Does Neck Stage Predict Local Control After Irradiation for Head and Neck Cancer?

The review by Mendenhall et al presents selected papers pertinentto the effect of metastatic nodes on local control in patientswith head and neck cancer. These data are retrospective and, asthe authors point out, do not resolve the matter.

Two Studies Not Included in the Review

The authors do not include a paper by Griffin et al in their review[1]. These researchers analyzed 997 patients with primary headand neck cancer who were entered into the Radiation Therapy OncologyGroup (RTOG) head and neck cancer registry and were treated byradiotherapy alone. Pertinent data, such as tumor site, T stage,and N stage, were all prospectively recorded.

Griffin et al found that T stage, N stage, primary site, and initialKarnofsky performance score were significant independent predictorsof primary tumor response. They designed a multivariate responsemodel using these factors to predict primary tumor response, thepredictive accuracy of which was very impressive. It was testedby comparing predicted and observed tumor clearance rates (O/Pratio) for each independent variable. For the 996 patients evaluated,the ratios according to N stage were: N0, 1.01; N1, 0.93; N2,1.03; and N3, 1.01.

In contrast to this is a very fine paper from Spain by Cerezoet al [2], who analyzed 492 squamous cell carcinomas of the headand neck with positive nodes. The paper focused on failure inpatients with clinically positive nodes and the identificationof prognostic factors for survival and control. The authors failedto show an effect on primary tumor control except in N3 disease,for which control was better than in N2 or N1 disease ina multivariate analysis. Cerezo et al suggest that this observationmay be due to N3 disease being more frequent in nasopharyngealcarcinoma. Elsewhere, they comment on the problems posed by retrospectivestudies.

How the Problem May Be Elucidated

A final paper by Johnson et al [3], which was included in thereview by Mendenhall et al, used an elegant method of quantitativetumor volume measurements from CT scans that indicated how thisproblem may be elucidated. They estimated the tumor volume (TV)and nodal volume (NV) and combined these as the total tumor volume(TTV) estimate. They were able to show that TTV was themost important outcome predictor of local control. The separateeffects of TV and NV could not be discerned in this study (51cases) but, obviously, could help resolve the controversy. Certainly,these tumor volume measurements help overcome the imprecisionof clinical staging.

Are We Losing Sight of the Forest for the Trees?

In conclusion, I am left with the feeling that we are losing sightof the forest for the trees. Our problem with radiotherapy, aspointed out by Cerezo et al, is recurrence at the primary site(20%), and N stage, particularly fixed nodes, is a major prognosticfactor for survival in head and neck cancer. The study by Johnsonet al demonstrates the importance of TTV. With these points inmind, I believe that more aggressive treatment, whatever the modalityemployed, would be justified in patients meeting these criteria,but obviously should be given in the setting of a clinical trial.

References:

1. Griffin T, Pajak T, Gillespie B, et al: Predicting the responseof head and neck cancers to radiation therapy with a multivariatemodeling system: An analysis of the RTOG head and neck registry.Int J Radiat Oncol Biol Phys 10:481-487, 1984.

2. Cerezo L, Millan I, Torre A, et al: Prognostic Factors forsurvival and tumor control in cervical lymph node metastases fromhead and neck cancer. A multivariate study of 492 cases. Cancer69:1224-1234, 1992.

3. Johnson C, Khandelwal S, Schmidt-Ullrich R, et al: The influenceof quantitative tumor volume measurements on local control inadvanced head and neck cancer using concomitant boost acceleratedsuperfractionated irradiation. Int J Radiat Oncol Biol Phys 32:635-641,1995.