Vinay, I appreciate your deep dive into the CABINET trial. As someone who runs clinical trials and treats neuroendocrine tumor (NET) patients, I agree that trial design matters, but I think the realities of NETs make this study more nuanced than your critique suggests.
Placebo control is always a tough call, but in NETs, there’s no clear sequencing standard after progression. While everolimus, sunitinib, and Lu-Dotatate are options, not every patient is eligible, and real-world practice often includes periods of observation. This wasn’t a “no treatment” placebo—patients could continue somatostatin analogues, and crossover ensured they weren’t left without access to therapy. Would a head-to-head trial against everolimus or sunitinib have been better? Absolutely. But a placebo control allowed a cleaner look at cabozantinib’s efficacy without confounding from another active drug.
PFS as a primary endpoint makes sense here. NETs are slow-growing, and OS is difficult to measure given long survival and post-progression treatment variability. The key question is whether a treatment can delay progression without worsening quality of life—and in this trial, it did. I disagree that this is toxic drug, if so QoL should have been significantly worse in the treatment arm. But it wasn’t. That’s an important finding, especially in a disease where keeping things stable for as long as possible can be a meaningful clinical goal. Dose reductions were common (~60%), but many patients continued on lower doses with good tolerance and maintained response.
The 2:1 randomization was an inefficient choice, and switching from central to local scan review does raise concerns. Investigator bias in calling progression earlier in the placebo arm is possible, but given the magnitude of the PFS difference, it’s unlikely to negate the treatment effect entirely.
I understand your frustration with how some NIH-funded trials are structured, but in rare cancers like NETs, public funding fills gaps where industry won’t. Without NIH support, this trial likely wouldn’t have happened, and we’d have even fewer data points to guide treatment. If anything, this should spark a conversation about designing better, more clinically relevant public trials—not a dismissal of the entire effort.
Was CABINET a perfect trial? No. Could the design have been stronger? Yes. But does it provide useful data for an unmet need? I think so. NETs require a different lens than many other metastatic cancers, and skepticism should push us toward better trials—not toward dismissing potentially meaningful progress for patients who don’t have many options.
Oncologists need to grow up and shelve Progression Free Survival. Patient's don't care about sizes of tumors on the scan and they want to know more about overall survival. Somehow over in Primary Care, people run 10-20 year trials and oncologists make every excuse under the sun to obscure the data and not report OS
I agree with you OS should be the primary endpoint of cancer studies and I get why some people say PFS isn’t important, especially if it doesn’t always lead to longer survival. But as an oncologist, I’ve seen how slowing cancer down can make a huge difference in a patient’s daily life. It’s not just about numbers. It’s about how people feel and what they’re able to do with the time they have.
I’ve had patients with lung tumors that made every breath a struggle, and when treatment shrank the tumor, they could finally breathe again, sleep through the night, and take walks with their family. I’ve seen people with bone metastases go from constant pain and barely moving to feeling strong enough to get out and enjoy life again. For patients with brain tumors, keeping the cancer from growing means they can stay sharp, independent, and present with their loved ones for longer.
For those who don’t have curative options, my priority is always their quality of life. If a treatment can give them more time feeling good, without pain or major life disruptions, that matters. PFS isn’t just a statistic. It represents real moments of better living for people who deserve every bit of that time.
What else is new. Government funded research is, at best, useless and most often corrupt. It is difficult to differentiate it from blatantly self-promoting Big Pharma research. In the same way that conventional news sources have been rendered obsolete by online sites and critics, so have the prestigious medical journals such as NEJM and Lancet proven to be mindless lackeys of the medical and pharmacological establishment. The adjective "peer reviewed" has become meaningless. Thank goodness for sites like Sensible Medicine that hold their feet to the fire.
You are exactly right about this bogus study. As a published researcher myself, I can tell you that there are 20 ways to design and conduct a study to pre-determine a favorable results. The FDA is complacent with these flawed studies because of a conflict of interest. 74% of the FDA's budget is paid for by Big Pharma, and 73% of their advisory committees have members of Big Pharma, that approve new drugs. Some of the common strategies of deception are to report relative benefits, rather than absolute benefits, use association studies to prove cause and effect, conduct non-inferiority studies (my new drug is no worst than another worthless drug), and hiding confounding variables. The NIH is just as bad as Big Pharma. Thanks for pointing out the hidden defects in the study. The poor cancer patients suffer.
Why is anyone surprised by this NIH funded study….this is no different than USAID funds….NO one are truly asking the question of effective use of OUR tax dollars!…. just no real oversight and it costs us all. THIS is what NIH has allowed…and if such similar unethical behavior to purposefully distort the results to make the study show what they wanted from the very beginning…gaming the results for money.
This article just flashed me back to a frightening and equally sad memory. I had a post transplant person in the addiction and chronic pain program, and addiction had occurred (despite many “studies” suggesting “pseudo addiction” back in the day when using large amounts of both short acting and long acting opioids were gorged into folks…). Patient had done well with opioid weaning and graduated successfully, however failed to continue addiction management and ultimately relapsed. I got a frantic call from the transplant surgeon telling me we had to keep this patient alive long enough to “complete the study (NIH funded).” That was the extent of the chat. I had to sit down. When studies (and poor ones at that) take precedence over people we know we are on the wrong side of the tracks. Hell I’m sometimes not even sure if the train will ever get back on em.
I don't understand much about the trial but I suspect it is similar to most drug trials...manipulated, fudged and costly on all fronts. We are supposed to trust the NIH? Trust big pharma? Trust the oncologists? Maybe if I am drugged enough.
This is very interesting. It is interesting that the NIH seems to ask a lot from researchers when granting funds which has been a big complaint of those seeking to do funded research, but at the same time the oversight seems dismal as in this case when the study is poor and very unethical. Great to be failing in all ways possible. I want us to have NIH funding research, but it makes me sad to hear about this level of dysfunction. BTW, I'm not in the medical field so it's really hard to understand all the nuances of this.
As a person who had a pancreatic NET (so far so good after resection) I read this with great interest. Only commenting here so you know that you have laypeople reading avidly.
The study is utterly inexcusable. And thank you for a clear, scientific analysis pointing out the many flaws. It is easy to get emotional over such sloppiness and poor judgement, but objective analysis is precisely what is needed. It's why I subscribe to this site.
This is far from my area, but a terrific exercise in appraisal with neutral priors.
I agree that placebo is appropriate only when it is on top of standard of care. If patients in the control arm of this study were denied therapy they otherwise would have gotten in their clinical situation, that is egregious and unethical.
I agree the primary endpoint seems ridiculous, unless patients in this scenario actually care about increases in tumor size on a scan. Lack of OS benefit over 4-5 yrs may not be surprising, if such malignancies are not “curable”. But this type of endpoint smells like something that reaches statistical significance but is clinically meaningless.
I’m not sure about the objection to crossover. As long as it is analyzed as ITT to maintain randomization.
Loss of central adjudication (esp part way through a trial) is a major red flag. I can’t think of any legitimate clinical reason to do this, with any study.
The funding question is interesting. There’s a certain appeal in suggesting public funding should be used only to answer clinical questions that serve no commercial interest (such that no corporate entity would even bother asking the question ). And further, that public funding should not be used to help create profits that accrue solely to a private entity. OTOH, we are always quick to cast at least a little aspersion on positive clinical trials that are pharma funded. That does seem a little bit of a catch 22. And the climate is worse now with stories of investigator malfeasance such that even independent CRO study conduct and monitoring, and independent data analysis and manuscript generation, do not absolve all innuendo of impropriety (when results one way or another can make the difference measure in billions). On the third hand, public funding did not prevent very questionable protocol design and study conduct in this case.
I wonder how this got past the IRB. When I was a resident, we had a research requirement. I intended to complete the requirement and go directly to private practice and never do another research project again. So, I chose a stupid survey to do intending to minimize my work load while meeting the requirement. And then my program director made me get IRB approval. He was a real jerk, but that is beside the point. The point is that I had to get IRB approval for a stupid, meaningless, voluntary survey where they questioned potential risks (zero!!!) to participants. How in the world does the IRB approve a study which compares a new drug to placebo (essentially withholding proven care)??!! The whole system needs to be reformed, starting with a serious look at IRB approval processes.
With second hand awareness, I would humbly suggest that you are making a gross overgeneralization when stating “this is the” type of cancer Steve Jobs died from. His specific condition was not fully publicized and, suffice to say, he was -like people in the NIH trial you review - in a fundamentally compromised immune-metabolic state after more than one conventional treatment. Perhaps what is most deadly in cancer diagnosis and treatment is the unremitting drive of those profiting from technological determinism as if ever more powerful, more specific or polypharmacy is generically better for everyone with a particular type of tumor.
Pardon me for asking but what if the tumor is an effect rather than a cause or key etiology? Drug trials seldom establish a coherent context for study.
No matter when clinical trials are only really staged for patented tech wherein the studies are business ventures in and of themselves. There’s always a way to spin numbers if only to give people access to something new and expensive somebody thought up as something new.
Your analyses are revealing. Too bad we ignore what proves effective because drug companies can’t patent them or make a big profit.
Vinay, I appreciate your deep dive into the CABINET trial. As someone who runs clinical trials and treats neuroendocrine tumor (NET) patients, I agree that trial design matters, but I think the realities of NETs make this study more nuanced than your critique suggests.
Placebo control is always a tough call, but in NETs, there’s no clear sequencing standard after progression. While everolimus, sunitinib, and Lu-Dotatate are options, not every patient is eligible, and real-world practice often includes periods of observation. This wasn’t a “no treatment” placebo—patients could continue somatostatin analogues, and crossover ensured they weren’t left without access to therapy. Would a head-to-head trial against everolimus or sunitinib have been better? Absolutely. But a placebo control allowed a cleaner look at cabozantinib’s efficacy without confounding from another active drug.
PFS as a primary endpoint makes sense here. NETs are slow-growing, and OS is difficult to measure given long survival and post-progression treatment variability. The key question is whether a treatment can delay progression without worsening quality of life—and in this trial, it did. I disagree that this is toxic drug, if so QoL should have been significantly worse in the treatment arm. But it wasn’t. That’s an important finding, especially in a disease where keeping things stable for as long as possible can be a meaningful clinical goal. Dose reductions were common (~60%), but many patients continued on lower doses with good tolerance and maintained response.
The 2:1 randomization was an inefficient choice, and switching from central to local scan review does raise concerns. Investigator bias in calling progression earlier in the placebo arm is possible, but given the magnitude of the PFS difference, it’s unlikely to negate the treatment effect entirely.
I understand your frustration with how some NIH-funded trials are structured, but in rare cancers like NETs, public funding fills gaps where industry won’t. Without NIH support, this trial likely wouldn’t have happened, and we’d have even fewer data points to guide treatment. If anything, this should spark a conversation about designing better, more clinically relevant public trials—not a dismissal of the entire effort.
Was CABINET a perfect trial? No. Could the design have been stronger? Yes. But does it provide useful data for an unmet need? I think so. NETs require a different lens than many other metastatic cancers, and skepticism should push us toward better trials—not toward dismissing potentially meaningful progress for patients who don’t have many options.
Oncologists need to grow up and shelve Progression Free Survival. Patient's don't care about sizes of tumors on the scan and they want to know more about overall survival. Somehow over in Primary Care, people run 10-20 year trials and oncologists make every excuse under the sun to obscure the data and not report OS
I agree with you OS should be the primary endpoint of cancer studies and I get why some people say PFS isn’t important, especially if it doesn’t always lead to longer survival. But as an oncologist, I’ve seen how slowing cancer down can make a huge difference in a patient’s daily life. It’s not just about numbers. It’s about how people feel and what they’re able to do with the time they have.
I’ve had patients with lung tumors that made every breath a struggle, and when treatment shrank the tumor, they could finally breathe again, sleep through the night, and take walks with their family. I’ve seen people with bone metastases go from constant pain and barely moving to feeling strong enough to get out and enjoy life again. For patients with brain tumors, keeping the cancer from growing means they can stay sharp, independent, and present with their loved ones for longer.
For those who don’t have curative options, my priority is always their quality of life. If a treatment can give them more time feeling good, without pain or major life disruptions, that matters. PFS isn’t just a statistic. It represents real moments of better living for people who deserve every bit of that time.
Nobody at NIH has read the most awesome book written Malignant.
What else is new. Government funded research is, at best, useless and most often corrupt. It is difficult to differentiate it from blatantly self-promoting Big Pharma research. In the same way that conventional news sources have been rendered obsolete by online sites and critics, so have the prestigious medical journals such as NEJM and Lancet proven to be mindless lackeys of the medical and pharmacological establishment. The adjective "peer reviewed" has become meaningless. Thank goodness for sites like Sensible Medicine that hold their feet to the fire.
You are exactly right about this bogus study. As a published researcher myself, I can tell you that there are 20 ways to design and conduct a study to pre-determine a favorable results. The FDA is complacent with these flawed studies because of a conflict of interest. 74% of the FDA's budget is paid for by Big Pharma, and 73% of their advisory committees have members of Big Pharma, that approve new drugs. Some of the common strategies of deception are to report relative benefits, rather than absolute benefits, use association studies to prove cause and effect, conduct non-inferiority studies (my new drug is no worst than another worthless drug), and hiding confounding variables. The NIH is just as bad as Big Pharma. Thanks for pointing out the hidden defects in the study. The poor cancer patients suffer.
Excellent summary. Should be required reading for all physicians and medical students.
Why is anyone surprised by this NIH funded study….this is no different than USAID funds….NO one are truly asking the question of effective use of OUR tax dollars!…. just no real oversight and it costs us all. THIS is what NIH has allowed…and if such similar unethical behavior to purposefully distort the results to make the study show what they wanted from the very beginning…gaming the results for money.
This article just flashed me back to a frightening and equally sad memory. I had a post transplant person in the addiction and chronic pain program, and addiction had occurred (despite many “studies” suggesting “pseudo addiction” back in the day when using large amounts of both short acting and long acting opioids were gorged into folks…). Patient had done well with opioid weaning and graduated successfully, however failed to continue addiction management and ultimately relapsed. I got a frantic call from the transplant surgeon telling me we had to keep this patient alive long enough to “complete the study (NIH funded).” That was the extent of the chat. I had to sit down. When studies (and poor ones at that) take precedence over people we know we are on the wrong side of the tracks. Hell I’m sometimes not even sure if the train will ever get back on em.
I don't understand much about the trial but I suspect it is similar to most drug trials...manipulated, fudged and costly on all fronts. We are supposed to trust the NIH? Trust big pharma? Trust the oncologists? Maybe if I am drugged enough.
This is very interesting. It is interesting that the NIH seems to ask a lot from researchers when granting funds which has been a big complaint of those seeking to do funded research, but at the same time the oversight seems dismal as in this case when the study is poor and very unethical. Great to be failing in all ways possible. I want us to have NIH funding research, but it makes me sad to hear about this level of dysfunction. BTW, I'm not in the medical field so it's really hard to understand all the nuances of this.
As a person who had a pancreatic NET (so far so good after resection) I read this with great interest. Only commenting here so you know that you have laypeople reading avidly.
The study is utterly inexcusable. And thank you for a clear, scientific analysis pointing out the many flaws. It is easy to get emotional over such sloppiness and poor judgement, but objective analysis is precisely what is needed. It's why I subscribe to this site.
Where was the IRB in this? How did they approve a trial that essentially withholds care by assigning people to a placebo arm?
Sad.
This makes my blood boil with anger.
This is far from my area, but a terrific exercise in appraisal with neutral priors.
I agree that placebo is appropriate only when it is on top of standard of care. If patients in the control arm of this study were denied therapy they otherwise would have gotten in their clinical situation, that is egregious and unethical.
I agree the primary endpoint seems ridiculous, unless patients in this scenario actually care about increases in tumor size on a scan. Lack of OS benefit over 4-5 yrs may not be surprising, if such malignancies are not “curable”. But this type of endpoint smells like something that reaches statistical significance but is clinically meaningless.
I’m not sure about the objection to crossover. As long as it is analyzed as ITT to maintain randomization.
Loss of central adjudication (esp part way through a trial) is a major red flag. I can’t think of any legitimate clinical reason to do this, with any study.
The funding question is interesting. There’s a certain appeal in suggesting public funding should be used only to answer clinical questions that serve no commercial interest (such that no corporate entity would even bother asking the question ). And further, that public funding should not be used to help create profits that accrue solely to a private entity. OTOH, we are always quick to cast at least a little aspersion on positive clinical trials that are pharma funded. That does seem a little bit of a catch 22. And the climate is worse now with stories of investigator malfeasance such that even independent CRO study conduct and monitoring, and independent data analysis and manuscript generation, do not absolve all innuendo of impropriety (when results one way or another can make the difference measure in billions). On the third hand, public funding did not prevent very questionable protocol design and study conduct in this case.
I wonder how this got past the IRB. When I was a resident, we had a research requirement. I intended to complete the requirement and go directly to private practice and never do another research project again. So, I chose a stupid survey to do intending to minimize my work load while meeting the requirement. And then my program director made me get IRB approval. He was a real jerk, but that is beside the point. The point is that I had to get IRB approval for a stupid, meaningless, voluntary survey where they questioned potential risks (zero!!!) to participants. How in the world does the IRB approve a study which compares a new drug to placebo (essentially withholding proven care)??!! The whole system needs to be reformed, starting with a serious look at IRB approval processes.
With second hand awareness, I would humbly suggest that you are making a gross overgeneralization when stating “this is the” type of cancer Steve Jobs died from. His specific condition was not fully publicized and, suffice to say, he was -like people in the NIH trial you review - in a fundamentally compromised immune-metabolic state after more than one conventional treatment. Perhaps what is most deadly in cancer diagnosis and treatment is the unremitting drive of those profiting from technological determinism as if ever more powerful, more specific or polypharmacy is generically better for everyone with a particular type of tumor.
Pardon me for asking but what if the tumor is an effect rather than a cause or key etiology? Drug trials seldom establish a coherent context for study.
No matter when clinical trials are only really staged for patented tech wherein the studies are business ventures in and of themselves. There’s always a way to spin numbers if only to give people access to something new and expensive somebody thought up as something new.
Your analyses are revealing. Too bad we ignore what proves effective because drug companies can’t patent them or make a big profit.
Thank you for this very useful critical analysis.