Now and then, a study grabs the interest of the editors and readers of Sensible Medicine to such a degree that we seem unable to let it rest. Here, Dr. Blase Polite pushes back against recent critiques of CHALLENGE, a trial examining the effects of exercise on colon cancer.
I wonder if the exercise protocol would improve OS for any cohort of patients with any health issues. 7 to 10% of absolute benefit seems plausible for that type of physical effort for couple of years. I agree with Dr. Cifu that priors are not so helpful in assessing rather novel approach.
I truly love the debate. Sensible Medicine makes life of a curious person a lot more interesting. Thank you all.
A cardiologist claiming since they didn't lose weight or inches that the results are implausible and hence study is flawed is just ridiculous. Should have stopped reading after that! And his whole pre-test probability is based on this faulty assumption. In fact, we have lot of research showing the issues with sedentary activity and how PA is so important independent of weight loss. Seriously, does he even recommends exercise to his patients unless they lose weight?
Also, pre-test probability should be taken with handful of salt. Imagine using pre-test probability for the CAST and HRT trials! Cifu has even written a medical reversal book that pretty much goes against using pre-test probabilities!
Problem with randomization could be easily spotted from differences in the baseline table for important prognostic variables and they all look fine! Also issues with randomization do not magically end in a positive result, unless he is claiming fraud.
Adherence is typically low for long duration exercise trials. and is expected! One way they could have improved adherence is by having a run in phase like pharma often does and only keep the motivated participants in the study! But nobody seems to have issues with that practice.
It is expensive, but nowhere near compared to the cancer drugs and also the side effects. Also the benefits are beyond just cancer and longevity: quality of life, physical function, heart benefits and so forth. If patients had to take an informed choice comparing the cost, benefits and risk, they wouldn't mind this cost!
I struggle with Dr Polite’s critique. If anything the Challenge trial probably shows that highly selected patients ( think here of how long it took to recruit ) are fitter and have a better performance status and hence tolerate current and future Colorectal cancer therapies.
The first thing I noticed is that there needs to be a link between a survival benefit in CRC patients and the proposed impact exercise has on that oncogenic process. Exercise improves cardiovascular health, fitter patients can tolerate treatment better , is that what the study has found ?
It would be great to get Dr Polite’s view on how exercise as described in the study has a specific mechanistic effect that is anti cancer ? What is the pathological mechanism that every other anti cancer treatment that is scrutinised from a scientific perspective must explain , what is the hypothesis ?
Also beware “ rolling in “ previous cohort studies with a RCT , they are completely different things.
If Dr. Polite truly believes that light exercise or for that matter any exercise has any effect on colon cancer progression, I have a bridge to nowhere to sell him. And for sure remind me not to consult Dr. Polite for any oncological advice. Of course exercise is good for one for all sorts of reasons, but even at face value the ~10% improvement is insignificant from any meaningful perspective. There is just no biological mechanism whereby exercise could be of any consequence. Now, if the good doctor is dealing with patients who can't climb up a flight of stairs with ease and are huffing and puffing, perhaps a small degree of exercise might improve their cardiovascular health and hence their overall survival but that has nothing to do with cancer progression. As far as colon cancer is concerned, claiming that exercise has an overall survival benefit is not different from claiming that a sugar pill increases OS, and I doubt even the good Dr would buy that.
"There is just no biological mechanism whereby exercise could be of any consequence."
That's a pretty bold statement, not backed up by evidence.
The authors offer a number of mechanisms: "Exercise may be an effective treatment for colon cancer micrometastases and prevention of second primary cancers through various mechanisms, including increased fluid shear stress, enhanced immune surveillance, reduced inflammation, improved insulin sensitivity, and altered microenvironment of major sites of metastases. In particular, exercise may affect metabolic growth factors such as insulin and insulin-like growth factors that promote cancer-cell proliferation and progression."
Planned analyses of serial blood samples will address questions related to possible mechanisms.
Yes. Multiple mechanisms, also some of them maybe unknown. That's the whole point of why we do RCT's. Also several observational studies and animal studies pointing int he same direction.
Not a bold statement at all. Just common sense. Why not compare cancer rates in those who exercise and those who don't. The differences are going to be tiny. To think that exercise "may be an effective treatment for colon cancer micrometastases and prevention of secondary cancers" is no different from believing in the tooth fairy. Ineed, it's no different from believing that oral vitamin C is an effective cancer-prevention regimen.
We have several observation studies and pre-clinical, animal data showing the benefits of exercise and cancer. It is nothing new. Read the National PA guidelines or the ACSM guidelines.
Dr Polite cites “10% improvement in 5 yr OS”. While seemingly “robust”: if 9 of 10 patients receive no OS benefit my guess is that when offered these options many patients may take a negative view and instead remain couch potato- like ??? Especially given the degree of effort/ manpower required to accrue this “robust” benefit.
I think this study is akin to STRONG-HF for heart failure. It is a laborious protocol that will have significant external validity challenges. But if it’s somehow something that the health system can provide (and the pt can stick to), there’s really no downside. In some regards it’s better than “strong hf”, since it’s just encouraging exercise and not a “rush to put 4 drugs on everybody”. This is nowhere near my field, and hence not something I would have much occasion to apply, but I “want” it to be correct.
I would note that a 2mL/kg/min improvement in MVO2 represents less than an increase of 1 MET (metabolic equivalent) which certainly does not portend going from NYHA class 3 to class 2 as suggested in the OP.
Also, the trial suggested an improvement of 25-30m for 6MWT (the threshold for meaningful clinical improvement for cardiopulmonary conditions is generally somewhere btw 25-50m). This would seem to be in that ballpark if the same metric applies in oncology. But I wouldn’t characterize that as being able to walk “1 or 2 more blocks”….at least not in the city I live in.
The point is not about focusing on the clinical significance of the improvements, especially when we don't have any threshold for this particular population and these are just secondary outcomes related to function. He is just trying to show that there were improvements in OBJECTIVE measures which Mandrola says there isn't and hence the study is 'flawed'!
I don’t want to speak for Dr. Mandrola, but I belief part of his critique is that some of those objective measures (like MVO2) showed only a small difference btw the 2 arms…so how believable is it to attribute something like improved cancer free survival to an objectively small improvement in fitness?
CV fitness is just one small part of the equation. There could be multiple mechanisms as authors pointed out and some of them could be just unknown, especially when we know there were cancer reductions. And this is exactly why we do RCT's!
Also RR reductions always looks impressive. AR reduction shows 6-7 people getting a benefit out of 100 people in 8 years. Not so impressive!
As I say, I “want to” believe the results. It’s just exercise, so no downside.
I also agree the result is the result, and that takes precedence over lack of an established mechanism. And to be sure, we wouldn’t (and shouldn’t) just accept biologic plausibility if it is not backed up by RCT evidence. But I’d prefer RCT that backs up biologic plausibility, over RCT that relies on a “black box” mechanism.
Remember there are several observational studies, animal data and invitro data that showed similar results and PA already recommended for cancer by federal guidelines, WHO and so forth. So these results appear nothing new for most folks. As authors mention, mechanisms could be due to effects of exercise on metabolic growth factors, inflammation, and immune function . Further analysis will give us some light on the mechanisms. Also, improvement in vo2 max and walk speed also part of the mechanisms for improved survival. He didn't even mention vo2 max improvement! So there is plausibility and clear objective improvements in fitness!
Also, IMO, Mandrola think this is a "massive" mortality benefit and hence his reluctance. But 8 out of 100 people getting a benefit in 8 years is no where close to massive! This a common mistake when folks use RR and tend to exaggerates the benefits.
I agree with your Blackbox comment. I also understand the criticism that we need to know more about the mechanisms behind cancer reduction and mortality benefit as shown here and we need more studies. But just outright saying the study is "flawed" beacuse the degree of improvement doesn't agree with his subjective calculations is just being very arrogant.
I wonder if the exercise protocol would improve OS for any cohort of patients with any health issues. 7 to 10% of absolute benefit seems plausible for that type of physical effort for couple of years. I agree with Dr. Cifu that priors are not so helpful in assessing rather novel approach.
I truly love the debate. Sensible Medicine makes life of a curious person a lot more interesting. Thank you all.
Spot on!
A cardiologist claiming since they didn't lose weight or inches that the results are implausible and hence study is flawed is just ridiculous. Should have stopped reading after that! And his whole pre-test probability is based on this faulty assumption. In fact, we have lot of research showing the issues with sedentary activity and how PA is so important independent of weight loss. Seriously, does he even recommends exercise to his patients unless they lose weight?
Also, pre-test probability should be taken with handful of salt. Imagine using pre-test probability for the CAST and HRT trials! Cifu has even written a medical reversal book that pretty much goes against using pre-test probabilities!
Problem with randomization could be easily spotted from differences in the baseline table for important prognostic variables and they all look fine! Also issues with randomization do not magically end in a positive result, unless he is claiming fraud.
Adherence is typically low for long duration exercise trials. and is expected! One way they could have improved adherence is by having a run in phase like pharma often does and only keep the motivated participants in the study! But nobody seems to have issues with that practice.
It is expensive, but nowhere near compared to the cancer drugs and also the side effects. Also the benefits are beyond just cancer and longevity: quality of life, physical function, heart benefits and so forth. If patients had to take an informed choice comparing the cost, benefits and risk, they wouldn't mind this cost!
Thank you for your sensible post!
Get 'em Blase!
I struggle with Dr Polite’s critique. If anything the Challenge trial probably shows that highly selected patients ( think here of how long it took to recruit ) are fitter and have a better performance status and hence tolerate current and future Colorectal cancer therapies.
The first thing I noticed is that there needs to be a link between a survival benefit in CRC patients and the proposed impact exercise has on that oncogenic process. Exercise improves cardiovascular health, fitter patients can tolerate treatment better , is that what the study has found ?
It would be great to get Dr Polite’s view on how exercise as described in the study has a specific mechanistic effect that is anti cancer ? What is the pathological mechanism that every other anti cancer treatment that is scrutinised from a scientific perspective must explain , what is the hypothesis ?
Also beware “ rolling in “ previous cohort studies with a RCT , they are completely different things.
If Dr. Polite truly believes that light exercise or for that matter any exercise has any effect on colon cancer progression, I have a bridge to nowhere to sell him. And for sure remind me not to consult Dr. Polite for any oncological advice. Of course exercise is good for one for all sorts of reasons, but even at face value the ~10% improvement is insignificant from any meaningful perspective. There is just no biological mechanism whereby exercise could be of any consequence. Now, if the good doctor is dealing with patients who can't climb up a flight of stairs with ease and are huffing and puffing, perhaps a small degree of exercise might improve their cardiovascular health and hence their overall survival but that has nothing to do with cancer progression. As far as colon cancer is concerned, claiming that exercise has an overall survival benefit is not different from claiming that a sugar pill increases OS, and I doubt even the good Dr would buy that.
"There is just no biological mechanism whereby exercise could be of any consequence."
That's a pretty bold statement, not backed up by evidence.
The authors offer a number of mechanisms: "Exercise may be an effective treatment for colon cancer micrometastases and prevention of second primary cancers through various mechanisms, including increased fluid shear stress, enhanced immune surveillance, reduced inflammation, improved insulin sensitivity, and altered microenvironment of major sites of metastases. In particular, exercise may affect metabolic growth factors such as insulin and insulin-like growth factors that promote cancer-cell proliferation and progression."
Planned analyses of serial blood samples will address questions related to possible mechanisms.
Yes. Multiple mechanisms, also some of them maybe unknown. That's the whole point of why we do RCT's. Also several observational studies and animal studies pointing int he same direction.
Not a bold statement at all. Just common sense. Why not compare cancer rates in those who exercise and those who don't. The differences are going to be tiny. To think that exercise "may be an effective treatment for colon cancer micrometastases and prevention of secondary cancers" is no different from believing in the tooth fairy. Ineed, it's no different from believing that oral vitamin C is an effective cancer-prevention regimen.
We have several observation studies and pre-clinical, animal data showing the benefits of exercise and cancer. It is nothing new. Read the National PA guidelines or the ACSM guidelines.
Dr Polite cites “10% improvement in 5 yr OS”. While seemingly “robust”: if 9 of 10 patients receive no OS benefit my guess is that when offered these options many patients may take a negative view and instead remain couch potato- like ??? Especially given the degree of effort/ manpower required to accrue this “robust” benefit.
I think this study is akin to STRONG-HF for heart failure. It is a laborious protocol that will have significant external validity challenges. But if it’s somehow something that the health system can provide (and the pt can stick to), there’s really no downside. In some regards it’s better than “strong hf”, since it’s just encouraging exercise and not a “rush to put 4 drugs on everybody”. This is nowhere near my field, and hence not something I would have much occasion to apply, but I “want” it to be correct.
I would note that a 2mL/kg/min improvement in MVO2 represents less than an increase of 1 MET (metabolic equivalent) which certainly does not portend going from NYHA class 3 to class 2 as suggested in the OP.
Also, the trial suggested an improvement of 25-30m for 6MWT (the threshold for meaningful clinical improvement for cardiopulmonary conditions is generally somewhere btw 25-50m). This would seem to be in that ballpark if the same metric applies in oncology. But I wouldn’t characterize that as being able to walk “1 or 2 more blocks”….at least not in the city I live in.
The point is not about focusing on the clinical significance of the improvements, especially when we don't have any threshold for this particular population and these are just secondary outcomes related to function. He is just trying to show that there were improvements in OBJECTIVE measures which Mandrola says there isn't and hence the study is 'flawed'!
I don’t want to speak for Dr. Mandrola, but I belief part of his critique is that some of those objective measures (like MVO2) showed only a small difference btw the 2 arms…so how believable is it to attribute something like improved cancer free survival to an objectively small improvement in fitness?
CV fitness is just one small part of the equation. There could be multiple mechanisms as authors pointed out and some of them could be just unknown, especially when we know there were cancer reductions. And this is exactly why we do RCT's!
Also RR reductions always looks impressive. AR reduction shows 6-7 people getting a benefit out of 100 people in 8 years. Not so impressive!
As I say, I “want to” believe the results. It’s just exercise, so no downside.
I also agree the result is the result, and that takes precedence over lack of an established mechanism. And to be sure, we wouldn’t (and shouldn’t) just accept biologic plausibility if it is not backed up by RCT evidence. But I’d prefer RCT that backs up biologic plausibility, over RCT that relies on a “black box” mechanism.
Remember there are several observational studies, animal data and invitro data that showed similar results and PA already recommended for cancer by federal guidelines, WHO and so forth. So these results appear nothing new for most folks. As authors mention, mechanisms could be due to effects of exercise on metabolic growth factors, inflammation, and immune function . Further analysis will give us some light on the mechanisms. Also, improvement in vo2 max and walk speed also part of the mechanisms for improved survival. He didn't even mention vo2 max improvement! So there is plausibility and clear objective improvements in fitness!
Also, IMO, Mandrola think this is a "massive" mortality benefit and hence his reluctance. But 8 out of 100 people getting a benefit in 8 years is no where close to massive! This a common mistake when folks use RR and tend to exaggerates the benefits.
I agree with your Blackbox comment. I also understand the criticism that we need to know more about the mechanisms behind cancer reduction and mortality benefit as shown here and we need more studies. But just outright saying the study is "flawed" beacuse the degree of improvement doesn't agree with his subjective calculations is just being very arrogant.
Thanks Steve!