Cellular, animal, and human epidemiological studies suggested that benzodiazepines increase the risk of cancer and cancer mortality. Obesity is also clearly linked to carcinogenesis. However, no human studies have examined benzodiazepine-associated carcinogenesis as assessed by changes in cancer biomarkers.
A total of 19 patients were recruited, and received a 6-week treatment of 0.5 mg lorazepam. The measured cancer biomarkers were angiopoietin-2 (ANG-2), soluble CD40 ligand, epidermal growth factor, endoglin, soluble Fas ligand (sFASL), heparin-binding EGF-like growth factor (HB-EGF), insulin-like growth factor binding protein, interleukin (IL)-6, IL-8, IL-18, plasminogen activator inhibitor (PLGF), placental growth factor, transforming growth factor (TGF)-α, tumor necrosis factor (TNF)-α, urokinase-type plasminogen (uPA), vascular endothelial growth factor (VEGF)-A, VEGF-C, and VEGF-D.
Six cancer biomarkers were significantly increased in all patients as a whole. The subgroup analysis revealed a distinct pattern of change. Overweight patients showed a significant increase in 11 cancer biomarkers, including ANG-2, sFASL, HB-EGF, IL-8, PLGF, TGF-α, TNF-α, uPA, VEGF-A, VEGF-C, and VEGF-D. However, normal-weight patients did not show any changes in cancer biomarkers.
Adiposity may have primed the carcinogenic potential, leading to lorazepam-associated carcinogenesis in overweight patients. Epidemiological studies addressing this issue should consider the potential modulator contributing to benzodiazepine-associated carcinogenesis.
Benzodiazepines refer to a chemical structure consisting of a benzene ring fused to a diazepine ring [
The adverse effects of benzodiazepine represent a substantial public health concern. These include daytime fatigue, decreased cognitive function, tolerance, dependence, impaired immune system, and increased risk of mortality. Benzodiazepine-associated excess mortality has been linked to the potential for carcinogenesis [
Importantly, carcinogenesis is a progressive chronic process that includes angiogenesis, cell proliferation, cell metastasis, cell adhesion, apoptosis, and inflammation [
Between January 2017 and June 2017, subjects who were hospitalized for treatment of depression and insomnia at the Beitou Branch of Tri-Service General Hospital, National Medical Defense Center, Taiwan, were considered eligible for participation in the study. Written informed consent was obtained in accordance with the National Health and Medical Research Council guidelines. All participants were fully informed about the aims and details of the study and were free to withdraw their consent at any time. The Institutional Review Board for the Protection of Human Subjects at the Tri-Service General Hospital approved the protocol (IRB No. 1-104-05-101).
To avoid potential confounders, patients were required to be male, drug-naïve, between the ages of 20 and 40, and in good health. The exclusion criteria were as follows: 1) having suffered from or suffering from cancer or a tumor; 2) history of a major medical disorder (e.g., hypertension, diabetes, or rheumatoid arthritis); 3) history of a neurological disorder (e.g., meningitis, epilepsy, or multiple sclerosis); 4) history of a major psychiatric disorder (e.g., bipolar disorder, schizophrenia, or mental retardation); 5) diagnosis of a substancerelated disorder; 6) previous head trauma with loss of consciousness; and 7) previous exposure to psychotropic agents (e.g., hypnotics, antidepressants, mood stabilizers, antipsychotics). Tobacco use and alcohol consumption were not allowed during the study period. Patients with major depression were excluded because the likelihood of polypharmacy during the study period was high. All were diagnosed with an adjustment disorder with depressed mood. The diagnosis was made by two certified psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition.
All patients received 0.5 mg of lorazepam at night for the treatment of insomnia for six weeks. If the patient’s depressive symptoms required pharmacotherapy, 20 mg of fluoxetine or 10 mg of escitalopram was given. Because epidemiological studies have shown that obesity is associated with increased risk of several cancer types [
The procedures for collection, preparation, freezing, and thawing of all serum samples used in this study were performed in a highly consistent manner, including the collection and handling of commercial samples. After fasting, peripheral venous blood samples were collected in the morning between 8:00 and 9:00 by venipuncture from patients prior to and after a 6-week treatment with lorazepam. For the preparation of the serum samples, 10 mL of peripheral blood was drawn into a Vacutainer Gel and Clot Activator tube and centrifuged to separate the serum, which was then aliquoted into NUNC-cryovial tubes and stored at -84°C. None of the serum samples had been previously thawed prior to the thawing for the Luminex assay.
The serum levels of human cancer markers were measured by using Luminex, which is a suspension assay that combines the principles of a standard sandwich immunoassay with flow cytometry, thus allowing a multiplex analysis of up to 100 individual biomolecules in a single microtiter plate well [
The participants were divided into normal-weight and overweight patients, whose BMI exceeded 25 kg/m2. Group differences in continuous variables were analyzed using independent sample t-tests. Group differences in baseline cancer markers were analyzed with a multivariate general linear model with covariate adjustment. The covariates were age, height, weight, educational levels, and the use of fluoxetine.
The paired sample t-tests were performed to examine within-group differences in cancer markers after lorazepam treatment. For the cancer markers that significantly changed after lorazepam treatment, another multivariate general linear model was created. In this model, changes in the levels of cancer markers were calculated as dependent variables, and antidepressant treatment (fluoxetine or escitalopram) was entered as a nominal variable. Changes in body weight also covaried. Identification of whether the patterns of cancer biomarker changes were different for the two groups was done by examining a multivariate general linear model with a covariate of weight changes. The analyses were considered statistically significant if the p values were less than or equal to 0.05 (two-tailed). All statistical analyses were conducted using the Statistical Package for the Social Sciences software for Windows ver. 22 (IBM Corp., Armonk, NY, USA).
Between January 2017 and June 2017, a total of 19 male inpatients were recruited, with a mean age of 26.1±3.4 years. The participants were divided into an overweight group (n=9) and a normal-weight group (n=10) according to the BMI cut-off point of 25 kg/m2.
This is the first human study to examine the carcinogenic potential of lorazepam through a simultaneous analysis of 18 human cancer biomarkers. Our data showed that the levels of sFASL, HB-EGF, IL-8, TGF-α, uPA, and VEGF-A were significantly increased after the 6-week treatment of lorazepam. However, the lorazepam-induced cancer biomarker changes were significantly different for the overweight and the normal-weight patients. In the normal-weight patients, lorazepam did not induce any changes in the cancer biomarkers; however, in the overweight patients, a total of 11 cancer biomarkers were significantly increased, including ANG-2, sFASL, HB-EGF, IL-8, PLGF, TGF-α, TNF-α, uPA, VEGF-A, VEGF-C, and VEGF-D. The patterns of change in the cancer biomarkers revealed that the slopes were significantly steeper in the overweight patients. Taking these findings together, lorazepam may not induce carcinogenic potential in normal-weight individuals, while lorazepamassociated carcinogenic potential may be determined in overweight patients.
We suggest that lorazepam may work in concert with adiposity, thereby promoting carcinogenesis. The rise in the numbers of overweight and obese individuals has been substantial and widespread over the past few decades. A large volume of epidemiological evidence points to an association between being overweight and increased risk of several cancer types [
Preclinical and animal research, randomized controlled trials, and epidemiological studies have all reported that benzodiazepines may increase the risk of carcinogenesis and cancer mortality [
Lorazepam-associated carcinogenesis may be related to the progression of tiny pre-existing cancers rather than from the effects of microscopic cancer initiation. One study analyzed all randomized, placebo-controlled trials for the four drugs-Z-drugs (zaleplon, eszopiclone, and zolpidem)-and ramelteon and reported the incidence of 12 cancers or tumors of uncertain malignancy among 6190 participants receiving the hypnotics and no cases among 2535 participants receiving placebo [
The findings of our study should be interpreted against the backdrop of the following limitations. First, our sample size was small, which might limit the widespread application of the findings. Furthermore, the probability of type 1 error may be increased. Using a more stringent alpha level of 0.01, we still found six cancer biomarkers that were significantly increased after lorazepam treatment in the overweight patients. Second, the participants also received antidepressant treatment, which may confound our findings. However, the multivariate general linear model confirmed that the use of antidepressants did not predict the changes in cancer biomarkers. Third, the study duration was short; therefore, whether the changes in cancer biomarkers are transient or gradually increased remains to be investigated. Fourth, changes in biomarkers need to be interpreted in clinical context, as cancer biomarkers can also change in other conditions. In addition, increasing biomarkers is not necessarily evidence of carcinogenesis. However, cancer biomarkers are quantifiable traits that may be helpful in early detection, diagnosis, and treatment. Finally, this study lacks a randomized control group, so the effects of time-in-hospital cannot be excluded. Therefore, further studies with prospective randomization to a control group are encouraged.
Evidence from basic research cannot be directly generalized to humans. To date, the link between benzodiazepines and carcinogenesis is suggestive but not conclusive. In our study, lorazepam is assumed work in concert with adiposity, triggering the carcinogenesis as evidenced by changes in cancer biomarkers. Epidemiological studies addressing this issue should consider the potential modulator contributing to benzodiazepine-associated carcinogenesis. Further research with a prospective follow-up design to explore the long-term effects of benzodiazepines on carcinogenesis is strongly encouraged.
This study was provided by the Civilian Administration Division of Beitou Branch, Tri-Service General Hospital, National Defense Medical Center (TSGH-BT-105-001).
Distinct pattern of caner biomarker changes between the overweight and the normal-weight patients. After six weeks of treatment with lorazepam, the nine cancer biomarkers were significantly increased in the overweight patients but not in the normal-weight patients. sFASL: soluble Fas ligand, HB-EGF: heparin-binding EGF-like growth factor, ANG-2: angiopoietin-2, VEGF: vascular endothelial growth factor, IL: interleukin, PLGF: placental growth factor, TGF: transforming growth factor, TNF: tumor necrosis factor.
Clinical characteristics in overweight and normal-weight patients
Total (N=19) | Overweight (N=9) | Normal-weight (N=10) | t | p |
|
---|---|---|---|---|---|
Age (y) | 26.1±3.4 | 26.9±3.3 | 25.3±3.5 | 1.021 | 0.322 |
Education (y) | 15.6±1.6 | 15.4±1.5 | 15.7±1.8 | 0.337 | 0.74 |
Height (cm) | 170.7±5.7 | 170.9±6.1 | 170.5±5.8 | 0.155 | 0.879 |
Weight (kg) | 70.7±12.8 | 81±9 | 61.5±7.7 | 5.09 | <0.001 |
BMI (kg/m2) | 24.2±3.8 | 27.7±2 | 21.1±1.7 | 7.729 | <0.001 |
differences in characteristics between groups were analyzed with independent-samples t-tests.
BMI: body mass index
Baseline cancer markers in overweight and normal-weight patients
Total (N=19) | Overweight (N=9) | Normal-weight (N=10) | F | p |
|
---|---|---|---|---|---|
ANG-2 | 583±177.8 | 485±149.2 | 671.2±158.7 | 0.019 | 0.894 |
sCD40L | 84.1±15.5 | 86.6±18.9 | 81.9±12.3 | 1.807 | 0.204 |
EGF | 24.9±5.5 | 25.9±6.6 | 23.9±4.4 | 0.855 | 0.373 |
Endoglin | 647.9±334.1 | 761.8±451.1 | 545.4±133.1 | 0.045 | 0.835 |
sFASL | 122.9±20.4 | 120.4±24.7 | 125.1±16.8 | 0.348 | 0.566 |
HB-EGF | 97.4±15.5 | 97±19.8 | 97.8±11.7 | 0.566 | 0.466 |
IGFBP | 1237±305 | 1227±354 | 1247±274 | 0.505 | 0.491 |
IL-6 | 23.4±4 | 23.6±4.5 | 23.1±3.7 | 0.056 | 0.817 |
IL-8 | 5.9±1.1 | 5.9±1.1 | 6±1 | 0.057 | 0.815 |
IL-18 | 57.2±25.3 | 66.4±30.9 | 49±16.6 | 0.016 | 0.903 |
PAI | 4046±1839 | 4749±2437 | 3413±689 | 0.078 | 0.785 |
PLGF | 26.2±7 | 26.4±7.4 | 26±7 | 0.337 | 0.572 |
TGF-α | 30±5.9 | 29.8±7.1 | 30.2±5 | 0.245 | 0.629 |
TNF-α | 15.5±3.3 | 15.3±3.6 | 15.8±3.2 | 0.002 | 0.967 |
uPA | 461.5±202.8 | 425.7±178.3 | 493.7±227 | 0.065 | 0.803 |
VEGF-A | 143.5±28.6 | 144.9±32.4 | 142.4±26.5 | 0.025 | 0.877 |
VEGF-C | 448.9±85.1 | 446.5±91.9 | 450.9±83.4 | 0.085 | 0.776 |
VEGF-D | 273.9±53.9 | 278.3±74.4 | 269.9±29.3 | 0.114 | 0.741 |
group differences in cancer markers were analyzed with a multivariate general linear model with covariates of age, height, weight, educational levels, and the use of fluoxetine.
ANG-2: angiopoietin-2, sCD40L: soluble CD40 ligand, EGF: epidermal growth factor, sFASL: soluble Fas ligand, HB-EGF: heparin-binding EGF-like growth factor, IGFBP: insulin-like growth factor-binding protein, IL: interleukin, PAI: plasminogen activator inhibitor, PLGF: placental growth factor, TGF: transforming growth factor, TNF: tumor necrosis factor, uPA: urokinase-type plasminogen activator, VEGF: vascular endothelial growth factor
Changes in cancer markers in all patients after a 6-week lorazepam treatment (N=19)
Baseline | Week 6 | t | p |
|
---|---|---|---|---|
ANG-2 | 583±177.8 | 595.4±158.6 | 0.561 | 0.582 |
sCD40L | 84.1±15.5 | 90.9±16.5 | 1.872 | 0.077 |
EGF | 24.9±5.5 | 27.7±7.9 | 1.741 | 0.099 |
Endoglin | 647.9±334.1 | 599.2±251.8 | 0.563 | 0.58 |
sFASL | 122.9±20.4 | 131.2±23.7 | 2.301 | 0.034 |
HB-EGF | 97.4±15.5 | 104.1±17.7 | 2.103 | 0.05 |
IGFBP | 1237±305 | 1379±429 | 1.869 | 0.078 |
IL-6 | 23.4±4 | 24.5±4.2 | 1.439 | 0.167 |
IL-8 | 5.9±1.1 | 6.5±1.3 | 2.766 | 0.013 |
IL-18 | 57.2±25.3 | 57.6±28.2 | 0.073 | 0.942 |
PAI | 4046±1839 | 3718±1124 | 0.634 | 0.534 |
PLGF | 26.2±7 | 28.2±7.4 | 1.922 | 0.071 |
TGF-α | 30±5.9 | 32.6±6.6 | 2.154 | 0.045 |
TNF-α | 15.5±3.3 | 16.5±3 | 1.625 | 0.122 |
uPA | 461.5±202.8 | 560.3±256.9 | 3.83 | 0.001 |
VEGF-A | 143.5±28.6 | 153.7±26.1 | 2.394 | 0.028 |
VEGF-C | 448.9±85.1 | 476.8±73.3 | 1.85 | 0.081 |
VEGF-D | 273.9±53.9 | 287.8±60.9 | 2.032 | 0.057 |
differences in cancer markers after six weeks of lorazepam treatment were analyzed with paired-samples t-tests.
ANG-2: angiopoietin-2, sCD40L: soluble CD40 ligand, EGF: epidermal growth factor, sFASL: soluble Fas ligand, HB-EGF: heparin-binding EGF-like growth factor, IGFBP: insulin-like growth factor-binding protein, IL: interleukin, PAI: plasminogen activator inhibitor, PLGF: placental growth factor, TGF: transforming growth factor, TNF: tumor necrosis factor, uPA: urokinase-type plasminogen activator, VEGF: vascular endothelial growth factor
Within-group changes in cancer markers in overweight and normal-weight patients after a 6-week lorazepam treatment
Baseline | Overweight (N=9) |
p |
Baseline | Normal-weight (N=10) |
p |
|||
---|---|---|---|---|---|---|---|---|
Week 6 | t | Week 6 | t | |||||
ANG-2 | 485±149.2 | 569.1±150.4 | 4.129 | 0.003 | 671.2±158.7 | 619.1±170 | 2.232 | 0.052 |
sCD40L | 86.6±18.9 | 94.8±13.2 | 1.363 | 0.210 | 81.9±12.3 | 87.5±19 | 1.213 | 0.256 |
EGF | 25.9±6.6 | 28.2±4.7 | 1.140 | 0.287 | 23.9±4.4 | 27.1±10.3 | 1.27 | 0.236 |
Endoglin | 761.8±451.1 | 578.3±226.3 | 1.134 | 0.290 | 545.4±133.1 | 618.1±283.5 | 1.157 | 0.277 |
sFASL | 120.4±24.7 | 137±24.2 | 3.823 | 0.005 | 125.1±16.8 | 125.9±23.1 | 0.187 | 0.856 |
HB-EGF | 97±19.8 | 110.4±14.1 | 3.445 | 0.009 | 97.8±11.7 | 98.3±19.3 | 0.135 | 0.896 |
IGFBP | 1227±354 | 1345±406 | 1.276 | 0.238 | 1247±274 | 1410±468 | 1.337 | 0.214 |
IL-6 | 23.6±4.5 | 26±3.4 | 2.235 | 0.056 | 23.1±3.7 | 23.2±4.5 | 0.098 | 0.924 |
IL-8 | 5.9±1.1 | 6.9±1.2 | 3.750 | 0.006 | 6±1 | 6.1±1.2 | 0.678 | 0.515 |
IL-18 | 66.4±30.9 | 63.3±37.9 | 0.315 | 0.761 | 49±16.6 | 52.4±15.8 | 0.621 | 0.55 |
PAI | 4749±2437 | 3692±1389 | 1.071 | 0.315 | 3413±689 | 3742±899 | 0.908 | 0.388 |
PLGF | 26.4±7.4 | 31.5±5.6 | 3.471 | 0.008 | 26±7 | 25.3±7.9 | 0.634 | 0.542 |
TGF-α | 29.8±7.1 | 35±5.4 | 3.804 | 0.005 | 30.2±5 | 30.4±7 | 0.09 | 0.93 |
TNF-α | 15.3±3.6 | 17.6±2.1 | 2.544 | 0.035 | 15.8±3.2 | 15.6±3.5 | 0.223 | 0.828 |
uPA | 425.7±178.3 | 535.6±202.2 | 2.678 | 0.028 | 493.7±227 | 582.4±307.5 | 2.61 | 0.028 |
VEGF-A | 144.9±32.4 | 160.7±23.3 | 2.320 | 0.049 | 142.4±26.5 | 147.5±28 | 1.011 | 0.338 |
VEGF-C | 446.5±91.9 | 509.1±42.8 | 2.806 | 0.023 | 450.9±83.4 | 447.8±84.6 | 0.198 | 0.848 |
VEGF-D | 278.3±74.4 | 309.7±77.2 | 3.275 | 0.011 | 269.9±29.3 | 268.1±34.9 | 0.261 | 0.8 |
differences in cancer markers after six weeks of lorazepam treatment were analyzed with paired-samples t-tests.
ANG-2: angiopoietin-2, sCD40L: soluble CD40 ligand, EGF: epidermal growth factor, sFASL: soluble Fas ligand, HB-EGF: heparin-binding EGF-like growth factor, IGFBP: insulin-like growth factor-binding protein, IL: interleukin, PAI: plasminogen activator inhibitor, PLGF: placental growth factor, TGF: transforming growth factor, TNF: tumor necrosis factor, uPA: urokinase-type plasminogen activator, VEGF: vascular endothelial growth factor