ASPIRIN and CANCER
Abstract: Aspirin taken regularly can reduce the risk of most types of cancer by up to 40%. But unlike its benefit for heart disease it probably produces benefits via a modest slowing of the growth of cancers. Thus it must be taken for at least 7 to 10 years for much benefit. Further, unlike the 75 - 81 mg amount needed for heart disease, a daily pill of 325 mg may be needed for its best benefit in reducing cancer risk. Also, the benefits of Aspirin appear less for breast cancer and other female organ cancers than for cancer at other sites. The actual data from 36 research comparisons included in 28 key studies located on aspirin and cancer are shown following. Aspirin sometimes causes internal bleeding and ulcers.
The much publicized benefit of aspirin for more than three decades has been its role in reducing risk of a heart attack. Research has now verified with more than adequate significance that aspirin also can reduce the risk of many and perhaps most types of cancer. Results of 36 research comparisons from tens of thousands of cases included in 28 studies of this found are shown in Table A following. These results that probably comprise the major research results on aspirin and cancer published to date develop an average risk factor of 0.73 for use of aspirin in reducing risks of all cancer types studied. This average may have a very narrow 5-95% margin of error of only about 0.68-0.78.
The role of aspirin in reducing cancer risk differs much from its role in reducing coronary disease. First as expected from the general biochemical development of cancer, any agent that reduces cancer risk probably will require a substantial duration of use for useful benefit. As found for cigarette smoking, best benefits for aspirin on cancer are obtained only after 20 years of use. Second, the amount of aspirin needed appears to be at least a full usual 325 mg pill each day. In contrast, only 75 -81 mg/day provided near full benefits for heart disease, and benefits for the heart develop near immediately with its use via reduction in blood clotting.
Best insight about the role of aspirin is developed in Giovannucci’s Study #5 in Table A following that verified the key effect of duration of aspirin use on benefit. The coefficient in a formula developed from this data shows that aspirin slowed development of colorectal cancer by a modest but highly significant 2.5% per year. This probably was a result of using four or more 325 mg tablets per week. Use of such a tablet daily should slow colorectal cancer by perhaps 3% per year. It is further possible as is suggested from Study #5 that a first 3 years of aspirin use will be of no benefit to cancer because of the time needed for new cancer precursors to develop.
Unfortunately and despite the striking finding of Giovannucci, no information about duration of aspirin use was included in most of the other studies of aspirin and cancer. The use of aspirin for preventing heart disease has been promoted since the late 1970’s, and many people have been taking a pill regularly for many years. Thus those taking aspirin in case control studies probably had been taking it for an average of 7 or more years. And those taking aspirin in prospective studies were analyzed for a usual longer duration of use. Thus durations in these studies were sufficient to obtain measurable benefits. A randomized clinical study of less than 10-15 years (5 -7 yr mean duration) probably would not be able to usefully measure a benefit of aspirin on cancer.
The benefits of aspirin to reduce risk of colorectal cancer are very consistent with the exception of those from Study #6 that found no benefit. The combination of low level of aspirin use and mean duration of 3 years would forecast a risk of only 0.96 for the clinical portion of this study, and this is statistically consistent with the 1.03 value found. The 1.07 value from the follow-up is not explained, but this value has a quite high error margin. This study was too short term to be able to identify a useful effect - as 20+ years were needed for full effect in Study #6. But Study #10 did find a definite change in progression of colorectal adenomas in patients that already had developed colorectal cancer.
The value of aspirin in reducing risk of breast cancer is less well established than that for colon cancer. Although the result of the largest study, that of the Nurses, was negative, all other studies showed a benefit in risk averaging 25%. Some factor clearly is missing here. Results for ovarian cancer similarly show only a very modest benefit if any, with an average risk factor of 0.9. The Nurse’s study found no effect of aspirin on this disease even to 20 years duration of use. A possible reason for this is that if benefits for aspirin were obtained via antioxidants this healthier than average group may have already been taking sufficient antioxidants, and the further amounts from aspirin were not effective. Also the Nurses averaged much younger than women in other research, and many were pre-menopausal with an antioxidant environment of higher estrogen. The Nurses group failed in other studies to obtain cancer benefit from antioxidants that was found by most other studies..
Aspirin reduced risk of other cancer types listed substantially and consistently. The present model used in Life Ahead that is based on the colon cancer data assumes that aspirin will reduce risk at 2.5% per year of use (factor of 0.975) of cancers other than those for female organs and the breast. This is for a use of 325 mg per day, and benefit computed will be reduced proportionately for use of lower amounts of aspirin. A single 325 mg daily tablet is highest amount accepted as useful, and the lowest risk factor accepted is 0.60. This method produces a conservative interpretation of the results in Table A. But some conservatism seems appropriate in displaying risk and benefit because improvements from aspirin may partly duplicate those from calcium and some other sources.
Aspirin does have side effects in promoting ulcers, gastritis and bleeding. A few can experience this even at the low 75 -81 mg dose. No quantitative basis was found for identifying how often this problem is obtained at the low doses suggested for disease prevention. Thus individuals must consult their doctors in event of any such problem as to desirability of considering alternates to aspirin. .
It seems evident that any health interested person should be taking aspirin regularly as a means to protect health from both cardiovascular diseases AND cancer provided this does not cause other serious problems. But again, the results on aspirin for cancer suggest that the daily 75 -81 mg or baby aspirin promoted for heart disease should be increased if possible to a usual daily 325 mg tablet for best overall protection. And that useful benefit on cancer will not be obtained until aspirin has been taken for 5-7 years or more.
Life Ahead shows that an average US man of age 50 that started taking 325 mg/day of aspirin each day at age 40 will gain about 1000 Well-Days or 2.8 years of healthy life. This assumes that aspirin is the only health measure taken. The contribution of Aspirin as a part of a combination of the other health improvements of diet and exercise usually indicated desirable by Life Ahead should be much smaller than this 2.8 years estimate and can be approximately identified in Life Ahead for any combination of user age, sex, habits and habit modifications.
|
|
TABLE A |
|||||||
|
|
ASPIRIN and CANCER |
|||||||
|
No |
Study |
Scope |
Sex |
Risk Ratio |
Error Margin |
Amt Diff |
Notes |
|
|
Colorectal Cancer |
||||||||
|
1 |
Thun MJ, N Engl J Med 1991, 325:1593 |
598 cases, 3058 controls of 662,000 |
M W |
0.60 0.58 |
0.40-0.89 0.37-0.90 |
|
Colon Cancer. Aspirin used more than 16 times per month |
|
|
2 |
Gann PH, J Natl Cancer Inst 1993, 85:1220 |
Part of 22,100 US physicians Randomized clinical trial |
M |
0.86 |
0.68-1.10 |
|
Combination of Cancer + Polyps. Risk Ratio for cancer alone was 1.15 Duration 5 years (avg exposure 2.5 only about 2.5 yrs), too short for full effect |
|
|
3 |
Giovannucci E, Ann Intern Med 1994, 121:241 |
251 Cases of 47,900 male health professionals |
M |
0.68 |
0.52-0.92 |
325 Est |
>=2 pills/week. 6 + years prospective |
|
|
4 |
Schreinemachers, DM, Epidemiology 1994, 5:138 |
1257 total cancer on 12,700, NHANES I |
M |
0.35 |
0.17-0.73 |
|
In younger men |
|
|
5 |
Giovannucci E, N Engl J Med 1995, 333:609 |
331 cases on 89,000 Nurses |
W |
1.06 0.84 0.70 0.56 |
0.78-1.45 0.55-1.28 0.41-1.20 0.36-0.90 |
4+ wk of 325 mg |
1-4 (2.5) years 5-9 (7) years 10-19 (14.5) years 20+ (25) years Key study showing effect of duration |
|
|
6 |
Sturmer T, Ann Intern Med 1998, 128:713 |
341 cases of 22,000 male physicians US |
M |
1.03 1.07
|
0.83-1.08 0.75-1.53 |
Avg 160 mg/day |
6 year randomized clinical study (3 yrs duration) 6 added years prospective study |
|
|
7 |
Sandler RS, Gastroenterology 1998, 11:441 |
210 cases, 199 controls |
M&W |
0.56 |
0.34-0.92 |
|
Adenoma formation (not cancer)
|
|
|
8 |
Garcia-Rodriguez LA, Epidemiology 2001, 12:88 |
2002 cases on 942,000 in Spain |
M&W |
0.60 |
0.4-0.9 |
300 mg/d |
No effect found for 75 and 150 mg/d of aspirin |
|
|
9 |
Bosetti, C, Eur J Cancer Prev 2002, 11:535 |
Review of Meta Studies |
M&W |
0.71 0.84 |
0.66-0.77 0.72-0.98 |
|
From Case Control Studies From Prospective Studies |
|
| 10 | Sandler, RS N Engl J Med 2003, 348:883 | 635 patients vs. 635 placebo, Randomised Study | M&W | 0.61 | 0.46-0.91 | 325 mg/d | Adenomas Only, but on patients with previous colorectal cancer. Clinical study, 1 year after start | |
|
|
|
|
|
|
|
|
|
|
|
Breast Cancer |
|
|
|
|
|
|||
|
11 |
Schreinemachers, DM, Epidemiology 1994, 5:138 |
1257 total cancer on 12,700, NHANES I |
W |
0.70 |
0.50-0.96 |
325 Est |
Followed 12.4 years |
|
|
12 |
Egan KM, J Natl Cancer Inst 1996, 88:988 |
2414 cases from 89,500 Nurses |
W |
1.03 1.05 |
0.95-1.12
|
325 mg |
2+ tablets per week, 12 years Heavy use, 2+ tablets per day |
|
|
13 |
Bosetti, C, Eur J Cancer Prev 2002, 11:535 |
Review of Meta Studies |
W |
0.70 0.79 |
0.61-0.81 0.59-1.06 |
|
Case Control Studies Prospective Studies |
|
|
14 |
Johnson, TW, Cancer Epidemiol Biomarkers Prev 2002, 11:1586 |
938 cases on 27,600 women, 6 years |
W |
0.80 0.71 |
0.67-0.95 0.58-0.87 |
|
All users of aspirin Users of 6 or more times/week |
|
|
|
|
|
|
|
|
|
|
|
|
Lung Cancer |
|
|
|
|
|
|||
|
15 |
Schreinemachers, DM, Epidemiology 1994, 5:138 |
1257 total cancer on 12,700, NHANES I |
M&W |
0.68 |
0.48-0.94 |
|
|
|
|
16 |
Moysich KB, BMC Cancer 2002, 2:31 |
868 cases vs 938 controls |
M&W |
0.57 |
0.41-0.78 |
|
Low cigarette use benefits most Same effect M & W |
|
|
17 |
Akhmedkhano,A, Br J Cancer 2002 Jul 1;87(1):49 |
88 Cases |
W |
0.66 |
0.84-1.08 |
|
|
|
|
|
|
|
|
|
||||
|
Ovarian Cancer |
|
|
|
|
||||
|
18 |
Cramer DW, Lancet 1998, 351:104 |
563 cases, 523 controls |
W |
0.75 |
0.52-1.01 |
|
|
|
|
19 |
Tavani A, Ann Oncol 2000, 11:1171 |
748 cases, 898 controls, Italy |
W |
0.93 |
0.53-1.62 |
|
Regular use for more than 6 months, High error margin |
|
|
20 |
Moysich KB, Cancer Epidemiol Biomarkers Prev 2001, 10:903 |
547 cases. 1094 controls |
W |
1.00 |
0.73-1.09 |
|
|
|
|
21 |
Bosetti, C, Eur J Cancer Prev 2002, 11:535 |
Review of Meta Studies |
W |
0.82 |
0.69-0.99 |
|
|
|
|
22 |
Fairfield, FM, Cancer Causes Control 2002, 13:535 |
333 cases on 76,800 Nurses |
W |
0.98 |
0.63-1.52 |
|
Duration up to 20 years also not associated with risk, risk ratio=0.99 |
|
|
|
|
|
|
|
|
|
|
|
|
Pancreatic Cancer |
|
|
|
|
|
|||
|
23 |
Anderson, KE, J Natl Cancer Inst 2002, 94:1168 |
28,300 Iowa Women, 7 years followup |
W |
0.57 |
0.36-0.90 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Esophageal cancer |
|
|
|
|
||||
|
24 |
Funkhouser EM, Cancer 1995, 76:1116 |
14,000 US NHANES I, 12-16 years |
M&W |
0.1 |
0.01-0.76 |
|
|
|
|
25 |
Corley, DA, Gastro-enterology 2003, 124:47 |
Review of Meta Analyses, 9 studies 1813 cancer cases |
M&W |
0.50 |
0.36-0.66 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Prostate Cancer |
|
|
|
|
|
|||
|
26 |
Habel, LA, Cancer Causes Control 2002, 13:427 |
2574 cases of 90,100 |
M |
0.76 |
0.60-0.98 |
|
More than 6 aspirins per week |
|
|
|
|
|
|
|
|
|
|
|
|
Gastric Cancer |
|
|
|
|
|
|||
|
27 |
Zaridze D, Int J Cancer 1999, 82:473 |
448 cases and 668 controls, Russia |
M&W |
0.49 |
0.31-0.77 |
|
Same risk for men and women |
|
|
28 |
Akre K, Br J Cancer 2001, 84:965 |
567 cases, 1165 controls in Sweden |
M& W |
0.7 |
0.60-1.00 |
|
Risk reduced with increasing frequency of aspirin use |
|
|
|
|
|
|
|
|
|
|
|