Prevalence, Incidence and Risk Factors of Tamoxifen-related Non-alcoholic Fatty Liver Disease

A Systematic Review and Meta-analysis

Bora Lee; Eun-Ae Jung; Jeong-Ju Yoo; Sang Gyune Kim; Cheon-Beom Lee; Young Seok Kim; Soung Won Jeong; Jae Young Jang; Sae Hwan Lee; Hong Soo Kim; Baek Gyu Jun; Young Don Kim; Gab Jin Cheon


Liver International. 2020;40(6):1344-1355. 

In This Article


This meta-analysis of 24 published documents demonstrated that the prevalence of tamoxifen-induced fatty liver was 40.25 per 100 patients and the incidence rate was 12.37 per 100 person-years. Moreover, tamoxifen increased the incidence of fatty liver by 3.12 times compared to control patients. Additionally, our results showed that patients with high BMIs and hypercholesterolaemia were more likely to develop tamoxifen-induced fatty livers.

Fatty liver is currently one of the main causes of cirrhosis and hepatocellular carcinoma, ranking second among the indications for liver transplantation.[34] Given the recent rapid increase in fatty liver worldwide, fatty liver might become the leading cause of various end-stage liver disease. Nevertheless, there is no current effective treatment. Thus, the prevention of fatty liver development is important. It is well known that tamoxifen causes fatty liver. However, little has been studied about its mechanism and specific characteristics. Considering the frequency of fatty liver due to tamoxifen, this topic has been almost neglected in the field of breast cancer treatment. An important contribution of this study is that it can alert physicians to the high frequency of fatty liver from the use of tamoxifen.

Reports of fatty liver incidence caused by tamoxifen vary between studies. The largest study was done by Bruno[4] in patients undergoing hysterectomy taking prophylactic tamoxifen. Since the study was not conducted on breast cancer patients, it was excluded from our meta-analysis. In this study, the incidence of tamoxifen-induced fatty liver was estimated to be 40% at one year. Considering that the prevalence of fatty liver in the general population is reported to be about 6.3%-33%, its prevalence is quite high. In our study, the prevalence of tamoxifen-induced fatty liver was 40.25 per 100 patients, quite similar to the results of Bruno's study. The IRR was reported as 2.0 in Bruno's[4] study, but the IRR in our study was higher, at 3.12. This is probably because the control groups were heterogeneous in our meta-analysis.

There can be differences in the prevalence of non-alcoholic fatty liver disease according to ethnicity, region and age. However, in our study, tamoxifen increased fatty liver regardless of region, east or west. Zhan et al reported no significant differences between African American and Caucasian women reporting tamoxifen side effects.[35] However, Flaws et al reported that black women tended to have high tamoxifen risk/benefit because tamoxifen is readily converted to N-desmethyltamoxifen in black women compared to white women.[36] Another study showed that Indian women seemed to tolerate tamoxifen therapy better than western women.[24] In our study, although not significant, the incidence of fatty liver was higher in the western region (IRR 5.80) compared to the eastern region (IRR 3.05). There is probably a reason for this. We think that the degree of CYP2D6 inhibition might be related to the side effects of tamoxifen. It is known that the frequency of the CYP2D6 allele varies among racial/ethnic groups.[37,38]

Nowadays, tamoxifen therapy has been extended for up to 10 years. At first, we hypothesized that the longer the duration of use, the higher the chance of developing a fatty liver. However, meta-regression analysis showed that tamoxifen duration and the incidence of fatty liver were not proportional. Therefore, we cautiously assume that the recent trend of the prolonged use of tamoxifen will not have any additional harmful effects on fatty liver development.

The risk factors identified in this study differed from previous reports. First, insulin resistance and diabetes have been reported to be associated with non-alcoholic fatty liver disease (NAFLD), but diabetes was not a risk factor for tamoxifen-induced fatty liver. In contrast, hyperlipidaemia was a risk factor in our study. Previous studies have reported that taking tamoxifen increased the hepatic fat content by increasing serum triglycerides.[39,40] Our study also supports these previous results. In addition, high BMI was analysed as a risk factor for tamoxifen-induced fatty liver. The increase in fatty liver incidence in people with high BMIs supports the previous second-hit hypothesis.[41] Regarding the BMI cut-off, although it differed among papers, there were reports of fatty liver protection at BMIs ≤22 kg/m2, with a significant increase in fatty livers at BMIs ≥25 kg/m2.[4,30] Therefore, clinical attention is needed for patients with high BMIs.

However, the incidence of fatty liver does not affect the long-term prognosis of breast cancer. Thus, it is better to pay more attention to fatty liver-related metabolic complications than discontinue the drug. Previous studies have shown that many patients do not take tamoxifen for fear of side effects.[42] Tamoxifen has been very effective in reducing breast cancer recurrence and lengthening survival, and therefore, has become increasingly widely used. Naturally, many clinicians should be concerned about the side effects of tamoxifen on fatty liver development. Furthermore, a treatment or solution should be formulated to increase patient drug compliance over the long term. A previous study suggested that exercising at least 150 minutes a week could reduce tamoxifen-induced fatty liver.[30] In animal experiments, Silybum marianum,[43] Rosa canina distilled water,[44] uridine[45] and tetradecylthioacetic acid[46] have been reported to be effective in treating tamoxifen-induced fatty liver, but none of these have not been tested in humans yet.

The strength of our study is that it was the first meta-study on the development of tamoxifen-induced fatty liver, and the results can alert physicians to the high frequency of fatty liver from tamoxifen. The limitations of our study are as follows. First, it did not include many high-quality randomized control studies. However, considering that the results were similar to a large-scaled RCT on the prophylactic use of tamoxifen by Bruno et al, we believe that the biases are relatively small and our results are valid. Second, the control group of each study was heterogeneous. Third, the diagnostic tools for fatty liver vary from study to study.

To overcome these limitations, three methods were utilized in this study. First, to overcome the heterogeneity of these papers by confounding factors, we added the results of a quality-effects model. Second, we assumed that age and duration of medication were important confounders, so we added meta-regression analysis by age and duration of medication. Third, we sub-analysed the results depending on the diagnostic tools.

In conclusion, tamoxifen significantly increased the incidence of fatty liver. High BMI was found to be a significant risk factor. Thus, patients with high BMI need active surveillance and prevention. More attention to this subject is needed. Well-designed prospective studies should be conducted in the future.