2018-19 Section 7-Neoplastic and Inflammatory Diseases of the Head and Neck eBook

|

CHERAGHLOU ET AL .

TABLE 2 (Continued)

Early-stage disease

Late-stage disease

Variables

HR

P value

HR

P value

Grade Low

1 [Reference]

1 [Reference]

. . .

. . .

Intermediate

1.756 3.102 1.750

.017

2.244 2.755 2.139

< .001 < .001 < .001

High

< .001

Unknown

.018

Surgical margins Negative

1 [Reference]

1 [Reference]

. . .

. . .

Positive

1.804 1.321

< .001

1.158 1.165

.002 .103

Unknown

.328

Treatment factors

Treatment combination Surgery only

1 [Reference]

1 [Reference]

. . .

. . .

Surgery 1 adjuvant therapy a

0.744

.001

0.647

< .001

Abbreviation: HR, hazard ratio. a Radiotherapy or chemoradiotherapy. Note: The facility type was initially included in the regression for each group but was excluded during stepwise removal. Additionally, race and histology were ini- tially included in the regression for the early-stage group but were excluded during the stepwise removal.

Added adjuvant treatment was not associated with improved survival on univariate Kaplan-Meier ( P 5 0.524) or multivariate analysis. This group was too small for propensity-score matched analysis. 3.4 | Late-stage disease with adverse features Patients with late-stage disease with adverse features were mostly over 60 years old (61.6%), men (65.3%), and white (82.8%; Table 1). Among these patients, 64.1% had nodal metastases and 82.9% had intermediate-/high-grade tumors. Adjuvant therapy was administered for the majority of the patients in this group, with 27.9% receiving no adjuvant treatment. The 5-year survival rate was 51.3% (SE 0.9). Among the patients in this group, administration of adju- vant therapy was associated with significantly improved sur- vival on univariate Kaplan-Meier analysis ( P < .001). The 5- year survival rate was 43.3% (SE 1.6) for patients who received surgery without adjuvant therapy and 54.3% (SE 1.0) for patients who received surgery with adjuvant therapy. Adjuvant radiotherapy was also associated with a survival benefit on multivariate (HR 0.647; P < .001; Table 2) and propensity-score matched (HR 0.688; P < .001) analyses. Within the propensity-score matched sample, the 5-year sur- vival rates for patients who did and did not receive adjuvant therapy were 54.1% (SE 1.6) and 43.8% (SE 1.6), respec- tively (Figure 2). On multivariate analysis of patients positive for specific adverse features, adjuvant therapy was associated

FIGURE 3 Hazard ratio of death (with 95% confidence intervals) associated with receipt of adjuvant therapy by presence of adverse features among patients with early-stage and late-stage disease on multivariate Cox survival analysis. The reference group is patients who did receive adjuvant therapy

52