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

Harris et al

Table 2. Cox Univariate Analysis of Recurrence. a

Table 1. Patient and Tumor Characteristics.

Factor

HR

P Value

Factor

n (%)

Female

0.97 1.28 0.93

.89 .32 .92 .38 .56 .47 .85 .76

Patients

212

Age . 70 y

Sex

Immunosuppressed

Male

185 (87.3) 27 (12.7)

Primary site b Ear

Female

Age, y 70

1.44 1.27 1.44 1.16 0.84 3.87 1.61

96 (45.3) 116 (54.7)

Cheek/temple

. 70

Lip

Immunosuppressed

7 (3.3)

Neck Nose

Primary site Ear

48 (22.6) 60 (28.3) 22 (10.4) 5 (2.4) 20 (9.4) 12 (5.7) 33 (15.6) 111 (52.4) 57 (26.9) 87 (41.0) 56 (26.4) 86 (40.6) 61 (28.8) 30 (14.2) 62 (29.2) 30 (14.2) 78 (36.8) 60 (28.3) 84 (39.6)

Periorbital Recurrent Diameter, cm b 2-4

.012

Cheek/temple

.06

Lip

Neck Nose

1.30 1.25

.41 .51

. 4

DOI, cm b 1-2

Periorbital

Scalp

1.32 1.02 1.56 1.90 1.11 1.21 1.22

.36 .95 .17 .70 .49 .44

. 2

Recurrent

LVI 1

Diameter, cm 0-2

PNI 1

.011

2-4 . 4

Regional disease

Poorly differentiated Adjuvant therapy

DOI, cm 0-1

Abbreviations: DOI, depth of invasion; HR, hazard ratio; LVI, lymphovascu- lar invasion; PNI, perineural invasion. a Bold indicates statistical significance, P \ .05. b Reference groups: age, 70 years; scalp diameter, 0-2 cm, DOI 0-1 cm.

1-2 . 2

Regional disease

LVI 1 PNI 1

Poorly differentiated

and tumors with PNI (HR = 1.73, P = .041) were signifi- cantly more likely to experience recurrence ( Table 2 ). Differences in DFS are depicted with Kaplan-Meier curves in Figure 2 . A Cox multivariate analysis was then per- formed to identify independent predictors of DFS. When controlling for other variables, PNI (HR = 2.62, P = .028) maintained significance as a predictor of DFS, and periorbi- tal primaries were no longer associated with increased DFS (HR = 2.4, P = .25). However, poorly differentiated tumors and recurrent tumors were significantly associated with DFS (HR = 2.21, P = .041). Interestingly, regional disease (HR = 1.31, P = .54) and adjuvant therapy (HR = 0.72, P = .39) were not associated with DFS ( Table 3 ). A multinomial logistic regression was performed to assess the association of patient and tumors factors with the presence of regional nodal disease. This demonstrated that ear primaries (odds ratio [OR] = 16.19, P = .016), cheek or temple primaries (OR = 15.04, P = .017), lip primaries (OR = 13.50, P = .037), and tumors with PNI (OR = 2.74, P = .82) were significantly more likely to present with regional nodal disease ( Table 4 ). Discussion Although factors associated with recurrence and survival have been consistently described for early-stage CSCC, the applicability of these findings to the subset of advanced

Adjuvant therapy

Abbreviations: DOI, depth of invasion; LVI, lymphovascular invasion; PNI, perineural invasion.

Figure 1. Survival outcomes: entire cohort. DFS, disease-free survival.

88