Cancer risk with folic acid supplements( 31.12.2011 )
3-page executive summary
Folic acid supplements have been considered as safe. A combined analysis from 2009 of two Norwegian randomized controlled clinical trials, with extended post-trial follow-up, demonstrated however an increased incidence of cancer among patients taking folic acid for homocysteine reduction as secondary prevention of cardiovascular events. In Norway folic acid is among the 10 most sold non-prescription drugs with 17.5 defined daily doses per 1000 inhabitants/day. Norwegian guidelines have since 1998 recommended supplements of folic acid 0.4 mg daily to women periconceptionaly in order to reduce the risk of neural tube defects.
The present systematic review explores whether there is an increased cancer risk associated with folic acid supplements given orally.
We have conducted a systematic review where we included randomized controlled trials (RCTs) and observational studies that compared groups taking folic acid supplements (≥0,4 mg) with control groups. The outcomes were cancer incidence, cancer mortality and/or precancers.
We performed a systematic liteerature search in May 2010 in the following databases: EMBASE (Ovid), Ovid MEDLINE, Cohrane Library, Centre for Reviews and Dissemination, INAHTA (International Network of Agencies for Health Technology Assessments), Clinical Evidence, ISI Web of Knowledge, NHS Evidence, AHRQ (Agency for Healthcare Research and Quality), SBU (Swedish Council on Health Technology Assessment), Dacehta (Danish Centre of Health Technology Assessment) and Finohta (Finish Office for Health Technology Assessment). For ongoing studies we searched in April 2011 in WHO’s International Clinical Trials Registry Platform (ICTRP) and ClinicalTrials.gov. An update search was conducted in November 2011. Two persons independently reviewed the results from the search. The selected full text articles that satisfied the inclusion criteria
were evaluated for quality of two independent persons. Three independent persons extracted data from the articles according to our predefined outcomes. Any disagreements between the independent assessments of the reviewers in literature selection, data extraction and quality assessment were dissolved by discussion or by a fourth reviewer. We used Intention to treat (ITT) data. When feasible, we pooled data by meta-analyses with Cochrane Collaboration software (RevMan5) and used random-effects model calculating relative risks (RR) with 95% confidence intervals (CIs). Sensitivity analyses according to pre-defined factors were conducted where e possible. The quality of the evidence for the individual endpoints was evaluated by use of GRADE (Grading of Recommendations, Assesment, Development, and Evaluation).
The data came from 10 RCTs from Canada, Europe and the USA with follow-up time from 20 to 88 months. The studies included a total of
38, 233 persons (65 % men) with average age of about 60 years, and where 93 % of the total population had cardiovascular disease or high risk for this (the others had history of colorectal adenoma). Meta-analyses and GRADE of the documentation showed:
- A borderline significant increase in cancer incidence in the group that received folic acid supplement compared to the placebo group: Relative risk of 1.07 (95 % CI 1.00-1.14). Seven of the 10 RCTs had a follow-up time ≥5 years (60-88 months), and meta-analysis of these also showed a borderline significant increased risk for cancer incidence in the group that got folic acid supplement as compared to the placebo group (RR of 1.09; 95 % CI 1.00 - 1.18). We evaluated the quality of the documentation to be high, which means that we trust the results. Three of the 10 RCTs have a follow-up time <5 years(20-38 months), and a meta-analysis of these showed no difference between the groups. We evaluated the quality of the documentation to be moderate, that means our confidence to the results is moderate.
Incidence of total cancer was not studied in the observational studies.
Incidence of different cancer types
The data came from nine RCTs and seven observational studies. The studies were done in Canada, Europe, China and the USA and included breast cancer, pancreatic cancer, colorectal cancer, hematological cancer, lung cancer and prostate cancer. The RCTs had a mean follow-up time from 36 to 88 months and included a total of 32,595 personer (63 % men) with a mean age of about 60 years. Smokers made up more than 30 % in one of the studies, otherwise from 7-15 %. Meta analyses and GRADE of the data showed:
- A significant increased risk for prostate cancer in the group that received folic acid supplements compared to the placebo group (RR of 1.24, 95 % CI 1.03-1.49). None of the other meta-analyses for incidence of different cancer types, whether from RCTs or observational studies, showed statistically significant differences between the groups. We evaluated the quality of the documentation to be high, which means that we trust the results.
Data for cancer mortality came from six RCTs with a mean follow-up time from 60-432 months, which included a total of 32,327 persons (59 % men). The mean age was 63 and 62 years respectively in the folic acid group and the control group for all the studies except for the study with pregnant women with a mean age of 26 years in both groups. In the intervention groups the percentage of smokers ranged from 11to 45 %, and in the control groups from 12to 46%. The studies were performed in Canada, Europe and the USA. Meta analyses and GRADE of the data showed:
- No statistically significant increase in cancer mortality in the folic acid groups as compared to the control groups (RR = 1,09; 95 % CI 0,92-1,0). We evaluated the quality of the documentation to be moderate.
- Mortality from different cancer types (breast-, colorectal--, hematological-, lung- and prostate cancer) showed no significant differences between the groups. We evaluated the quality of the documentation to be low for breast cancer and moderate for the others. That means that we have little confidence in theresults regarding breast cancer.
Cancer mortality was not studied in the observational studies.
Incidence of precancers
Data came from 12 RCTs and two cohort studies.
Meta analyses and GRADE of the data showed:
- Any colorectal adenoma : No significant difference between the groups (RR=0.97;95 % CI 0.83-1.14). We evaluated the quality of the documentation from the six RCTs to be moderate. The results from the observational study had very low quality so for this we could not draw conclusions..
- Advanced colorectal damage, without cancer: No significant difference between the groups. The quality of the documentation was low, so we have low confidence to the results.
- Precancer to lung- cervix-aesophagus and gastric cancer. Uncertain, impossible to conclude.
Our thorough systematic search makes it plausible that we have retrieved all relevant studies andit is reasonable to conclude that we have not overlooked any important data. The majority of the RCTs had high quality and a low risk of bias. This means generally that we can have confidence in that the studies were adequately randomized and that we can trust the results.
A limitation in our review is the relative short follow-up time (from 20 to 432 months), because the time to develop cancer can exceed the follow-up time.
A major limitation in our review is the highly selected population, mainly men above 60 years with cardiovascular disease.
Folic acid supplement compared to control gave a borderline increased incidence of cancer and significantly increased incidence of prostate cancer. The data for cancer incidence were of high quality for studies with a follow-up time ≥5 years and of moderat quality for studies with a follow-up time<5 years. The data for prostste cancer had high quality. A major limitation in our review is the highly specific population, mainly men above 60 years with cardiovascular disease. This review found insufficient documentation to conclude about cancer risk for fertile women that are recommended folic acid periconceptionally in order to reduce the risk of neural tube defects.
Need for further research
There is a need to explore possible factors that can infuence the effect of folic acid supplement such as age, smoking, dose, organ specificity as well as studies with longer follow-up times.
Studies in this area involving pregnant women should be epidemiological studies; RCTs are not usable due to the protective effect of folic acid for neural tube defects.
1-page key messages
>3-page executive summary