyzing genetic heterogeneity therefore also provides a concrete and straightforward way with which to test hypotheses regarding genetic heterogeneity and outcomes in other types of cancer. The results of the current study indicate that the type of genetic heterogeneity captured by MATH values is related with HNSCC outcomes. Future controlled studies will determine the clinical usefulness of MATH as a prognostic biomarker in HNSCC and in other types of cancer.

FUNDING SUPPORT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Supported by The National Institute of Dental and Craniofacial Research (R01 DE022087 and RC2DE020958), the National Cancer Institute (R21 CA119591), the Cancer Prevention Research Institute of Texas (RP100233), and the Bacardi MEEI Biobank Fund.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Massachusetts General Hospital has filed a patent application based on subject matter discussed in this article, with Dr. Mroz and Dr. Rocco listed as inventors.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES