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

|

CHERAGHLOU ET AL .

FIGURE 1 Selection of the study population. NCDB, National Cancer Data Base; pN, pathologic N classification; pT, pathologic T classification

margins. We defined patients who underwent any surgical procedure, including local tumor destruction, local tumor excision, partial parotidectomy or removal of the major sali- vary gland, total parotidectomy or removal of the major sali- vary gland, radical parotidectomy or removal of the major salivary gland, parotidectomy, not otherwise specified (NOS), and surgery, NOS, as having received surgery, and patients who had no surgery of the primary site as not having had surgery. Patients were defined as having received chemotherapy if they received any chemotherapy, regardless of the type or number of agents. We considered patients to have received radiotherapy if they received external-beam radiation with a regional dose of at least 44 Gy. Patients were included in the surgery only group if they received surgery but no radiother- apy or chemotherapy; in the surgery and radiotherapy group if they received surgery along with starting radiotherapy within 90 days after the date of their surgery, but no chemo- therapy; and in the surgery and chemoradiotherapy group if they received surgery along with starting radiotherapy within

90 days after the date of their surgery and starting chemo- therapy within 14 days of their first radiation treatment. Although we aimed to use data on extracapsular extension (ECE) as well, this information was available for only 37% of the node-positive sample and was therefore excluded from analysis. Of the 230 patients with node-positive disease with known ECE status treated with surgery alone, 73 (31.74%) were ECE-positive. Of 803 patients with node-positive dis- ease with known ECE status treated with surgery and adju- vant therapy, 352 (43.84%) were ECE-positive. Kaplan-Meier analyses stratified by stage/adverse feature grouping and treatment combination were performed. Propensity-score matching was conducted for patients who did or did not receive adjuvant therapy after surgery within groups. For patients with late-stage disease with adverse fea- tures, matching was also conducted between patients who received adjuvant radiotherapy versus those who received adjuvant chemoradiotherapy. Scores were calculated using a logit model with all factors outlined in Table 1. Treatment and control groups were generated using a one-to-one nearest

47