|Home | About | Journals | Submit | Contact Us | Français|
This study was conducted to compare the effectiveness of Cyproheptadine (CY) use in patients with different stages of HCC who received different therapeutic modalities; such a comparison has not been conducted by previous large, prospective, randomized studies. We conducted a cohort study using the Taiwan Cancer Registry Database for analysis. We included patients diagnosed as having HCC from January 1, 2002, to December 31, 2011. The patient cohort comprised those who received different treatments, and we compared patients who received CY with those who did not. In total, 70,885 patients were included, and the mean follow-up duration was 1.95 years. The adjusted hazard ratio (aHR) of all-cause deaths significantly decreased in all stages in the patients who received palliative treatments with CY use compared with those who received palliative treatments without CY use (all P < 0.0001 and aHR = 0.76, 0.80, 0.66, and 0.66 for stages I, II, III, and IV, respectively). Among the patients who received no treatment, CY use alone reduced the risk of all-cause deaths in stages I-IV (all P < 0.0001 and aHR = 0.61, 0.57, 0.54, and 0.52 for stages I, II, III, and IV, respectively). Among the patients with clinical stage I-II HCC (as determined by the American Joint Committee on Cancer) who received curative treatments, CY use did not reduce all-cause deaths. CY use might improve survival in patients with HCC receiving palliative treatments or no treatment regardless of clinical stages.
Hepatocellular carcinoma (HCC) is an malignant tumor that frequently occurs from long-term chronic liver disease and cirrhosis. Sub-Saharan Africa, Hong Kong, and Taiwan are high prevalence of HCC regions with > 15 cases per 100,000 population per year . The incidence of HCC is 24.20 per 100,000 in Africa and 35.50 per 100,000 in East Asia . HCC remains the second leading cause of cancer death in Taiwan . As we know, surgical resection is the mainly curable therapy for HCC patients. However, most HCC patients are not suitable for resection because of underlying liver dysfunction or HCC extent [4,5]. For patients who are not qualified for surgical resection, treatment options include local nonsurgical methods such as percutaneous ethanol injection (PEI), radiofrequency ablation (RFA), radiation therapy (RT), transarterial chemoembolization (TACE), and systemic therapy [4,5].
Systemic therapy for HCC is a developing field [6-8]. In terms of conventional cytotoxic chemotherapy (CT), the efficacy is unsatisfactory, and the beneficial duration of CT is limited [6-8]. There is no single regimen superior to othrer combined regimen and improved survival definitely, even some randomized trials were carried out [9-11]. Sorafenib monotherapy could be considered as standard single regimen for advanced HCC. Unfortunately, the efficacy of sorafenib monotherapy was poor response rate and mild survival benefits in Taiwan .
Cyproheptadine (CY) has been used for reducing all-cause deaths or deaths due to HCC in Taiwan [13,14]. However, there were no large, prospective, randomized studies for comparison of the effectiveness in CY with that of different therapeutic modalities among HCC patients with different clinical AJCC stages. The effects of CY should be further investigated in patients with different AJCC clinical stages of HCC receiving different treatment modalities to determine the most appropriate indications of CY use for patients with HCC. In this study, we assessed the therapeutic effects of CY on HCC patients with different stages receiving different treatment modalities. To date, no epidemiologic or clinical data have clarified the effect of CY combined with curative treatments such as surgical resection, RFA, or PEI; palliative treatments such as RT, CT, and TACE; and no treatments.
The Taiwan Cancer Registry Database was used for the analysis. Patients with HCC who were treated from January 1, 2002, to December 31, 2011, were included in the study. The follow-up duration was from the index date to December 31, 2013. Our protocols were reviewed and approved by the Institutional Review Board of Taipei Medical University (TMU-JIRB No. 201402018). The cancer registry database includes information regarding the clinical stage, treatment modalities, pathological data, and death because of specific diseases [15-18]. We enrolled patients newly diagnosed as having HCC with the absence of other cancers or distant metastases. Inclusion criteria included a diagnosis of HCC (identified according to the International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] codes 070.2, 070.3, and V02.61; ICD-9-CM codes 155.0 and 155.2; and the International Classification of Diseases for Oncology, Third Edition codes C22.0 and C22.1), age > 20 years, and medical procedure received (namely undergoing curative treatments such as surgical resection, RFA, or PEI; palliative treatments such as RT, CT, or TACE; or observation with no treatment). The index date was the date of HCC diagnosis. Exclusion criteria comprised a diagnosis of other cancers before HCC diagnosis, liver transplantation, missing information on sex, distant metastasis, hepatitis B virus (HBV) infection, hepatitis C virus (HCV) infection, CY treatment since 6 months before HCC diagnosis, and age < 20 years. In total, 70,885 patients with HCC were enrolled. We categorized these patients into CY users and non-CY users. The index date of CY use was the date of HCC diagnosis. This study investigated the effects of CY use on patients with HCC who received curative treatments such as surgery, RFA, or PEI; palliative treatments such as TACE, RT, or CT; and no treatments. The primary endpoint of this study was overall death. Secondary endpoints were survival benefits after CY was used in combination with curative treatments, palliative treatments, or no treatment. The defined daily dose (DDD), recommended by the WHO, is a measure of the prescribed drug amount. The DDD is the assumed average maintenance dose per day of a drug consumed for its main indication in adults (12). Patients who received < 28 cumulative DDD (cDDD) were defined as non-CY users.
The possible confounding factors of comorbidities included age; sex; the American Joint Committee on Cancer (AJCC) clinical cancer stage; Charlson comorbidity index (CCI); and use of metformin, aspirin, statins, and sorafenib were adjusted for or stratified in the analysis. We included comorbidities observed within 6 months before and after the index date of HCC, according to the main diagnosis codes for the first admission or more than two repeated main diagnosis codes in an outpatient department.
The cumulative incidence function of death was estimated using the Kaplan-Meier method, and CY use and non-CY use among different treatment modalities were compared using the log-rank test. A time-dependent Cox proportional hazard model was used to calculate hazard ratios (HRs) for death among patients with HCC undergoing different treatment modalities with or without CY use. The HRs were adjusted for age; sex; aspirin, metformin, statins, and sorafenib; CCI score; and stage in the multivariate analysis. A stratified analysis was conducted to compare the effect of CY use among patients with the same stage of HCC receiving curative, palliative, or no treatment. All statistical analyses were conducted using SAS software, Version 9.3 (SAS, Cary, NC, USA). A two-tailed P value of < 0.05 was considered significant.
In total, we enrolled 70,885 patients with HCC, and the mean follow-up duration in the study was 1.95 years. Of these patients, 5,198 were CY users, and 65,687 were non-CY users (Table 1). Among the 65,687 non-CY users, 24,995 (38.05%) received curative treatments, 23,813 (36.25%) received palliative treatments, and 16,879 (26.70%) did not receive any treatment. Among the 5,198 CY users, 2,422 (46.59%) received curative treatments, 2,044 (39.32%) received palliative treatments, and 732 (14.08%) did receive no treatment. Furthermore, the proportion of the patients who received curative treatments was higher in the CY user group than in the non-CY user group. The mean age of the patients was higher and the follow-up duration was longer in the CY user group than in the non-CY user group. Aspirin, statins, and metformin were more frequently used in the CY user group. Regarding comorbidities, the CCI score was higher in the CY user group, with a higher proportion of the patients having CCI scores of ≥ 3. Furthermore, the proportion of the patients with advanced-stage (III-IV) HCC was higher in the non-CY user group. Deaths due to HCC (disease-specific survival) in the non-CY user group were significantly higher than those in the CY user group. In addition, we assessed the risk of all-cause deaths and HCC-specific deaths according to the CY use and HCC treatments stratified by the AJCC clinical stage (Table 2).
As shown in Table 2, patients who received palliative treatments or no treatment with CY use had a lower all-cause death rate per 1,000 person-years in all stages than did patients who received palliative treatments or no treatment without CY use. The HCC death rate per 1,000 person-years in all stages also showed the same trend, and the HCC death rate was lower in the patients who received palliative treatments or no treatment with CY use (Table 2). However, the all-cause and HCC death rates were similar in the patients with stage I-II HCC who received curative treatments with or without CY use. To investigate the effect of CY use in patients with HCC receiving different treatments, the patients were categorized into no treatment, curative treatment, and palliative treatment groups (Table 3). Our data demonstrated that after the time-dependent Cox proportional hazard model was adjusted for age, sex, drug use (aspirin, metformin, statins, and sorafenib), CCI scores, and stages, the all-cause and HCC death rates significantly improved in the no treatment, curative treatment, and palliative treatment groups with CY use. The adjusted HRs (aHRs; 95% confidence intervals [CIs]) of all-cause and HCC-specific deaths decreased in the patients who received no treatment with CY use compared with those who received no treatment without CY use (P < 0.0001; aHRs = 0.56 and 0.59 for all-cause and HCC-specific deaths, respectively). Figure 1A presents Kaplan-Meier survival curves for the patients who received no treatment with or without CY use. Our data also showed that all-cause and HCC-specific death rates in CY users were significantly superior to those in non-CY users who received curative treatments (P < 0.0001 and P < 0.01, aHRs = 0.80 and 0.81 for all-cause and HCC deaths, respectively). In the palliative treatment group, CY users also had lower death rates compared with other non-CY users. The aHRs for all-cause and HCC-specific deaths decreased sequentially from curative treatment (aHRs = 0.80 and 0.81, respectively) to palliative treatment (aHRs = 0.71 and 0.63, respectively), and to no treatment (aHRs = 0.56 and 0.59, respectively) (Table 3). Figure 1B shows better overall survival in the patients who received curative treatments with CY use compared with those who received curative treatments without CY use (P < 0.001); however, Figure 2B and and2E2E demonstrates that overall survival rates were statistically equal in patients with stage I-II HCC who received curative treatments with or without CY use (P = 0.2422 and 0.1529 for clinical stages I and II, respectively).
We estimated the risks of HCC-specific and all-cause deaths according to both CY use status and HCC treatment stratified by the AJCC stage (Table 4). Compared with the patients who received palliative treatments without CY use, the patients who received palliative treatments with CY use had a significant overall survival benefit in all stages. Furthermore, compared with the patients who received palliative treatments without CY use, the aHRs of all-cause deaths significantly decreased in all stages in the patients who received palliative treatments with CY use (all P < 0.0001 and aHR = 0.76, 0.80, 0.66, and 0.66 for stages I, II, III, and IV, respectively). Among the patients who received no treatment, CY use alone reduced the risk of all-cause death in stages I-IV (all P < 0.0001 and aHR = 0.61, 0.57, 0.54, and 0.52 for stages I, II, III, and IV, respectively). The aHRs of all-cause death decreased in stages III and IV in the patients who received curative treatments with CY use compared with those who received curative treatments without CY use (P < 0.05 and P < 0.0001, respectively, and aHR = 0.86 and 0.50 for stages III and IV, respectively). Among the patients with stage I-II HCC who received curative treatments, CY use did not reduce all-cause deaths (Table 4; Figure 2B and and2E).2E). However, among the patients with stage III HCC who received curative treatments, CY use reduced all-cause deaths but did not reduce HCC-specific deaths (Table 4 and Figure 2H). Among the patients with stage IV HCC who received curative treatments with CY use, overall and disease-specific survival benefits were observed (Table 4 and Figure 2K).
HCC-specific deaths were lower in all stages among the patients who received palliative treatments with CY use, compared with those who received palliative treatments without CY use (P < 0.0001, P < 0.01, P < 0.0001, and P < 0.0001 and aHR = 0.60, 0.67, 0.64, and 0.59 for stages I, II, III, and IV, respectively; Table 4). HCC-specific deaths did not significantly decrease in early HCC stages (I-III) among the patients who received curative treatment with CY use (Table 4 and Figure 2). However, compared with the patients who received curative treatments without CY use, HCC-specific deaths significantly decreased in stage IV in the patients who received curative treatments with CY use (P < 0.05 and aHR = 0.62). HCC-specific deaths were significantly lower in all HCC stages (I-IV) in the patients who did not receive any treatment but used CY (Table 4 and Figure 2).
Figure 1 illustrates the Kaplan-Meier curves of overall survival for the patients in the three treatment arms. The patients who received curative treatments with CY use exhibited the highest overall survival rate (log-rank test, P < 0.0001). The 2-year overall survival rates were 24.23%, 44.93%, and 84.44% in the CY users who received no treatment, palliative treatments, and curative treatments, respectively. By contrast, the 2-year overall survival rates were 8.22%, 19.97%, and 72.24% in the non-CY users who received no treatment, palliative treatments, and curative treatments, respectively.
CY is a histamine and serotonin antagonist that has been observed to cause weight gain in observational studies of patients with advanced cancers . CY appears to be effective in patients with carcinoid syndrome who have anorexia or cachexia [19-21]. In such patients, CY presumably acts by directly counteracting increased serotonin activity . Recently, CY showed an anticancer effect in various cancer cells such as human colon carcinoma cells (HT29), acute lymphoblastic leukemia cells, human breast cancer cells (MCF-7), and HCC (HepG2 and Huh-7) [22-25]. In an in vivo study, CY exhibited antitumor activity in urothelial carcinoma by targeting GSK3β to suppress mTOR and β-catenin signaling pathways . CY inhibits the proliferation of HCC cells by blocking cell cycle progression through the activation of p38 MAP kinase . In addition, CY promotes G2/M arrest and increases the susceptibility of cancer cells to radiation-induced DNA damage . In 2012, a case report from Taiwan revealed dramatic changes in patients with HCC who had complete remission after CY treatment . Subsequently, a small retrospective study conducted in Taiwan also demonstrated the anti-HCC effect of CY in 52 sorafenib-treated patients with advanced HCC . The aHR of all-cause deaths was 0.24 (95% CI: 0.10-0.58) during the follow-up duration of 17 months . However, a small sample size is a critical limitation of this study, which may limit the scientific validity of the conclusion. By contrast, the current study may provide detailed information regarding CY use in different stages and treatment modalities because of the inclusion of a large cohort.
Several observational studies have found that statin use is associated with a lower risk of HCC and might improve survival in patients with HCC [27-32]. In patients with HCC who tolerated palliative treatments, the combined use of statins reduced mortality . The use of only statins might possibly reduce all-cause deaths in patients with HCC receiving no treatments . In addition, statin use improved the survival outcomes of curative treatments in patients with HCC . Metformin use in Taiwan appeared to be associated with lower mortality in diabetic patients with HCC (aHR = 0.24; 95% CI: 0.07-0.80, P = 0.020) . A matched-pair analysis in China demonstrated that aspirin use combined with palliative treatments reduced all-cause deaths in patients with HCC (aHR = 0.50; 95% CI: 0.28-0.80, P = .018) . Sorafenib is the preferred therapy for HCC [35,36]. Molecular-targeted therapies offer the potential for prolonged survival, and the aHR in the sorafenib group was 0.69 (95% CI: 0.55-0.87; P < 0.001); nevertheless, objective tumor remissions are rare . Taken together, the aforementioned drugs might cause a bias in the analysis; thus, we made adjustments for these known drugs in our analysis.
In the current study, we separated different stages and treatment modalities because stages and curative or no treatments majorly influence the survival of patients with HCC [4,38]. A previous study demonstrated the effects of CY combined with sorafenib in patients with advanced-stage HCC without other treatments; however, this study involved only a small sample size and a short follow-up duration . There are no clear suggestions of CY use for clinical practice in patients with HCC who receive curative or palliative treatments or cannot tolerate any treatments. CY inhibited tumor growth without significant toxicity, and in clinical trials, the drug was well tolerated without hematological toxicity [19,20,39]. The cost of CY is cheap. In patients with HCC who could not tolerate any treatment, CY might be an alternative choice (Tables 3 and and4).4). By contrast, CY use might not be beneficial in patients with stage I-II HCC who could tolerate curative treatments (Table 4). However, in our study, the patients with HCC who received palliative treatments combined with CY had decreased all-cause and HCC-specific deaths, irrespective of stages (Tables 3 and and44).
Although the CY user group had a higher number of elderly patients with HCC and higher CCI scores (Table 1), the death rate was lower in this group than that in the non-CY user group (Table 2). Older age and higher CCI scores are poor independent prognostic factors for patients with HCC [40,41]. Therefore, because of differences between the CY user and non-CY user groups, the effect of CY use would be underestimated and lead to bias toward the null hypothesis [42-44].
The patients with HCC who received > 28 cDDD of CY were defined as CY users. This is because in a previous retrospective study, 32 patients with HCC enrolled in the CY user group were administered CY for > 4 weeks . Compared with the aHR in our study, the aHR was higher in the previous study, with the aHR of all-cause deaths being 0.24 (95% CI: 0.10-0.58; Tables 3 and and4).4). The reason for this finding might be the difference among selected patients with HCC, different therapeutic modalities, smaller sample size, shorter follow-up duration, and continuous CY administration for > 4 weeks in the previous study. In our study, the patients received > 28 cDDD of CY rather than a continuous dosage. The two prerequisites for the development of cancer are cell proliferation and the inhibition of apoptosis [45-47]. Preclinical studies have revealed that CY-induced apoptosis is preceded by the activation of the mitochondrial pathway of caspase activation , reduction of D-cyclin expression , inhibition of PI3K/AKT signaling , blocking of cell cycle progression through the activation of p38 MAP kinase , or targeting of GSK3β to suppress mTOR and β-catenin signaling pathways . A continuous CY-induced apoptosis effect is necessary to reduce tumor burden and cancer cell proliferation [45-47]. This is the reason why the aHRs in our study were not higher than those reported in the study of Feng et al. In future clinical trials, we suggest continuous administration of CY rather than intermittent administration, because CY is well tolerated and already approved in multiple countries for clinical use as an antihistamine and an appetite stimulant; thus, it could be moved directly into clinical trials for its anticancer effect .
The aHRs for all-cause and HCC-specific deaths decreased sequentially as follows: curative treatment (aHRs = 0.80 and 0.81, respectively), palliative treatment (aHRs = 0.71 and 0.63, respectively), and no treatment (aHRs = 0.56 and 0.59, respectively) (Table 3). This phenomenon reflects the existence of different levels of tumor burden in the therapeutic modalities of no treatment, palliative treatments, and curative treatments. CY use was associated with more distinct survival benefits, along with more residual tumors in different treatments with more tumor burden. Consequently, HCC-specific death in CY users was statistically significant only in patients with stage IV HCC who received curative treatments; this is because even curative treatments might still have residual tumor in patients with advanced HCC (Table 4) . However, the patients who received palliative treatments or no treatment still exhibited a large amount of tumor burden . The reduction trends of aHRs associated with CY use regarding all-cause deaths or HCC-specific deaths were proportional to AJCC clinical stages (Table 4 and Figure 2), because more advanced stages of HCC were reported to involve a larger amount of tumor burden . CY use could induce apoptosis of residual HCC cells to improve survival [25,26,39,48].
This study has several strengths. To our knowledge, this study has the largest sample size. In addition, clinical stages and various therapeutic modalities were differentiated. The follow-up duration was also longer in this study. The results suggest that CY can be used as an alternative drug to improve survival in patients with HCC receiving palliative or no treatments, regardless of clinical stages. In patients with stage I-II HCC receiving curative treatments, CY use is not beneficial. This study is the first to indicate the optimal therapeutic decisions combined with CY use for patients with HCC according to cancer stages. Palliative or no treatments with CY use are more suitable; this finding should be considered in future clinical studies.
The limitations of this study are that the actual CY dose and duration were not measured; thus, the actual dose-response effect was unclear. The compliance of CY use was not clear in the analysis, and the potential adverse effects of CY use are unknown. Ultimately, obtaining this crucial information necessitates conducting a large randomized trial with a suitable regimen in well-selected patients that compares standard approaches. Additionally, the diagnoses of all comorbidities were dependent on ICD-9-CM codes. Nevertheless, the Taiwan Cancer Registry Administration randomly reviews charts and interviews patients to verify the accuracy of the diagnoses, and hospitals with outlier chargers or practices may undergo an audit and subsequently receive heavy penalties if malpractice or discrepancies are identified . Finally, there was no information on tobacco use, alcohol consumption, dietary habits, socioeconomic status, laboratory data, or the body-mass index in the database, which may also be risk factors for the survival analysis. However, these limitations are unlikely to have compromised the results, given the magnitude and statistical significance of the observed effects in the current study .
Taipei Medical University & Wan Fang Hospital funding (105TMU-WFH-08).
Conception and Design: Mao-Chih Hsieh, Szu-Yuan Wu*. Financial Support: Taipei Medical University & Wan Fang Hospital. Collection and Assembly of Data: Mao-Chih Hsieh, Alexander TH Wu, Jyh-Ming Chow, Chia-Lun Chang, Kevin Sheng-Po Yuan, Szu-Yuan Wu. Data Analysis and Interpretation: Mao-Chih Hsieh, Szu-Yuan Wu. Administrative Support: Szu-Yuan Wu*. Manuscript Writing: All authors. Final Approval of Manuscript: All authors.