Abstract
Background:
The presence of comorbidities may play an important role in the way treatment plans are developed for management of cancer. The aim of this study was to investigate the prevalence and impact of comorbidities in a retrospective nationwide population-based study of patients with oral cavity cancers.
Material and methods:
A total of 16,676 oral cavity cancer patients diagnosed in the period 2007–2010 were identified through the Taiwan Cancer Registry database. By linking this data set to the National Health Insurance Research Database and the death registry, information on comorbidity and cause of death was obtained and adapted to the Charlson’s comorbidity index. The influence of comorbidity on overall survival and cancer-related death was evaluated by Kaplan–Meier estimates and Cox regressions.
Results:
In total, 21.3% of patients had some comorbidity during the year before cancer diagnosis. Diabetes mellitus was found to be the most common comorbidity (10%). Methods of management were significantly different in patients with stages 2, 3 and 4 of the disease. Those with severe comorbidity conditions received less surgical interventions (5–13%) compared with the patients without any comorbidity. Differences in the survival rates were only significant in all-cause death, but not in cancer-related death in the early stages (stages 1 and 2). Two-year survival rates for patients with severe comorbidity conditions were lower by 10–22% given the same stage of disease and treatment modality.
Conclusion:
The survival rates of oral cancer patients in early stages are lower in patients with comorbidity conditions and this may be due to the less aggressive cancer treatment offered to this group of patients. Comorbidities affect the decision on cancer treatment modality and thereby outcomes.
Introduction
Cancer patients’ comorbidity condition may play an important role in treatment decisions during their management. Patients with comorbidity consequently may tend to have less aggressive non-curative treatments. 1 Many lifestyle factors have been reported to be associated with occurrence of cancers. These factors simultaneously also result in other medical conditions, and, therefore, comorbidity could be a common condition for cancer patients. Particularly in oral cancer patients, while cigarette smoking, alcohol drinking and betel quid chewing are well-known risk factors, these lifestyles are also associated with several systematic diseases. The effects of these diseases on selection of cancer treatments and effectiveness of treatments are crucial issues in patients’ prognoses. 1
Comorbidity increases mortality in oral cancer patients, and this fact may due to less aggressive treatment being offered or more treatment complications. 2 Information on comorbidity may be obtained from self-reports of patients, from the review of their medical records and also from electronic databases, which were initially created for insurance reimbursement. Different resources would be required to obtain comorbidity information. Currently, there are several measurements of comorbidity in use. The Charlson’s comorbidity index (CCI) and the adult comorbidity evaluation 27 are commonly used in evaluating head and neck cancer patients. 2
Current reports from large-scale databases have primarily focused on head and neck cancers. 3,4 Head and neck cancer generally includes oral cavity, pharynx/larynx and thyroid cancer. The risk factors and treatment modalities are often different between these different types of cancer. Therefore, it is desirable for these anatomical subsites to be investigated separately in terms of association between mortality and comorbidity. The study objectives were to investigate the management of cancer patients in different comorbidity conditions, and further to compare their prognosis based on overall survival and disease-specific survival, with those without comorbidities in terms of corresponding treatment decisions.
Materials and methods
Data source
This study cohort was constructed based on the Taiwan Cancer Registry (TCR), death registry (DR) and the National Health Insurance Research Database (NHIRD). The TCR was implemented in 1979 and is organized and funded by the Ministry of Health and Welfare, Taiwan. To further promote the cancer care quality, a long-form database was initiated in 2002 to include the American Joint Committee on Cancer staging for selected major cancers and was further extended in 2007 to include date of initial diagnosis, primary site, histology, clinical Tumor-Node-Metastasis (TNM), pathological TNM, metastasis, surgical procedure of primary site, chemotherapy and radiotherapy. 5 Patients with multiple cancers of different organs were documented as multiple records. A total of 80 hospitals are using the long-form database for reporting, which account for 90% of major cancers. The DR database was used for identifying any available death dates and cause of death for cancer patients.
The NHIRD is provided by the Bureau of National Health Insurance (NHI) and managed by Department of Statistics, Minister of Health and Welfare. The NHI programme in Taiwan was initiated in 1995. This programme covers all medical services provided with diagnosis, drugs prescribed and the costs with various co-payment rates from the patients. The NHI database of this programme contains inpatient and outpatient physician claims and prescription drug claims as well as demographic data on all beneficiaries. The coverage rates of NHI programme were 93.1% in 1996 to 99.6% in 2010 of Taiwan’s total population.
The three population-based databases were linked by encrypted personal identification, and the whole data management and statistical analyses were conducted in a restricted workplace in Health and Welfare Data Science Center, Department of Statistics, Minister of Health and Welfare for the personal information privacy protection.
Study cohort
Oral cavity cancer patients diagnosed during 2007–2010 were identified from TCR database using the International Classification of Diseases for Oncology (ICD-O): C00-C06, or the ICD, Ninth Revision, Clinical Modification (ICD-9-CM): 140–141 and 143–145. Information on cancer stage and treatment modality was also obtained from TCR. Since the time NHI programme was implemented in Taiwan, it is very unlikely that any curative treatment attempts were not provided after a cancer diagnosis. The decision of no curative treatments or receiving palliative management only might largely be made by patients and their families. In this study, we focus on outcome of treatment managements by medical service; therefore, patients not receiving any surgery, radiotherapy or chemotherapy treatments were further excluded from the study cohort. Palliative management is not considered as a treatment modality in this study. Mortality of cancer patients was confirmed by DR on death dates and causes of death. This study was approved by the Internal Review Board of Koahsiung Medical University Hospital (KMUHIRB-EXEMPT(I)-20160031).
Analysis variables
The primary cancer sites were determined by the first reported cancer site of oral cavity; categories were lip, tongue, gum, floor of mouth and other unspecified parts of oral cavity. Treatment modality was identified by reported information of primary cancer treatments in long-form database, which included date of cancer diagnosis, date chemotherapy started, date radiotherapy started, date of first surgical procedure, date of most definite surgical resection of the primary site and surgical procedure of primary site. Initially, seven treatment modalities were defined as (1) radiotherapy only, (2) chemotherapy only, (3) radiotherapy and chemotherapy, (4) surgery only, (5) surgery with radiotherapy, (6) surgery with chemotherapy and (7) surgery with radiotherapy and chemotherapy. Palliative management was not considered as a treatment modality in this study. In the analyses of stage/comorbidity subgroups, due to small sample sizes, the first three treatment groups were combined together as radiotherapy and/or chemotherapy; similarly, the last two groups were combined as surgery with radiotherapy and/or chemotherapy.
NHIRD was used to identify comorbidities of cancer patients during the year before cancer diagnosis. Each separate disease specified in the CCI 6 was identified from the diagnosis code of inpatients and outpatients records using ICD-9-CM. 7,8 To prevent any diagnosis codes given for screening purposes and to assure that a patient had been diagnosed and treated for a comorbidity, we specified that a disease which was considered as a comorbidity should have had three or more outpatient records or one hospital admission of a given condition. The CCI was further categorized into three groups (0 = no comorbidity, 1 = mild-to-moderate comorbidity, 2 = severe comorbidity) in statistical analyses. Age at cancer diagnosis was used for defining age groups.
Statistical analysis
The survival analysis with Kaplan–Meier estimates of survival rates and Cox regressions were the primary statistical methods of analysis in this study. The events and dates of all-cause death or cancer death were obtained from 2007 to 2012 of DR database. For those patients without death occurrence, the time from first diagnosis to December 31, 2012 (censored date), was considered as the follow-up time for censored patients. A p value >0.05 was considered as statistically significant. The statistical software SAS V9.3 (SAS Institutes, Inc., Cary, North Carolina, USA) was used to conduct analyses.
Results
Initially, 17,028 newly diagnosed oral cavity cancer patients were identified from TCR during 2007–2010. Among them, seven patients aged less than 18 years old and 345 patients without receiving any surgery, radiotherapy or chemotherapy treatments were further excluded. These 345 patients accounted for 1.0, 1.4, 1.9 and 3.2% of the initial patients in stages 1–4 (16,676 + 345), and 1.9, 2.1 and 3.5 of the initial patients in none, mild-to-moderate and severe comorbidity conditions. There were finally 16,676 patients included in the analysis. The study cohort comprised 15,268 males (94.56%) and 1408 females (8.44%). Mean (±SD) age of diagnosis was 53.53 (±11.81) with overall median follow-up time of 2.98 years (range: 3.29 months to 6 years). There were 38.29% (n = 6385) of patients in stage 4 and 45. Thirty-two percent (n = 7558) receiving surgery as the only treatment. The average of CCI was 0.35 (±0.84) and 21.34% of patients were recorded to have one or more comorbidities (CCI ≥ 1) during the year before diagnosis (Table 1). Diabetes mellitus was found to be the most common comorbidity (9.85%).
Characteristics of oral cavity cancer patients.
CCI: Charlson’s comorbidity index; AIDS: acquired immune deficiency syndrome; HIV: human immunodeficiency virus.
We investigated patients’ prognoses by their comorbidity condition (CCI = 0, 1, ≥2) for each cancer stage, Table 2 lists survival rates and univariate hazard ratios (HRs) for all-cause death and cancer-related death. The overall survival rates were significantly different among comorbidity condition within stages (p < 0.05). Univariate HRs were all significantly higher for the group with severe comorbidity (CCI ≥ 2) versus no comorbidity. Significantly different survival rates were only found in stages 3 and 4 for cancer-related death (p = 0.0086, <0.0001).
Survival rates and univariate hazard ratios for all-cause death and cancer-related death.
A comparison of cancer management is shown in Table 3. Patients with severe comorbidity tended to have lower percentages in the treatment group having surgery with radiotherapy/chemotherapy when compared with patients without comorbidities: stage 1: 2.8% (12.2% vs. 15.0%), stage 2: 7.3% (23.4% vs. 30.7%), stage 3: 7.0% (52.3% vs. 59.3%) and stage 4: 13.4% (49.0% vs. 62.4%). The log-rank tests on survival rates of different management types within each combination of stage/comorbidity condition were all significant.
Comparison of treatment management in stages with comorbidity condition.
In the adjusted HRs of using the treatment group of radiotherapy/chemotherapy as reference (Table 4), surgery alone had significantly lower risks in all combinations (HRs: 0.23–0.53) except for severe comorbidity in stage 1 (HR = 0.44; 95% CI = 0.16, 1.24; p = 0.1208) and surgery with radiotherapy/chemotherapy had significantly lower risks in all combination (HRs: 0.41–0.57) except for mild-to-moderate (HR = 0.48; 95% CI = 0.21, 1.09; p = 0.0773) and severe (HR = 0.98; 95% CI = 0.33, 2.94; p = 0.9675) comorbidity in stage 1, severe comorbidity (HR = 0.62; 95% CI = 0.32, 1.19; p = 0.1530) in stage 2 and mild-to-moderate comorbidity (HR = 0.76; 95% CI = 0.46, 1.25; p = 0.2741) in stage 3.
Adjusted hazard ratios for treatments in different stage and comorbidity condition.
HR: hazard ratio; CI: confidence interval; CCI: Charlson’s comorbidity index.
aThe HRs are also adjusted by age and gender. For the HRs of all stages, the model also includes stages as an adjusted variable.
When considering all stages together (data not shown in table), the adjusted HRs of using the treatment group of radiotherapy/chemotherapy as reference, surgery alone had an HR of 0.27 (95% CI = 0.25, 0.30; p < 0.0001), surgery with radiotherapy/chemotherapy also had significantly lower risks (HR = 0.44; 95% CI = 0.41, 0.47; p < 0.0001). The HRs of comorbidity conditions are 1.00 (95% CI = 0.93, 1.08; p = 0.9252) for mild-to-moderate and 1.40 (95% CI = 1.29, 1.53; p < 0.0001) for severe condition using no-comorbidity as the reference. When we restricted our study patients as surviving for at least 2 years, the corresponding estimates were HR = 0.31 (95% CI = 0.26, 0.37; p < 0.0001) for surgery alone, HR = 0.44 (95% CI = 0.41, 0.47; p < 0.0001) for surgery with radiotherapy/chemotherapy, HR = 1.09 (95% CI = 0.95, 1.26; p = 0.2231) for mild-to-moderate comorbidity condition and 1.30 (95% CI = 1.07, 1.53; p = 0.0074) for severe condition. In addition, to compare the comorbidity condition to general population (aged 18 years old or older), we used the same procedure to identify the comorbidity diseases and condition from the whole NHI study samples in the year of 2010. The proportions of mild-to-moderate and severe comorbidity conditions were 8.5% and 6.4%, respectively. In addition, there were 5.4% with diabetes, 1.1% with diabetes with chronic complication, 1.9% with mild liver disease, 0.1% with moderate or severe liver disease, 2.5% with chronic pulmonary disease and 2.2% with cerebrovascular disease.
Discussion
Comorbidity has an important role for deciding a treatment modality for any patient diagnosed with cancer. It is pertinent to note that cigarette smoking, excessive alcohol drinking and betel quid chewing are the major causes of oral cancer, and these habits are also generally associated with other medical conditions. In this study, we incorporated three nationwide databases and found that overall survival rates were significantly associated with patients’ comorbidity condition within each cancer stage. In terms of the stage of disease, for those in late stages (stages 3 and 4), the survival rates tend to be significantly lower in patients with any comorbidity, and both all-cause death and cancer-related death were significantly different. Patients in severe comorbidity condition (CCI ≥ 2) had significantly higher HRs. Our study results are consistent with several recent clinical studies reporting that comorbidity is associated with poor prognosis in advanced head and neck cancer patients following major surgery 9 and in subsets of nasopharyngeal, oropharyngeal, oral cavity and laryngeal cancers. 10
Nevertheless, in our study, survival rates were only significant for all-cause death, but not for cancer-related death in the early stages of the disease (stages 1 and 2). This result suggested that the mortality of patients in early stages may also be due to challenges in the management of comorbidity conditions. A Danish population-based study analysed from the Danish Head and Neck Cancer Group (DAHANCA) database showed that comorbidity has a negative prognostic impact on overall survival in head and neck cancer patients, but cancer-specific death was not significantly affected by comorbidity suggesting that patients die from or of their comorbidities rather than their cancer. 3
A recent clinical cross-sectional study from India investigated changes in therapeutic decision-making, and among 139 patients, there were no changes made from institutional evidence-based guidelines. There was significant positive correlation between comorbidity and change in therapeutic decision-making.1 Treatment managements were significantly different at stages 2, 3 and 4, with patients in severe comorbidity tending to have less curative treatment (surgery alone or surgery with radiotherapy and/or chemotherapy). For stages 2–4, patients who did not receive surgery were 5, 8 and 13% more in severe comorbidity condition than in patients without comorbidity. Consequently, the 2-year survival rates within each combination of stage and comorbidity condition were 13–37% lower, indicating a survival benefit from aggressive treatment management. In addition, while potential covariates were adjusted, the lower HRs suggested protection of mortality for patients in late stages and severe comorbidity condition. Previous results also from electronic databases showed that curative-intent aggressive treatment improved survival by 63–67% (HRs = 0.37–0.33) in various degrees of comorbidity condition for locally advanced (stages 3 and 4) head and neck squamous cell carcinoma patients. 11
Reported prevalence rates of comorbidity often vary due to the use of various definitions. The CCI is a simple and validated instrument 2,12 and can be conveniently derived from ICD-9 or ICD-10 codes. 7,8 While the severity of disease was considered as a potential limitation for CCI, we have used the criterion of three or more outpatient diagnosis or one hospitalization in the year before as a surrogate of treated medical conditions at the time of cancer diagnosis. Our study revealed that 78.66% of oral cancer patients were free of any comorbidity. While a similar study reported from Taiwan 11 had a lower comorbidity-free proportion of 32.36%, but the period for extracting records of comorbidity was limited to a 6 months period before and after cancer diagnosis. A Danish study using the DAHANCA database using up to 10 years of discharge records showed a 64% of no comorbidity. 3,4
Kim et al. suggested that risk of any comorbidity condition in recurrent stage 3 and 4 head and neck squamous cell carcinoma (HNSCC) should be justified in the decision for salvage surgery, which may have worse quality of life. 13 Another clinical study also indicated that comorbidities, particularly lung diseases, appeared to have an important and unique influence on health-related quality of life scores among HNSCC patients treated with curative intent. 14 While the tendency of poor prognosis in patients with comorbidity may be due to less aggressive cancer treatment, our results showed that there were 10–22% lower 2-year survival rates for patients with severe comorbidity condition given the same disease stage and treatment modality. Therefore, in addition to arriving at decisions on cancer treatment modality, it is also crucial to emphasize the management of other medical conditions in cancer patients.
Majority of our study patients were males (94.56%). It is the consequence that most of the risk factors (betel quid chewing, cigarette smoking and alcohol drinking) are male dominant. 15 Our study also has limitations on some confounders of cancer prognosis, including clinical and pathological characteristics during cancer treatment, use of betel quid chewing, cigarette smoking and alcohol drinking as well as other lifestyles. These factors were not recorded in any of the three databases used in this study, and might have effects on the estimates of risk, if they were included in regression analyses. We suspect that the effects might be minimal since these factors could affect in both directions. We used the ICD-9-CM to identify comorbid diseases. Since sometimes disease codes may be given for screening purposes, to assure that a patient had been diagnosed and treated for a comorbidity disease, we restricted that a disease had three or more outpatient records or one hospital admission of a given condition to minimize any false positives. While we tried to accumulate as many patients as possible in this study, the minimum follow-up time was 3.29 months (median follow-up time 2.98 years), to investigate whether any differences in estimates out sensitivity analyses show estimates from whole study patients and patients with 2 or more years of survival were not largely different. Therefore, the effect of follow-up could be minimal.
Footnotes
Conflict of interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is supported by Koahsiung Medical University Research Fund (KMU-M103014).