Better PFS but not better OS?
I would like to understand more about Progression Free Survival vs. Overall Survival for stage iv. Recently on the topic "P13K Single Therapy..." Bestbird said, "What I'm beginning to notice with Afinitor and Aromasin, and IBRANCE and Letrozole (for example), is a pattern whereby PFS is significantly extended, but OS is not significantly extended." So does this mean that generally patients did not live a longer timespan, but that more of that time was with NED or stable mets? Wouldn't that mean that eventually things went bad fast near the end of that timespan? And does it mean that even though the patients did not live longer than they would on the AI alone, there was better QOL for a longer time? More time when mets were not causing too much pain or disability?
Comments
-
Great question - I too have been wondering why PFS is extended but OS is not! It's a true conundrum.
My understanding of PFS is that the disease has not become worse. Hence it is considered stable or receding, and it's also possible that the patient becomes NED.
Extending PFS is certainly nothing to be sneezed at, especially for symptomatic patients and/or patients experiencing pain. But OS is also considerably important.
Why would OS not also be significantly extended along with PFS? Possibly drug toxicity adversely impacted the patient over time, and/or the cancer became unusually aggressive towards the end. I've directed the question to an expert to see what she has to say!
-
I think the difference might lie in clinical trials having to satisfy two entities. Clinical studies use progression free survival statistics with an arbitrary starting point, such as the beginning of treatment. Scan results, etc. are all pooled to reach a quantative result. The FDA uses (and demands) overall survival statistics, which include deaths by treatment toxicity, etc. in addition to the stats for progression free survival. A real grey area based on the Mcnamara Fallacy employed by the FDA.
https://en.m.wikipedia.org/wiki/McNamara_fallacy#I...
-
Just realized that might be an overly complicated way to explain it and I could very well be wrong, but this is the explanation I got from my onc.
Progression free survival - the point being you "got there"
Overall survival - how you "got there".
-
Thanks for your reply, Bestbird. I see that you understand my question, and I'll be eager to hear if you learn anything more.
-
Leggo, that is a very helpful link about the McNamara fallacy. So to make it personal, if Ibrance gives me more time without severe symptoms or treatments, even if I don't ultimately live longer, it is worth it to me. That's more time before I have to give up my dancing that I love, etc. (It seems to me that the same fallacy operates in the "It doesn't make any difference to catch mets early" idea.)
But I still don't get why longer PFS wouldn't mean longer OS.
Your onc's explanation seems backward to me. Wouldn't PFS be how you got there, if survival time is the "got there"?
-
Sorry, I deleted my posts when I saw yours. I didn't think my explanation answered your question. Probably better for Bestbird to explain once she's spoken to an expert. My apologies for the intrusion.
-
Leggo, please no apologies! I'm glad you joined the discussion! I addressed my first reply only to Bestbird because I hadn't seen yours yet--we were apparently composing our comments at the same time. When I saw you had posted, I was concerned that you would think I was ignoring you, and went on to compose a reply to you, hoping you realized we had posted at the same time. Ah, the joys of internet communication!
-
This is a really good question and something I've thought about a lot over the past 5+ years. The short answer, in my opinion, is YES to all the questions and comments you raise. Improved PFS means better quality of life longer on less toxic drugs, even if overall survival is not extended.
Interestingly, FDA does not consider improved PFS a "clinically meaningful endpoint" for new drugs. I tried to argue this back when we were fighting to keep Avastin available, that improved PFS equals improved quality of life, but they didn't buy it.
You raise a really good point, one that I think is lost on the general public. How many times have you read a headline that says something like, "XYZ drug costs $100,000 and only increases life expectancy by 2 months". The general public doesn't understand that the 2 months refers to improvement in median overall survival versus the standard of care overall survival. For any given patient with MBC, depending on how far along they are in their treatment, expected overall survival could easily be years. It isn't like they are going to die within 2 months, which is how people interpret the headlines. Not sure if I'm making sense, but I just think the stats are more nuanced and complex than they are simplistic.
-
Musa Mayer, a breast cancer expert, kindly responded to my inquiry with the following:
Two things come to mind, Anne.
First, we don't have the mature results of the palbociclib trials, so we are lacking OS data. Nothing can be said about this drug so far regarding whether it extends survival. I'd have to check, but I think the same is still true about the Afinitor Phase III trials. Phase II trials should be ignored, IMO, because they aren't sufficiently powered for this endpoint.
Second, this is particularly problematic for first line trials in hormone-sensitive MBC, because of the confounding effect of all the subsequent treatments. MDAnderson biostatistician Don Berry did an interesting statistical paper on this phenomena several years ago that made the case that the longer the survival postprogression, the less likely an OS benefit could ever be shown, even if there was one. I'm attaching this paper, which is called, "Detecting an Overall Survival Benefit that Is Derived From Progression-Free Survival"
Note: Musa's attachment is a long and very challenging paper to read. There is an excerpt that I'm still trying to fully comprehend at: http://www.pubfacts.com/detail/19903805/Detecting-....
The excerpt states, "Addressing Survival Post Progression (SPP) is important in understanding treatment effects. For clinical trials with a PFS benefit, lack of statistical significance in OS does not imply lack of improvement in OS, especially for diseases with long median SPP. Although there may be no treatment effect on SPP, its variability so dilutes the OS comparison that statistical significance is likely lost. OS is a reasonable primary endpoint when median SPP is short but is too high a bar when median SPP is long, such as longer than 12 months."
-
I have questioned this issue as well with respect to affinitor and ibrance. The ibrance data is still out and I am hoping it performs better than affinitor in this respect.
It doesn't seem median overall survival tells the whole story. If a drug only works for 40% of patients and significantly improves overall survival for those 40%, this may have zero impact on the median overall survival. I would like to see more data on overall survival (such as an average survival, survival range and more detailed stats than median). -
Bestbird explains things better than I do but I'll wade into this one. There are a couple of possibilities for why PFS and OS differ. There are likely other possibilities and I can ask some people.
The most likely possibility [to me] is that usually these studies test A vs A+B. They do NOT test A vs A-then-B. Do you see the difference? A+B is usually better than A alone, ergo PFS survival is extended. However, that may not mean that people live longer. I have brought this up with my clinical trialist onc and a couple of my editors. Both shrug and say "true but no way to know".
Another possibility is, and I've noticed this following the palbociclib trials, that the big PFS advantage touted, is in a Phase 2 study. When you get to a bigger study (Phase III), have broader inclusion criteria and many more women in the study, the end results almost never look as good. This has definitely been true with Ibrance.
A third possibility is that these are median PFS and OS. Maybe the people who fail to respond to the first drug, and have less PFS take a different drug next and respond much better. Ergo their OS is at least as long. We do know that there are magic bullets for people. There are so many variables after people stop taking any particular drug. . .
Different learned people think differently about whether PFS or OS is the best thing. Our previous oncology editor felt that if a drug didn't show an overall survival advantage it wasn't a good drug. [This is my prejudice because I listened to him for 10 years]. On other words, we do not yet know whether Ibrance is a good drug, because the overall survival data have not yet been published. Our current oncology editor loves all new drugs. If it extends PFS he is on board. They're both very smart and thoughtful people.
It's all a conundrum.
Edited to add: I cross posted with the two posts above. We are all in agreement on the Phase II/Phase III story on palbociclib.
-
I have wondered about this also. The paragraph Bestbird quotes actually makes sense to me. I read it to say that just because the OS does not appear to have an improvement this is not necessarily true for scenarios in that the OS is a longer time period. That there are too many other factors over a longer period of time after the trial drug has failed to effectively understand the impact the trial drug had on it. Additionally Musa Mayer wrote this is particularly problematic for first line trials in hormone-sensitive MBC, because of the confounding effect of all the subsequent treatments.
I take away that I am thrilled with the new drug options we are getting that increases the PFS and will not put too much emphasis on the murky OS numbers.
-
Bestbird (Anne), thank you so much for the information from Musa Meyer. Her explanations articulate some of the random thoughts that bounce around in my head. Tell her thank you!
Specifically, she says, "OS is a reasonable primary endpoint when median SPP is short but is too high a bar when median SPP is long, such as longer than 12 months."
Exactly.
-
At some point I read (though I haven't kept up) that letrozole did not provide an overall survival advantage compared to tamoxifen for early breast cancer although disease-free survival (DFS) was longer. In this case it was the greater toxicity of letrozole to blame. The solution was to switch from AI to Tamoxifen or vice-versa to reduce cumulative toxicity. Letrozole is a good drug so, personally, I'd go for PFS/DFS in general (like pajim's current onc editor) and switch before you have too much toxicity or cellular adaptations. We need to develop biomarkers for the latter, pronto!
-
I second/third the idea that when you are talking about last-ditch therapy, Overall Survival is more important than when you are talking about first line therapy. I sort of said that above, but hadn't thought of it in exactly those terms.
There's so much confounding between first progression and the end of life. I've noticed even on these boards that we may each have the same treatments but not necessarily in the same order nor for the same length of time.
Biomarkers are tricky. Right now we think of breast cancer as 4-5 (or a dozen depending on who you talk with) diseases. In reality it's probably hundreds of different diseases. Now if there were a way to easily measure cellular adaptations. . .except that tumors are heterogeneous even within themselves. If you take a biopsy, it's totally dependent on the cells they happen to take. The cells right next door might have different markers and actually behave differently, but you wouldn't know it. Sigh.
-
Wow, this discussion is making my brain hurt. But, it's a very important discussion, one I've thought a lot about. I am faced with a difficult decision in choosing where to go with my treatment after a significant progression. I always ask about PFS vs OS and this discussion has brought to light some nuances that I hadn't previously taken into account. I have rejected treatments with no or little OS benefit when the toxicity is more than I want to accept. This makes me reconsider. Wow, I wish there was a more black and white answer to this question. Thanks for starting this discussion, Shetlandpony, I've learned something here!
-
My brain hurts too! lol This is a very interesting thread. Heidihill's post makes me thing that avoiding cellular adaptation is why my onc stopped letrozole "to keep out in front of things." Now to go back and re-read and see what else I can absorb :-)
Thank you, all.
-
To Heidihill's excellent point, I often wonder whether it may be worth switching from one therapy to the next before the therapy has failed. I have 2 overall thoughts that I'm hoping people might weigh in on:
1) Perhaps it would be more efficacious to cycle through a set of therapies in a continuous loop before any single therapy fails. As we're aware, the current conventional strategy entails trying one therapy, failing it, then trying another, failing that, and so forth until the list is exhausted. Alternatively, cycling through a given set of therapies and then starting all over again in a continuous loop might keep the cancer totally confused and not provide it with the opportunity to develop the resistance and escape mechanisms it's infamous for.
2) Since bc/mbc is a heterogeneous disease and it is possible that some parts of the tumor (or tumors in other parts of the body) have a different hormonal and/or HER2 profile than the biopsied section, what is the harm in starting everyone on hormonal therapy unless they are significantly ill from the disease (in which case chemo would be called for because it is faster-acting). Perhaps the same should apply to Herceptin for HER2- at some point. Again, just musing!
-
I'm
I was just wondering today something along these same lines. Was going to talk to my onco about it but they screwed up and did not have an appt made for me. I was so cheesed off. I really wanted to see him.
But, I digress.lol when they publish facts about perjeta increasing survival time by 15 months...is that only for ladies that reach a state of NED
-
Bestbird your points are very valid, but as you know we are all limited by insurance.To your second point, I had read some years ago that even women without positive hormone receptors may be helped by Tamoxifen or AIs. But the problem is getting it approved.
Same thing with the Her2nu. Several years ago my onc wanted to test my bone mets to see if they were still ER/PR+ since as you know that can change over time. The biopsy did still show positive for ER/PR, but did not show Her2nu. My onc told me he was certain that I was still Her2nu+ since that typically does not change from+ to - but can change from - to +. Most times Her2nu does not show up in bone mets but will typically show up in organs. When I was trying to get Perjeta it was initially denied since my latest biopsy did not show Her2nu. I had to appeal and then try to get my original report from 1997. Not an easy task. So unforntunately we are all limited by insurance.
-
Bestbird, Your points are entirely valid.
It is known that some women who appear to be HER2- or ER/PR- to benefit from Herceptin or hormonal treatment. The general feeling at my workplace is that they are actually positive, but that the lab result was a false negative. It could easily be heterogeneity of the tumor(s) instead.
As for cycling through the hormonals, I'm not sure. On the plus side, isn't progression all about drug resistance? Sort of like antibiotic resistance?The end result of the true hormonals (I'm not thinking about the new drugs) is to block the estrogen receptor. Does it matter how you block it? My answer is partially. Isn't it generally true that patients get the "most" out of the first hormonal, and the rest work for a progressively less time? I made have just made a bunch of false statements here. Maybe one of you knows the answer?
I wonder if anyone has tried the cycling approach? You'd think someone would try but the $$ might not be there for a full clinical trial. A quick search of PubMed turned up zilch.
I heard a talk at the MBC Conference last year where the speaker what eventually happens is the tumor cell grows so many estrogen receptors that you can't stop them all. That's why giving estrogen might work after all the hormonal therapies fail. There are so many estrogen receptors on the tumor cell that you can overwhelm it and the tumor cell 'explodes'.
I'm due to have a philosophical discussion with my oncologist next week. I'll ask him. He loves discussing this kind of question.
-
About statistics and PFS and OS:
Man, if I had known I would get cancer, I would have taken statistics in college! I conclude from Berry that statistics probably can't tell us if a particular drug that increases PFS also increases OS, when survival post-progression is long (over twelve months). So I say, just because you can't show something with statistics does not mean it is not true. Common sense would say improved PFS would mean improved OS unless the drug in question has effects toxic enough to outweigh its benefits. Also, as pajim and JFL point out, median survival may not tell the whole story. A subset of patients who are helped a lot may get lost in the numbers. Thank you, everyone, for your interesting comments! SonnyB, I like your summary.
About the idea of cycling treatments:
I have been thinking about this and I think alternating the type of treatment might be good. After tamoxifen failed me (stage 1 and I think it likely my ILC was de novo resistant) my onc recommended taxol as the first treatment (extensive liver mets). She said we would try to put the cancer into remission with chemo, then put me on an AI. This strategy worked. If/when I have progression, I am going to ask if we can use the same strategy again: knock it back with chemo, then hold the line with hormone therapy. (Different ones than before.) It seems to me that alternating the two types of treatment would address resistance and heterogeneity.
-
ShetlandPony, I have been thinking along the same lines as far as alternating chemo and hormonal therapy. I feel an additional reason would be giving our bodies a rest between chemos. My onc feels that when I progress, if it's in bones only to try another hormonal, if soft tissue to attack it with chemo. Although I hope that discussion is in the distant future I feel more secure in having thought this out in advance.
-
http://elifesciences.org/content/2/e00747
According to the linked article above solid cancer drug resistance is not like antibiotic resistance. These guys believe there are already resistant cells in solid tumors before the start of treatment. They grow as others that are not resistant either die or develop resistance. This scenario, however, does not apply to immunotherapies because the immune system can "replicate and evolve". Hey, just like cancer. So there's hope there.
Something similar to the mathermatical model for resistance used here might be developed for cycling through different therapies. Maybe the authors could come up with a predictive model to guide when a switch should be made. The model and article shown may take me a while to dissect together with a bottle of aspirin.
Yes, this is my current combination therapy, which is supposed to be better than sequential therapy.
Kidding aside, I wish they would get the data for breast cancer.For my n of 1 case I did cycle through 7 different drugs without any progression (taxotere, cytoxan, adriamycin, zometa in combination as first line therapy and femara, tamoxifen, arimidex, tamoxifen again, sequentially). I also took drug holidays every once in a while, to get back some quality of life and some estrogen. Maybe the latter also killed some cancer cells or at least resensitized them to antihormonals.
-
http://www.sciencedaily.com/releases/2015/07/150708151227.htm
This is an example of a potential biomarker.
-
I asked my onc about cycling therapies. He said that they tried it some time ago (read many years I'd guess), but it didn't seem to help. I didn't ask for references. He thinks it's because while tumors are heterogeneous, they are more the same than they are different.
In other words, while the hormonal therapies attack the estrogen pathway differently, they all attack the estrogen pathway. Chemo kills fast growing cells. It may do so from different angles but the end result is the same.
He, too, is anxious for immunotherapy to reach breast cancer.
-
I have some ideas about PFS (progression free survival) and OS (overall survival) that I think are share worthy.
First, as a long term cancer survivor who's read lots of medical and research literature, I'm a medical observer and patient, not a practitioner. Please don't take any of this as medical advice, just my thoughts on the topic. Feel free to question me and to talk with your own oncologists.
During nearly 25 years with breast cancer, I've seen many promising ideas, treatments and approaches come and go. The media is filled with forward looking news releases about them, but reading the research, many of them are based on laboratory (in vitro) or animal studies and far too many don't develop into workable, achievable treatments/cures.
Even reports of clinical trial results are often over-inflated about positive results and don't reflect detrimental or long-term effects of treatment. Phase I and Phase II trials are often under-powered and lack statistical significance for PFS or OS (the studies that led to approval of Ibrance (palbociclib) are).
Yet, patients and those who care for us can be caught up in the hype and hope.
I think there may be an element of wishful thinking and the placebo effect* at work. If it's newer, it must be better, right? Maybe.
Some of the things I've seen come and go include radical mastectomies, removal of many underarm lymph nodes (I lost 35 in 1991) instead of sentinel nodes, Megace, DES**, high dose chemotherapy with stem cell transplant in breast cancer, Avastin which worked well for some but not all, radiation for bone mets*** and many other treatments that looked promising in studies and trials, but turned out to be worthless or more dangerous in practice.
I'm concerned that both Afinitor (everolimus) and Ibrance (palbociclib) may turn out to be more of the same.
The thing about PFS is that it doesn't always equal good quality of life. It measures the cancer, not the unwanted, possibly toxic, effects. These may worsen our quality of life in the short or long term.
Just because we achieve PFS or even NED (no evidence of disease), doesn't mean that the quality or length of our lives have improved.
Thanks for letting me think about this with you all.
Warm healing wishes, Stephanie
References:
* Here's a fascinating idea – a possible study on a study of open label placebos for cancer fatigue:
https://www.uab.edu/mix/stories/how-open-label-placebos-turn-fake-pills-into-real-treatment
** DES and lower dose estrogens in metastatic breast cancer
http://jama.jamanetwork.com/article.aspx?articleid=184425
*** On single dose rather than five doses of radiation for bone mets:
http://www.jhartfound.org/blog/adventures-in-choosing-wisely/?print=print
xox
Two more accepted practices being re-evaluated:
Lumpectomy leads to equal or better survival compared to mastectomy and reconstruction
Rethinking cancer screening (hundreds of articles and even a few books are devoted to the nuances of cancer screening)
xox
P.S. I've written more on the topic of cancer as a biological, ecological and evolutionary process at:
https://www.smartpatients.com/conversations/5140
(free registration required)
-
Thank you for sharing, Stephanie. I appreciate the honesty of the post.
Stefanie (spelled different 🙂
-
Well said, Stephanie! Thank you for sharing your experience with all this.
Being do in 2014, Ibrance was the first one I've seen--the next big thing, breakthrough drug--but not as successful for many of us (though still good to have another weapon, don't get me wrong, it's just that the hype seems to have been a bit overstated). It's hard not to get your hopes up that something new will be the thing we've all been looking for. But right now, having had no success with Ibrance or Afinitor, I'm putting my hopes in my generic Xeloda, hoping "tried and true" will be the next big thing for me for now
-
Stephanie,
I so appreciate your wisdom regarding quality of life in regards to treatment. That is my current struggle. I'm currently NED after a mastectomy, crainotomy, and GammaKnife radiation. I'm struggling to take Tamoxifen as it so terribly effects my quality of life. I see my onc tomorrow to discuss this and recent blood work. I recognize that I need to SOMETHING to (hopefully) remain NED but when the treatment is worse than the disease...I don't know what that "something" may be.
Healing thoughts,
Carrie
Categories
- All Categories
- 679 Advocacy and Fund-Raising
- 289 Advocacy
- 68 I've Donated to Breastcancer.org in honor of....
- Test
- 322 Walks, Runs and Fundraising Events for Breastcancer.org
- 5.6K Community Connections
- 282 Middle Age 40-60(ish) Years Old With Breast Cancer
- 53 Australians and New Zealanders Affected by Breast Cancer
- 208 Black Women or Men With Breast Cancer
- 684 Canadians Affected by Breast Cancer
- 1.5K Caring for Someone with Breast cancer
- 455 Caring for Someone with Stage IV or Mets
- 260 High Risk of Recurrence or Second Breast Cancer
- 22 International, Non-English Speakers With Breast Cancer
- 16 Latinas/Hispanics With Breast Cancer
- 189 LGBTQA+ With Breast Cancer
- 152 May Their Memory Live On
- 85 Member Matchup & Virtual Support Meetups
- 375 Members by Location
- 291 Older Than 60 Years Old With Breast Cancer
- 177 Singles With Breast Cancer
- 869 Young With Breast Cancer
- 50.4K Connecting With Others Who Have a Similar Diagnosis
- 204 Breast Cancer with Another Diagnosis or Comorbidity
- 4K DCIS (Ductal Carcinoma In Situ)
- 79 DCIS plus HER2-positive Microinvasion
- 529 Genetic Testing
- 2.2K HER2+ (Positive) Breast Cancer
- 1.5K IBC (Inflammatory Breast Cancer)
- 3.4K IDC (Invasive Ductal Carcinoma)
- 1.5K ILC (Invasive Lobular Carcinoma)
- 999 Just Diagnosed With a Recurrence or Metastasis
- 652 LCIS (Lobular Carcinoma In Situ)
- 193 Less Common Types of Breast Cancer
- 252 Male Breast Cancer
- 86 Mixed Type Breast Cancer
- 3.1K Not Diagnosed With a Recurrence or Metastases but Concerned
- 189 Palliative Therapy/Hospice Care
- 488 Second or Third Breast Cancer
- 1.2K Stage I Breast Cancer
- 313 Stage II Breast Cancer
- 3.8K Stage III Breast Cancer
- 2.5K Triple-Negative Breast Cancer
- 13.1K Day-to-Day Matters
- 132 All things COVID-19 or coronavirus
- 87 BCO Free-Cycle: Give or Trade Items Related to Breast Cancer
- 5.9K Clinical Trials, Research News, Podcasts, and Study Results
- 86 Coping with Holidays, Special Days and Anniversaries
- 828 Employment, Insurance, and Other Financial Issues
- 101 Family and Family Planning Matters
- Family Issues for Those Who Have Breast Cancer
- 26 Furry friends
- 1.8K Humor and Games
- 1.6K Mental Health: Because Cancer Doesn't Just Affect Your Breasts
- 706 Recipe Swap for Healthy Living
- 704 Recommend Your Resources
- 171 Sex & Relationship Matters
- 9 The Political Corner
- 874 Working on Your Fitness
- 4.5K Moving On & Finding Inspiration After Breast Cancer
- 394 Bonded by Breast Cancer
- 3.1K Life After Breast Cancer
- 806 Prayers and Spiritual Support
- 285 Who or What Inspires You?
- 28.7K Not Diagnosed But Concerned
- 1K Benign Breast Conditions
- 2.3K High Risk for Breast Cancer
- 18K Not Diagnosed But Worried
- 7.4K Waiting for Test Results
- 603 Site News and Announcements
- 560 Comments, Suggestions, Feature Requests
- 39 Mod Announcements, Breastcancer.org News, Blog Entries, Podcasts
- 4 Survey, Interview and Participant Requests: Need your Help!
- 61.9K Tests, Treatments & Side Effects
- 586 Alternative Medicine
- 255 Bone Health and Bone Loss
- 11.4K Breast Reconstruction
- 7.9K Chemotherapy - Before, During, and After
- 2.7K Complementary and Holistic Medicine and Treatment
- 775 Diagnosed and Waiting for Test Results
- 7.8K Hormonal Therapy - Before, During, and After
- 50 Immunotherapy - Before, During, and After
- 7.4K Just Diagnosed
- 1.4K Living Without Reconstruction After a Mastectomy
- 5.2K Lymphedema
- 3.6K Managing Side Effects of Breast Cancer and Its Treatment
- 591 Pain
- 3.9K Radiation Therapy - Before, During, and After
- 8.4K Surgery - Before, During, and After
- 109 Welcome to Breastcancer.org
- 98 Acknowledging and honoring our Community
- 11 Info & Resources for New Patients & Members From the Team