The use of adjunctive traditional Chinese medicine therapy and survival outcome in patients with head and neck cancer: a nationwide population-based cohort study.

The use of adjunctive traditional Chinese

medicine

therapy and survival outcome in patients

with head

and neck cancer: a nationwide

population-based

cohort study

Hung-Che Lin

, Cheng-Li Lin

, Wen-Yen Huang

, Wei-Chuan

Shangkuan

,

Bor-Hwang Kang

, Yueng-Hsiang Chu

, Jih-Chin Lee

,

Hueng-Chuen Fan

and Chia-Hung Kao

Introduction

Head and neck cancer (HNC) is an important cause of death and morbidity worldwide. It is the eighth most frequent type of cancer in USA with nearly 53 000 new cases diagnosed annually, accounting for 11 520 deaths.1 More than 550 000 cases of HNC

occur annually worldwide.2 The current treatment modalities for patients with HNC include surgery, radiotherapy, chemotherapy and targeted biological therapies.3 However, the prognosis and 5-year survival rates for patients with HNC remain poor, particularly when the cancer occurs in the oropharynx and hypopharynx.1,4 Traditional Chinese medicine (TCM) is one of the widely used alternative medicine therapies among patients in China, Hong Kong and Taiwan. Some studies have proposed that TCM can be beneficial in the treatment of cancer,

including breast cancer, hepatocellular carcinoma, gastric cancer, lung cancer, colorectal cancer, prostate cancer, superficial bladder cancer and leukemia.5–11 However, clinical studies on

the therapeutic effects of TCM in patients with HNC are scarce and have limited sample sizes.11–13 In Taiwan, TCM is a widely accepted form of medical treatment for many diseases. It is covered by the Taiwanese National Health Insurance (NHI) program.

Using the population-based National Health Insurance

Research Database (NHIRD), we investigated whether the combination of TCM and contemporary cancer treatments affected

the survival of patients with HNC.

Methods

Data source

The NHI program was implemented in 1995 and covers >99% of the 23.74 million Taiwan residents.14 The NHI program is a mandatory health insurance program that offers comprehensive medical care coverage, including outpatient, inpatient,

emergency and TCM services as well as prescription drugs for all insurants. In the NHI program, insurants with any of the 30 categories of catastrophic illness specified by the Bureau of NHI (including cancer) can apply for catastrophic illness certificates [Registry of Catastrophic Illnesses Patient Database (RCIPD)]. If the insured has major diseases such as cancer, he or she can apply for a catastrophic illness certificate. To reduce the financial hardship associated with catastrophic illness, beneficiaries are exempted from copayments. The issuance of certificates is validated through careful review of medical records, pathological reports, and imaging studies by at least two specialists.

The diseases investigated in this study were identified according to the diagnosis codes in the RCIPD, based on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. All personally identifiable information in the RCIPD is scrambled to conform to the Personal Information Protection Act. This study was approved by the Institutional Review Board of China Medical University (CMU-REC-101-012).

Sampled participants

Data extracted from the RCIPD were used for this retrospective cohort study. This study comprised patients aged >20 years who were diagnosed with primary HNC (ICD-9-CM codes 140-149) between 2001 and 2011. Patients diagnosed with HNC who used TCM for >90 days were defined as TCM users, and those

who used TCM for fewer than 30 days were considered TCM non-users. The date of the 90th day of TCM treatment was used as the index date. To assemble a comparison cohort, controls were randomly selected and frequency matched with patients in the TCM cohort at a ratio of _1:1, based on age group (in 10-year span); sex; comorbidities of hypertension, diabetes, congestive heart failure (CHF), stroke, chronic obstructive pulmonary disease (COPD) and liver cirrhosis; treatment for HNC;

and index year, using the same exclusion criteria during the same period (Figure 1).

Outcome

The study outcome was all-cause mortality during the 11-year follow-up. The identification of death events was determined according to the RCIPD. The study patients were followed from the index date until their deaths, withdrawal from the NHI program, or the end date of the database (31 December 2011).

Variables of exposure

Premium-based income was categorized into three levels: <NT$15 000, NT$15 000–19 999 and _NT$20 000 (US$1*NT$30). The urbanization level was based on the population density of the residential area, population ratio of elderly people, the number of agricultural workers, educational level and the number of physicians per 100 000 people. Urbanization was categorized into four levels, with Level 1 being the most urbanized and Level 4 the least urbanized. We also assessed patients who had at least three claims for ambulatory care or hospitalization visits at the baseline with principal or secondary diagnoses of hypertension (ICD-9-CM 401-405), diabetes (ICD-(ICD-9-CM 250), CHF (ICD-(ICD-9-CM 428), stroke (ICD-9-CM 430-438), COPD (ICD-9-CM 490-496) or liver cirrhosis (ICD-9-CM 571), which were identified from the RCIPD as

baseline comorbidities. The treatment of HNC was divided into six groups, according to the treatment status: (i) surgery alone; (ii) surgery and adjuvant therapy; (iii) radiotherapy alone; (iv) chemotherapy alone; (v) radiotherapy and systemic therapy and (vi)

others or no treatment (Table 2).

Statistical analysis

The distributions of sociodemographic status and comorbidities were expressed as a frequency (with a percent) or a

mean6standard deviation (SD). The categorical variables were analyzed using a chi-square test, and the continuous variables of the baseline characteristics of TCM users and non-users were analyzed using a Student t-test. The Kaplan–Meier method was used to depict the curves of the event-free rate between the two cohorts, and the log-rank test was used to examine the difference between the curves. Univariate and multivariate Cox proportional hazards regression analyses were used to estimate

the hazard ratio (HR) with a 95% confidence interval (CI) for mortality. The variables that were significant in the univariate

Cox analysis were included in the multivariable Cox proportional hazards model to identify the independent predictors

of mortality. To verify the dose–response relationship between TCM use and mortality, the TCM use category was treated as a continuous variable to calculate the P value of the linear trend. The herbal prescription patterns and herbal formulae were described. The Statistical Analysis System (SAS),

Version 9.3 (SAS Institute, Cary, NC) computer software program was used to perform all statistical analyses. Comparison results with a P value of <0.05 were considered statistically significant.

Results

The TCM and comparison cohorts comprised data for 2966 and 2670 patients, respectively. Table 1 shows sociodemographic, comorbidity and treatment data for the TCM and control patients. No significant differences in age distribution were found

between the TCM and comparison cohorts (mean age 51.3610.8 vs. 51.7611.2 years, respectively). Approximately 45.6% of patients

in the TCM cohort were <49 years old. Additionally, the majority of the patients in the TCM cohort were male (78.6%).

Participants in both cohorts had monthly income levels of NT$15 000–19 999 and tended to live in more urbanized areas (57.9 vs. 55.7% for urbanization Levels 1 and 2, respectively). Patients who were treated with TCM had a higher prevalence of diabetes, CHF, stroke and COPD than the patients in the comparison cohort. The mean follow-up time was 3.63 years (SD¼2.51) in the TCM cohort and 3.19 years (SD¼2.54) in the comparison cohort (data not shown). Figure 2 shows the survival curve for the two cohorts and indicates that the

incidence curve of mortality was significantly lower in the TCM cohort than in the comparison cohort (log-rank test, P<0.001).

The univariate andmultivariate Cox regression analyses demonstrated a strong association between the use of TCM and lower

mortality (Table 2). Compared with TCM non-users, the mortality of TCM users was 31% lower (crude HR, 0.69; 95% CI¼0.62–0.75). After adjustment for sociodemographic variables, comorbidities and treatment of HNC, TCM users had lower mortality by 32%. Further analysis demonstrated a dose–response relationship between TCM use and mortality (Table 3). The adjusted HRs were

0.57 (95% CI¼0.49–0.67) and 0.30 (95% CI¼0.25–0.36) for HNC patients with TCM use of 251–550 days and >550 days, respectively. Therefore, a longer duration of TCM use was associated

with a lower mortality rate (P for trend <0.001).

Discussion

To our knowledge, this is the first large-scale nationwide cohort study investigating the association between adjunctive TCM therapy and the survival of patients with HNC. We found that TCM user had better survival outcomes (adjusted HR¼0.68). A longer duration of TCM is associated with a lower mortality rate (P for trend <0.001). In the multivariable model, increasing age, male sex, lower income, CHF and stroke were significantly associated with an increased risk of mortality. Sex was a strong and

independent prognostic factor, and women had higher survival rates than men did. Our data indicated that higher socioeconomic status was associated with increased survival. This is

likely attributable to greater exposure to tobacco, alcohol and betel nuts, which are all established indicators for poor outcomes in HNC patients, among patients with low socioeconomic than those with high socioeconomic status.15 In Taiwan, betel-nut chewing is common and associated with all-cause mortality.16

Approximately 2 million people habitually chew betel nuts in

Taiwan (10% of the population).17 A hospital-based case–control study of the incidence of oral cancer was computed to be 123-fold

higher in patients who smoked, drank alcohol and chewed betel nuts than in non-users in Taiwan.18 Chen et al. reported that the most prevalent site of oral squamous cell carcinoma (SCC) in

Taiwan was the buccal mucosa (37.4%), which contrasted with data from other countries, such as USA and Northern Norway.19

Buccal cancer accounted for only 2% of all oral SCC in USA and 14% of all oral SCC in Northern Norway.19–21

TCM users had a higher prevalence of comorbidities than the control group, but the TCM group had a higher survival rate than the comparison cohort, reflecting that the superiority of survival among TCM users relative to the cohort patients might be underestimated.

Treatment modalities also had an impact on the risk of mortality. Treatment with radiotherapy and systemic therapy was

shown to cause the greatest risk for mortality. This increased risk for mortality likely occurred among patients using this combined modality because this combined modality was used for patients who had advanced unresectable cases. Patients who were treated with surgery and adjuvant therapy also had higher risks of mortality. This is likely attributable to the fact that most of these patients had locally advanced cases. The lowest risk for mortality occurred among patients who received radiotherapy alone, which was typically used for early stage patients for organ preservation, such as early stage cancers of the

larynx, base of tongue or hypopharynx.

The TCM prescribed to treat patients with HNCs according to the TCM functional classifications are shown in Table 4. In recent years, research on the use of TCM for cancer treatment has accelerated with the advancement of molecular biology. TCM affects cancer treatment by (i) targeting apoptosis pathways in cancer; (ii) reducing inflammatory and infectious complications in the tissues surrounding the carcinoma; (iii) enhancing immunity and body resistance; (iv) improving patients’ general

conditions and qualities of life as well their cancer related fatigue to prolong their life spans; (v) increasing antioxidant activities; (vi) targeting the tumor cellular proteasome and nuclear

factor-kappaB (NF-jB) pathway5,10,22–25 TCM often involves mixtures of several Chinese herbs, and some studies have proposed that the pharmacological advantages of TCM may be derived from the potentiating action of these multiple bioactive components and the advancement of individualized therapy.26,27

The most popular herbal formula of the compound prescription in our study was Gan-Lou-Yin, which consists of Radix

Rehmanniae, Radix Rehmanniae Preparata, Herba Dendrobii, Radix Asparagi, Radix Ophiopogonis, Radix Scutellariae, Herba Artemisiae Scopariae, Fructus Aurantii, Folium Eriobotryae and Radix Glycyrrhizae Preparata.

Herba Dendrobii has been shown to act as a NF-jB inhibitor. Radix Glycyrrhizae Preparata has been reported to induce the autophagy of cancer cells through the suppression of Bcl-2 expression and the mTOR pathway. Radix Scutellariae targets cancer cells through reactive oxygen species (ROS)-mediated mechanisms, which can potentiate the cytotoxicity of anticancer medications by depleting glutathione (GSH). The GSH is a crucial factor in antioxidant defense. Moreover, Scutellariae can also inhibit the activities of extracellular signal-regulated kinases (ERK), serine/thronine kinase (AKT), cyclooxygenase-2 (COX-2) and NF-jB to enable it to arrest tumor cell cycle.5,10,23,28

The most popular herbal formula of the single prescription in our study was Bai-Hua-She-She-Cao, which can attenuate toxicity and enhance the efficacy of allopathy, improving

phagocytosis.9,29

TCM is currently considered effective in treating several cancers other than HNC. Jiang et al.11 reported that TCM was beneficial in treating unresectable hepatocellular carcinoma in

combination with transcatheter arterial chemoembolization (TACE). They revealed improved survival in patients treated with TCM and TACE. TCM is also effective as an adjuvant therapy for several cancers, including colorectal cancer, advanced hepatocellular carcinoma, non-small cell lung cancer (NSCLC), gastric cancer and advanced breast cancer. TCM can reduce gastrointestinal toxicity and enhance the tumoricidal effect of chemotherapy in patients with advanced colorectal cancer. Additionally, TCM can reduce the adverse effects of cytotoxic

drug capecitabine in patients with advanced hepatocellular carcinoma and prolong the survival of patients with NSCLC. In patients

with breast cancer, TCM can alleviate bone marrow

inhibition and cellular immunity suppression caused by chemotherapy and prolong the survival of these cancer patients. TCM

also seemed to prolong the survival of patients with progressive

gastric cancer.5 Ling et al.9 reported that TCM could reduce tumor recurrence and metastasis in patients with NSCLC, colorectal

cancer, hepatocellular carcinoma, superficial bladder cancer and gastric cancer. Zhou et al.6 also reported that TCM had synergistic antitumor effects in patients with leukemia and prostate cancer.

The large sample size obtained from our nationwide database lends statistical strength to our examination of the association between TCM and survival outcomes in patients with

HNC. The patients in our study displayed a wide range of demographic characteristics, which might render our results more

applicable to the general population. Moreover, this allowed us to perform stratified analyses according to age, sex, income and urbanization level. However, this study has several limitations. First, the NHI program covers only for TCM prescribed by TCM physicians; therefore, the NHI data do not include

overthe-counter TCM. In other words, the use of TCM may be underestimated. Second, the NHIRD is a claims-based database;

therefore, no detailed clinical information regarding cancer staging or biochemical data are recorded. It is the major limitation of this study. Besides, there may be some unmeasured

confounder correlates with TCM use that is the causative agent for better outcomes. Therefore, we recommend a randomized controlled trial to test the effects of adjunctive TCM therapy in HNC patients is necessary. In addition, we do not analyze survival outcomes in different types of TCM. Many participants had mixed types of TCM at the same time or sequentially used different types of TCM. This is very common in TCM prescription. Therefore, we could not compare the effects of different types of TCM. Finally, the NHIRD documents only the date of death, not the cause of death. The effect of TCM on HNCspecific mortality therefore cannot be analyzed.

Conclusions

Our study showed that adjunctive therapy with TCM is associated with higher survival outcome. However, this finding

should be interpreted with caution. The major limitation is the lack of information of patient cancer stage. In addition, causality

cannot be assessed with the retrospective design. A randomized controlled trial to test the effect of adjunctive TCM therapy in HNC patients is needed.9