The Laryngoscope C 2014 The American Laryngological, V
Rhinological and Otological Society, Inc.
Prognostic Value of Hypoxia-Associated Markers in Advanced Larynx and Hypopharynx Squamous Cell Carcinoma Jonathan M. Bernstein, MD, FRCS; Timothy D. Andrews, FRCPath; Nick J. Slevin, FRCR; Catharine M. L. West, PhD; Jarrod J. Homer, MD, FRCS Objectives/Hypothesis: To determine the prognostic value of hypoxia-associated markers carbonic anhydrase-9 (CA-9) and hypoxia-inducible factor-1a (HIF-1a) in advanced larynx and hypopharynx squamous cell carcinoma (SCCa) treated by organ preservation strategies. Study Design: Retrospective cohort study. Methods: Pretreatment CA-9 and HIF-1a expression, clinicopathologic data, and tumor volume were analyzed in a series of 114 patients with T3–4 SCCa larynx or hypopharynx treated by (chemo)radiation. Results: Adverse prognostic factors for locoregional control were T4 classification (P 5 0.008), and for disease-specific survival were CA-9 positivity (P 5 0.039), T4 classification (P 5 0.001), larger tumor volume (P 5 0.004), N1–3 classification (P 5 0.002), and pretreatment hemoglobin < 13.0 g/dl (P 5 0.014). With increasing CA-9 expression, there was a trend to increasing tumor recurrence (Ptrend 5 0.009) and decreasing survival (Ptrend 5 0.002). On multivariate analysis, independent variables were T4 classification (hazard ratio [HR] 13.54, P 5 0.01) for locoregional failure, and CA-9 positivity (HR 5 8.02, P 5 0.042) and higher tumor volume (HR 5 3.33, P 5 0.007) for disease-specific mortality. Conclusion: This is the first study to look specifically at T3 and T4 SCCa larynx and hypopharynx for a relationship between hypoxia-associated marker expression and clinical outcome. Pretreatment immunohistochemical CA-9 expression is an adverse prognostic factor for disease-specific survival, indicating that CA-9 expression may confer a more aggressive tumor phenotype. Key Words: Head and neck squamous cell carcinoma, biomarker, tumor hypoxia, CA-9, HIF-1a, tumor volume, larynx, hypopharynx. Level of Evidence: 4. Laryngoscope, 125:E8–E15, 2015
INTRODUCTION Laryngeal squamous cell carcinoma (SCCa) accounts for 2.4% of cancer cases worldwide.1 Concomitant chemoradiation therapy became a leading modality of treatment for T3 SCCa larynx following the Ann Arbor Veterans Affairs and Radiation Therapy Oncology Group 91-11 studies.2,3 Chemoradiation therapy is larynx-sparing, although swallowing and speech are significantly affected in over 20% of patients.4 For those who develop local recurrence, salvage laryngectomy is usually offered. In T4 SCCa larynx and T3–4 SCCa From the University Department of Otolaryngology–Head & Neck Surgery (J.M.B., J.J.H.), Manchester Royal Infirmary and Manchester Academic Health Science Centre; the Translational Radiobiology Group, Institute of Cancer Sciences (J.M.B., N.J.S., C.M.L.W., J.J.H.), Manchester Academic Health Science Centre, University of Manchester; the Department of Head and Neck Clinical Oncology, The Christie NHS Foundation Trust (N.J.S.), Manchester; and the Department of Pathology, Royal Liverpool University Hospital (T.D.A.), Liverpool, United Kingdom. Editor’s Note: This Manuscript was accepted for publication August 25, 2014. This work was supported by funding from Experimental Cancer Medicine Centre (ECMC) Network. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Jonathan M. Bernstein, MD, FRCS, University Department of Otolaryngology–Head and Neck Surgery, Manchester Royal Infirmary, Oxford Road, Manchester, M13 9WL, U.K. E-mail: [email protected] DOI: 10.1002/lary.24933
Laryngoscope 125: January 2015
E8
hypopharynx, there is a trend toward greater benefit from surgery and postoperative radiation, but some patients are selected for treatment without surgery.5 In some centers, an initial response to induction chemotherapy is used to select patients for organ-preservation therapy, with nonresponders managed by total laryngectomy.6 There is a need to identify those patients in whom advanced SCCa larynx or hypopharynx is likely to progress despite chemoradiation. This study looks for a way to improve the selection of treatment for advanced cancer of the larynx and hypopharynx. There is consensus but not total unanimity among studies of head and neck squamous cell carcinoma (HNSCC) as a whole that the expression of hypoxia-associated markers is associated with adverse outcome.7–15 Tumor hypoxia is likely to represent an emerging hallmark of HNSCC and is associated with malignant progression and resistance to radiation and cytotoxic chemotherapy16 (although there is limited evidence suggesting that hypoxia may adversely affect the outcome after surgery as well17). Hypoxic cells are more resistant to damage from ionizing radiation because oxygen promotes the generation of free radicals.18 Chemotherapy drugs must diffuse from an adequate blood supply, and hypoxic cells spend less time in the growth phase when chemotherapeutic agents are most effective.19 Hypoxia-inducible factor-1a (HIF21a) is a transcription factor, which in normoxia is degraded after binding Bernstein et al.: Hypoxia-Associated Biomarkers
with von Hippel-Lindau protein and ubiquitin. In hypoxia (pO2 < 20 mm Hg), HIF-1a translocates to the nucleus and binds with HIF-1b and the hypoxiaresponsive elements of hypoxia-regulated genes, such as the gene encoding carbonic anhydrase-9 (CA-9). In hypoxia, HIF-1a induces the expression of CA-9.20 This is the first study to look specifically at locoregionally advanced SCCa larynx and hypopharynx for a relationship between the expression of hypoxia-associated markers HIF-1a and CA-9 and clinical outcome after organ preservation therapy. Tumor volume, an independent marker of prognosis, was measured because it has been suggested that the volume of hypoxic tumor may be an important consideration.21
ing high and low marker expression and running sections from these biopsies with each batch.
Histologic Analysis Nuclear HIF-1a and plasma membrane CA-9 staining were scored according to published scoring methods by a histopathologist (T.D.A.) and researcher (J.M.B.) who were blinded to outcome.23,24 Percentage scores in the single 4003 field containing SCCa with greatest staining were recorded. Sections were categorized as negative where staining was less than 1% and positive where 1% or more. Dysplastic or necrotic areas were excluded. Where more than one biopsy sample containing invasive carcinoma was available, the mean score was recorded. Necrosis was determined as either present or absent on hematoxylin and eosin-stained sections.
MATERIALS AND METHODS
Tumor Volume Measurement
Patients with American Joint Committee on Cancer (AJCC, 7th ed.) T3-4 SCCa larynx or hypopharynx treated with curative intent by (chemo)radiation were identified using the radiation therapy database at The Christie NHS Foundation Trust, United Kingdom. Patients with T4 tumors included in the study were deemed by the multidisciplinary team to be unfit for surgical extirpation and were therefore treated with curative intent by organ preservation strategies. All patients were diagnosed between June 2005 and December 2010. Clinicopathologic and outcomes data were collected retrospectively. Approval for the study was obtained from an ethics committee (ref: 11/NW/0284) and institutional research and development board. Pretreatment biopsy formalin-fixed paraffin-embedded tumor blocks were requested. SCCa was confirmed by a consultant histopathologist (T.D.A.). The REMARK reporting recommendations for prognostic biomarker studies were used.22
Primary tumor volume was determined using pretreatment computed tomography in patients with digital image archiving. The largest two perpendicular measurements in a single axial slice and the maximum coronal measurement were incorporated into the formula p/6 3 (length) 3 (width) 3 (height) to determine the ellipsoid volume.25
Immunohistochemistry Marker expression was examined in 4-lm sections. HIF-1a was detected using a Bond-Max autostainer (Leica Microsystems Ltd., Milton Keynes, U.K.) and the Bond Polymer Refine Detection system (Leica Biosystems, Newcastle-on-Tyne, U.K.) involving a biotin-free polymeric horseradish-peroxidase (HRP) linker antibody conjugate system to detect the antigen followed by 3,3’-diaminobenzidine (DAB) tetrahydrochloride for visualization and hematoxylin as counterstain. After deparaffinization, rehydration, and microwave epitope retrieval, antigen detection was undertaken using mouse monoclonal anti-human HIF-1a diluted 1:50 (61099, BD Biosciences, Oxford, U.K.) at pH9 at room temperature for 15 minutes. Mouse IgG1 (X0931, Dako, Ely, U.K.) was used as a negative control. CA-9 was detected using a Biogenex i6000 autostainer (Launch Diagnostics Ltd, Longfield, U.K.) and the EnVision Plus HRPconjugated polymer system (K4007, Dako). After dewaxing and rehydration, endogenous peroxidase was quenched with 0.3% hydrogen peroxide (Sigma, Poole, U.K.), and nonspecific staining was blocked with 10% casein solution (Vector Laboratories Inc., Peterborough, U.K.). The primary antibody was mouse monoclonal antihuman CA-9 diluted 1:50 (clone M75; gift from Profs. Pastorekova and Pastorek, Slovak Academy of Sciences, Bratislava, Slovakia), and sections were incubated at room temperature for 30 minutes. Mouse IgG2B (X944, Dako) was used as a negative control and DAB was used for visualization (K3467, Dako). Sections were removed from the processor and counterstained with Gill’s hematoxylin and dehydrated. Batchto-batch variation was assessed by choosing two sections show-
Statistical Analyses SPSS version 19.0 (IBM, Armonk, NY) was used. Time to locoregional control and disease-specific survival from the date of biopsy were estimated using the Kaplan-Meier method. Survival curves were compared in univariate analyses using the Mantell-Cox log-rank test; categorical variables were compared using the X2 test; and multivariate analyses were undertaken using the Cox proportional hazards model. P values < 0.05 were considered statistically significant. The log-rank test was used to determine a trend to adverse clinical outcome with increasing CA-9 and HIF-1a expression, with each marker categorized into four groups (0%, 1%–9%, 10%–29%, 30%). CA-9 and HIF-1a expression, T- and N-classification, tumor volume, hemoglobin level, tumor differentiation, performance status, gender, age, site, and alcohol and smoking history were analyzed for prognostic significance.
RESULTS One-hundred-and-fourteen patients were identified (86 male, 28 female) with T3–4 SCCa larynx or hypopharynx treated by (chemo)radiation with curative intent. Patient characteristics are given in Table I. Eighty-nine patients (78.1%) had SCCa larynx (51 glottic, 37 supraglottic, 1 subglottic), and 25 patients (21.9%) had SCCa hypopharynx (15 piriform sinus, 7 postcricoid, 3 posterior pharyngeal wall). One hundred patients had T3 tumors and 14 had T4 tumors. Tumor blocks for 104 patients were obtained of which 103 blocks contained tumor. These tumors included 83 cases of T3 SCCa larynx. Either IMRT or 3D-conformal radiation (4–6 MeV linear accelerator) was used. Eighty-two (71.9%) performance status 0 or 1 patients received chemoradiation, and 11 of these received induction chemotherapy due to a high tumor burden (Table I). No patients were treated by surgery following induction chemotherapy during the period of study.
TABLE I. Patient Characteristics. Characteristic
No. of Cases
Percentage of Cases (%)
Male Female
86 28
75.4 24.6
AJCC T-classification
T3
100
87.7
Tumor differentiation
T4 Well differentiated
14 25
12.3 21.9
Moderately differentiated
41
36.0
Poorly differentiated Undifferentiated
40 2
35.1 1.8
Gender
Not recorded Tumor site
Larynx
Glottis Supraglottis
Hypopharynx
Smoking
Alcohol
1
0.9
Piriform sinus Postcricoid hypopharynx
15 7
13.2 6.1
Concurrent chemotherapy regimen Cetuximab
3
2.6
0 1
89 10
78.1 8.8
2a
3
2.6
2b 2c
7 4
6.1 3.5
3
1
0.9
Never Ex 1 year
6 26
5.3 22.8
Ex < 1 year
37
32.5
Current Never heavy
45 60
39.5 52.6
Previous heavy
32
28.1
Current heavy 0
22 46
19.3 40.4
1
42
36.8
2 3
12 6
10.5 5.3
Not recorded
8
7.0
1–2 3–4
3 8
2.6 7.0
Capecitabine
62
54.4
WHO performance status
Number of induction chemotherapy cycles with TPF
Cisplatin Radiation therapy alone Radiation therapy dose
5.3 44.7 32.5
Subglottis
Posterior pharyngeal wall AJCC N-classification
6 51 37
11
9.6
22
19.3
55.0 Gy in 20 fractions
19 92
16.7 80.7
62.5–66.0 Gy in 30 fractions
22
19.3
Characteristic
Mean/Median Value
Age (years)
63 (mean)
CA-9 score (%)
5 (median)
HIF-1a score (%)
10 (median)
Range
37–86 0–80 0–80 Tumor volume (mL)
5.70 (median) 0.37–48.71
AJCC 5 American Joint Committee on Cancer; CA-9 5 carbonic anhydrase-9; HIF-1a 5 hypoxia-inducible factor-1a; TPF 5 docetaxel, cisplatin, and 5flurouracil; WHO 5 World Health Organization.
Laryngoscope 125: January 2015
E10
Bernstein et al.: Hypoxia-Associated Biomarkers
Fig. 1. Vertical scatter plots showing percentage marker staining from two or more separate biopsies from individual tumors for CA-9 in (a) and HIF-1a in (b).
CA-9 staining was negative in 25 (24.3%) cases and positive in 78 (75.7%) cases. HIF-1a staining was negative in 18 (17.5%) cases and positive in 85 (82.5%) cases. There was a correlation between the expression of CA-9 and HIF-1a (Spearman’s rho, 0.58; P < 0.0001). More than one tumor block was available in 14 patients. In these cases, there was variation mainly in the range 0 to 20% (Fig. 1). The results were identical in five cases for CA-9 and three cases for HIF-1a. In patients where there was one negative-scoring biopsy, recording the mean score resulted in six cases assigned CA-9 positive and three HIF-1a positive. Thirty-two of 114 patients developed locoregional recurrence, and a further seven had residual tumor. Of these 39 patients, 15 underwent salvage (pharyngo)laryngectomy usually with bilateral neck dissection, 19 were considered to be unfit for salvage surgery, and five were designated with recurrence because they died without another clear cause. In total, there were 42 deaths, of which 34 were deemed to be from the index HNSCC and eight were from causes such as pancreatic cancer or pneumonia in patients without recurrence. Seven of the patients who underwent salvage surgery died subsequently. As determined by Kaplan-Meier analysis after a median follow-up of 44.8 months, 5-year locoregional control was 63.5%; disease-specific survival was 65.0%; and overall survival was 55.4%. Estimated 5-year locoregional control was not affected by primary tumor site (P 5 0.498), but estimated disease-specific survival was better for carcinoma of larynx (70% at 5 years) than hypopharynx (45.1% at 5 years) (P 5 0.017) (Table II). T4-classification was an adverse prognostic factor for locoregional tumor recurrence (P 5 0.008). Adverse prognostic factors for disease-specific survival were T4classification (P 5 0.001), CA-9 positivity (P 5 0.039), Laryngoscope 125: January 2015
higher N-classification (P 5 0.002), and pretreatment hemoglobin level 13.0 13.0
76 38
58.5 49.6
33.1–77.1 30.0–66.5
0.074
75.0 47.2
62.0–84.1 27.5–64.6
0.014*
Hb2 (g/dL)
> 13.6
56
70.5
55.9–81.0
0.418
76.9
61.9–86.6
0.135
13.6
58
56.9
41.4–69.7
54.9
38.3–68.8
*Statistically significant result (P < 0.05). † Mantel-Cox log-rank test. ‡ Tumor volume stratified by the median (5.70 mL). Hb stratified by WHO definition of anemia in males in Hb1 and the median for this series in Hb2. AJCC 5 American Joint Committee on Cancer (7th Ed.); CA-9 5 carbonic anhydrase-9; CI 5 confidence interval; DSS 5 disease-specific survival; Hb 5 hemoglobin level; HIF-1a 5 hypoxia-inducible factor-1a; LRC 5 locoregional control; WHO, World Health Organization.
DISCUSSION Outcomes in HNSCC after radical radiation therapy are adversely influenced by tumor hypoxia measured by a number of approaches, including polarographic electrodes and immunohistochemical markers.26–29 In the present study, the first to look specifically at the prognostic effect of hypoxia markers in locoregionally advanced SCCa larynx or hypopharynx, the expression of CA-9 was prognostic for disease-specific mortality after therapy with organ preservation. CA-9 expression was associated with disease-specific mortality in about half the cases; and CA-9 negativity was associated mostly with disease-specific survival. CA-9 expression was associated inversely with microvessel density and treatment response in locally advanced HNSCC treated by chemoradiation in a study by Koukourakis et al.8 CA29 expression was associated with poor locoregional control and poor survival in HNSCC in the Continuous Hyperfractionated Accelerated Radiotherapy study, in which CA29 and HIF-1a showed prognostic potential in 198 pretreatment biopsy specimens.9 In the study by Schrijvers et al., among 91 patients with T1–2N0 SCCa glottic larynx, CA-9 and HIF-1a were associated with reduced disease-free survival; but a further study by the same authors of 60 patients with T1–2 SCCa supraglottic larynx showed a lack of prognostic significance of CA-9 and HIF-1a.10,15 A meta-analysis of 16 studies including 1,470 patients showed that high CA29 expression was associated with reduced survival (P < 0.0001) and reduced disease-free survival (P 5 0.0008) in HNSCC.11 There was no analyLaryngoscope 125: January 2015
E12
sis according to tumor site within the head and neck, and only one article on laryngeal cancer that concerned T1–T2 tumors. The results of the present study are in keeping with the meta-analysis in that CA-9 expression is prognostic in HNSCC. Not all studies have shown a relationship between CA9 expression and worse outcome. In the Danish Head and Neck Cancer Group (DAHANCA) 5 study of 320 patients with SCCa supraglottic larynx and pharynx treated by radiation, there was no prognostic or predictive potential for CA9.12 Half these patients had T1–2 tumors, and the other half were T3–4; both groups were about 70% positive for CA-9 expression, and the survival data were not analyzed according to tumor stage. Our own work has focused largely on SCCa larynx. In T1–2 SCCa glottic larynx, CA-9 expression was found in 72% of tumors, and there was no statistically significant association with outcome.7 The present study concerns T3–4 SCCa larynx and hypopharynx. In the whole group, there was positive CA-9 expression in 76% of cases (73% of T3 and 80% of T4 tumors), and CA-9 expression was associated with disease-specific mortality. The results differ from the negative results of the DAHANCA 5 study, and support the previous findings in advanced HNSCC in general that CA-9 expression is associated with disease-specific mortality.11 In the present study, biomarker analysis of the T3 SCCa subgroup lost statistical significance while the trends were maintained, indicating the reduction in patient numbers for this subanalysis. We found CA-9 expression to be a statistically significant prognostic marker of disease-specific mortality but not locoregional failure, a finding that is in Bernstein et al.: Hypoxia-Associated Biomarkers
33
Fig. 2. Kaplan-Meier plots of 103 cases of T3–4 squamous cell carcinoma of the larynx or hypopharynx for locoregional control as a function of (a) CA-9 and (b) HIF-1a expression, and disease-specific survival as a function of (c) CA-9 and (d) HIF-1a expression. CA-9 5 carbonic anhydrase-9; DSS 5 disease-specific survival; HIF-1a 5 hypoxia-inducible factor1a; LRC 5 locoregional control.
agreement with previous studies of HNSCC and cervix cancer.11,24 CA-9 is a better marker of chronic than acute hypoxia30; whereas both acute and chronic hypoxia confer radioresistance, and hence impaired locoregional control. Given the association of CA-9 with disease-specific mortality but not locoregional control, Loncaster et al. suggested that CA-9 expression in cervix cancer might reflect a more aggressive tumor phenotype rather than radioresistance.24 In that case, surgery might not be a suitable alternative treatment option in patients identified as having a high risk of disease-specific mortality following radiation therapy. Further studies are warranted. The expression of HIF-1a predicted an adverse prognosis in a range of HNSCC.9,10,13,14 However, in this series, HIF-1a expression was not significantly associated with locoregional recurrence or disease-specific mortality. In the study by Aebersold et al., the extent of HIF-1a expression was adversely associated with outcome in 98 patients with SCCa oropharynx.13 In the CHART study, the expression of HIF-2, a HIF-1a homologue, was associated with poor locoregional control and poor survival.9 In the Dutch study of 91 early glottic SCCa, HIF-1a expression was high in 51% of cases and was associated with worse locoregional control and worse overall survival.10 In our study of T1–2 glottic larynx SCCa, there was HIF-1a overexpression in 54% of tumors, but with no association between HIF-1a expresLaryngoscope 125: January 2015
sion and outcome.7 In the present study, HIF-1a positivity was found in 81.6% (79.5% of T3 and 90% of T4 tumors), but was not associated with poor locoregional control or disease-specific survival in the 103 cases. High-risk human papillomavirus (HPV) status was not determined because HPV positivity does not affect rates of hypoxia and expression of CA-9 or HIF-1a.31,32 A very recent study found that HPV expression confers a better prognosis in SCCa larynx33, so the exclusion of any HPV-positive cases might have further strengthened our findings. However, the effect would be small given that the prevalence of high-risk HPV subtypes in SCCa larynx and hypopharynx is only 3% to 10%.34–36 Tumor volume is known to be an adverse prognostic factor in SCCa larynx and hypopharynx, and the present findings confirm this.37–39 A larger tumor volume was sufficient explanation for the worse prognosis of T4 cancers over T3. It has been suggested that the aggregate volume of hypoxic tumor segments has more prognostic significance than the total volume.21 In the present study, tumors were categorized by the most intense region of staining in the initial diagnostic biopsy, and we were not in a position to determine the proportion of each tumor that was hypoxic because these were not resections. The meta-analysis of CA-9 in HNSCC suggested that average CA-9 scores were higher in advanced tumors (T3–4) than in early tumors (P 5 0.01).10 The present study shows that CA-9 positivity Bernstein et al.: Hypoxia-Associated Biomarkers
E13
and tumor volume are independent risk factors, which is in agreement with a study of carcinoma of the cervix in which hypoxia was measured using oxygen electrodes.40 In the cases in which more than one biopsy was available, variation in marker expression was seen mostly in the range 0% to 20%. This opens the question as to how far heterogeneity in areas of presumed hypoxia affects resistance to radiation therapy and outcome, and may explain the differences in the findings with CA-9 and HIF-1a. The heterogeneity of tumor hypoxia might be addressed further by voxel-by-voxel mapping with imaging modalities such as dynamic contrast-enhanced magnetic resonance imaging, although it is not yet clear how to utilize the complex data produced by perfusion imaging.
CONCLUSION HNSCCs have many inherent properties that influence prognosis. In the present study, pretreatment CA-9 expression was prognostic for disease-specific mortality. Hypoxia-associated marker expression and tumor volume might be combined with a prognostic nomogram to help tailor therapy for patients with laryngeal cancer in the multidisciplinary setting.41 Hypoxia modification during radiation therapy may be beneficial in hypoxic laryngeal cancers.42 It also would be particularly helpful to know when to go straight to surgery, because primary surgery is thought to have a better outcome than salvage laryngectomy after failed chemoradiation therapy, where the 5-year disease-specific survival rate is about 58%.43 It has yet to be established whether first-line surgical extirpation would offer a better outcome in hypoxic tumors, and the association of CA-9 expression with disease-specific mortality but not with a more direct effect on radioresistance (locoregional control) might imply that CA-9 expression indicates a more aggressive phenotype.
Acknowledgment The authors acknowledge the support of Mr. Ric Swindell, Clinical Research Statistician, The Christie NHS Foundation Trust.
BIBLIOGRAPHY 1. Bray F, Sankila R, Ferlay J, Parkin DM. Estimates of cancer incidence and mortality in Europe in 1995. Eur J Cancer 2002;38:99–166. 2. Induction chemotherapy plus radiation compared with surgery plus radiation in patients with advanced laryngeal cancer. The Department of Veterans Affairs Laryngeal Cancer Study Group. N Engl J Med 1991;324: 1685–1690. 3. Forastiere AA, Goepfert H, Maor M, et al. Concurrent chemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med 2003;349:2091–2098. 4. Lambert L, Fortin B, Soulieres D, et al. Organ preservation with concurrent chemoradiation for advanced laryngeal cancer: are we succeeding? Int J Radiat Oncol Biol Phys 2010;76:398–402. 5. Pignon JP, Bourhis J, Domenge C, Designe L. Chemotherapy added to locoregional treatment for head and neck squamous-cell carcinoma: three meta-analyses of updated individual data. MACH-NC Collaborative Group. Meta-Analysis of Chemotherapy on Head and Neck Cancer. Lancet 2000;355:949–955. 6. Lefebvre JL, Pointreau Y, Rolland F, et al. Induction chemotherapy followed by either chemoradiotherapy or bioradiotherapy for larynx preservation: the TREMPLIN randomized phase II study. J Clin Oncol 2013; 31:853–859.
Laryngoscope 125: January 2015
E14
7. Douglas CM, Bernstein JM, Ormston VE, et al. Lack of prognostic effect of carbonic anhydrase-9, hypoxia inducible factor-1alpha and bcl-2 in 286 patients with early squamous cell carcinoma of the glottic larynx treated with radiotherapy. Clin Oncol (R Coll Radiol) 2013;25:59–65. 8. Koukourakis MI, Giatromanolaki A, Sivridis E, et al. Hypoxia-regulated carbonic anhydrase-9 (CA9) relates to poor vascularization and resistance of squamous cell head and neck cancer to chemoradiotherapy. Clin Cancer Res 2001;7:3399–3403. 9. Koukourakis MI, Bentzen SM, Giatromanolaki A, et al. Endogenous markers of two separate hypoxia response pathways (hypoxia inducible factor 2 alpha and carbonic anhydrase 9) are associated with radiotherapy failure in head and neck cancer patients recruited in the CHART randomized trial. J Clin Oncol 2006;24:727–735. 10. Schrijvers ML, van der Laan BF, de Bock GH, et al. Overexpression of intrinsic hypoxia markers HIF1alpha and CA-IX predict for local recurrence in stage T1-T2 glottic laryngeal carcinoma treated with radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:161–169. 11. Peridis S, Pilgrim G, Athanasopoulos I, Parpounas K. Carbonic anhydrase9 expression in head and neck cancer: a meta-analysis. Eur Arch Otorhinolaryngol 2011;268:661–670. 12. Eriksen JG, Overgaard J. Lack of prognostic and predictive value of CA IX in radiotherapy of squamous cell carcinoma of the head and neck with known modifiable hypoxia: an evaluation of the DAHANCA 5 study. Radiother Oncol 2007;83:383–388. 13. Aebersold DM, Burri P, Beer KT, et al. Expression of hypoxia-inducible factor-1alpha: a novel predictive and prognostic parameter in the radiotherapy of oropharyngeal cancer. Cancer Res 2001;61:2911–2916. 14. Hui EP, Chan AT, Pezzella F, et al. Coexpression of hypoxia-inducible factors 1alpha and 2alpha, carbonic anhydrase IX, and vascular endothelial growth factor in nasopharyngeal carcinoma and relationship to survival. Clin Cancer Res 2002;8:2595–2604. 15. Wachters JE, Schrijvers ML, Slagter-Menkema L, et al. Prognostic significance of HIF-1a, CA-IX, and OPN in T1-T2 laryngeal carcinoma treated with radiotherapy. Laryngoscope 2013;123:2154–2160. 16. Bernstein JM, Bernstein CR, West CM, Homer JJ. Molecular and cellular processes underlying the hallmarks of head and neck cancer. Eur Arch Otorhinolaryngol 2013;270:2585–2593. 17. Winter SC, Shah KA, Han C, et al. The relation between hypoxiainducible factor (HIF)21alpha and HIF-2alpha expression with anemia and outcome in surgically treated head and neck cancer. Cancer 2006; 107:757–766. 18. Thomlinson RH, Gray LH. The histological structure of some human lung cancers and the possible implications for radiotherapy. Br J Cancer 1955;9:539–549. 19. Teicher BA. Hypoxia and drug resistance. Cancer Metastasis Rev 1994;13: 139–168. 20. Sowter HM, Raval RR, Moore JW, Ratcliffe PJ, Harris AL. Predominant role of hypoxia-inducible transcription factor (Hif)21alpha versus Hif2alpha in regulation of the transcriptional response to hypoxia. Cancer Res 2003;63:6130–6134. 21. Dunst J, Stadler P, Becker A, et al. Tumor volume and tumor hypoxia in head and neck cancers. The amount of the hypoxic volume is important. Strahlenther Onkol 2003;179:521–526. 22. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM. Reporting recommendations for tumour MARKer prognostic studies (REMARK). Br J Cancer 2005;93:387–391. 23. Silva P, Slevin NJ, Sloan P, et al. Prognostic significance of tumor hypoxia inducible factor-1alpha expression for outcome after radiotherapy in oropharyngeal cancer. Int J Radiat Oncol Biol Phys 2008;72:1551–1559. 24. Loncaster JA, Harris AL, Davidson SE, et al. Carbonic anhydrase (CA IX) expression, a potential new intrinsic marker of hypoxia: correlations with tumor oxygen measurements and prognosis in locally advanced carcinoma of the cervix. Cancer Res 2001;61:6394–6399. 25. Wapnir IL, Wartenberg DE, Greco RS. Three dimensional staging of breast cancer. Breast Cancer Res Treat 1996;41:15–19. 26. Janssen HL, Haustermans KM, Balm AJ, Begg AC. Hypoxia in head and neck cancer: how much, how important? Head Neck 2005;27:622–638. 27. Brizel DM, Dodge RK, Clough RW, Dewhirst MW. Oxygenation of head and neck cancer: changes during radiotherapy and impact on treatment outcome. Radiother Oncol 1999;53:113–117. 28. Gatenby RA, Kessler HB, Rosenblum JS, et al. Oxygen distribution in squamous cell carcinoma metastases and its relationship to outcome of radiation therapy. Int J Radiat Oncol Biol Phys 1988;14:831–838. 29. Nordsmark M, Overgaard J. A confirmatory prognostic study on oxygenation status and loco-regional control in advanced head and neck squamous cell carcinoma treated by radiation therapy. Radiother Oncol 2000;57:39–43. 30. Wykoff CC, Beasley NJ, Watson PH, et al. Hypoxia-inducible expression of tumor-associated carbonic anhydrases. Cancer Res 2000;60:7075–7083. 31. Hong A, Zhang M, Veillard AS, et al. The prognostic significance of hypoxia inducing factor 1-alpha in oropharyngeal cancer in relation to human papillomavirus status. Oral Oncol 2013;49:354–359. 32. Kong CS, Narasimhan B, Cao H, et al. The relationship between human papillomavirus status and other molecular prognostic markers in head and neck squamous cell carcinomas. Int J Radiat Oncol Biol Phys 2009; 74:553–561. 33. Shaughnessy JN, Farghaly H, Wilson L, et al. HPV: a factor in organ preservation for locally advanced larynx and hypopharynx cancer? Am J Otolaryngol 2014;35:19–24.
Bernstein et al.: Hypoxia-Associated Biomarkers
34. Chernock RD, Wang X, Gao G, et al. Detection and significance of human papillomavirus, CDKN2A(p16) and CDKN1A(p21) expression in squamous cell carcinoma of the larynx. Mod Pathol 2013;26:223–231. 35. Halec G, Holzinger D, Schmitt M, et al. Biological evidence for a causal role of HPV16 in a small fraction of laryngeal squamous cell carcinoma. Br J Cancer 2013;109:172–183. 36. Anderson CE, McLaren KM, Rae F, Sanderson RJ, Cuschieri KS. Human papilloma virus in squamous carcinoma of the head and neck: a study of cases in south east Scotland. J Clin Pathol 2007;60:439–441. 37. Chen SW, Yang SN, Liang JA, Tsai MH, Shiau AC, Lin FJ. Value of computed tomography-based tumor volume as a predictor of outcomes in hypopharyngeal cancer after treatment with definitive radiotherapy. Laryngoscope 2006;116:2012–2017. 38. Lee WR, Mancuso AA, Saleh EM, Mendenhall WM, Parsons JT, Million RR. Can pretreatment computed tomography findings predict local control in T3 squamous cell carcinoma of the glottic larynx treated with radiotherapy alone? Int J Radiat Oncol Biol Phys 1993;25:683–687.
Laryngoscope 125: January 2015
39. Mancuso AA, Mukherji SK, Schmalfuss I, et al. Preradiotherapy computed tomography as a predictor of local control in supraglottic carcinoma. J Clin Oncol 1999;17:631–637. 40. Fyles A, Milosevic M, Hedley D, et al. Tumor hypoxia has independent predictor impact only in patients with node-negative cervix cancer. J Clin Oncol 2002;20:680–687. 41. Egelmeer AG, Velazquez ER, de Jong JM, et al. Development and validation of a nomogram for prediction of survival and local control in laryngeal carcinoma patients treated with radiotherapy alone: a cohort study based on 994 patients. Radiother Oncol 2011;100:108– 115. 42. Janssens GO, Rademakers SE, Terhaard CH, et al. Accelerated radiotherapy with carbogen and nicotinamide for laryngeal cancer: results of a phase III randomized trial. J Clin Oncol 2012;30:1777– 1783. 43. van der Putten L, de Bree R, Kuik DJ, et al. Salvage laryngectomy: oncological and functional outcome. Oral Oncol 2011;47:296–301.
Bernstein et al.: Hypoxia-Associated Biomarkers
E15
Rhinological and Otological Society, Inc.
Prognostic Value of Hypoxia-Associated Markers in Advanced Larynx and Hypopharynx Squamous Cell Carcinoma Jonathan M. Bernstein, MD, FRCS; Timothy D. Andrews, FRCPath; Nick J. Slevin, FRCR; Catharine M. L. West, PhD; Jarrod J. Homer, MD, FRCS Objectives/Hypothesis: To determine the prognostic value of hypoxia-associated markers carbonic anhydrase-9 (CA-9) and hypoxia-inducible factor-1a (HIF-1a) in advanced larynx and hypopharynx squamous cell carcinoma (SCCa) treated by organ preservation strategies. Study Design: Retrospective cohort study. Methods: Pretreatment CA-9 and HIF-1a expression, clinicopathologic data, and tumor volume were analyzed in a series of 114 patients with T3–4 SCCa larynx or hypopharynx treated by (chemo)radiation. Results: Adverse prognostic factors for locoregional control were T4 classification (P 5 0.008), and for disease-specific survival were CA-9 positivity (P 5 0.039), T4 classification (P 5 0.001), larger tumor volume (P 5 0.004), N1–3 classification (P 5 0.002), and pretreatment hemoglobin < 13.0 g/dl (P 5 0.014). With increasing CA-9 expression, there was a trend to increasing tumor recurrence (Ptrend 5 0.009) and decreasing survival (Ptrend 5 0.002). On multivariate analysis, independent variables were T4 classification (hazard ratio [HR] 13.54, P 5 0.01) for locoregional failure, and CA-9 positivity (HR 5 8.02, P 5 0.042) and higher tumor volume (HR 5 3.33, P 5 0.007) for disease-specific mortality. Conclusion: This is the first study to look specifically at T3 and T4 SCCa larynx and hypopharynx for a relationship between hypoxia-associated marker expression and clinical outcome. Pretreatment immunohistochemical CA-9 expression is an adverse prognostic factor for disease-specific survival, indicating that CA-9 expression may confer a more aggressive tumor phenotype. Key Words: Head and neck squamous cell carcinoma, biomarker, tumor hypoxia, CA-9, HIF-1a, tumor volume, larynx, hypopharynx. Level of Evidence: 4. Laryngoscope, 125:E8–E15, 2015
INTRODUCTION Laryngeal squamous cell carcinoma (SCCa) accounts for 2.4% of cancer cases worldwide.1 Concomitant chemoradiation therapy became a leading modality of treatment for T3 SCCa larynx following the Ann Arbor Veterans Affairs and Radiation Therapy Oncology Group 91-11 studies.2,3 Chemoradiation therapy is larynx-sparing, although swallowing and speech are significantly affected in over 20% of patients.4 For those who develop local recurrence, salvage laryngectomy is usually offered. In T4 SCCa larynx and T3–4 SCCa From the University Department of Otolaryngology–Head & Neck Surgery (J.M.B., J.J.H.), Manchester Royal Infirmary and Manchester Academic Health Science Centre; the Translational Radiobiology Group, Institute of Cancer Sciences (J.M.B., N.J.S., C.M.L.W., J.J.H.), Manchester Academic Health Science Centre, University of Manchester; the Department of Head and Neck Clinical Oncology, The Christie NHS Foundation Trust (N.J.S.), Manchester; and the Department of Pathology, Royal Liverpool University Hospital (T.D.A.), Liverpool, United Kingdom. Editor’s Note: This Manuscript was accepted for publication August 25, 2014. This work was supported by funding from Experimental Cancer Medicine Centre (ECMC) Network. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Jonathan M. Bernstein, MD, FRCS, University Department of Otolaryngology–Head and Neck Surgery, Manchester Royal Infirmary, Oxford Road, Manchester, M13 9WL, U.K. E-mail: [email protected] DOI: 10.1002/lary.24933
Laryngoscope 125: January 2015
E8
hypopharynx, there is a trend toward greater benefit from surgery and postoperative radiation, but some patients are selected for treatment without surgery.5 In some centers, an initial response to induction chemotherapy is used to select patients for organ-preservation therapy, with nonresponders managed by total laryngectomy.6 There is a need to identify those patients in whom advanced SCCa larynx or hypopharynx is likely to progress despite chemoradiation. This study looks for a way to improve the selection of treatment for advanced cancer of the larynx and hypopharynx. There is consensus but not total unanimity among studies of head and neck squamous cell carcinoma (HNSCC) as a whole that the expression of hypoxia-associated markers is associated with adverse outcome.7–15 Tumor hypoxia is likely to represent an emerging hallmark of HNSCC and is associated with malignant progression and resistance to radiation and cytotoxic chemotherapy16 (although there is limited evidence suggesting that hypoxia may adversely affect the outcome after surgery as well17). Hypoxic cells are more resistant to damage from ionizing radiation because oxygen promotes the generation of free radicals.18 Chemotherapy drugs must diffuse from an adequate blood supply, and hypoxic cells spend less time in the growth phase when chemotherapeutic agents are most effective.19 Hypoxia-inducible factor-1a (HIF21a) is a transcription factor, which in normoxia is degraded after binding Bernstein et al.: Hypoxia-Associated Biomarkers
with von Hippel-Lindau protein and ubiquitin. In hypoxia (pO2 < 20 mm Hg), HIF-1a translocates to the nucleus and binds with HIF-1b and the hypoxiaresponsive elements of hypoxia-regulated genes, such as the gene encoding carbonic anhydrase-9 (CA-9). In hypoxia, HIF-1a induces the expression of CA-9.20 This is the first study to look specifically at locoregionally advanced SCCa larynx and hypopharynx for a relationship between the expression of hypoxia-associated markers HIF-1a and CA-9 and clinical outcome after organ preservation therapy. Tumor volume, an independent marker of prognosis, was measured because it has been suggested that the volume of hypoxic tumor may be an important consideration.21
ing high and low marker expression and running sections from these biopsies with each batch.
Histologic Analysis Nuclear HIF-1a and plasma membrane CA-9 staining were scored according to published scoring methods by a histopathologist (T.D.A.) and researcher (J.M.B.) who were blinded to outcome.23,24 Percentage scores in the single 4003 field containing SCCa with greatest staining were recorded. Sections were categorized as negative where staining was less than 1% and positive where 1% or more. Dysplastic or necrotic areas were excluded. Where more than one biopsy sample containing invasive carcinoma was available, the mean score was recorded. Necrosis was determined as either present or absent on hematoxylin and eosin-stained sections.
MATERIALS AND METHODS
Tumor Volume Measurement
Patients with American Joint Committee on Cancer (AJCC, 7th ed.) T3-4 SCCa larynx or hypopharynx treated with curative intent by (chemo)radiation were identified using the radiation therapy database at The Christie NHS Foundation Trust, United Kingdom. Patients with T4 tumors included in the study were deemed by the multidisciplinary team to be unfit for surgical extirpation and were therefore treated with curative intent by organ preservation strategies. All patients were diagnosed between June 2005 and December 2010. Clinicopathologic and outcomes data were collected retrospectively. Approval for the study was obtained from an ethics committee (ref: 11/NW/0284) and institutional research and development board. Pretreatment biopsy formalin-fixed paraffin-embedded tumor blocks were requested. SCCa was confirmed by a consultant histopathologist (T.D.A.). The REMARK reporting recommendations for prognostic biomarker studies were used.22
Primary tumor volume was determined using pretreatment computed tomography in patients with digital image archiving. The largest two perpendicular measurements in a single axial slice and the maximum coronal measurement were incorporated into the formula p/6 3 (length) 3 (width) 3 (height) to determine the ellipsoid volume.25
Immunohistochemistry Marker expression was examined in 4-lm sections. HIF-1a was detected using a Bond-Max autostainer (Leica Microsystems Ltd., Milton Keynes, U.K.) and the Bond Polymer Refine Detection system (Leica Biosystems, Newcastle-on-Tyne, U.K.) involving a biotin-free polymeric horseradish-peroxidase (HRP) linker antibody conjugate system to detect the antigen followed by 3,3’-diaminobenzidine (DAB) tetrahydrochloride for visualization and hematoxylin as counterstain. After deparaffinization, rehydration, and microwave epitope retrieval, antigen detection was undertaken using mouse monoclonal anti-human HIF-1a diluted 1:50 (61099, BD Biosciences, Oxford, U.K.) at pH9 at room temperature for 15 minutes. Mouse IgG1 (X0931, Dako, Ely, U.K.) was used as a negative control. CA-9 was detected using a Biogenex i6000 autostainer (Launch Diagnostics Ltd, Longfield, U.K.) and the EnVision Plus HRPconjugated polymer system (K4007, Dako). After dewaxing and rehydration, endogenous peroxidase was quenched with 0.3% hydrogen peroxide (Sigma, Poole, U.K.), and nonspecific staining was blocked with 10% casein solution (Vector Laboratories Inc., Peterborough, U.K.). The primary antibody was mouse monoclonal antihuman CA-9 diluted 1:50 (clone M75; gift from Profs. Pastorekova and Pastorek, Slovak Academy of Sciences, Bratislava, Slovakia), and sections were incubated at room temperature for 30 minutes. Mouse IgG2B (X944, Dako) was used as a negative control and DAB was used for visualization (K3467, Dako). Sections were removed from the processor and counterstained with Gill’s hematoxylin and dehydrated. Batchto-batch variation was assessed by choosing two sections show-
Statistical Analyses SPSS version 19.0 (IBM, Armonk, NY) was used. Time to locoregional control and disease-specific survival from the date of biopsy were estimated using the Kaplan-Meier method. Survival curves were compared in univariate analyses using the Mantell-Cox log-rank test; categorical variables were compared using the X2 test; and multivariate analyses were undertaken using the Cox proportional hazards model. P values < 0.05 were considered statistically significant. The log-rank test was used to determine a trend to adverse clinical outcome with increasing CA-9 and HIF-1a expression, with each marker categorized into four groups (0%, 1%–9%, 10%–29%, 30%). CA-9 and HIF-1a expression, T- and N-classification, tumor volume, hemoglobin level, tumor differentiation, performance status, gender, age, site, and alcohol and smoking history were analyzed for prognostic significance.
RESULTS One-hundred-and-fourteen patients were identified (86 male, 28 female) with T3–4 SCCa larynx or hypopharynx treated by (chemo)radiation with curative intent. Patient characteristics are given in Table I. Eighty-nine patients (78.1%) had SCCa larynx (51 glottic, 37 supraglottic, 1 subglottic), and 25 patients (21.9%) had SCCa hypopharynx (15 piriform sinus, 7 postcricoid, 3 posterior pharyngeal wall). One hundred patients had T3 tumors and 14 had T4 tumors. Tumor blocks for 104 patients were obtained of which 103 blocks contained tumor. These tumors included 83 cases of T3 SCCa larynx. Either IMRT or 3D-conformal radiation (4–6 MeV linear accelerator) was used. Eighty-two (71.9%) performance status 0 or 1 patients received chemoradiation, and 11 of these received induction chemotherapy due to a high tumor burden (Table I). No patients were treated by surgery following induction chemotherapy during the period of study.
TABLE I. Patient Characteristics. Characteristic
No. of Cases
Percentage of Cases (%)
Male Female
86 28
75.4 24.6
AJCC T-classification
T3
100
87.7
Tumor differentiation
T4 Well differentiated
14 25
12.3 21.9
Moderately differentiated
41
36.0
Poorly differentiated Undifferentiated
40 2
35.1 1.8
Gender
Not recorded Tumor site
Larynx
Glottis Supraglottis
Hypopharynx
Smoking
Alcohol
1
0.9
Piriform sinus Postcricoid hypopharynx
15 7
13.2 6.1
Concurrent chemotherapy regimen Cetuximab
3
2.6
0 1
89 10
78.1 8.8
2a
3
2.6
2b 2c
7 4
6.1 3.5
3
1
0.9
Never Ex 1 year
6 26
5.3 22.8
Ex < 1 year
37
32.5
Current Never heavy
45 60
39.5 52.6
Previous heavy
32
28.1
Current heavy 0
22 46
19.3 40.4
1
42
36.8
2 3
12 6
10.5 5.3
Not recorded
8
7.0
1–2 3–4
3 8
2.6 7.0
Capecitabine
62
54.4
WHO performance status
Number of induction chemotherapy cycles with TPF
Cisplatin Radiation therapy alone Radiation therapy dose
5.3 44.7 32.5
Subglottis
Posterior pharyngeal wall AJCC N-classification
6 51 37
11
9.6
22
19.3
55.0 Gy in 20 fractions
19 92
16.7 80.7
62.5–66.0 Gy in 30 fractions
22
19.3
Characteristic
Mean/Median Value
Age (years)
63 (mean)
CA-9 score (%)
5 (median)
HIF-1a score (%)
10 (median)
Range
37–86 0–80 0–80 Tumor volume (mL)
5.70 (median) 0.37–48.71
AJCC 5 American Joint Committee on Cancer; CA-9 5 carbonic anhydrase-9; HIF-1a 5 hypoxia-inducible factor-1a; TPF 5 docetaxel, cisplatin, and 5flurouracil; WHO 5 World Health Organization.
Laryngoscope 125: January 2015
E10
Bernstein et al.: Hypoxia-Associated Biomarkers
Fig. 1. Vertical scatter plots showing percentage marker staining from two or more separate biopsies from individual tumors for CA-9 in (a) and HIF-1a in (b).
CA-9 staining was negative in 25 (24.3%) cases and positive in 78 (75.7%) cases. HIF-1a staining was negative in 18 (17.5%) cases and positive in 85 (82.5%) cases. There was a correlation between the expression of CA-9 and HIF-1a (Spearman’s rho, 0.58; P < 0.0001). More than one tumor block was available in 14 patients. In these cases, there was variation mainly in the range 0 to 20% (Fig. 1). The results were identical in five cases for CA-9 and three cases for HIF-1a. In patients where there was one negative-scoring biopsy, recording the mean score resulted in six cases assigned CA-9 positive and three HIF-1a positive. Thirty-two of 114 patients developed locoregional recurrence, and a further seven had residual tumor. Of these 39 patients, 15 underwent salvage (pharyngo)laryngectomy usually with bilateral neck dissection, 19 were considered to be unfit for salvage surgery, and five were designated with recurrence because they died without another clear cause. In total, there were 42 deaths, of which 34 were deemed to be from the index HNSCC and eight were from causes such as pancreatic cancer or pneumonia in patients without recurrence. Seven of the patients who underwent salvage surgery died subsequently. As determined by Kaplan-Meier analysis after a median follow-up of 44.8 months, 5-year locoregional control was 63.5%; disease-specific survival was 65.0%; and overall survival was 55.4%. Estimated 5-year locoregional control was not affected by primary tumor site (P 5 0.498), but estimated disease-specific survival was better for carcinoma of larynx (70% at 5 years) than hypopharynx (45.1% at 5 years) (P 5 0.017) (Table II). T4-classification was an adverse prognostic factor for locoregional tumor recurrence (P 5 0.008). Adverse prognostic factors for disease-specific survival were T4classification (P 5 0.001), CA-9 positivity (P 5 0.039), Laryngoscope 125: January 2015
higher N-classification (P 5 0.002), and pretreatment hemoglobin level 13.0 13.0
76 38
58.5 49.6
33.1–77.1 30.0–66.5
0.074
75.0 47.2
62.0–84.1 27.5–64.6
0.014*
Hb2 (g/dL)
> 13.6
56
70.5
55.9–81.0
0.418
76.9
61.9–86.6
0.135
13.6
58
56.9
41.4–69.7
54.9
38.3–68.8
*Statistically significant result (P < 0.05). † Mantel-Cox log-rank test. ‡ Tumor volume stratified by the median (5.70 mL). Hb stratified by WHO definition of anemia in males in Hb1 and the median for this series in Hb2. AJCC 5 American Joint Committee on Cancer (7th Ed.); CA-9 5 carbonic anhydrase-9; CI 5 confidence interval; DSS 5 disease-specific survival; Hb 5 hemoglobin level; HIF-1a 5 hypoxia-inducible factor-1a; LRC 5 locoregional control; WHO, World Health Organization.
DISCUSSION Outcomes in HNSCC after radical radiation therapy are adversely influenced by tumor hypoxia measured by a number of approaches, including polarographic electrodes and immunohistochemical markers.26–29 In the present study, the first to look specifically at the prognostic effect of hypoxia markers in locoregionally advanced SCCa larynx or hypopharynx, the expression of CA-9 was prognostic for disease-specific mortality after therapy with organ preservation. CA-9 expression was associated with disease-specific mortality in about half the cases; and CA-9 negativity was associated mostly with disease-specific survival. CA-9 expression was associated inversely with microvessel density and treatment response in locally advanced HNSCC treated by chemoradiation in a study by Koukourakis et al.8 CA29 expression was associated with poor locoregional control and poor survival in HNSCC in the Continuous Hyperfractionated Accelerated Radiotherapy study, in which CA29 and HIF-1a showed prognostic potential in 198 pretreatment biopsy specimens.9 In the study by Schrijvers et al., among 91 patients with T1–2N0 SCCa glottic larynx, CA-9 and HIF-1a were associated with reduced disease-free survival; but a further study by the same authors of 60 patients with T1–2 SCCa supraglottic larynx showed a lack of prognostic significance of CA-9 and HIF-1a.10,15 A meta-analysis of 16 studies including 1,470 patients showed that high CA29 expression was associated with reduced survival (P < 0.0001) and reduced disease-free survival (P 5 0.0008) in HNSCC.11 There was no analyLaryngoscope 125: January 2015
E12
sis according to tumor site within the head and neck, and only one article on laryngeal cancer that concerned T1–T2 tumors. The results of the present study are in keeping with the meta-analysis in that CA-9 expression is prognostic in HNSCC. Not all studies have shown a relationship between CA9 expression and worse outcome. In the Danish Head and Neck Cancer Group (DAHANCA) 5 study of 320 patients with SCCa supraglottic larynx and pharynx treated by radiation, there was no prognostic or predictive potential for CA9.12 Half these patients had T1–2 tumors, and the other half were T3–4; both groups were about 70% positive for CA-9 expression, and the survival data were not analyzed according to tumor stage. Our own work has focused largely on SCCa larynx. In T1–2 SCCa glottic larynx, CA-9 expression was found in 72% of tumors, and there was no statistically significant association with outcome.7 The present study concerns T3–4 SCCa larynx and hypopharynx. In the whole group, there was positive CA-9 expression in 76% of cases (73% of T3 and 80% of T4 tumors), and CA-9 expression was associated with disease-specific mortality. The results differ from the negative results of the DAHANCA 5 study, and support the previous findings in advanced HNSCC in general that CA-9 expression is associated with disease-specific mortality.11 In the present study, biomarker analysis of the T3 SCCa subgroup lost statistical significance while the trends were maintained, indicating the reduction in patient numbers for this subanalysis. We found CA-9 expression to be a statistically significant prognostic marker of disease-specific mortality but not locoregional failure, a finding that is in Bernstein et al.: Hypoxia-Associated Biomarkers
33
Fig. 2. Kaplan-Meier plots of 103 cases of T3–4 squamous cell carcinoma of the larynx or hypopharynx for locoregional control as a function of (a) CA-9 and (b) HIF-1a expression, and disease-specific survival as a function of (c) CA-9 and (d) HIF-1a expression. CA-9 5 carbonic anhydrase-9; DSS 5 disease-specific survival; HIF-1a 5 hypoxia-inducible factor1a; LRC 5 locoregional control.
agreement with previous studies of HNSCC and cervix cancer.11,24 CA-9 is a better marker of chronic than acute hypoxia30; whereas both acute and chronic hypoxia confer radioresistance, and hence impaired locoregional control. Given the association of CA-9 with disease-specific mortality but not locoregional control, Loncaster et al. suggested that CA-9 expression in cervix cancer might reflect a more aggressive tumor phenotype rather than radioresistance.24 In that case, surgery might not be a suitable alternative treatment option in patients identified as having a high risk of disease-specific mortality following radiation therapy. Further studies are warranted. The expression of HIF-1a predicted an adverse prognosis in a range of HNSCC.9,10,13,14 However, in this series, HIF-1a expression was not significantly associated with locoregional recurrence or disease-specific mortality. In the study by Aebersold et al., the extent of HIF-1a expression was adversely associated with outcome in 98 patients with SCCa oropharynx.13 In the CHART study, the expression of HIF-2, a HIF-1a homologue, was associated with poor locoregional control and poor survival.9 In the Dutch study of 91 early glottic SCCa, HIF-1a expression was high in 51% of cases and was associated with worse locoregional control and worse overall survival.10 In our study of T1–2 glottic larynx SCCa, there was HIF-1a overexpression in 54% of tumors, but with no association between HIF-1a expresLaryngoscope 125: January 2015
sion and outcome.7 In the present study, HIF-1a positivity was found in 81.6% (79.5% of T3 and 90% of T4 tumors), but was not associated with poor locoregional control or disease-specific survival in the 103 cases. High-risk human papillomavirus (HPV) status was not determined because HPV positivity does not affect rates of hypoxia and expression of CA-9 or HIF-1a.31,32 A very recent study found that HPV expression confers a better prognosis in SCCa larynx33, so the exclusion of any HPV-positive cases might have further strengthened our findings. However, the effect would be small given that the prevalence of high-risk HPV subtypes in SCCa larynx and hypopharynx is only 3% to 10%.34–36 Tumor volume is known to be an adverse prognostic factor in SCCa larynx and hypopharynx, and the present findings confirm this.37–39 A larger tumor volume was sufficient explanation for the worse prognosis of T4 cancers over T3. It has been suggested that the aggregate volume of hypoxic tumor segments has more prognostic significance than the total volume.21 In the present study, tumors were categorized by the most intense region of staining in the initial diagnostic biopsy, and we were not in a position to determine the proportion of each tumor that was hypoxic because these were not resections. The meta-analysis of CA-9 in HNSCC suggested that average CA-9 scores were higher in advanced tumors (T3–4) than in early tumors (P 5 0.01).10 The present study shows that CA-9 positivity Bernstein et al.: Hypoxia-Associated Biomarkers
E13
and tumor volume are independent risk factors, which is in agreement with a study of carcinoma of the cervix in which hypoxia was measured using oxygen electrodes.40 In the cases in which more than one biopsy was available, variation in marker expression was seen mostly in the range 0% to 20%. This opens the question as to how far heterogeneity in areas of presumed hypoxia affects resistance to radiation therapy and outcome, and may explain the differences in the findings with CA-9 and HIF-1a. The heterogeneity of tumor hypoxia might be addressed further by voxel-by-voxel mapping with imaging modalities such as dynamic contrast-enhanced magnetic resonance imaging, although it is not yet clear how to utilize the complex data produced by perfusion imaging.
CONCLUSION HNSCCs have many inherent properties that influence prognosis. In the present study, pretreatment CA-9 expression was prognostic for disease-specific mortality. Hypoxia-associated marker expression and tumor volume might be combined with a prognostic nomogram to help tailor therapy for patients with laryngeal cancer in the multidisciplinary setting.41 Hypoxia modification during radiation therapy may be beneficial in hypoxic laryngeal cancers.42 It also would be particularly helpful to know when to go straight to surgery, because primary surgery is thought to have a better outcome than salvage laryngectomy after failed chemoradiation therapy, where the 5-year disease-specific survival rate is about 58%.43 It has yet to be established whether first-line surgical extirpation would offer a better outcome in hypoxic tumors, and the association of CA-9 expression with disease-specific mortality but not with a more direct effect on radioresistance (locoregional control) might imply that CA-9 expression indicates a more aggressive phenotype.
Acknowledgment The authors acknowledge the support of Mr. Ric Swindell, Clinical Research Statistician, The Christie NHS Foundation Trust.
BIBLIOGRAPHY 1. Bray F, Sankila R, Ferlay J, Parkin DM. Estimates of cancer incidence and mortality in Europe in 1995. Eur J Cancer 2002;38:99–166. 2. Induction chemotherapy plus radiation compared with surgery plus radiation in patients with advanced laryngeal cancer. The Department of Veterans Affairs Laryngeal Cancer Study Group. N Engl J Med 1991;324: 1685–1690. 3. Forastiere AA, Goepfert H, Maor M, et al. Concurrent chemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med 2003;349:2091–2098. 4. Lambert L, Fortin B, Soulieres D, et al. Organ preservation with concurrent chemoradiation for advanced laryngeal cancer: are we succeeding? Int J Radiat Oncol Biol Phys 2010;76:398–402. 5. Pignon JP, Bourhis J, Domenge C, Designe L. Chemotherapy added to locoregional treatment for head and neck squamous-cell carcinoma: three meta-analyses of updated individual data. MACH-NC Collaborative Group. Meta-Analysis of Chemotherapy on Head and Neck Cancer. Lancet 2000;355:949–955. 6. Lefebvre JL, Pointreau Y, Rolland F, et al. Induction chemotherapy followed by either chemoradiotherapy or bioradiotherapy for larynx preservation: the TREMPLIN randomized phase II study. J Clin Oncol 2013; 31:853–859.
Laryngoscope 125: January 2015
E14
7. Douglas CM, Bernstein JM, Ormston VE, et al. Lack of prognostic effect of carbonic anhydrase-9, hypoxia inducible factor-1alpha and bcl-2 in 286 patients with early squamous cell carcinoma of the glottic larynx treated with radiotherapy. Clin Oncol (R Coll Radiol) 2013;25:59–65. 8. Koukourakis MI, Giatromanolaki A, Sivridis E, et al. Hypoxia-regulated carbonic anhydrase-9 (CA9) relates to poor vascularization and resistance of squamous cell head and neck cancer to chemoradiotherapy. Clin Cancer Res 2001;7:3399–3403. 9. Koukourakis MI, Bentzen SM, Giatromanolaki A, et al. Endogenous markers of two separate hypoxia response pathways (hypoxia inducible factor 2 alpha and carbonic anhydrase 9) are associated with radiotherapy failure in head and neck cancer patients recruited in the CHART randomized trial. J Clin Oncol 2006;24:727–735. 10. Schrijvers ML, van der Laan BF, de Bock GH, et al. Overexpression of intrinsic hypoxia markers HIF1alpha and CA-IX predict for local recurrence in stage T1-T2 glottic laryngeal carcinoma treated with radiotherapy. Int J Radiat Oncol Biol Phys 2008;72:161–169. 11. Peridis S, Pilgrim G, Athanasopoulos I, Parpounas K. Carbonic anhydrase9 expression in head and neck cancer: a meta-analysis. Eur Arch Otorhinolaryngol 2011;268:661–670. 12. Eriksen JG, Overgaard J. Lack of prognostic and predictive value of CA IX in radiotherapy of squamous cell carcinoma of the head and neck with known modifiable hypoxia: an evaluation of the DAHANCA 5 study. Radiother Oncol 2007;83:383–388. 13. Aebersold DM, Burri P, Beer KT, et al. Expression of hypoxia-inducible factor-1alpha: a novel predictive and prognostic parameter in the radiotherapy of oropharyngeal cancer. Cancer Res 2001;61:2911–2916. 14. Hui EP, Chan AT, Pezzella F, et al. Coexpression of hypoxia-inducible factors 1alpha and 2alpha, carbonic anhydrase IX, and vascular endothelial growth factor in nasopharyngeal carcinoma and relationship to survival. Clin Cancer Res 2002;8:2595–2604. 15. Wachters JE, Schrijvers ML, Slagter-Menkema L, et al. Prognostic significance of HIF-1a, CA-IX, and OPN in T1-T2 laryngeal carcinoma treated with radiotherapy. Laryngoscope 2013;123:2154–2160. 16. Bernstein JM, Bernstein CR, West CM, Homer JJ. Molecular and cellular processes underlying the hallmarks of head and neck cancer. Eur Arch Otorhinolaryngol 2013;270:2585–2593. 17. Winter SC, Shah KA, Han C, et al. The relation between hypoxiainducible factor (HIF)21alpha and HIF-2alpha expression with anemia and outcome in surgically treated head and neck cancer. Cancer 2006; 107:757–766. 18. Thomlinson RH, Gray LH. The histological structure of some human lung cancers and the possible implications for radiotherapy. Br J Cancer 1955;9:539–549. 19. Teicher BA. Hypoxia and drug resistance. Cancer Metastasis Rev 1994;13: 139–168. 20. Sowter HM, Raval RR, Moore JW, Ratcliffe PJ, Harris AL. Predominant role of hypoxia-inducible transcription factor (Hif)21alpha versus Hif2alpha in regulation of the transcriptional response to hypoxia. Cancer Res 2003;63:6130–6134. 21. Dunst J, Stadler P, Becker A, et al. Tumor volume and tumor hypoxia in head and neck cancers. The amount of the hypoxic volume is important. Strahlenther Onkol 2003;179:521–526. 22. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM. Reporting recommendations for tumour MARKer prognostic studies (REMARK). Br J Cancer 2005;93:387–391. 23. Silva P, Slevin NJ, Sloan P, et al. Prognostic significance of tumor hypoxia inducible factor-1alpha expression for outcome after radiotherapy in oropharyngeal cancer. Int J Radiat Oncol Biol Phys 2008;72:1551–1559. 24. Loncaster JA, Harris AL, Davidson SE, et al. Carbonic anhydrase (CA IX) expression, a potential new intrinsic marker of hypoxia: correlations with tumor oxygen measurements and prognosis in locally advanced carcinoma of the cervix. Cancer Res 2001;61:6394–6399. 25. Wapnir IL, Wartenberg DE, Greco RS. Three dimensional staging of breast cancer. Breast Cancer Res Treat 1996;41:15–19. 26. Janssen HL, Haustermans KM, Balm AJ, Begg AC. Hypoxia in head and neck cancer: how much, how important? Head Neck 2005;27:622–638. 27. Brizel DM, Dodge RK, Clough RW, Dewhirst MW. Oxygenation of head and neck cancer: changes during radiotherapy and impact on treatment outcome. Radiother Oncol 1999;53:113–117. 28. Gatenby RA, Kessler HB, Rosenblum JS, et al. Oxygen distribution in squamous cell carcinoma metastases and its relationship to outcome of radiation therapy. Int J Radiat Oncol Biol Phys 1988;14:831–838. 29. Nordsmark M, Overgaard J. A confirmatory prognostic study on oxygenation status and loco-regional control in advanced head and neck squamous cell carcinoma treated by radiation therapy. Radiother Oncol 2000;57:39–43. 30. Wykoff CC, Beasley NJ, Watson PH, et al. Hypoxia-inducible expression of tumor-associated carbonic anhydrases. Cancer Res 2000;60:7075–7083. 31. Hong A, Zhang M, Veillard AS, et al. The prognostic significance of hypoxia inducing factor 1-alpha in oropharyngeal cancer in relation to human papillomavirus status. Oral Oncol 2013;49:354–359. 32. Kong CS, Narasimhan B, Cao H, et al. The relationship between human papillomavirus status and other molecular prognostic markers in head and neck squamous cell carcinomas. Int J Radiat Oncol Biol Phys 2009; 74:553–561. 33. Shaughnessy JN, Farghaly H, Wilson L, et al. HPV: a factor in organ preservation for locally advanced larynx and hypopharynx cancer? Am J Otolaryngol 2014;35:19–24.
Bernstein et al.: Hypoxia-Associated Biomarkers
34. Chernock RD, Wang X, Gao G, et al. Detection and significance of human papillomavirus, CDKN2A(p16) and CDKN1A(p21) expression in squamous cell carcinoma of the larynx. Mod Pathol 2013;26:223–231. 35. Halec G, Holzinger D, Schmitt M, et al. Biological evidence for a causal role of HPV16 in a small fraction of laryngeal squamous cell carcinoma. Br J Cancer 2013;109:172–183. 36. Anderson CE, McLaren KM, Rae F, Sanderson RJ, Cuschieri KS. Human papilloma virus in squamous carcinoma of the head and neck: a study of cases in south east Scotland. J Clin Pathol 2007;60:439–441. 37. Chen SW, Yang SN, Liang JA, Tsai MH, Shiau AC, Lin FJ. Value of computed tomography-based tumor volume as a predictor of outcomes in hypopharyngeal cancer after treatment with definitive radiotherapy. Laryngoscope 2006;116:2012–2017. 38. Lee WR, Mancuso AA, Saleh EM, Mendenhall WM, Parsons JT, Million RR. Can pretreatment computed tomography findings predict local control in T3 squamous cell carcinoma of the glottic larynx treated with radiotherapy alone? Int J Radiat Oncol Biol Phys 1993;25:683–687.
Laryngoscope 125: January 2015
39. Mancuso AA, Mukherji SK, Schmalfuss I, et al. Preradiotherapy computed tomography as a predictor of local control in supraglottic carcinoma. J Clin Oncol 1999;17:631–637. 40. Fyles A, Milosevic M, Hedley D, et al. Tumor hypoxia has independent predictor impact only in patients with node-negative cervix cancer. J Clin Oncol 2002;20:680–687. 41. Egelmeer AG, Velazquez ER, de Jong JM, et al. Development and validation of a nomogram for prediction of survival and local control in laryngeal carcinoma patients treated with radiotherapy alone: a cohort study based on 994 patients. Radiother Oncol 2011;100:108– 115. 42. Janssens GO, Rademakers SE, Terhaard CH, et al. Accelerated radiotherapy with carbogen and nicotinamide for laryngeal cancer: results of a phase III randomized trial. J Clin Oncol 2012;30:1777– 1783. 43. van der Putten L, de Bree R, Kuik DJ, et al. Salvage laryngectomy: oncological and functional outcome. Oral Oncol 2011;47:296–301.
Bernstein et al.: Hypoxia-Associated Biomarkers
E15
