Topic: What would you do?

Hmmmm....how odd that they would ask for that?? I would think that they would be clammering for people to help out. I also live in an area where there are several large hospitals. If you want I would be happy to give them my information. Tell me how you got in touch with them, and I will give it a whirl and let you know what I find out.
Kimber

Well, sometimes I wonder what in the world has happened to my brain. Why didn't I think to call the local office in the first place? (slaps self upside the head. . .)

Anyway, they were very nice. There is a program in my area and I am waiting for the contact person to get back with me. I'll post anything that I find out. I'm dreaming a little here, but I'd love to do the training, work in the established program for a while, and then bring it back to my county, which is small but does have a community hospital. I'd love for this to be available outside the big centers. Well, we'll see.

Thanks so much for putting that article up, Kimber!

I don't know how the ACS is about patient-related services, but I was not very impressed by the way the ACS treated LCIS in their MRI screening paper last year. http://caonline.amcancersoc.org/cgi/content/full/57/2/75


Maybe its because it was written by a committee, but with regard to LCIS, I found it inconsistent, including inconsistent with the only journal article it cited for LCIS.


I am not a scientific researcher.

LCIS = WAIT.

Love it, Kimber!

Leaf, I agree with you. It sounds like the ACS is just like so many of us--they find LCIS to be a tad confusing! I wish they'd just admit it and say that more research is needed. One of my doctors (can't remember which one, now) told me that my lifetime risk would be about 30%. I'm not sure where the ACS is getting that 10-20% statistic. Also, this is the first place I've seen both LCIS and ALH described as lobular neoplasia. That's sort of interesting. I always thought LN was the "new and improved" term for LCIS, and that ALH was simply known as ALH. Live and learn, I guess.

I'm so sorry that they wouldn't let you volunteer for the study. They seem pretty rule-bound, don't they? From my dealings so far with the ACS on this patient navigator thing, it seems that they have a fairly hard time just answering a question. They have to hunt around for somebody who might know the answer, and they aren't always sure who that might be. I wish they were a little more organized.

Thank you so much for gathering all the information that you do. You're our LCIS Answer Chick! Your postings are always very helpful.

Have a great weekend, everybody.

Personally, I think it's  ridiculous to try and clump LCIS and ALH into one catagory (an "umbrella heading" as one study referred to lobular neoplasia) as they are two different stages along the bc spectrum much like DCIS, only the potential for invasiveness is less (normal--hyperplasia--atypical hyperplasia (ALH)--insitu bc (LCIS)---invasive bc (ILC).  ALH has an increased risk of 3x to 5x, where LCIS has double that (7x to 10x).  Even if you figure that times the lowest "base risk" of 5 to 6%, that is a risk range of 35 to 60%.  (leaf--so I wonder how that one place came up with a low of only 10% for you).  Bessie (and any studies I've seen) has said that ADH/ALH has a risk of about 20 to 25%; it would follow that LCIS probably has a risk of 40 to 50%.  (my onc figure it out as 36.6% for me, although I'm not quite sure how he came up with that number as I also have family history, but he said he didn't really know and honestly said he was estimating.  Of course I'm hoping it's lower, but realistically I think it's probably on the higher side. I'm praying that the tamoxifen does it's job and the fact they I've lost the ovaries doesn't help a great deal (per my oncologist since I wasn't under 40), but it sure can't hurt.

Anne 

Hi Peaches!

This first part will only be about mathematics and statistics. Its NOT about LCIS populations.

If you have a constant risk of breast cancer of 1% per year, the first year you have a risk of 1%, The 2nd year you have a risk of 1%, and the 3rd year you have a risk of 1%. So for the entire 3 years you have a risk of 1+1+1=3% for 3 years.

So, just using mathematics and nothing else, if your risk was constant and you didn't die from a car accident or anything else, over 10 years your risk would be 10%.

Now, most women with LCIS are diagnosed in their 40s or 50s. To make it simple, lets say you were diagnosed at age 50. The average lifespan of a woman in the US is something like 83 years old. Lets round it off to 80. That is probably fairly valid for LCIS because probably not a lot of people die as a consequence of LCIS. (Overall, about 65% of the women who DO get breast cancer die of something else.) Now, not everyone will live to be 83. About half of the women in the US will live to 83, and half won't. Most of the people who die before reaching 83 will die of cardiac disease, if they're like the general population.

So if you have a risk of 1% per year, are diagnosed at age 50, and live to age 80, 80-50 = 30, or 30%.

But not everyone lives to be 80. So you have to take into account the people that die of something else. If you die of a heart attack, its hard to die of breast cancer. Or at least if you die of a heart attack, you can't later die of breast cancer. So the risk might not be 30%. Maybe its 25%, (I'm just guessing, I don't have the actuarial tables, and even if I did I wouldn't know how to calculate it correctly.)

Also, if you get diagnosed initially with LCIS at age 50, clearly it is impossible for you to get initially diagnosed with LCIS at age 30. So I think you subtract some risk due to that. You can't get breast cancer from a known LCIS diagnosis at age 35 if you are initially diagnosed with LCIS at age 50. (I don't know exactly how to mathematically account for this though.)

If a chart or doctor says your risk is 7 times higher than the normal risk , then that does NOT NOT NOT NOT mean that your risk is 10 x 7 = 70%.

The risk of the AVERAGE woman in the US is about 10-12%. That hypothetical average woman in the US has some risk factors.

You multiply your LCIS risk (say its 7x) by the risk of breast cancer *for a person who has NO OBVIOUS risk factors (besides being a woman)*, which is about 3% (per Dr. Love's Breast book p. 143), taken from an article in Cancer, 1988, 62:1695, according to the citation in her book. So this would be about 7 x 3 =21% (assuming these numbers are accurate.)

Now, even if you do statistics, you must realize that different studies differ and give you different values. So it may not be 7x, or 5x, or 0.5% per year, or 1% per year.

When I stick my numbers into the Hall calculator, I, too, get a value as high as 83%. (I can't fit my numbers in the calculator exactly because I took hormones for a time, then stopped, then started again, and the calculator doesn't allow for that.)
I have been told I have dense, but not extremely dense breasts.



So far, this discussion could just as soon be the chance of me picking a green jelly bean in a jar as opposed to a red jelly bean.

*******
OK, so now I'm going to add some of the real mysteries of breast cancer, and the even more mysterious LCIS.

First, one problem that studies can have is that since there are few LCIS women, there may be differences between how YOUR pathologist diagnosed your LCIS and how my pathologist diagnosed MY LCIS. They may use different yardsticks.


Now, one of the really important things I think the Port et al study did, was that its the only study that I have come across that actually gave specific data on the INDIVIDUALS with LCIS who got breast cancer. They looked at 252 LCIS patients and 126 atypical (ALH and/or ADH) patients over 6 years.


14 patients got breast cancer in that time (discovered by whatever means). 3 of these 14 had atypia, and 11 had LCIS. So if you take the LCIS patients, just plugging in the numbers, 11 / 252 = 0.043. This means that each LCIS woman had a 0.043 chance of getting breast cancer over 6 years, or 0.043 x 100 = 4.3%. So for each year, that would be about 4.3/6 = about 0.7% per year. (One LCIS patient got DCIS in one breast, and a 1.2 cm tumor in the other, and another LCIS patient got 1 tumor in each breast also, so I don't know how you want to count that. I just counted each person as 1 person in the 11 above.) None of these women with LCIS (or atypia) who got breast cancer got anything worse than stage II.

As a whole, 2 of these LCIS women who got breast cancer had a first degree relative with breast cancer, one person has an 'unknown' status (I assume that person was adopted or lost their parents or something), and the rest did NOT have a first degree relative with breast cancer. So 2 or 3 out of the 11 women with LCIS who got breast cancer had a stronger family history. (I'm sure that 2 or 3 out of 11 is not statistically different from 10-20% - the rate of breast cancers that are genetically linked in the general population of women who get breast cancer.) 4 out of the 11 LCIS patients took tamoxifen or raloxifene.

OK, so that about 0.7% fits with the 0.5-1% per year figure. It might be higher if you count each individual breast cancer. But I'm just doing it roughly and I don't know how to do it more accurately.


Undoubtedly, SOME of these 252 LCIS patients have dense breasts, a family history, or some of the other risk factors that you or I have. I have dense, but not super dense breasts, and you have super dense breasts. And, ballpark, lets say not more than 10 or 15 women with LCIS post here, and I don't know how many of them have dense breasts.

Now, clearly, the incidence of breast cancer in that LCIS group of 252 LCIS women was CLEARLY NOT 90%, even over a lifetime of 30 more years. So that's why its IMPORTANT that Dr. Hall cautioned that his calculator had NOT been compared to populations. The population of women with LCIS and ANYTHING ELSE (risk factors or not) is puny. So we don't know. I've read papers that Guesstimate the incidence of LCIS without anything worse in the general population at age 50 might be something on the order of 1:10,000 women-years. This is a real rough estimate, because they don't know how many women have LCIS and don't know it because for most women, LCIS is an incidental finding on a biopsy, and not everyone has breast biopsies, and even if they did they can't biopsy ones' entire breast. (They might miss the area with LCIS.)

Certainly, to my knowledge, there have been NO studies that have looked at LCIS patients with dense breasts, to see if they really have a higher risk than LCIS patients without dense breasts.

So this means that we don't know if you have LCIS *AND* dense breasts, whether the dense breasts risk factor ADDS any risk to the LCIS risk. It may, but it may not. Its hard to tell without knowing how many of the LCIS patients had these individual risk factors. This study was designed to find out if MRI screening was of use for LCIS women, and not to estimate the incidence of breast cancer in LCIS women with or without other risk factors. But *IF* half of the 252 LCIS patients had dense breasts, and their risk for breast cancer was 90%, then I'm guessing that the yearly incidence for breast cancer would be higher that 0.7% per year. So I think its less probable that common factors immensely increase one's risk beyond the LCIS risk that one already has. But that's just a guess.

There probably will not be a study in my lifetime because there are too few LCIS women, and, frankly, as a whole on average, their risk for breast cancer is lower than that of BRCA women.

****************

OK, so we don't know if risk factors are additive, even if we assume there is one number you can assign to each risk.

There is an additional problem with these risk calculators.

Now, Dr. Hall admits that his calculator has NOT been peer reviewed or compared to populations. He's honest. Maybe its valid for Ms. Average Woman in the US who has dense breasts, no biopsies, and has been on birth control pills for 10 years.

The top part of Dr. Hall's calculator mimicks the modified Gail breast cancer risk model, which was named after the scientist Gail, who wrote a paper about it which was peer reviewed.

Now the modified Gail model posted on the NCI website http://www.cancer.gov/bcrisktool/
specifically excludes LCIS (because they don't have enough data about LCIS.)

OK. But lets just PRETEND that the Gail model DID hold true for LCIS patients. There is an additional problem. Now, unless you have BRCA or at least a VERY STRONG family history, I think the Gail model is about the only model we have. I did find a reference to another model, I haven't heard that model mentioned anywhere else. I don't know if this other model is only for genetically-linked breast cancers.

One study looked at how well the Gail model worked, and whether other risk factors, such as breast density, would do a better job of prediction. They looked at how many women got breast cancer in Florence, Italy. They found the Gail model actually did a pretty fine job of predicting how many women in the population got breast cancer. They also tried an Italian model that added more factors like breast density. The Gail model predicted there would be 180 women with breast cancer, the Italian model predicted there would be 186 women, and the actual figure of women who got breast cancer was 194. Excellent. They will know how much chemo to order, how many radiation machines they need, etc.


HOWEVER, there is ANOTHER aspect to assessing a model. You not only want to know how accurate it is for the population, you also want to know how accurate it is for an individual. This is called the concordance statistic.

You and I are not so terribly interested if there are 1589 women in your state with LCIS or 743, but (I assume) we are QUITE interested what is the risk of breast cancer for US as an individual. (Or I assume so.) In other words, ASSUMING that we, with LCIS, have a 25% lifetime risk is our risk (and we don't know that), how confident can we be that YOUR lifetime risk for breast cancer is 25% and MY lifetime risk for breast cancer is 25%? How good are these models at predicting breast cancer for an INDIVIDUAL?

What they do is they separate their subjects into two groups. One group is the women who were diagnosed with breast cancer, and the other group is the women who were not. They randomly take 1 woman from each group, and calculate their Gail model (or Italian model) score. They see which individual woman had a higher score.

Now if you had the perfect prediction model, that was 100% accurate all of the time, then on every sampling, the woman with breast cancer would get a higher rating than the woman without breast cancer. If your model was only as good as the toss of a dice, then 50% of the time the woman with breast cancer would have a higher score, and 50% of the time, the woman without breast cancer would have a higher score. The better model has the higher score, and the scores range between 0.5 (= 50% if its as good as a toss of the dice) and 1 (=100%, if the model is a perfect predictor.)

Well, it ended up that both the Gail and Italian models had a concordance value of 0.59. This means that 59% of the time the Gail or Italian models gave the woman with breast cancer a higher value, but it also means that 41% of the time, the Gail or Italian models gave the woman WITHOUT breast cancer a higher score.

This means that these models are only slightly better than the roll of a dice for predicting which INDIVIDUAL woman will get breast cancer. We simply don't know.

That's even IF the Gail model worked, which it doesn't for LCIS (according to the NCI website.)

So I think the best information we have is information like the Port, et al study, which gave about a 0.7% chance of getting breast cancer/year. (maybe more if you want to count individual breast cancers.)

Now one professional subscription service which I have access, says (of LCIS) the absolute risk of developing invasive carcinoma after biopsy alone in multiple studies with long term followup ranges from 3.3 to 37%, and the risk of LCIS is 7-18 times higher than in the general population. (Note in the Port study they counted DCIS as cancer, so the definitions are different.) It also opines that efforts to predict which LCIS women are more at risk have been unsuccessful. In one report, they thought the women who had LCIS and a bad family history had a higher risk, but on re-analysis they found a positive family history increased the risk beyond that of LCIS ONLY FOR women who were under age 40 at diagnosis. So since most LCIS is diagnosed in one's 40s or 50s, for most women, family history will NOT be a predictive factor, at least according to this physician's opinion.

And, yup, this major institution I got my consult at in July 2007 is an NCI designated cancer center. They said, when I pushed them, that surely my risk was between 10 and 60%, and if I really pushed them for a number, it was close to 10%. (I do take tamoxifen, and I don't know if they were counting that too.) And, yup, when you stick my numbers in the modified Gail model (ignoring my LCIS - I do have ALH and I am including that) - I get about 22%. Go figure. I wish they'd just say they don't know.

Any questions, please feel free to ask. I'm NOT a statistics expert, but I'll answer what I can.

Having been through many years of biopsies, close surveillance, and numerous pathology results...I looked at the LCIS diagnosis as a question of how well onecould deal with uncertainty. There is ultimately no source or model to predict invasive cancer risk with this diagnosis with accuracy, probably because there are many discrete changes in cells that can lead to " cancer" as we usually understand it. And those changes probably follow more than one single path, with variations along the way.

My advice is to have comprehensive evaluation with several oncologists, to get their estimates of your risk. A Gail model isn't much help. I got 3 separate evaluations and when they all agreed , I decided to accept the number that fell out. They based their assessments on LCIS, all the other pathologies, my breast density, and the speed at which imaging on MRI seemed to change from scan to scan on my breasts. I also participated in a surveillance study that took regular cell samples.

After more than 12 years I did ultimately have mastectomies. The decision was based on MRI scans that had widespread features of invasive disease, and my risk was so high by that point it seemed the right thing. I doubt I would have opted for the surgery at an earlier point.Counselling was helpful, though I was not someone who was thinking about LCIS all the time....just right before imaging or a biopsy. If the diagnosis is really affecting your life when you do not have invasive disease, it seems any move you can make to free yourself up to enjoy your life fully might be a good idea.

I hope this helps in some way.
Moogie

You're quite welcome. But please do realize I have a lot of self-interest in this too. For me, learning is power. My emotions do not always follow my head, however.

Its very hard to get a 'big picture' on LCIS and breast cancer. When I first got diagnosed with LCIS, I thought I'd never meet anyone with LCIS. I am so happy I have women like you who have helped me and shared your experience, humanity, and knowledge with me. Thank you again.