Multiple Primary Tumours in Oral Cancer: Patient Characteristics and Survival Patterns

Introduction

Oral cancer is a significant public health concern globally. It is the sixth most common cancer in the world, with an annual incidence of approximately 300 000 reported cases, about two-thirds of which are in developing countries.¹ According to the Canadian Cancer Society, more cases of oral cancer are diagnosed every year than ovarian or cervical cancer, and its average 5-year survival rate is only 64%.² Oral cancer is a broad term that includes 2 subcategories: oral cavity cancer (OCC) and oropharyngeal cancer (OPC). Despite the decline in tobacco consumption and a decreasing incidence of other head and neck cancers, the incidence of oral cancer has increased, mainly because of an increase in HPV-associated OPC.³ Patients with HPV-positive OPC have been found to be younger and present clinically with an advanced nodal stage; however, they respond favourably to treatment and have better survival rates than patients with HPV-negative OPC.⁴

According to an earlier study conducted in Alberta,⁵ 47.9% of patients diagnosed with oral cancer died, indicating a high mortality rate associated with oral cancer. The suggested reasons included late presentation of advanced disease and the risk of MPT development.⁶ MPTs are defined as the second, third, fourth or even more primary tumours and are not the same as residual/recurrent tumours.⁷ To identify multiple primary tumours, SEER (surveillance, epidemiology and end results program) considers histologic findings, site, laterality and time since the diagnosis of the primary tumour.⁸ Population-based studies have documented the incidence of MPTs in oral cancer patients ranging from 3% to 21%.⁹

An overall poor prognosis and a significant decrease in 5-year survival rates from 69% to 32% have been documented in patients with MPTs compared with patients without MPTs.¹⁰ According to a Canadian study, the impact of MPTs on 5-year relative survival is greatest in bladder cancer (−2.4% age-standardized) followed by oral cancer (−1.9%).¹¹ Considering this significant effect of MPTs on the survival rate of oral cancer patients, this study aimed to identify risk factors of MPT development in oral cancer patients and to examine the survival rates of oral cancer patients with MPTs and associated factors. The findings will help inform health care professionals about the group of oral cancer survivors who are at higher risk for developing MPTs and help them carry out preventive strategies including close monitoring and follow-up appointments.

Materials and Methods

Study Population

Using a retrospective, population-based cohort design, we obtained data from the Alberta Cancer Registry (ACR) for Alberta residents diagnosed with primary oral cancer (POC) at age ≥18 years between 2005 and 2020. The data were retrieved on 31 July 2020. The ACR, a population-based registry certified by the North American Association of Central Cancer Registries, maintains comprehensive data on all new cancer diagnoses and deaths in Alberta.

The location of oral cancers was based on the International Classification of Diseases for Oncology (3rd edition, ICD-O-3).¹² Patients with MPT were identified from the original cohort.

MPTs were defined according to Warren and Gates criteria.¹³ Patients who developed second, third, fourth or even more tumours were identified as patients with MPT. MPTs were categorized as synchronous if diagnosed within 6 months of the index tumour and metachronous if diagnosed later.

Data Extraction

Sociodemographic and Clinical Characteristics: Baseline demographics and clinical variables were gathered from patients’ records. Demographic factors included age at diagnosis, gender, year of diagnosis, age at death, vital status and household income. Clinical characteristics included number of SPTs, duration of development of MPT from POC, site distribution of tumour according to ICD-O-3 grading,¹² tumour stage as reported in pathology reports according to the American Joint Committee on Cancer, 7^th and 8^th editions (AJCC 7 and AJCC 8)^14,15 and history of medical conditions. Treatment modality factors included surgery, chemotherapy, radiation, immunotherapy and combinations of treatment.

The patients were then divided into 2 groups: patients with POC who developed MPTs and patients who did not develop MPTs. A comparative analysis was carried out to study the demographic and clinicopathological characteristics of both groups. Regression analysis was carried out to determine whether there was an association between risk factors and MPT development.

Outcome Variable (Survival): The vitality status of oral cancer patients was recorded as alive or deceased. The time from diagnosis of POC until date of death was used to calculate survival rate for patients with and without MPTs.

The survival period was up to 15 years if POC was diagnosed in 2005; however, the study endpoints consisted of 5-, 10- and 15-year overall survival rates based on the year of POC diagnosis. Disease-specific survival was calculated for 5 years, comparing the survival rates of patients who developed MPT to those who did not.

Statistical Analyses

Descriptive statistics for demographic and clinical characteristics for all patients and each group were computed. Categorical variables were summarized in frequency and percentages and numerical variables were expressed as mean ± standard deviation (SD) or median interquartile range (IQR) for each group. Comparisons of the demographic and clinical characteristics were made between MPT and non-MPT patients. Chi-squared tests and Fisher exact tests were employed where appropriate for categorical variables. For numerical variables, an independent Student t test was used. To determine the significant prognostic factors associated with the development of MPT, binary logistic regression was used at univariable and multivariable levels. The primary outcome measure was 5-year disease-specific survival. Kaplan-Meier survival curves were plotted to estimate the overall survival of patients with and without MPT. Unadjusted and adjusted hazard ratios were computed using Cox proportional hazard models to present the risk of death. Cox regression analysis is a specific regression model to evaluate the influence of multiple independent variables on survival time (a time to event outcome). Time from a start point (e.g., disease diagnosis) to an event (e.g., death due to the disease) is the main component of Cox regression models.

By using multivariable regression analyses, the relative odds ratios of MPTs for each possible related factor was estimated by adjusting for possible confounding variables. A p value < 0.05 was considered statistically significant. All statistical analyses were performed using statistical software SPSS v. 22.

Ethics Approval

This study was conducted with the approval of the Health Research Ethics Board of Alberta Cancer Committee (HREBA.CC-17-0370) and the University of Alberta. All data were fully anonymized to ensure patient confidentiality.

Results

Sociodemographic Characteristics

Records of 3549 POC patients between 2005 and 2020 were retrieved from the ACR database; 71.8% were males (Table 1). The patients were 18–97 years of age (mean 61 ± 12.5 years). Of the original cohort, 513 (14.5%) patients developed MPTs.

Comparative analyses showed no significant difference in gender distribution between MPT and non-MPT patients (p = 0.36); more males were in both MPT (69.8%) and non-MPT (71.8%) groups. For both groups, the average age at first diagnosis was 61 years (61 ± 12.5 and 61 ± 11.7, respectively). There was no significant difference in body mass index (BMI) of patients in the 2 groups (p = 0.067). However, we found a significant difference in average annual income (p = 0.019): the largest proportion of non-MPT patients (47.2%) had an income >$75 000, while the largest proportion of MPT patients (48.7%) reported an average income of $45 000–75 000. No significant differences between MPT and non-MPT patients were found based on geographic location of diagnosis (p = 0.730).

Clinical Characteristics

Occurrence and Distribution of MPTs: Of the 513 patients with MPTs, 82.8% (n = 425) were diagnosed with 1 MPT 15.6% (n = 80) with 2 MPTs and 1.6% (n = 8) with more than 2 MPTs. It was observed that 77.9% of the MPTs were metachronous and 22.03% were synchronous. The mean time to occurrence of MPTs following the diagnosis of POC was 4.2 ± 3.8 years.

Of 3549 patients with POC, 1712 (48.2%) had OCC, while 1837 (51.8%) had OPC. The proportion of OCC patients who developed MPTs (53.4%) was slightly higher than that of OPC patients who developed MPTs (46.6%) (Table 2a). Of the primary OCC subsites in the study, tongue tumours (37.6%) showed a higher probability of developing MPTs followed by floor of mouth tumours (21.9%). Of the primary OPC subsites in the study, tonsil (48.7%) was the most predominant site to develop MPTs followed by base of tongue (37.6%). The most common site for development of MPT was the oral region (31.8%) followed by lung/bronchus (19.1%), digestive system (12.3%) and head and neck sites other than the oral region (11.3%). A similar pattern was observed for 3rd, 4th and 5th primary tumours (Table 2b).

History of Medical Conditions: The proportion of patients with any comorbid condition was found to be higher among those with MPTs (51.0%) compared with those without MPTs (40.0%). The most prevalent and statistically significant comorbid conditions in MPT patients were chronic obstructive pulmonary disease (COPD, 24.6%, p < 0.001), diabetes (18.7%, p = 0.004) and cardiovascular disease (11.5%, p = 0.040) (Table 3).