Trish that tool isn’t for bilateral disease. There’s no way to plug it all in.
I’m all for trying to understand and I appreciate all I’ve learned here. I’m not uncommon but mixed type cancer which is better than straight IDC Er+ but I’ve read a study that it has a higher percent of 2nd primary. Hopefully years down the road when I’m dead from another cause. To have a 4 cm tumor with ductal and lobular features with one node a micro met - that sucker was in there growing for quite a while. For me, the only reason it wasn’t found earlier is because i was never screened!! I’m lucky and I hope fear of it appearing again will keep getting better with time.
Thanks for the thread. Interesting stuff.
Trishyla, there's not much I can add to what moth and runor have said. There is not a lot out there on situations like yours, and what is out there isn't clear and consistent.
As moth said, my understanding is that when there are multiple tumors, prognosis is based on the largest/most aggressive tumor. But there is concern that this might under-estimate risk. However adding together the tumor sizes appears to over-estimate risk. No easy answer. My question is the same as moth's - did your MO talk to you about your recurrence risk?
Review on Practical Approach in Multiple Breast Carcinomas: Does Each Focus Matter? https://www.revistachirurgia.ro/pdfs/2017-4-418.pdf
runor, no apologies necessary. I am neither choking on my coffee (my wine, actually; I just can't drink coffee after noon or it keeps me awake at night), nor am I rolling my eyes. Everything you wrote makes perfect sense to me. It's not all about stats. Clinical and pathological factors are very important. It's why I take Oncotype scores with a grain of salt and why I often suggest the people ask their MO to run the Oncotype RSPC model, which adjusts Oncotype recurrence risks by incorporating patient age and clinical/pathology factors.
**Trishyla, I just saw your latest post as I was finishing off this post. This might not make sense (and statisticians might roll their eyes), but can you use something like the PREDICT model and input your largest ER+/PR+ cancer, and look at the 5 year and 15 year results? If you subtract the 5 year results from the 15 year results, this might tell you what risk you face from your ER+ cancer from next year (year 5) through the next 10 years. Maybe ask your MO if something like this would be a reasonable way to assess your future risk?
Question on a tangential topic: what percent of women (with "Western" diets and ethnicity, as US and Canada) are diagnosed with breast cancer in their lifetime? I have long seen a number that was about 12%. But does that include women diagnosed with "Stage 0" forms such as DCIS and LCIS? Or only those with invasive forms?
Thanks in advance for any insight.
Great question MountainMia. I hope Beesie will chime in.
ha. I answered my own question:
Another thing I wondered about was the number of women diagnosed with non-invasive DCIS or LCIS. This answer is from another bco page:
"There are about 266,120 new cases of invasive breast cancer and 63,960 cases of non-invasive breast cancer this year in American women. " https://www.breastcancer.org/risk/factors/woman
Since this thread popped back up, I want to thank everyone who answered my query about how to calculate my risk. I think you're right, Beesie. I have to treat them as separate cancers, and run the numbers consecutively rather than concurrently.
So, I know some people will be grinding their teeth about this thread being bumped up but I was JUST sent this Medscape article and I have to say..... I was thinking that 30 was a big over-estimate but now I'm doubting myself on that.
Have we talked about this yet? I didn't see it on the thread but maybe I just missed it https://www.medscape.com/viewarticle/849644#vp_1
"The Mystery of a Common Breast Cancer Statistic"
"Dr Brawley worked with two ACS epidemiologists to examine the issue. They looked at breast-cancer-specific mortality (as identified on death certificates) in 12 health districts in the United States from 2008 to 2012. They were surprised by the finding: "28% of the women who died of breast cancer during that time period had localized disease at diagnosis," said Dr Brawley.
The result was unexpected. "We all thought 30% was too high," said Dr Brawley." (it's on page 3 of the Medscape article).
So.... that's a surprising turn of events for me. Maybe 30% is not wrong after all? I wish we actually had more numbers on this. It seems absurd that we don't.
Moth, you are right. I was told from my MO that my recurrence rate is 30% with 3mm cancer and 2 nodes positive. That's why I get worried when I have strange pains in strange places.
Moth, I think it's important to keep everyone informed! I was stage 1, node negative, oncotype score of 23 in 2008 (44 years old, no family history). At that time, that oncotype score was in the "gray area" to determine if there was a benefit from chemo. I did do chemo - weekly CMF chemo for 6 months, lumpectomy/radiation and 7 years of femara. I was told that all this treatment brought my risk to around 10% (or even less?) of having a recurrence yet - here I am, dx stage IV 11 years later (10 years after radiation was complete - almost to the day). Personally, I wonder if I had stayed on femara longer would I be here but who knows? At the time, I was told that my risk from bone loss/osteoporosis was greater than my risk of a breast cancer recurrence so they had me stop taking it after 7 years.
Yup, I'm grinding my teeth that this thread has yet again been resuscitated.
I don't think anyone has disagreed that the 30% stat used to be correct. But it's not correct anymore. That's the reason why this thread should be left to the archives.
What we have to remember when we look at mortality data is that there is a time lag. moth, the article you've linked is from 2015, and it looked at cause of mortality for women over the period of 2008 to 2012. So these are people who died during the period from 9 years ago to 13 years ago. These women died of breast cancer during this time, but they had started off with a localized diagnosis. So when were they first diagnosed and treated? Since the 5 year survival rate of those with a localized diagnosis is in the high 90% range, and since most people survive 2-3 years with mets, with some exceptions, the most recent diagnosis dates would be between 2001 to 2005, i.e. 16 to 20 years ago. Of course for many with early stage disease, particularly those with ER+/PR+ disease, the time to mets is longer, sometimes as long as 15 or even 20 years. And the length of survival after the diagnosis of mets is also longer for many patients. This means that some of those who died of metastatic breast cancer between the years of 2008 and 2012 might have been diagnosed as long ago as 1988, if not earlier. In other words, for the most part this study reflects what happened to those diagnosed with breast cancer somewhere between the years of 1988 and 2005.
Having been diagnosed in 2005, just from my observations I can say with 100% certainty that treatment has changed, and survival rates have improved since that time. Just for a start, there is Herceptin, there is the broad use of Tamoxifen for early stagers, there is approval of AIs for early stagers....
So let me again copy and paste the information I provided previously, which compares current rates of metastatic recurrence to the rates during earlier time periods, such as the time period covered by the study linked above:
From Examination of a paradox: recurrent metastatic breast cancer incidence decline without improved distant disease survival: 1990–2011:
"The observed fifty percent relative decline in distant breast cancer recurrence over time may be related to both improved treatment for initial disease at diagnosis decreasing recurrence risk (hormonal therapy, polychemotherapy, dose-dense chemotherapy, taxanes, and trastuzumab) and stage shift to more early and less late stage disease at diagnosis with improved screening technology and screening program participation. Distant recurrent disease incidence decline over time differed by phenotypic characteristics. We observed a 70% decline in distant disease recurrence among hormone receptor-positive patients and a 50% decline among hormone receptor-negative patients over time. We observed a 60% decline in distant disease recurrence among HER2 positive patients from 1999 to 2006. The differential decline associated with phenotypic subtype creates a new profile of recurrent metastatic breast cancer with fewer HR and HER2-positive cases and relatively more TNBC cases. Ten-year cumulative incidence comparisons to accommodate the longer interval to distant recurrence among HR-positive versus HR-negative disease did not significantly differ from 5-year rates." Source: Examination of a paradox: recurrent metastatic breast cancer incidence decline without improved distant disease survival: 1990–2011https://link.springer.com/article/10.1007/s10549-018-05090-y
From Differential presentation and survival of de novo and recurrent metastatic breast cancer over time: 1990–2010, here's a graph showing, over 3 time periods, the percent who have developed recurrent MBC, by stage at time of initial diagnosis:
So, maybe 30% is not wrong after all? No, it is wrong. Once upon a time, for those diagnosed about 20-30 years ago, the 30% figure was correct. But for those diagnosed over the past 5 years and those diagnosed today, it is wrong.
aprilgirl, the problem with stats is that even when the stats are favorable, there will be those who fall on the bad side of the odds. Your risk of mets at time of diagnosis was estimated to be 10%. The fact that you developed mets doesn't mean that this original estimate was wrong - and that maybe your risk really was more in the range of the 30%. It only means that you had the crappy luck of being one of the 10 out of 100 who developed mets, rather than one of the 90 of 100 who didn't develop mets. I'm sorry you developed mets despite having a favorable diagnosis and doing all the treatments you could. Unfortunately that's the risk that we all face, even when the odds are overwhelmingly in our favor. But the flip side of that is that there are no doubt some people reading this who are facing a very high risk of recurrence, and yet fortunately some of them will never recur. To repeat the common saying around here, it's a crap shoot.
Beesie - yes, I agree. One correction - I was told that my risk was 10% or less after completing treatment. I believe the risk without chemo was around 20%? Regardless, there are cases like mine, which does show that it is a bit of a crap shoot.
Thankfully, I am a good advocate and when I had a couple of odd symptoms for me , I pushed and pushed to get a CT (took me 2.5 months). I was repeatedly patted on the head and told I had a virus causing my "laryngytis" and other symptoms. I caught it fairly early, for stage IV and have so far responded well to treatment .
There is no way to predict if a stage 1 patient will be the 1 out of 10 to develop mets so I think it's important that all patients KNOW this and don't get complacent. I should not have had to fight to get a CT and have both my GP and Seattle Cancer Care Alliance dismiss my concerns because "your risk of recurrence is really low" and "insurance won't cover a petscan when it could be a virus" after I offered to pay out of pocket for the scan. Really frustrating.
Beesie, I was diagnosed less than 18 months ago and I was told from MO that my recurrence is 30% without letrozole and 15-20% with letrozole. I guess MO uses some other data when they come up with these numbers.
See I'm not convinced that 30 is wrong. Why did they get this result ". They looked at breast-cancer-specific mortality (as identified on death certificates) in 12 health districts in the United States from 2008 to 2012. They were surprised by the finding: "28% of the women who died of breast cancer during that time period had localized disease at diagnosis,"
This isn't 20-30 yrs ago so I don't get it.
aprilgirl, I couldn't agree more with your comment about those with low risk not becoming complacent. Everyone should understand that no matter how low their risk, it could happen to them. If they develop symptoms that could be suggestive of mets, get them checked out! My risk is low, but I remember going to my MO after I had a breast MRI that had an incidental finding of a spot on my spine. The description of the spot could have indicated something benign but it also could have described mets. My MO, who I hadn't seen in years, saw me very quickly but I could tell when I walked in that he wondered why I was there. But as soon as he read the report he became very serious. Fortunately it did turn out to be nothing serious. Yup, no one should think that low risk is no risk because that isn't the case at all.
LillyIsHere, did you have an Oncotype test done? Otherwise your MO might have used Adjuvant Online (not sure if that's the current name) or even PREDICT, which patients can access themselves. Have you used either PREDICT or CancerMath? Both are easy to use on-line models that estimate metastatic risk with and without various treatments
One thing I want to clarify. My beef is with the 30% figure as an average for all early-stagers, Stage IA to Stage IIIA. I simply do not believe that the 30% figure is accurate any longer, although I know when I joined here in 2005, the digging I did back then did seem to support the 30% figure. But a lot has changed in terms of breast cancer treatment over 15 years. However I am not saying that some early stagers might not have a risk that is 30% or even higher, and certainly those with more aggressive cancers may face risks in that range if they don't have chemo. My two issues with the 30% stat are that: 1) the 30% figure is outdated; I don't know what the current "average metastatic risk" would be for all early stagers, but I would guess probably closer to 20% than 30%; and 2) whatever the correct number is, an average risk figure that groups together such a huge population with such a wide range of diagnoses is completely meaningless for any one individual who wants to understand her own risk. This "average risk" figure is probably very useful for those planning breast cancer treatment resources and funding, and those assessing long term mortality trends, but it just doesn't apply to individuals.
moth, if someone diagnosed with early stage breast cancer died between 2008 and 2012 (between 9 and 13 years ago), when would she have been initially diagnosed and treated? How long is the time between diagnosis to development of mets to mortality? By my calculations (which are just my own estimates), I came up with the years between 1988 to 2005 for initial diagnosis (between 16 to 33 years ago). But even if everyone in the group passed away within 5 to 15 years of initial diagnosis, we would be looking at a group of people who were all diagnosed and initially treated between 1993 and 2007. My point about treatment advances and mortality reduction since those years doesn't change. The 30% figure was valid 10 years ago (the approx. time frame of the article) reflecting the treatment of those diagnosed 20 years ago. The 30% figure is not valid today for those diagnosed now, reflecting the treatment available today.
Thank you for your patience with us Beesie . In my case, I don't have an Oncotype number because it wasn't enough material for the test. The 3mm cancer tissue was sent out but came back as not enough tissue to be studied. However, MO at that time thought it was a low number of Oncotype because it is ILC. Yes, I tried PREDICT and it gave me amazing numbers, with or without letrozole it was the same extremely low recurrence. I wish I could go with PREDICT, I would have stopped letrozole immediately.
Thank you for all the info here and other forums.
I know I haven't spent as much time immersing myself in this literature as some of you, but I am a biostatistician by training.
When I first started getting up to speed ~4 months ago, I had a hard time reconciling all sorts of seemingly contradictory stats that appear in reputable, peer-reviewed venues. I feel like I'm starting to have partial success at making sense of it all. Of course some of the ambiguity will never go away, because the different studies all have different strengths/weaknesses and are looking at things from a different angle.
I think the thing Beesie points out is really significant here, which is ... we can't really say what the lifetime probability of distant recurrence is for someone diagnosed today. By definition. Because the relevant population hasn't experienced enough follow-up for us to quantify it.
But we can put a bound on it because I'm quite sure that "we" are getting better at breast cancer detection and treatment. Someone diagnosed today is likely to be diagnosed at an earlier stage and receive a treatment that is more effective re: preventing recurrence than someone diagnosed 5, 10, or 20, years ago.
The math Beesie does about the year of diagnosis for those dying in 2008 - 2012 makes sense to me: they represent a population initially diagnosed at least 8 years ago and, in most cases, many more than 8 year ago. So I interpret the 28% from death certificates as an upper bound on the distant recurrence probabiliy for those diagnosed today. And probably a very pessimistic upper bound, i.e. the true number is substantially lower now.
(Also that figure makes me twitch (although it's very interesting and I basically believe it,). The bar for those diagnosed as Stage 1 from 2005 - 2010 is clearly wrong. It reports a proportion of 98% and therefore should be higher than all other bars. Clearly some glitch happened in this publication, which is of course going on all the time in these papers.)
So we're putting all the benefit on early detection? Because tamoxifen and herceptin came on board just before 2000 so those benefits should be seen in those earlier 2008-12 numbers, shouldn't they?
Perjeta is too new to show results yet but it's the only new early stage thing isn't it? Someone asked me on Twitter what else is new that would make us think things are better now? Same surgery, rads and chemo for early stage (I can think of new mets tx but not early stage).
mostly I'm just so so so frustrated that recurrent mbc is not being properly counted. We're sifting tea leaves here looking for signs. It's ridiculous
sasamat, thanks for your input and for verifying that my thought process makes sense (at least a bit!).
moth, take a look at the 20 year survival stats from the SEER data. Note that the most recent 20 year data that is available is for those diagnosed in 1997. If you look at the 4 periods that I estimated to be within the range of the study you linked, you'll see the following 20 year survival rates:
1990-1994: 70.3% (75.3%)
1995: 71.9% (76.9%)
1996: 72.4% (77.4%)
1997: 74.6% (79.6%)
The SEER data includes all stages. Stage IV de novo consistently represents 5%-6% of all diagnoses. So the number in brackets represents the 20 year survival rate for all breast cancer patients diagnosed in that year, when you exclude the 5% who were Stage IV de novo. That doesn't quite match "early stage", since it includes Stages IIIA and IIIB. But just looking at these figures, what you can see is that if 30% of early stagers were to eventually develop mets (i.e. if the long term survival rate for early stagers is 70%), this means that even in 1990-1994, 5% of early stage patients who would eventually succumb were surviving longer than 20 years. By 1997, when the 20 year mortality rate was 79.6%, it means that almost 10% of those patients were still alive at 20 years. This is why I assumed that those who passed away between 2008 and 2012 were diagnosed as much as 20 years earlier. If anything, I was conservative in that estimate and should have assumed that those who passed away in 2008 might have been diagnosed as much as 30 years earlier.
moth, to your comment, "Because tamoxifen and herceptin came on board just before 2000 so those benefits should be seen in those earlier 2008-12 numbers, shouldn't they?", the answer is No.
Herceptin was approved for metastatic breast cancer in 1998. It was approved for node-positive non-metastatic cancer in 2006 and it was approved for early stage node negative after that. The same is true of AIs - they were barely in use and I believe not approved for early stage back when I was diagnosed in 2005. Even Tamoxifen was considered the new wonder drug for early stagers back in 2005 - it was approved but as I recall not all insurance companies paid for it yet. So if most of those who passed away between 2008 and 2012 were initially diagnosed in the early 2000s or prior to 2000, most would not have had access to any of these drugs. The other significant change would be 3rd generation chemos. So in fact there have been hugely significant treatment changes between the group discussed in the study you linked, and those diagnosed today.
Edited to correct the spelling of sasamat's name. Sorry for the error!
May I ask you how did you find your 3 mm lobular tumor? I mean which device was able to catch it? Did your biopsy take the whole tumor?
I posted this elsewhere so I know Beesie already saw it :-) but this article might be a valuable contribution to this thread. I personally find the 30% statistic to be plausible when considering lifetime metastatic recurrence risk.
The article (which is recent) states that "[a]lthough the incidence of distant relapse has been shown to be decreasing and survival times for patients with recurrent disease have improved, 20–30% of patients with early breast cancer still die of metastatic disease."
This article talks about many interesting things. One of which is that a very complicating factor in counting lifetime metastatic recurrence risk is that many HR+ cancers recur decades after initial diagnosis. DECADES! Personally, I would have loved to have decades before my HR+ metastatic recurrence, but that's another story. Hopefully for all those out there who recur later, better treatments and cures will be found by then.
Anyway, the article is an excellent read for those who want to delve into the why and how of metastatic recurrence. There is still a lot we do not know.
The Lingering Mysteries of Metastatic Recurrence
Could you please help me with some positive statistics regarding young women (30-40) and AI treatment? I read results from one study about benefit of the ovarian suppression+AI to prevent cancer reoccurrence. But they also mentioned that this treatment did not increase overall survival rate because of increased mortality from other issues. I assume heart problems were the major cause. I want to find something more optimistic.
Beesie and Sasamat thank you for the excellent information. To Moth's point, there is such a need to collect, evaluate, and disseminate data in fighting this disease.
Buttonsmarch - that is an interesting article. Thank you for posting.
buttonsmachine, yes, that is a very interesting article.
However to the quote "20–30% of patients with early breast cancer still die of metastatic disease.", here is what I replied in the other thread in which buttonsmachine posted this same information *** I suspect that the 20% figure is probably fairly accurate, if not low - but 20% is 33% less than 30%, which to me is a huge difference. Based on the number of people diagnosed every year with early stage breast cancer in North America, over 10 years the difference between 20% and 30% would be an additional 250,000 people who survive and never develop mets. And note that while the article is recent, the sources for the "between 20% and 30%" stat are from 2000 and 2005, which means that current treatments are not reflected in these figures.***
And for additional clarification, my discussion here (and elsewhere) about the 30% stat is talking about lifetime metastatic risk, not the 5 or 10 or 15 year risk that we see reflected in the Oncotype score and the PREDICT and CancerMath models. Note the SEER chart that I posted above, which shows 30 year mortality rates. Of course, you can only have 30 year mortality figures for those who were diagnosed more than 30 years ago. So to sasamat's point, there is simply no way to know the lifetime probability of distant recurrence for someone diagnosed today.
What we do know, as buttonsmachine points out, is that some recurrences don't happen until decades after the initial diagnosis. Take a look here (this is a chart that I have posted previously in this thread). It shows recurrence patterns/timelines for each major subtype of breast cancer.
It appears that almost all recurrences for HER2 positive disease happen within the first 6 years. 80% of recurrences of TN disease happen within 5 years. ER+/HER2- disease is the one that tends to have a larger percent of late recurrences, but even here, it appears that about 80% of recurrences happen within the first 10 years, and about 95% occur within 20 years. This information helps frame this discussion about lifetime recurrence risk. If we can get a handle on 10 year or 15 year recurrence rates (as the Oncotype and PREDICT and CancerMath models tell us), we can estimate lifetime recurrence risk from there.
All that said, we have to remember that these figures are nothing more than estimates, even when we are looking at recurrence risks based on our own diagnoses, as determined by Oncotype and PREDICT and CancerMath. As for the much disputed 30% stat (which I believe today would be closer to 20%), which is a generalized stat that's been calculated to reflect outcomes based on an average of hundreds of thousands of people with every possible "early stage" diagnosis, well, that is absolutely meaningless for any one individual. And that, I suppose, is the point of all my posts and why I hate it so much whenever this 30% stat is brought up again. Too often, that stat is misinterpreted to reflect individual risk, and in that context, it is misleading and irrelevant and too often creates unnecessary fear and angst.
MikaMika, I'm sorry but I can't help you on your question. It's not something that I have ever read up on.
Edited to correct typos (I was in a bit of a rush this morning).
Beesie, thank you as always. My feeling is that MO takes into account other conditions like in my case probably since I am PR-, small tumor that has spread to nodes.
Mika, 3mm was discovered after BMX biopsy. 2 positive nodes were found at that time as well.
Here are a couple of very simple stats to consider:
Yes, the vast majority of those who died were first diagnosed in a different year. How long ago? Ten years ago or less? Probably. Has the rate of new diagnoses changed a lot? Not much, about .3% a year, according to this.
So maybe the risk of death is somewhere between 15-20% across all invasive cases, from early to de novo diagnosis.
Stats in this comment are from this link. https://www.breastcancer.org/symptoms/understand_b...
'28% of deceased from breast cancer were early stage', isn't equal to 28% of early stagers go on to have a metastatic recurrence, is it? Aren't they comparing percentages from two different populations?
My dad's best friend is a very notable oncologist. I didn't get to use him since I live in another state but he helped me a lot through my treatment. When the BCI test came out he was very quick to tell me to have it done because most studies showed an accuracy rate far more than 99.9%. At the time I had it done (I came back high risk of recurrence, albeit the very low end, but low benefit from continuing Tamoxifen which is very uncommon.) he told me that the oncology community across the board feels the advances made in breast cancer treatment in the last 5 years are equal to the advances made in the 40 years before that. He also told me that most of them think they'll have a "cure" for breast cancer in my lifetime (I'm 57) so fingers crossed. And when I was first diagnosed he told me that I would probably outlive my peers because due to my cancer, I would receive much better medical care than most women with my physical stats. It's been quite a delight to pick his brain.