Aspirin’s potential in cancer prevention, treatment remains compelling yet complicated

Several recent studies have added to the evidence that aspirin may help prevent or treat a variety of cancers.

However, many questions remain, including the duration of use and ideal dosage required to reduce cancer incidence and/or mortality.

David H. Ilson, MD, PhD, a medical oncologist with Memorial Sloan-Kettering Cancer Center’s Gastrointestinal Oncology Service and a professor of medicine at Weill Cornell Medical College, led a round table discussion during the HemOnc Today section editors’ retreat in which participants reviewed current data, explored aspirin’s potential mechanisms of action and discussed optimal dosage.

The round table discussion follows.

Roundtable Participants

• 

• Moderator

• David H. Ilson, MD, PhD

• Memorial Sloan-Kettering Cancer Center’s Gastrointestinal Oncology Service and Weill Cornell Medical College

• 
• Joseph R. Bertino, MD

• Associate Medical Editor of HemOnc Today

• 
• A. Koneti Rao, MD
• Temple University School of Medicine

• 
• Barbara Burtness, MD
• Fox Chase Canter Center

• 
• Harry S. Jacob, MD, FRCPath(Hon)

• Chief Medical Editor of HemOnc Today

• 
• Ralph Green, MD

• UC Davis Medical Center

Ilson: The issue of using aspirin as a therapeutic is a rich and broad topic.

The issue of using aspirin as a therapeutic is a rich and broad topic.

To me, it’s always appealing to think that the drug started as a natural product from chewing tree bark to get pain relief.

We certainly have a much better understanding of the molecular biology of how these drugs work, the pathways of Cox and their involvement in inflammation, angiogenesis invasion and metastasis of tumors. Then there are the cardiovascular preventive aspects.

This is a story that has been developing over decades, and there is abundant literature.

The first study probably was the Nurses’ Health Study in the 1980s. That large data set suggested that regular aspirin use of more than 20 years was associated with up to a 50% reduction in the risk of colorectal cancer mortality. You only tended to see this benefit after long-term use.

Similar results were reported in the Physicians’ Health Study, which indicated that long-term aspirin use was associated with up to a 30% to 50% reduction in colorectal cancer mortality.

There also is the aspect of polyp prevention. Cox-2 inhibitors have shown the ability to reduce the number of polyps in high-risk patients, the familial adenomatous polyposis patients who develop thousands of polyps.

So it’s a complicated story. You not only have the cardiovascular preventive benefits, but you have the cancer story, which you have to separate into primary prevention and secondary prevention.

Primary prevention would be patients who don’t have any obvious risk factors, but if they’re taking aspirin as a preventive, theoretically you’re preventing some precancerous lesion from evolving into a cancer. This includes patients with Barrett’s esophagus, or patients with intestinal metaplasia in the stomach from Helicobacter pylori.

Secondary prevention involves patients who already had a treated cancer and take aspirin as a way to reduce recurrence risks.

There has been a lot of recent interesting literature, which include both successes and failures.

The identification that Cox-2 inhibition could potentially reduce polyp formation led to the study of more specific Cox-2 inhibitors, and those large-scale studies did achieve proof of principle.

High-dose celecoxib and high-dose rofecoxib did result in reduced polyp formation in some of the larger studies, but there were cardiovascular risks associated with the high-dose therapies, so several of those studies were closed prematurely. Yes, you had a 4% lower risk of developing polyps, but you had a doubling or tripling of the risk of heart attack, stroke and vascular events. These were patients who were being screened for polyps already, so why subject them to a potentially harmful treatment when they’re going to be screened and followed anyway?

A paper in The Lancet that came out at the end of last year included fascinating data about Lynch syndrome patients. In this multinational study, patients were randomly assigned to placebo vs. 300 mg aspirin twice daily.

There were two endpoints. One was to see if you could reduce polyp formation in the Lynch families. The other was whether it had impact on cancer development.

The first endpoint failed to be accomplished. Aspirin therapy had no impact on polyp formation. They actually were going to close the study because the primary endpoint was not reached, but they continued to follow the patients and they found that 7, 8 or 9 years later, they started to see up to a 30% reduction in Lynch-related cancer with aspirin therapy.

That’s now translated into a treatment recommendation. If someone carries one of the DNA mismatch repair genes, they should take aspirin daily. There is a significant reduction in the risk of colorectal cancer development with aspirin therapy, but you don’t see the benefits early.

The same group is now planning a dose-finding study to see whether low-dose aspirin is comparable to this higher 300-mg [twice daily] dosing.

Other interesting data came from the analysis of the CALGB irinotecan adjuvant study. This trial was a comparison of FOLFIRI vs. 5-FU, but several other secondary analyses looked at outcomes in patients with various lifestyles.

One was the McDonald’s diet vs. the healthy diet. If you ingest the McDonald’s diet, you had a much higher risk of developing recurrence compared with a healthy diet. If you had regular exercise as part of your regimen, you had a reduced risk of recurrence, and if you were an aspirin user after treatment, you had a significant reduction in the risk of recurrence of colon cancer.

That’s where we get into the issue of secondary prevention. It has become a standard recommendation in our armamentarium: Individuals who have curatively treated colorectal cancer should probably take at least a low-dose aspirin daily.

So it is a rich, complicated and multifaceted area. The data seem to be the strongest and most compelling in colorectal cancer, but the issues of primary prevention and secondary prevention certainly extend to several other cancers.

Mechanism of action

Joseph R. Bertino, MD: So what’s the mechanism?

Ilson: We think it’s Cox-2. I’m not an expert in this whole pathway, but there’s a lot of active research looking at how these drugs have an impact. There were studies that tried to combine Cox-2 inhibitors with chemotherapy to see whether we could get enhanced effects. That turned out to be a failed strategy. The preclinical data suggested that with Cox-2 inhibitors, you got synergy with chemotherapy and radiation, but to date, some of the limited studies have not shown an added benefit.

A. Koneti Rao, MD: Aspirin is a remarkable drug. Armand J. Quick, MD, PhD, who developed the prothrombin time, wrote in one of his monographs that aspirin is God’s gift to man for pain. If you look at what aspirin costs and what it does, we still don’t know its full efficacy range. This is probably going to keep evolving. If you compare its effectiveness to cost, it’s a drug that no other drug would be able to beat easily.

Having said that, mechanism-wise, the strongest evidence, indeed, is in colorectal cancers. It may be the effect on Cox-2, but studies suggest there is an effect in other tumors as well, and it might not be Cox-2 alone. In fact, it may be a combination of effect on Cox-1 and Cox-2, and other effects that we don’t know as yet.

We know Cox-1 in platelets is inhibited effectively by aspirin, and there’s evidence going back 40 years that platelets play a role in tumor metastasis.

More recently, studies by Stone and colleagues published in The New England Journal of Medicine looked at thrombocytosis in association with ovarian cancer. Through an elegant set of studies that included mouse models and human studies, they showed higher platelet counts are associated with worse prognosis.

They went on to show that interleukin-6 (IL-6) is produced by tumors and stimulates platelet production. In mouse models, they showed for ovarian tumors that if you added an agent to block IL-6 and prevent an increase in platelet count, you get an effect on tumor models. Aspirin has a platelet inhibitory, and the larger effect on platelets probably comes from Cox-1 inhibition rather than of Cox-2. So mechanisms are not known, and each tumor might behave differently.

Barbara Burtness, MD: So you believe aspirin is not only the cheapest and safest, but potentially the most effective relative to the non-steroidals and the Cox-2 inhibitors?

Rao: We don’t know. My point was that no matter what new agent we come up with, if you stack it against aspirin, taking cost into account, it’s going to be tough.

Ilson: The experience with the high-dose celecoxib and causing strokes and heart attacks diminished interest in using that class of drugs as a potential primary and secondary preventive agent. Aspirin has the added benefit of cardiovascular protection.

The guidelines now are, if you have a cardiovascular risk, you should probably be taking a baby aspirin. In the colon cancer realm, if you either have a history of colon polyps or treated colon cancer, you should take a baby aspirin. But it shouldn’t be a global recommendation for everyone.

Ideal dosage

Harry S. Jacob, MD, FRCPath(Hon): We’re comfortable that once-a-day aspirin will keep your platelets irreversibly blocked for the 10 days of their expected life span. This may be useful in terms of the metastatic potential, but I have a feeling that multiple doses of aspirin in a day might be, in fact, more efficacious. We have to get off the cardiovascular thinking here if we’re talking about colon cancer. I’d like to hear from others whether that impression is correct.

That brings me to another concept. Considering that aspirin’s effect is on the tumor cells — and/or their precursors, in the case of adenomas — I have no idea whether anybody has looked at the bacteria of the colon in terms of its use of an aspirin-inhibited toxin, for instance.

I need not tell you about the studies by Burkitt and others in which the bacterial flora of the gut is highly suspected of being a factor in carcinogenesis of the colon, increasing flow of bacteria and so on. Bacteria turn over rapidly and, if we’re doing anything to the bacteria with aspirin, it would impact on this idea of once a day. It might not be enough, if we’re basically dealing with the toxin producer that might be causing cancer.

So the question I have is, what about multiple dosing?

Ilson: The argument for the high-dose Cox-2s is, you want to get complete inhibition of the enzyme. But all of the epidemiologic data we have is for relatively low doses, so I don’t suspect there’s any demonstrable dose effect that we’ve seen. When you start pushing doses, that’s when you get into the side effect issues, such as ulcer development and bleeding risks.

The issue of bacteria flora is an evolving field. In terms of Helicobacter, the cause of chronic inflammation and irritation of the stomach leading to gastric tumor genesis, there seems to be a fit that inhibiting the inflammatory pathways might have some preventive benefit.

Another fascinating study came out in which researchers got tissue on the cases of colon cancer in the Physicians’ Health Study, and they queried the tumors for Cox-2 expression by immunohistochemistry. They only saw the survival benefit in use of aspirin in tumors that had high Cox-2 mRNA expression, so they actually could identify a biomarker that might predict the benefit from aspirin in that population. So they actually do have some mechanistic evidence that it is the Cox-2 pathway, and Cox-2 is overexpressed in Barrett’s esophagus, so it looks to be a viable target. It may be it has an issue in platelet inflammation, thrombosis, metastasis invasion or development.

There still is an active interest in developing novel Cox-2 inhibitors that avoid the cardiovascular toxicities. The beauty of aspirin is that it actually is cardiovascular protective, so you’re getting that secondary benefit of cardiovascular protection in addition to the potential primary and secondary cancer preventive benefits.

Platelets vs. tumor cells

Rao: The story with platelets and cardiovascular disease has been low-dose aspirin — once a day is fine, but that concept has been re-thought, even in cardiovascular diseases.

We know that if there is rapid turnover of platelets, single dose is not effective simply because the plasma half-life of aspirin is 20 minutes. It’s gone rapidly. By the next day, thromboxane production is back much faster and is not fully inhibited in people who have rapid platelet turnover.

Strong evidence for this comes from studies by Dragani and colleagues. They evaluated patients with essential thrombocythemia and showed that, 24 hours after aspirin dose, patients were making a lot more thromboxane than normal individuals.

In a recent paper in Blood, the same group showed that if you increased the dose from 100 mg to 200 mg you can’t suppress it, but if you give them double dosing — one in the morning and one in the evening — you can knock down the thromboxane production. We know that reticulated platelets are increased in diabetics. They are increased in people with cardiovascular disease. Therefore, we need to re-visit whether twice daily is better than single dose.

In the case relevant to the tumors at the moment, a meta-analysis by 
Peter M. Rothwell, MD, PhD, FRCP, 
suggest that the effect was seen more when the dose was higher. This may be because higher doses are required to inhibit Cox-2.

Lastly, two recent trials did not show efficacy in terms of cancer events. One is the Physicians’ Health Study and the other is the Women’s Health Study. These were two large studies, and neither showed a decrease in overall incidence or mortality with aspirin.

Ilson: In the short term.

Rao: Both of these studies gave aspirin every other day. One gave 325 mg every other day and the other was 100 mg every other day. Therefore, those studies do not fall into the same group as the studies that Dr. Rothwell brought together, where everybody got aspirin every day.

Jacob: You’re suggesting, and I would agree, that probably three- or four-times-a-day aspirin would make some sense at this point. I do believe that.

Platelets don’t turn over particularly quickly, but tumor cells are going to be turning over rapidly. Then you need to have it dosed three or four times a day rather than just once a day, and possibly adult aspirin instead of baby 
aspirin.

Rao: The paper in The New England Journal of Medicine by Stone and colleagues showed that patients with ovarian cancer with high platelet counts have a worse prognosis. In animal models, they show if you knock down the platelet count, you get an antitumor effect. So they’re all linked. In the case of tumors, we probably need the higher dose of aspirin, and we may need it more than once for optimal effect.

Additional studies

Ralph Green, MD: The Physicians’ Health Study has been mentioned several times. There were several different limbs to the study.

It’s perhaps an object lesson in overambitious design and unintended consequences. In the JAMA paper, the startling news came out that one group was given vitamin supplements, the intention being administration of folate — folic acid, specifically — as a preventive agent in order to assist in DNA repair. It actually accelerated the progression of polyps in terms of their number and degree of histological aggressiveness. The authors concluded that folic acid supplementation might increase the risk of colorectal neoplasia and the study was stopped.

Ilson: There are active studies looking at both aspirin and celecoxib in the setting of cancer treatment.

One is the ASPECT trial, which is a British study that involves more than 3,000 patients with long-segment Barrett’s esophagus who get randomly assigned to 300 mg once a day, twice a day or placebo. The study is looking at whether daily aspirin use has any impact on evolution to invasive cancer. That will be an interesting data set. The safety data suggests that it’s safe to administer aspirin to a large population, that if you’re going to get bleeding events, you see them very early on and it’s a very low incidence.

The other trial in the United States, which actually is in danger of failing, is the CALGB Trial 80702. It’s looking at daily celecoxib for 3 years as an adjunctive to adjuvant chemotherapy.

Patients with stage III colon cancer get standard adjuvant chemotherapy, and then some get randomly assigned to receive an intermediate dose of celecoxib. It’s a 100-mg twice-daily dosing, and the impact is to see whether they can improve disease-free survival at 3 years. That trial actually is in danger. It’s a 2,500- or 3,000-patient trial. They’ve only recruited a few hundred patients.

There’s an interesting secondary aspect of that trial, which is part of a global effort. In addition to the celecoxib question, there’s a second randomization to compare 3 months vs. 6 months of adjuvant treatments. That seems to be what’s hurting the study in the United States. American patients, for whatever reason, think more is better and they don’t want to get randomly assigned to the shorter therapy.

Burtness: This isn’t a subset analysis that I’ve ever seen in any adjuvant trial, but if we’re saying that some patients take aspirin — either voluntarily or with direction from their doctor — and there is a significant imbalance between arms in use of post-treatment completion aspirin, it’s potentially a significant variable.

Ilson: In the CALGB IFL Adjuvant Study 89803, they had a whole correlative component where they had patients answer questionnaires, and that’s how they collected the lifestyle and drug use data, exercise levels and so forth. But that’s going to be confounding because many of us feel compelled to recommend daily low-dose aspirin in these patients.

Burtness: I’m not just thinking colon cancer. It may well have some impact in other cancers as well, right? So we’re always looking for 3% and 4% differences with adjuvant treatment, and if we’re not collecting a potential confounder like this …

Rao: People with malignancies have other risk factors for cardiovascular disease, so many of them — if not most — are already on aspirin. In these large studies that Rothwell and colleagues are looking at, one of the issues has been how carefully were the mortality causes established in each one of those trials? Those trials were actually cardiovascular-driven. So is it a non-vascular death or is it cancer death? This needs to be established. This is one reason why it may have not been possible to include some other such trials in the analyses.

Ilson: The Rothwell series actually suggested an earlier benefit. The Physicians’ and Nurses’ health studies had suggested you need to be on aspirin for 20 years to derive a benefit, and this study suggested the benefit as early as 2 to 5 years for aspirin therapy in terms of reducing cancer mortality.

But this is all retrospective analysis, so another confounder is that patients who were taking daily aspirin might have a greater propensity to bleed. It may come down to finding the cancers and making earlier diagnosis rather than preventing them. That’s the naysayer critique, that discovering the cancers earlier resulted in a better outcome, and that it wasn’t actually aspirin impacting on the natural history of the cancer.

References:

For more information:

Disclosure: Bertino, Burtness, Green, Ilson, Jacob and Rao report no relevant financial disclosures.