Volumetric stratification of cT4 stage head and neck cancer

Background Locoregionally advanced stage head and neck cancer (HNC) is known for unfavorable outcome with only ~ 40–50 % 3-year overall survival (OS). Clinical T4 stage includes a wide range of tumor burden. The lack of further nonsurgical subgrouping of cT4 stage makes intercenter outcome of irradiated cohorts difficult. Aim of this analysis was to further stratify cT4 stage HNC using volumetric staging. Material and methods Between January 2002 and January 2013, a total of 201 cT4 stage squamous cell cancer (SCC) HNC patients referred to our center for curative definitive radiation were consecutively irradiated. Radiation was performed using modulated techniques. Total gross tumor volumes (tGTV: primary + nodal tumor volume) of all patients have retrospectively been stratified using a prospectively evaluated volumetric staging system which bases on 3 cut-offs (15/70/130 ml), translating into 4 prognostic subgroups [V1: 1–15 ml (n = 15), V2: 16–70 ml (108), V3: 71–130 ml (62), V4: > 130 ml (16)]. OS, disease-free survival (DFS), locoregional control (LRC), and distant metastasis-free survival (DMFS) rates were calculated. Results The mean/median follow-up was 31/23 months (range 1–116 months). The 3-year OS, DFS, LRC, and DMFS rates of the entire cohort were 63, 44, 48, and 77 %, respectively. Volumetric staging revealed its potential to prognostically statistically significantly divide the cT4 cohort into 4 volume subgroups (V1/2/3/4): OS: 90 %/72 %/58 %/18 %; DFS: 83 %/50 %/39 %/10 %; LRC: 81 %/53 %/47 %/15 %; DMFS: 93 %/90 %/70 %/41 %, all p < 0.0001. Conclusion Volumetric staging allowed a highly statistically significant stratification of cT4 HNC stages into prognostic subgroups, which offers the chance of better intercenter comparability of irradiated advanced stage HNC cohorts.

Advanced stage head and neck cancer (HNC) is known for generally unfavorable outcome with only ~40-50% 3-year overall survival [1,2,3]. Clinical T4 stage includes a wide range of tumor volumes. The lack of further nonsurgical subgrouping of cT4 stage makes intercenter comparison of outcome results in irradiated cT4 patient cohorts difficult. The estimation of operability (cT4a versus cT4b) is sometimes quite dependent of a surgeon's individual opinion and experience. In addition, the in-or exclusion of very advanced cT4 any NM0 into curatively aimed treatment regimens remains quite subjective.
The aim of this analysis was to further stratify cT4 stage squamous cell HNC disease using volumetric staging. This was performed with the help of a formerly prospectively tested and published volumetric scoring system [4,5,6,7]. Using this scoring system, we previously demonstrated that volumetric staging was superior compared to the standard TN/AJCC systems regarding predictive power of disease control and survival of our irradiated cohorts.
Included in the presented analysis were all cT4 stage primary squamous cell cancer (SCC) HNC patients referred for definitive radiation.
All patients underwent modulated radiation therapy using simultaneously integrated boost techniques [SIB-IMRT/ SIB-volumetric modulated arc therapy (SIB-VMAT)]. In 84%, concomitant cis-platin chemotherapy (40 mg/m 2 /radiation week) or cetuximab (loading dose 400 mg/m 2 , followed by concomitant doses of 2250 mg/m 2 /radiation week) was administered. In 36 patients with very advanced disease of questionably curable stage, TPF (docetaxel, cis platin, 5-fluorouracil)-based induction chemotherapy was given as a decision aid to add or not curatively intended radiation. The remaining 16% of patients were treated with radiation only because of age or substantial comorbidity.
All GTVs were contoured or reviewed by at least one of the authors on all relevant axial computerized images without using interpolation; in most cases the contouring was also reviewed by a third staff physician. In addition, the wide volumetric ranges (cut-offs 15/70/130 ml) render the system quite robust with respect to interindividual contouring differences. Volumetric three-dimensional measurements (cm 3 ) of contoured structures were

Outcome prediction by volumetric scoring
Between January 2002 and January 2013, a total of 201 cT4 stage SCC HNC patients were curatively treated at our department. The mean/median follow-up was 31/23 months (range 1-116 months). In all, 67% of all patients were alive at last follow-up, and 49% had no signs of disease. Of the 33% of patients who had died, 24% died due to disease-related reasons. The 3-year OS, DFS, LRC, and DMFS rates of the entire cohort were 63, 44, 48, and 77%, respectively.

Additional parameters with potential impact on disease control and OAS
The following parameters were tested in univariate analysis:

Discussion
Aim of this work was to assess the potential of volumetric stratification of our cT4 SCC HNC cohort into different prognostic subgroups. We found volumetric stratification highly statistically significant in predicting outcome for different volume subgroups in the assessed cT4 HNC cohort. The volumetric system itself is considered robust with respect to interobserver GTV contouring, as its cut offs values differ markedly (15 ml/70 ml/130 ml) [4,7]. The potential benefit of the assessed stratification lays in its more precise prediction of disease control in irradiated cT4 patient cohorts, and therefore more accurate characterization of cT4 cohorts for intercenter comparison purposes. A weakness of this study is its retrospective stratification approach, which however applied a prospectively tested staging system [4,5,6,7]. In addition, the assessed cohort includes different unbalanced tumor sites as well as unbalanced volume subgroups (. Tab. 1).

Keywords
Volumetric staging · cT4 stage tumors · Head and neck neoplasms · Neoplasm staging · Prognosis

Literature on head and neck cancer (HNC) outcome prediction based on volumetric classifications
own system included. All but two analyses showed significant difference in outcome between larger vs smaller tumor volumes. Been et al. [34] failed to demonstrate statistical significance between pGTV and locoregional outcome, perhaps due to not considering the nodal tumor volume which may significantly impact locoregional outcome. Mendenhall et al. [8] found no outcome difference in tumors of the hypopharynx/base of tongue/ posterior tonsillar pillar when using a cut off value of 6 ml. This cut-off may have been too low. The data presented here are derived from a cohort treated with IMRT techniques, with previous careful staging (in most cases using PET-CT) [36,37].

Conclusion
Volumetric staging was shown to allow for highly statistically significantly stratification of cT4 stage SCC HNC into different prognostic subgroups, offering the option of better comparability of irradiated advanced stage HNC cohorts.