There are many partially answered questions about the best ways to do things in breast oncology.  Several questions had great answers provided at this year’s San Antonio Breast Cancer Symposium (SABCS). Below is the newest information on how long to use adjuvant Herceptin (trastuzumab) and whether you can rely on a sentinel node biopsy after chemotherapy.

How long should a woman take trastuzumab (Heceptin)?

Trastuzumab, also known by the brand name, Hereptin, is the “poster child” for the future of targeted, individualized cancer treatment. 

About one-fifth of breast cancers have an overactive HER2 gene that makes the cancer more aggressive.  Trastuzumab literally reverses the effects of overactive HER2.  I’ve had patients with spread of cancer to their lungs and other places who’ve had what can only be described as a miraculous response to trastuzumab. 

More recently it has been shown that, if a woman’s breast cancer is tested and found to overexpress the HER2 gene, she will have better survival if she receives a year of trastuzumab before she has any evidence of spread of cancer.  This treatment before any evidence of metastasis is called “neoadjuvant treatment.”

The question has been whether one year of neoadjuvant trastuzumab is optimal.

Parallel tests of more or less traztizumab.

Two, back-to-back reports at the SABCS answered this question.  One study looked at whether 6 months of trastuzumab was not inferior to 1 year of treatment – a so-called “non-inferiority” test.  That’s’ a backwards way to ask if 6 months is as good as 12 months.  The researchers could not show that 6 months was as good as 12 months.

The other report looked at 2 years compared to 1 year of traztuzamab.  Many oncologists had assumed that, if one year worked, 2 years would be better.  It did not happen that way.  There was no improvement with the second year of trastuzumab. 

Answer:  One year of trastuzumab is just right.

Does sentinel lymph node biopsy work after chemotherapy?

Whether or not cancer has spread to nodes is a good guide in determining the best course of treatment to recommend.  For example, cancer in the nodes is a strong reason to consider chemotherapy, especially in younger women.

Using sentinel lymph node biopsy, also called SLNB, has helped doctors learn whether cancer is in nodes by evaluating only one or two or at most a few nodes.  This, in turn, has helped reduce the number of lymph nodes that are removed for the treatment of most women with early breast cancer.   The benefit of removing fewer nodes is a reduction in side effects of surgery such as lymphedema.  

Sometimes cancer is more advanced – possibly involving the nodes already – and it is desirable to give chemotherapy before surgery in order to shrink the cancer.  This can reduce the amount of tissue that must be removed from the breast, and thus improve cosmetic results.

The question has remained whether it is possible to rely on SLNB if the patient had chemotherapy before surgery.  It turns out there are three answers depending on what was done before chemotherapy began.  

Sentinel node when there has not been a node biopsy.

The easy situation is when a woman only had a biopsy of her breast to confirm that cancer was present, and then had chemotherapy without any node biopsy.  In this situation, SLNB is a reliable way to determine if there is cancer remaining in the nodes after the chemotherapy.  This has been known.  What has not been known is whether SLNB is reliable if a biopsy of a node was done to prove there was cancer in a node before chemotherapy was begun. 

Node surgery before chemotherapy can make sentinel node biopsy less reliable.

At SABCS, another pair of back-to-back presentations demonstrated that reliability of SLNB depends on the type of node biopsy that was done before chemotherapy.

If a SLNB was done as a surgical procedure – with isotope or blue dye node tracking, anesthesia, and a skin incision – then an attempt at a second SLNB after chemotherapy is not very reliable.  It finds the sentinel node only 61 percent of the time, and 52 percent of the time it was negative when another, different node had cancer – a so-called “false negative result.”  This is probably because the surgical SLNB scars the normal lymphatic channels that are needed for SLNB to work.

In contrast, if only a needle biopsy was done of a node before the chemotherapy, the results of SLNB are much more reliable.  There is a false negative result of only 13 percent.  If the surgeon uses two different ways to find the sentinel node, this false negative rate drops to 11 percent.  This is probably because using a needle to biopsy a node is much less disruptive of the lymphatic vessels.

In practice, this means that if the oncologist wants to know if a node is positive before administering chemotherapy, the pre-chemotherapy biopsy should be done with a needle rather than an open surgical biopsy.  That way, SLNB – with its benefits of less surgery and fewer side effects – can be used to assess the patient after the chemotherapy is complete. 

Answer:  Sentinel lymph node biopsy is reliable after chemotherapy if the nodes were not surgically biopsied before chemotherapy.   Remember, this information applies to the specific situation when chemotherapy is given before surgery to shrink a cancer and to improve the results of surgery.

More issues from SABCS in the next Perspectives on Women’s Health.


Hearing presentations at the San Antonio Breast Cancer Symposium (SABCS) is kind of like watching your friend’s children grow up in their annual Christmas cards.   An idea will show up one year as a single presentation.  If it’s a good idea, there will be a few more papers the next year, and the following year, there will be a flood.  We’ve seen this with BRCA1 and 2 breast cancer genes, sentinel node surgery, trastuzumab (Herceptin), and efforts to map how genes cause cancer.  As ideas mature, they are applied to help patients. 

BPA arrived on the podium this year 

In 2011, my colleagues and I presented our work looking at the adverse effects of BPA (bisphenol-A) on normal human breast cells.  Ours was the only paper I had ever seen at the SABCS on any xenoestrogen. 

This year, the keynote speaker who opened the entire meeting (Dr. Cathrin Brisken from Switzerland) began her talk discussing how estrogens and progesterone contribute to cancer.  Then, midway through her talk, she changed her focus and spent the rest of her time saying that her own mice experiments have convinced her that BPA exposure can contribute to cancer development.  She was especially concerned about lifelong effects after BPA exposure of the fetus during pregnancy.

The next day, in a separate session on chemical modifications that alter how genes work (so called epigenetic modifications), the speaker (Dr. Tim Huang from Texas) reported BPA could cause ongoing changes in gene function without mutating the actual genes in cancer cells.  

It is exciting that the oncology community is realizing that real prevention of breast cancer is part of our job.  

Now, on to other topics. 

10 years of tamoxifen treatment is a little better than 5 years 

Since a 1991 report that more than 5 years of tamoxifen treatment might harm women, physicians in the United States have prescribed tamoxifen for only 5 years.  The British, however, were skeptical and did their own trial of 10 versus 5 years of tamoxifen.  Their study found that 10 years of tamoxifen treatment increased survival – a little bit. 

The effects of tamoxifen persist after the drug is discontinued

The interesting thing is that taking tamoxifen for 10 years did not have an effect while women took the drug between year 5 and year 10.  It only had an effect between year 10 and year 15.  This is puzzling until you realize that 5 years of tamoxifen has carryover effects, which is to say that the benefits of tamoxifen last 5 years after the drug is discontinued – whether or not a woman is still taking the drug.  Stated another way, taking tamoxifen between year 5 and year 10 did not have an effect because the patient already has the carryover effect, from the first 5 years, helping between year 5 and year 10.  You can’t make good be perfect. 

The reason 10 years of tamoxifen helps is that it has its own carryover effects that continue after treatment is stopped at 10 years.   Specifically, tamoxifen seems to have a benefit that lasts 5 years after it is stopped, whether it is stopped at 5 or 10 years.  

How big is the benefit from 10 years of tamoxifen?

This is fascinating biology, but the benefit is not huge.  On average, there is a 3% gain in survival for 10 years of tamoxifen versus 5 years.  It is important to understand, however, that this is the average gain obtained by combining results for high-risk and low-risk women.  For a woman with high risk – for example, a big tumor or multiple positive nodes – the benefit would be larger than average because her risk was larger to begin with.  For a woman who had a small tumor with negative nodes, the average benefit will be less than 3%.

What is the “cost” of tamoxifen? 

In the tamoxifen trial, women who took tamoxifen for 5 years had a 0.2 % chance of dying from cancer of the uterus caused by the tamoxifen itself.  For a woman who took tamoxifen for 10 years, this increased to 0.4 % chance of dying from cancer of the uterus caused by tamoxifen.   In both situations, a lot more women had to be treated for uterine cancer, but did not die from it.

These complications of tamoxifen occur independent of whether or not a woman gets a benefit.

Should women on tamoxifen take it for 10 years?

For a woman who had a high-risk cancer the English study supports keeping her on tamoxifen at least 7 years and probably 10 years.  [Or changing to an aromatase inhibitor if she has gone through menopause.]  

In contrast, when a woman has a lower risk, she has less to gain – but the same to lose from side effects – so she may decide not to take long term tamoxifen.  The risks of 10 years of tamoxifen may not be justified, because anything she gains might be offset by her chance of having an adverse side effect from the tamoxifen itself.

Other observations from SABCS to follow.


According to a new law in California, radiologists who read mammograms will soon need to inform patients if their breasts are dense as seen with the mammograms.  The reason is that breast cancer occurs more often in women with dense breasts than in women with low-density or fatty breasts.  What should you do if you get a letter like this? First, please don’t panic!

Dense breasts are the normal starting point.

Virtually all breasts start out dense.  We don’t usually do mammograms on young women, but when we do, they are invariably dense.  It is how breasts grow during adolescence.  For this reason, breast density is better understood as persistence of how breasts grew in the first place than as a change.   The reason is that, as women age, the gland tissue is often replaced with fat, even though the breasts can stay the same size.  Fat is less dense on mammograms, so breasts often become less dense after breastfeeding, after menopause, or after stopping hormone replacement therapy at menopause.  Whatever the cause, it is the low-density, fatty breast that has changed, not the other way around. Persistent breast density is literally the lack of change.

Cancer starts more easily where there is more gland tissue.

Density of breast tissue does not make the tissue bad – it’s just that dense breasts have more tissue that can go bad because of other factors.  When breast tissue is examined, there are some types of fibrocystic change that predict cancer, and some types of fibrocystic change that indicate no increase in risk. Years ago, we did a study to see if dense breasts had more of the kind of fibrocystic change that predicted cancer.  We found that dense breasts do not have more high-risk fibrocystic change.

So, you might ask, why is breast density related to breast cancer risk?  The answer is that breast cancer has to start in gland tissue, and when there is more gland tissue – for example, when gland tissue persists and makes the breast dense – there is more tissue where cancer can begin. We’ve actually known for a long time that cancer occurs where there is more breast gland tissue:  if a woman is lying down, there is much more tissue in the part of her breast toward her armpit, and that part toward the armpit is also where the most cancers occur.  So it is pretty logical that cancer would be more likely if breast gland tissue persists.  And similarly, there will be less cancer if there is less gland tissue in which cancer can begin, like in a fatty replaced breast. 

The risk is real, but it often decreases over time.

What is the risk that a woman with no family history of breast cancer and no history of any breast biopsy – two factors that independently predict higher risk of breast cancer – will develop breast cancer in the next five years?  This table shows the percent of women who develop breast cancer in the next five years, for women without a family history and who have not had a breast biopsy*.  Women with family history of breast cancer, or who have had a biopsy, have an increased risk even if they do not have dense breasts.  The most important thing to remember when looking at this table is that breast density tends to decrease over time, so a woman who has dense breasts at age 45 can have less dense breasts, and thus less risk, when she reaches age 65.

Understanding the risk with dense breasts.

To understand the effect of density from this table, compare the percent of women with fatty breasts who get breast cancer to the percent of women with very dense breasts who get cancer.

For example, the women who are age 60 to 64 with fatty breasts get cancer 0.7% of the time over five years, but the women age 60 to 64 with very dense breasts get cancer 2.7% over the same time.  This can be summarized two ways, and both are correct:  1. women with fatty breasts have about one-quarter the risk of women with very dense breasts (0.7 ÷ 2.7 = 0.26); or 2. women with dense breasts have about four times the risk of women with fatty breasts (2.7 ÷ 0.7 = 3.9). It all depends on which perspective you want to take. 

The table is also useful to compare relative risk and absolute risk.  If you look at the same women who are age 60 to 64, the absolute difference in risk is 2 percent (2.7 – 0.7 = 2 percent).  This is the largest difference for any age group in the table.  Thus, it is correct to say that the risk of cancer is “four times higher” with dense breasts, but the absolute difference is only 2 percent. This is a real difference – especially for the women who are affected – but emphasizing the “four times higher” relative risk description may cause more anxiety than fits the facts.

An important additional observation is that many women lose density over time so that a person who is at higher risk now, will tend to lose gland tissue (it gets replaced by fat) and thus experience reduced risk over time. 

Fatty (low-density) breasts are not risk free.

Breast cancer is more likely in dense breasts, but fatty, low-density breasts are not guaranteed to stay cancer free.  As the table shows, we still find cancer in fatty breasts, and it is important for everyone to be vigilant.  The best advice is to be careful, get your mammograms, and hopefully there will be improved ways to screen for cancer in women whose breasts remain dense.

* These percentages were published in 2008 in an article by Tice et al, “Using Clinical Factors and Mammograpic Density to Estimate Breast Cancer Risk: Development and Validation of a New Predictive Model.”  Annals of Internal Medicine, 2008; 148: 337-347, Table 5.  Percentages apply only to women who do not have the risk factors of family history and/ or a previous breast biopsy.


MRI is the most sensitive technology currently available to detect breast cancer.  However, in a situation akin to the proverbial million-dollar “hammer” where everything starts to look like a nail, there is a temptation to think of MRI as the solution to all breast diagnostic situations.  The result is an epidemic of radiology reports that suggest a breast MRI “if clinically indicated” without reference to specific clinical guidelines.

Screening MRI works for high-risk women

Breast MRI has a defined role in the evaluation of high-risk women.  When compared to mammograms and clinical breast exam, it can identify cancers that would otherwise be missed.  In a sample of recent screening studies, MRI identified additional cancers after a negative mammogram in the contralateral breast (the breast without cancer on the other side) of 2.9 to 5.4% of women with primary breast cancer at the time of diagnosis.  In studies where patients were followed with repeat MRIs over time, MRI detected breast cancer for 4% of women with history of chest radiation for treatment of other cancers (such as Hodgkin’s Disease), for 5.4% of women diagnosed with lobular carcinoma in situ, and for 1.1 to 1.4% of women with a genetic predisposition to breast cancer.  In a separate study of high-risk women, mammograms found as many cancers as MRI, but the MRI found the cancers at a smaller size and earlier stage.  [It is important to note that mammograms and MRI are complementary.  In each of these studies, mammograms found cancers that were missed by MRI.]

Groups of women at high-risk

Based on these observations, several organizations* have published guidelines for use of screening breast MRI for high-risk women.  The agreed high-risk groups for routine screening breast MRI include:

1.  Women predisposed to breast cancer because they carry a deleterious mutation in the “breast cancer gene” BRCA 1 or 2

2.  Women predisposed to breast cancer because they carry a deleterious mutation in another gene associated with several different types of cancer such as PTEN (Cowden’s Syndrome) or p53 (Li-Fraumeni Syndrome)

3.  Women with an estimated lifetime risk of breast cancer of 20 % or more by a standard risk model based on family history such as the Gail Model, etc. 

4.  Women with previous radiation exposure of the breasts as with treatment of Hodgkin’s disease or monitoring tuberculosis or scoliosis (particularly if multiple x-rays were done during adolescence).

Some organizations also include:

5. Women with a previous biopsy showing lobular carcinoma in situ or atypical ductal or lobular hyperplasia, but this is often redundant since these conditions are frequently associated with a 20% lifetime risk in the Gail model above.

Screening breast MRI can be falsely positive and falsely negative

MRI has not been recommended for general screening because of the high false positive rate.  Even in recent studies published in 2010 through 2012, MRI was reported as abnormal about five times more often than women actually had breast cancer.  MRI also misses cancer: in the screening studies that reported all cancers – including those found between screening tests – about five percent of cancers were diagnosed in the interval after a negative MRI.

Should women with dense breasts have MRIs?

Recently, California Governor Jerry Brown signed a bill requiring radiologists, “…to include in the summary of the written [mammogram] report that is sent to the patient a prescribed notice on breast density.”  The bill also states, “Nothing in this section shall be deemed to create a duty of care or other legal obligation beyond the duty to provide notice as set forth in this section.”  This creates the uncertain situation where women will receive this information, but as yet, there is no good way to advise women on what to do about it.

One possibility is to do screening breast MRIs for these women because they have dense breasts on mammograms, but this is an untested option.  In fact, I have found only one study reporting MRI in women with dense breasts, but in that study, all women had an additional factor – in addition to dense breasts on mammograms – that increased their risk, such as family history, etc.  Even in the 612 women in this study who had factors in addition to dense breasts that increased their risk, 43 biopsies based on MRI alone were needed to find 9 cancers.  If the other factors were not present, it is likely even more biopsies would have been needed to find this number of cancers which means that even more women would have had negative biopsies that did not help them. There is study underway in the Netherlands using breast density as a criteria for MRI screening, but there is no information from that study yet.  Given the significant false-positive rate of screening breast MRI – and the real possibility of unreliable reassurance from a falsely negative study – this is an untested option. 

This is a field with much active research, and we all hope for better answers as soon as possible.  For the present, screening MRI should be used in accordance with guidelines offered by various professional organizations.   All these guidelines are essentially the same and suggest screening MRI for women at high-risk.  There is no consensus whether mammographically dense breasts meet this criteria for women who do not have other risk factors.


*  Guidelines for screening breast MRI from various organizations:

American Cancer Society

American College of Radiology

American Society of Breast Surgeons

National Comprehensive Cancer Network You will need to create a log in for the NCCN Guidelines.




A group of doctors at Stanford Medical School has done us the favor of reviewing 237 research studies that looked for possible benefits of organic foods.  It’s a great article, but my first reaction concerns the title, “Are Organic Foods Safer or Healthier Than Conventional Alternatives?”  Right up front – in the title – these doctors assume that foods produced with fertilizers, ammonia, pesticides, herbicides, and antibiotics are “conventional.”  I’ll come back to this.

Did the doctors find that organic food is more nutritious?

Not much, if any.  Most research found the nutritional value of “conventional” food and organic food pretty similar for micronutrients such as calcium and vitamins.  There are a little more phosphorus and total phenols in organic foods, but it is difficult to say what that might mean.  There was also a strong suggestion that organic milk and organic chicken had more omega-3 fatty acids, but there was no demonstrated health benefit.

What is different about organic food?

Researchers identified no difference in the percent of foods with bacterial contamination.  There was even a suggestion that some types of organic food might have more bacteria than “conventional” foods.  However, “conventional” foods had a 50 percent higher chance (or as they state, organic food had a 33 percent lower chance) that the bacteria in the food would be resistant to 3 or more antibiotics.  

The number of people studied were too small to detect risks from these drug resistant bacteria, but many experts believe that use of antibiotics to make “conventional” animals grow faster has been a driver of the resistance to antibiotics found in many bacteria that commonly infect people.  [Antibiotics shift intestinal bacteria so animals absorb more nutrients from the intestine and gain weight faster.  A heavier animal earns more money for the farmer, etc.] 

More troubling, however, 38 percent of “conventional” foods contained pesticides, but – I am disappointed to note – so did 7 percent of organic foods. It seems that eating organic doesn’t totally protect one from pesticides.

Nevertheless, in a couple of small studies, children who ate organic food had less pesticide residue in their urine tests.  The authors commented that “…organic fruits and vegetables may significantly reduce pesticide exposure in children, [but the studies] were not designed to assess the link between the observed urinary pesticide levels and clinical harm.”  That doesn‘t make pesticides safe, it only means no one has looked.

What is conventional?

Reliance on pesticides, herbicides, and antibiotics is a new feature in American farming. My great grandfather and his neighbors raised corn, cows, chickens, and pigs without the presumed benefits of insecticides and antibiotics for his animals. DDT, the first widely used pesticide, was a military secret and unavailable to civilians, e.g. farmers, until late 1945. Erythromycin, a commonly used animal feed antibiotic, wasn’t invented until the early 1950’s.  What we call “organic” was actually conventional until relatively recently.

The doctors who wrote the article might argue that herbicides, pesticides, and antibiotics have been adopted so widely that they have become conventional as defined in the dictionary, “Developed, established, or approved by general usage; customary.”  But there is also a legal definition, “Based upon consent or agreement; contractual.” 

I’d like to ask, when did we consent or agree that chemical farming would be “conventional”? I don’t remember being consulted, but glyphosate, AKA Roundup, became the leading herbicide used in food production at the end of the 20th Century when I was old enough to have an opinion. 

[Just for reference, Monsanto sells the weed killer glyphosate under the name Roundup. They also sell seeds for genetically engineered soy plants that contain a gene that allows the soy plant to survive the herbicide.  Since the seed is “ready” for the herbicide, it is called “Roundup Ready.” The problem is that weeds are developing the same resistance and becoming “roundup ready” on their own – so Monsanto is working on stronger herbicides even as you read this.]

Why we eat may guide how we choose to eat.

If the purpose of food is only to shovel in nutrients, buying organic probably doesn’t make much difference.  If, however, you believe we are what we eat, then organic food exposes us and our children to fewer chemicals and fewer bacteria that are resistant to common antibiotics. 

Compared to much of the world, I believe that American food is pretty safe, but the prevailing sentiment seems to be that, if safety of chemicals in food hasn’t been looked at, then it isn’t a problem.  I worry that the safety of “conventional” food is a lot like the problems of BPA in our food – until you look really carefully, you don’t know if it’s safe – only that you don’t know.

We owe a big thanks to the Stanford doctors.  They’ve highlighted how much we still need to find out before we can accept many current farming methods as “conventional.”



This website is for educational purposes only. Your personal situation may vary.

Consult your physician before making your healthcare decisions.


Vitamin-D, that is…

Several weeks ago there was a newspaper report that a specific change (technically called a polymorphism) in the gene for the vitamin-D receptor (VDR) was almost twice as common in women at high risk for breast cancer who live in Marin County.  Without specifically stating it, the article implied that these women were less responsive to vitamin-D because they had a less effective receptor.  The authors then suggested that this might explain breast cancer in Marin and it might be a rationale to give high-risk women supplemental vitamin-D.  Interesting idea.  Where do we go with this?

For years it has been known that “sunshine vitamin-D” was a vitamin because it was necessary for our bodies to absorb calcium to make our bones healthy.  Without adequate vitamin-D, we will not absorb calcium. If this happens to a growing child, the child’s bones will be weak, and the child’s legs will grow with a curve from being bent under their own weight.  This is called rickets. 

More recently, it has been found that vitamin-D levels are lower in persons with breast and colon cancer when compared to persons without these cancers. It is not known, however, whether the lower level of vitamin-D caused the cancer, or if the low level of vitamin-D and cancer coexist by chance, or if they are both caused by some unknown, third factor.  The possibility of a relation is supported, but not proved, by research showing that vitamin-D has its own pathway by which it can control cell growth.  [For a discussion of how “pathways” work, see my post, “Whither breast cancer in 2013?”.]

In this context, it is interesting that Marin County women at higher risk of breast cancer are more likely to have an abnormal vitamin-D receptor gene (called A2/A2).  At first thought, one might assume this explains more breast cancer in Marin.  However, on further consideration, this inherited change in the VDR is not likely to explain breast cancer in Marin for two reasons. 

Marin County is a diverse group of persons.  It is difficult to believe that somehow a selected group of women with this specific mutation selectively moved to Marin.  It is more likely that any woman living anywhere with breast cancer would be more likely to have an abnormal VDR. Such a general association of vitamin-D receptor changes and breast cancer is supported by a 1999 study that found changes in the VDR in Australian women with breast cancer.

The other limitation is that the difference reported in Marin was found by comparing women believed to be at high risk by a theoretical model to women thought to be at low risk by the same model.  What the authors did not emphasize was that when they looked at women who actually had breast cancer 62 percent had the A2/A2 change in the VDR gene, but so did 58 percent of women without cancer. Not a big difference.

Despite these limitations, the possible role of vitamin-D and vitamin-D receptors is an exciting new area of research.  We just don’t know what to do with it yet. 

One might argue to put everyone on high doses of vitamin-D – just to be safe – but that is untested, and it might have adverse consequences.  Remember the excitement that so-called “antioxidants” such as vitamin-E were going to prevent cancer? Epidemiology and laboratory work showed how this was certain to work.  But when it was tested, vitamin-E not only did not decrease cancer, it may have increased lung cancer in smokers and prostate cancer.  High doses of vitamin-E did the opposite of what was hoped.

So how does one decide what to do about vitamin-D in 2012? 

Answering this question is made especially difficult because different studies gather information different ways so we can’t easily compare one study to another.  For example, one study found no relationship of vitamin-D level to developing breast cancer, but three-quarters of their subjects were vitamin-D deficient.  It is not easy to look for a benefit of vitamin-D if most of the subjects are deficient.  In another study, researchers found a benefit of very high levels compared to anything lower, but they did not evaluate mid range levels that might be safer (see below). 

Until we know more about using vitamin-D to prevent cancer, our best guide is research on osteoporosis.  Low levels of vitamin-D are associated with more fractures, and giving vitamin-D reduces the risk of fractures. 

The most practical way to address this is to start with a blood test to find out what your vitamin-D level is.  For example, when I first encountered research on vitamin-D and cancer in 2006, doctors were already recommending calcium with vitamin-D, but the researchers presented data that a lot of women were vitamin-D deficient.  I started testing whether my patients had adequate vitamin-D levels and a quarter of them were deficient.  My own blood level was the third from the lowest I have ever seen.  Why?  I’m indoors all day long, and I use sunscreen when I’m outside.

What vitamin-D level to aim for?  Less than 20 nanograms per milliliter of blood (< 20 ng/ml) is clearly too low.  Between 20 and 30 ng/ml used to be considered low – but not terrible – but now the lowest acceptable level is greater than 30 ng/ml.  How much higher is still debated.

A recent report in the New England Journal of Medicine observed fewer fractures when a woman started out with a vitamin-D level that was over 30 ng/ml.  There was a possible added benefit going from 30 to 60 ng/ml, but no benefit being over 60 ng/ml. 

Some doctors take a more-the-merrier attitude and push vitamin-D levels above 60 ng/ml, but there is not a lot known about what level of vitamin-D might be toxic.  For example, one study in 2010 suggested that levels above 50 ng/ml might be associated with increases in otherwise rare cancers (recall how vitamin-E increased some cancers).

For this reason, at this time – and for the foreseeable future – I suggest that the wise choice is to start by finding out your own vitamin-D blood level.  If it is over 40 ng/ml, that is probably sufficient.  If it is 30 to 40 ng/ml, I would try to boost that to at least 40 ng/ml.  If it is anything under 30 ng/ml, you definitely need to correct that.   Sunshine can help, but beware of sun damage to your skin.  Furthermore, sunscreen blocks the UV light that usually makes vitamin-D in our skin, so sunscreen makes it so that sunlight cannot always make our own vitamin-D.

If you take a vitamin-D supplement, remember it is a fat-soluble vitamin.  That means it must dissolve in fat (or oil) so it can be absorbed into your body with the fat. If you take a solid vitamin-D pill with water and there is no fat in your stomach – for example, on an empty stomach or after a non-fat meal – the vitamin-D is poorly absorbed.  If you are trying to boost a low vitamin-D blood level, I suggest using a gel cap in which the vitamin-D is already dissolved in a drop of oil and thus ready to be absorbed from your intestine (with the oil) to your blood where it can help you.

This is an area of active research, and there will be more information as current studies are completed.  For example, today I read a study from July 2012 Archives of Pediatrics and Adolescent Medicine that looked at stress fractures in adolescent young women who played high-impact sports.  The amount of calcium they had in their diet did not affect their rate of stress fractures, but those who had more vitamin-D in their diet had significantly fewer broken bones from stress fractures.  This seems to reinforce the importance of vitamin-D intake.

As I said, there is new information all the time. 

Please stay tuned.

This website is for educational purposes only. Your personal situation may vary. Consult your physician before making your healthcare decisions.


When I return from breast cancer meetings, I am often asked, “What’s new?”

Sometimes there is a spectacular new treatment.  At other times, there is new information that has not yet led to new treatments, but the information still raises hope because it means there are new ways to think about cancer that will help us design new and better treatments.

This year, the 11th International Congress on the Future of Breast Cancer was the second kind of meeting.  There were new ideas on the verge of becoming the basis of better treatments.

These ideas – like a lot of ideas – are not surprising when you think about them, but you ask yourself, “Why didn’t we think of that before?”

For example:

Cancer Dominoes

Genes are the units in chromosomes that carry the directions or codes for making proteins.  Each gene has the code for one protein.   When a specific gene is turned on, it makes the single protein that is specific for that gene.

What gets overlooked in the discussion is that these genes don’t act as one gene that starts (or stops) working and then you have cancer. 

The metabolism of cells is always controlled by a series of steps.  One gene makes one protein. That protein triggers another gene, and that second gene makes a second protein. That second protein triggers a third gene to make a third protein, and on it goes.  One step triggers another, and that triggers another, like a series of dominoes. 

The analogy of dominoes works pretty well, but it is important to go back and remember that it is not a direct “one gene triggers another gene” event.  In cell metabolism, an activated gene makes a protein, and it is this protein, made by the activated gene, that activates the next gene in the series.

A series of this kind of individual steps that works as a unit is called a “pathway”.

One of the “pathways” that got a lot of attention at the meeting was the PI3K/ AKT/ mTOR pathway.  In this pathway, the PI3K gene makes the PI3K protein that triggers the AKT gene to make the AKT protein which in turn triggers the mTOR gene.  Once mTOR is triggered, it triggers more genes that eventually lead to other specific genes that start cells growing, one of the major hallmarks of cancer.

A major difference between cell metabolism and the game dominoes is that in normal cells there are some genes that slow down the pathway.  In contrast, there is no easy way for a game domino to slow down a series of dominoes that are falling.

PTEN and p53 are two major genes that normally slow cells down. If either the PTEN or the p53 gene is damaged in a cell, that cell is more likely to become cancer because it has lost its “brakes” that should normally control that cell’s growth.


Doctors have known about metabolic pathways for a long time.  What is new is the realization that more than one pathway can keep a cancer growing. 

For example, it is often the case that the PI3K/ AKT/ mTOR pathway may not be the initial pathway in the cancer.  However, when drugs treat the pathway that started the cancer – and make that pathway inactive – instead of the cancer being stopped completely, the PI3K/ AKT/ mTOR pathway can take over and keep the cancer growing. 

Several researchers at the meeting compared this to Whack-a-Mole in the amusement park.  Push down one pathway, and another will pop up to take its place. 

This is a groundbreaking concept, and clinical scientists are working hard to figure out how to use this to treat cancer.  Instead of the old idea of targeting one pathway, the new idea is to figure out how to target two or perhaps even more pathways at the same time.

This is progress, not in the form of a specific new treatment, but rather in the form of a new way to plan therapy.  In the next few years there will be a lot of new information on how to use this. 

There is already a new plan to treat cancer that was responsive to hormone-based treatment but then stopped responding to treatment.  In the new two-drug approach, doctors treat two pathways.  One drug attacks the same initial hormone pathway with a different drug, while at the same time a different drug blocks the PI3K/ AKT/ mTOR pathway.  This is like hitting two Whack-a-Moles at the same time.

The Role of Environmental Chemicals

An important realization to me from this research is that the PI3K/ AKT/ mTOR pathway may not be a common pathway to cause cancer to take off, but it seems to be sufficient to keep cancer cells growing when the other pathways are blocked by treatment.  This is important because our research found that the environmental chemical bisphenol-A (BPA) activates the PI3K/ AKT/ mTOR pathway.

We are just beginning to understand how this affects treatment of women with cancer, but for now it seems prudent to avoid dietary BPA.


The San Francisco Chronicle and the New York Times recently reported research that taking calcium supplements might increase the risk of heart attack.  The most recent study in Germany and Switzerland  recorded calcium supplements used by 23,980 women without prior heart disease over an average time of 11 years.  Taking supplements of calcium alone, i.e. a calcium supplement but no other supplement, was associated with twice the rate of heart attacks.  However, there was only a slight and not significant increase in cardiovascular related death in women taking calcium supplements, and the risk was based on observation of the small subgroup of 256 women who took calcium supplements alone, less than 2 percent of the entire study group.

A limitation of this paper is that the authors gave no explanation for why their results are the opposite of results from the Iowa Women’s Health Study published in 2011.   The Iowa study followed 38,772 women from 1986 to 2008 and found that daily use of a calcium supplement was associated with a statistically significant 9 percent decrease in the chance of death from all causes. 

With conflicting results, what should a prudent person do in 2012? 

First, find out what you need.  The Institute of Medicine still recommends 1000 to 1200 milligrams of calcium each day.  You need it for proper function of muscles, your heart for instance, and you need it for bone health.  Remember that osteoporosis (weakening of your bones by chronic loss of calcium) contributes to hip fractures, and it is still true that 10 percent of persons, over 65 years of age, who have a hip fracture, die of a various complications within 6 months.

Second, figure out how much calcium you get already.  Most of us get some calcium in our regular diet.  You can estimate your calcium intake from the US Department of Agriculture website.  This website shows that the largest amount of calcium is in dairy products and foods prepared with dairy products.  Some vegetables have moderate amounts.  Look at what you eat in a typical day and add it up. It is likely that you get a lot of your needs met already, but unless you take milk by the glass, cheese by the ounce, or yogurt by the cup, every day, you are likely to fall short.

Third, makeup the difference between what you get and what you need.  No one knows for certain the best way to make up the difference between what you get and what you need.  For now, I suggest taking the lead from the reasons people speculate the calcium supplements might cause harm.  A calcium pill, taken with water, on an empty stomach, can cause a calcium spike in your blood.  Calcium in food does not cause a spike, so the logical thing is to mimic the calcium in food – take the supplement you need with food.

To put this all together, unless you eat dairy regularly, you may not to take in enough calcium.  If you need a calcium supplement, the best guess in 2012 is to take only what you need, and take it with food. 

Obviously, this is an important topic.  There is conflicting information, but there is also certain to be more information over the next few years. 


“The Food and Drug Administration said Friday that it lacks scientific information to ban the chemical bisphenol A from food containers.”

Wall Street Journal, March 31, 2012

“…FDA supports reasonable steps to reduce exposure of infants to BPA in the food supply.  …FDA will support changes in food can linings and manufacturing to replace BPA to minimize BPA levels…”
US Food and Drug Administration website, March 30, 2012.

While it is technically correct that the FDA did not ban bisphenol-A (BPA), the Wall Street Journal misled readers when it ignored the clear FDA statements that we need to reduce BPA exposure of our most vulnerable citizens.  This is a prototype example of the spin-versus-spin arguments that characterize most of the environmental chemical debate.  These arguments are not about truth.  They are about control of public perception.  It is time for willful efforts to skip the spin, outline the challenges, and answer the questions that affect the safety for our children and grandchildren. 

The BPA debate centers around three types of data:

1.  BPA promotes breast cancer in animals.  Animals exposed to BPA, either in utero or through nursing from dams fed BPA, develop breast epithelial abnormalities and are more susceptible to breast cancer induction by the carcinogen dimethylbenzanthracene (DMBA). Like natural estrogen, BPA stimulates growth of human breast cancer cells transplanted into mice.  These results parallel the effects of estrogens, which is not unexpected because BPA has been known to act as an estrogen since the 1930’s.  Note, the concern is not that BPA by itself causes cancer but rather that it is an estrogen that promotes cancer.

2.  In cell culture studies, BPA causes non-malignant human breast epithelial cells (obtained by fine needle aspiration from women volunteers) to grow faster and evade apoptosis (programmed cell death) a necessary step for radiation and drugs to kill tumor cells.  These changes are sometimes called hallmarks of cancer because they are prerequisites for cancer to occur.  BPA by itself will induce these changes.  It is especially relevant that these changes occur in non-malignant cells because these are the cells that must be changed for breast cancer to occur.

3.  Although most of us ingest BPA daily, it is metabolized quickly.  Recent studies have shown that BPA in canned soups and similar products causes a spike in serum BPA, followed by a rapid fall back to pre-ingestion levels.  Some critics conclude from this that BPA cannot be around long enough to cause harm.  The opposing argument has two parts:  First, clearance may not be functionally permanent.  Most clearance is by conjugation (or chemical attachment) to glucuronate.  This is a common way the body uses to clear hormones, chemicals, and drugs from the blood.  However, the enzyme glucuronidase can free BPA from glucuronate, and biologically significant glucuronidase is present in many tissues.  In the case of the breast, biologically active glucuronidase is present in mother’s milk and contributes to neonatal jaundice by deconjugating the baby’s bilirubin.  It is difficult to expect that the BPA known to be in human breast milk would not be activated the same way.

Second, it is argued that BPA rises and falls too quickly to cause harm.  However, endocrinologists have known for years that reproduction responds to spikes in hormone levels as much as to constantly elevated hormone levels.  It may be that spikes, such as from a pre-adolescent eating BPA-rich food, would have a greater rather than a lesser effect.

These studies to not imply that BPA by itself causes breast cancer, but rather the estrogenic activity of BPA causes changes that promote breast cancer.  The cancer promoting properties of estrogens are the reason millions of women discontinued menopausal hormone replacement therapy after the Women’s Health Initiative showed that combination estrogen plus progestin therapy increased breast cancer.  Unfortunately, women living in the United States cannot decide to discontinue their exposure to BPA.  BPA is present in food packaging, house dust, cash register receipts, etc. such that 95 percent of Americans and 90 percent of Canadians test positive for BPA in their urine.

Thirty years ago, Lippman and Bolan recognized the importance of cell culture “…in the analysis of oestrogen action in non-malignant tissue…” Working with non-malignant cells, we have catalogued six different major cell tests that are altered by BPA in ways that promote hallmark, cancer-like behavior in previously benign cells.  The irony in this battle of spin is that if we had a drug that turned off cancer cells in culture as much as BPA turns them on, if we had data showing a mechanism of action like we have for BPA, and if we had the animal exposure data that already exists for BPA, the hypothetical drug would be lined up for clinical trials.  Unfortunately, all of us are unwittingly already in the clinical trial.

It is time for responsible parents and grandparents to demand an end to the battle of spin.  Calm minds must sort through the facts objectively and design and perform the studies still needed to resolve the questions.