Association between elcatonin use and cancer risk in Japan: A follow-up study after a randomized, double-blind, placebo-controlled study of once-weekly elcatonin in primary postmenopausal osteoporosis.

OBJECTIVES: On July 20, 2012, the European Medicines Agency (EMA) provided a recommendation that limits the long-term use of calcitonin. Based on this recommendation, we investigate the presence or absence of a cancer diagnosis in subjects who participated in the ongoing clinical trial of elcatonin.
METHODS: When the EMA gave this recommendation, we were conducting "a 3-year placebo-controlled clinical study for elcatonin" (hereinafter, referred to as "the original study"). In accordance with the recommendation of EMA, we performed an intermediate analysis on the subjects of the original study to assess whether the study could be safely continued. We also added a 2-year follow-up study to investigate the risk of carcinogenesis for 5 years from the start of administration. We compared the risk of carcinogenesis estimated by person-year method in elcatonin group with that in placebo group.
RESULTS: In the original study, there were 433 subjects in the elcatonin group and 437 in the placebo group, of whom 322 and 323, respectively, agreed to participate in the additional follow-up study. The average cancer incidence rate per 100 person-years 5 years from the start of administration was 1.02 in the elcatonin group and 1.08 in the placebo group, respectively, and there was no clear difference.
CONCLUSIONS: Since the number of cases in this study was small, we cannot completely deny the cancer risk due to long-term administration of this drug. However, the results do not suggest that once-weekly administration of 20 units of elcatonin increases the carcinogenic risk.

Introduction

On July 20, 2012, the European Medicines Agency (EMA) provided a recommendation that limits the long-term use of calcitonin products. This was because some results of clinical trials showed that calcitonin significantly increased the carcinogenic risk compared with a placebo [1]. Considering the increase in carcinogenic risk due to long-term administration and its limited benefits of preventing vertebral fracture in patients with postmenopausal osteoporosis, the EMA concluded that the benefit of calcitonin products does not exceed the risk [1].

In contrast, elcatonin is a synthetic eel calcitonin derivative in which the S–S bond of eel calcitonin is changed to a C–C bond and differs from salmon calcitonin products widely used in Europe. Their dosage regimens are also different. There have been no reports from clinical trials and postmarketing studies that elcatonin increases the risk of carcinogenesis.

In this study, we followed the recommendation of the EMA about calcitonin products, and confirmed whether there is any problem in the safe continuation of the original study from the viewpoint of protecting the subjects in the phase III clinical study (the original study [2], involving once-weekly administration of 20 units of elcatonin) that had already been conducted. We established an independent data-monitoring committee and conducted an open-label interim analysis of the cases diagnosed as cancer; however, investigators and the sponsor were blinded. Based on the criteria for continuity set forth by the committee prior to code breaking, it was judged that there was no problem in continuation of the original study. Based on this judgment, we continued the original study, and according to the suggestion of the committee, we conducted a 2-year follow-up study after the end of the original study which lasted for 3 years so that maximum data on the carcinogenic risk of elcatonin could be acquired. For 5 years, we examined the presence of cancer and the health statuses of the subjects.

Methods

Study design

This study is a follow-up study after the phase III trial involving once-weekly elcatonin administration. The original study was a “multicenter, randomized, double-blind, placebo-controlled phase III study of once-weekly elcatonin in primary postmenopausal osteoporosis” [2], which was conducted between March 2006 and August 2013. After the termination or discontinuation of the original study, we obtained informed consent from the subjects of the original study for the follow-up study which was conducted from April 2013 to September 2015. The subjects were interviewed every year for 5 years to determine if they had any cancer diagnosis and to establish their survival status. Details of patients diagnosed with cancer were obtained from the medical institution where the cancer was diagnosed. The study was designed by H.O. and T.S. (the investigators) and by Asahi Kasei Pharma Corporation (the sponsor). The sponsor was responsible for data collection, data management, and reliability of the results of the analysis. The investigators and sponsor conducted analyses for manuscript submission, and all authors contributed equally to the manuscript and approved of its content. The original trial was registered with clinicaltrials.jp (https://www.clinicaltrials.jp/cti-user/trial/ShowDirect.jsp?clinicalTrialId=11928).

Subjects

In the original study, we investigated postmenopausal women, aged ≥65 years, who were diagnosed with primary osteoporosis based on the diagnostic criteria for primary osteoporosis (2000 revision) in Japan and were able to ambulate independently. The original study was conducted in accordance with the tenets of the Declaration of Helsinki and Good Clinical Practice and was approved by the Institutional Review Board of each participating medical institution. Written informed consent was obtained from all study subjects prior to administration of the treatment drug.

After completion or discontinuation of the original study, the investigators explained the content of this follow-up study to the subjects and obtained written informed consent on the basis of freedom of participation. Written informed consent was obtained from the legal guardians of subjects who were unable to provide informed consent themselves. The investigators kept records of the written informed consent obtained.

This follow-up study was also conducted in accordance with the tenets of the “Declaration of Helsinki: Ethical principles for medical research involving human subjects,” “Ministerial Ordinance on Good Clinical Practice for Drugs,” and related laws and regulations. Careful attention was paid particularly to the protection of the privacy of the subjects.

Treatment

In the original study, subjects were randomly divided into 2 groups based on the dynamic allocation method using “number of existing vertebral fractures,” “spontaneous pain at rest (chronic pain),” and “age” as stratification factors. The subjects received 20 units of elcatonin or placebo once weekly for 144 weeks via intramuscular injection. In addition, all subjects took calcium (400 mg) and native vitamin D (400 IU) supplements once a day. In addition, the follow-up study was conducted under general practice settings without treatment restrictions.

Diagnosis of cancer

During the 5-year period from the start of the original study, the investigators examined the presence or absence of cancer, the type of cancer, and the timing of the diagnosis if present. The investigators interviewed the legal guardians of subjects who were unable to visit the hospital or were confirmed to have died, regarding the subjects’ statuses or their statuses until the time of death. For subjects diagnosed with cancer, the investigators obtained the name of the medical institutions at which they were diagnosed and information about physicians in charge, etc. from the subjects or their legal guardians, and also obtained detailed information from the physicians. The investigators received periodical medical expert’s review based on detailed information of cancer. The timing was at the start, the fourth year from the start (from the start to the fourth year), and the fifth year from the start (from the fourth year to the fifth year).

Confirmation of living status

Physicians in each institution confirmed the subjects’ health status at the start of the follow-up study and contacted subjects who were unable to visit study sites via telephone to confirm their status.

Statistical analysis

The rates of diagnosed cases of cancer in each group were estimated in person-years method. The difference of the hazard and the hazard ratio and 95% confidence interval of these were calculated. SAS ver. 9.4 (SAS Institute, Cary, NC, USA) was used for statistical analysis.

Results

In the original study, there were 433 and 437 subjects in the elcatonin (EL) group and placebo (P) group, respectively. Among those, informed consent was provided by 322 and 323 subjects or legal guardians in the follow-up study, respectively (Fig. 1).

Fig. 1

Disposition of subjects. This study is a 5-year survey combining the original study (3 years) already underway and a 2-year follow-up survey. EL, elcatonin group; P, placebo group.

The distributions of the EL and P groups in the follow-up study were 0.3% and 0.0% (<3 years), 0.6% and 1.9% (≥3 and < 4 years), 34.5% and 31.0% (≥4 years and <5 years), and 64.6% and 67.2% (≥5 years), respectively. There was no obvious difference between the 2 groups, and most patients were successfully followed for at least 4 years. The mean age, height, and weight of the subjects in the follow-up study at the start of the study were 74.9 ± 5.7 years, 147.6 ± 6.1 cm, and 51.0 ± 7.9 kg in the EL group, respectively; and 74.9 ± 5.7 years, 147.7 ± 6.3 cm, and 50.7 ± 7.8 kg in the P group, respectively (Table 1). No difference was found between both groups.

Table 1

Patient characteristics at baseline and at the start of follow-up study.

Variable	Baseline		At the start of follow-up study
	EL (N = 433)	P (N = 437)	EL (N = 322)	P (N = 323)
Age, yr	75.5 ± 5.7	75.5 ± 5.7	74.9 ± 5.7	74.9 ± 5.7
Height, cm	147.4 ± 6.0	147.3 ± 6.6	147.6 ± 6.1	147.7 ± 6.3
Weight, kg	50.5 ± 8.0	50.3 ± 7.7	51.0 ± 7.9	50.7 ± 7.8

Values are presented as mean ± standard deviation.

EL, elcatonin group; P, placebo group.

The incidence rates of cancer after the start of the study drug in the EL and P groups, respectively, were: 0.0% and 0.7% (less than half a year); 0.2% and 0.5% (more than half a year, less than 1 year); 1.0% and 0.8% (more than 1 year, less than 2 years); 2.3% and 1.0% (more than 2 years, less than 3 years); 0.9% and 1.2% (more than 3 years, less than 4 years); 0.9% and 0.6% (more than 4 years, less than 5 years); and 0.0% and 1.4% (more than 5 years). There was no notable difference between both groups (Table 2). There was no difference in the age-specific incidence of cancers between the EL and P groups (Table 3). There was no specific location where cancer frequently occurred in the EL group, and the types of cancer were varied in both groups (Table 4). Furthermore, the estimation of the hazard ratios of cancer occurrence in person-years showed no clear difference between both groups (Table 5).

Table 2

Incidence of cancer in each period after the first administration.

Group	Period after first dose, yr
	<0.5	≥0.5, <1	≥1, <2	≥2, <3	≥3, <4	≥4, <5	≥5
EL	0/433 (0.0)	1/412 (0.2)	4/404 (1.0)	9/385 (2.3)	3/321 (0.9)	3/319 (0.9)	0/208 (0.0)
P	3/437 (0.7)	2/415 (0.5)	3/394 (0.8)	4/386 (1.0)	4/323 (1.2)	2/317 (0.6)	3/217 (1.4)

Values are presented as number of patients diagnosed with cancer/total number (%).

EL, elcatonin group; P, placebo group.

Table 3

Incidence of cancer by age.

Age, yr	EL (N = 433)		P (N = 437)
	No. of patients diagnosed with cancer	Total No. of cases	No. of patients diagnosed with cancer	Total No. of cases
≥65, <70	2	63	2	65
≥70, <75	3	134	5	131
≥75, <80	7	135	9	142
≥80, <85	5	73	5	68
≥85	3	28	0	31

EL, elcatonin group; P, placebo group.

Table 4

Incidence of cancer by site.

Variable	EL (N = 433)	P (N = 437)
Diagnosed with cancer	20 (4.6)	21 (4.8)
Stomach	2 (0.5)	4 (0.9)
Colon	3 (0.7)	3 (0.7)
Rectum	1 (0.2)	0 (0.0)
Malignant mesothelioma of the abdomen	1 (0.2)	0 (0.0)
Hepatocarcinoma	0 (0.0)	1 (0.2)
Pancreas	3 (0.7)	1 (0.2)
Malignant neoplasm of lung	2 (0.5)	0 (0.0)
Adenocarcinoma of lung	1 (0.2)	1 (0.2)
Epidermoid carcinoma	0 (0.0)	2 (0.5)
Breast	2 (0.5)	3 (0.7)
Uterus	1 (0.2)	0 (0.0)
Bladder	0 (0.0)	1 (0.2)
Transitional cell carcinoma of kidney pelvis and ureter	0 (0.0)	1 (0.2)
Kidney	1 (0.2)	0 (0.0)
Thyroid	0 (0.0)	1 (0.2)
Lymphoma	1 (0.2)	0 (0.0)
B-cell lymphoma	0 (0.0)	1 (0.2)
Plasma cell myeloma	1 (0.2)	0 (0.0)
Recurrent, metastasis
Bladder	1 (0.2)	0 (0.0)
Lung with distant metastasis	0 (0.0)	1 (0.2)

Values are presented as number (%).

EL, elcatonin group; P, placebo group.

Table 5

The rate of diagnosed cases of cancer (in person-years).

Group	Research period (person-years)	The rate of diagnosed cases of cancer (/yr/100)		Hazard ratio (=EL/P)		Difference of hazard (=EL-P)
		Estimate	95% CI	Estimate	95% CI	Estimate	95% CI
EL (n = 20)	1956.8	1.0221	0.5741–1.4700	0.949	0.514–1.750	−0.055	−0.698–0.587
P (n = 21)	1949.5	1.0772	0.6165–1.5379	–	–	–	–

EL, elcatonin group; P, placebo group; CI, confidence interval.

Discussion

In accordance with the recommendation of EMA for calcitonin products, we assessed the presence of cancer and the health statuses of the subjects over the course of 5 years from the start of the original study. There were no differences in subjects’ demographics between both groups during the periods of these 2 studies and at the start of the original study. There were no clear differences between the 2 groups in the incidence of cancer, the incidence including cancer and death, the timing of onset of cancer, the age-specific incidence of cancer, the cancer site, and the annual average incidence of cancer in person-years. When considering the increase in carcinogenic risk due to long-term administration and its limited benefits of preventing vertebral fracture in postmenopausal osteoporosis patients, the EMA concluded that the benefit of calcitonin products for patients with osteoporosis does not exceed the risk [1]. Several meta-analyses have been reported regarding the carcinogenicity of salmon calcitonin. Of 18 randomized trials of salmon calcitonin products targeting patients with osteoporosis that evaluated patients for 6 months or more, 15 trials showed that calcitonin groups had a higher incidence of cancer than placebo groups [3]. A meta-analysis of 17 randomized controlled trials for intranasal salmon calcitonin products targeting patients with osteoporosis reported that the groups that received calcitonin for more than 6 months had slightly higher carcinogenic risk than the placebo groups [4]. However, in a meta-analysis of 22 trials that added new trials on oral salmon calcitonin products and the like, the risk was smaller than reported in the previous papers. While the relationship between calcitonin administration and the carcinogenic risk cannot be completely denied, skeptical reports on the causal relationship were also found [5]. Furthermore, in a case-control study, it was reported that the risk of liver cancer among females was higher, but the risk of breast cancer was lower in the group that received intranasal salmon calcitonin products compared to the control group. It has also been shown that the risk varies according to the cancer type [6].

Our results suggest that once-weekly administration of 20 units of elcatonin bears no definite risk of cancer. Differences in carcinogenic risk could not be evaluated due to the low incidence rates of the individual cancers.

From our results, we consider that the lack of increase in carcinogenic risk due to elcatonin administration was because elcatonin is a synthetic eel calcitonin derivative and is different from the salmon calcitonin widely used in Europe, and its dosage regimen (20 units elcatonin once-weekly) is lower than that of intranasal salmon calcitonin products (200 units daily) or oral products (100–400 units daily). In clinical trials and postmarketing studies, there has been no report that elcatonin increases the risk of carcinogenesis, and our results support these findings.

Our study has some limitations. First, it was not designed from the beginning to verify the carcinogenic risk of calcitonin, and some background information on patients’ lifestyles in relation to cancer risk (such as drinking and smoking) was therefore not available. Second, it was not a meta-analysis as indicated in the report of the EMA (calcitonin group: 2258 cases and placebo group: 976 cases). Therefore, it is impossible to completely deny the carcinogenic risk due to long-term administration of elcatonin because of the small number of cases. Third, cancer examinations were not actively performed, and the possibility of cases with undiagnosed cancer could therefore not be ruled out. The age-specific incidence rates of cancer among women according to epidemiological data in Japan (per 100 population, in 2013) were 0.8 (65–69 years), 1.0 (70–74 years), 1.2 (75–79 years), and 1.5 (80–84 years) [7]. Despite these limitations, the cancer incidence was in agreement with Japanese epidemiological data (Table 5). Fourth, the sponsor (not independent third party) was responsible for data collection, management, and analysis. Deviation or bias in cancer reports cannot be ruled out.

Conclusions

Our findings did not suggest an increase in the risk of carcinogenesis with administration of 20 units of elcatonin weekly.

CRediT authorship contribution statement

Hiroaki Okamoto: Formal analysis, Writing - original draft. Nayumi Shibazaki: Formal analysis, Writing - original draft. Takeshi Yoshimura: Writing - original draft. Toyonobu Uzawa: Writing - original draft. Toshitsugu Sugimoto: Formal analysis, Writing - original draft.

Conflicts of interest

H.O. received consulting fees from Asahi Kasei Pharma Corporation as well as research grants from Takeda Pharmaceutical, MSD, Ono Pharmaceutical, Astrazeneca, Merck, Chugai Pharmaceutical, Taiho Pharma, Bristol-Myers Squibb, Eli Lilly and Daiich Sankyo. N.S., T.Y., and T.U. are employees of Asahi Kasei Pharma. T.S. has received research grants from Astellas Pharma, Eisai, Daiichi-Sankyo, Chugai Pharmaceutical, and Eli Lilly Japan, as well as consulting and/or lecture fees from Asahi Kasei Pharma and Daiichi-Sankyo.