Two interesting studies published in early 2020 are relevant to our use of fasting during cancer treatment. Stefanie de Groot and her team of Dutch researchers reported in Nature Communications that in a randomized trial of 129 human epidermal growth factor receptor-2 (HER-2) negative breast cancer patients, about half of whom followed a fast-mimicking diet (FMD) for three days prior to and during chemotherapy, that the intervention significantly increased the likelihood of a complete or partial response to treatment, presumably because the caloric restriction had lowered IGF-1 levels.¹

In late April, Yiwei Tong’s team in China reported surprising, though somewhat confusing, information about insulin-like growth factor-1 (IGF-1) in HER-2 positive breast cancer patients.² Tong’s results contrasted what many of us would have predicted. It will be important to integrate their data into our understanding as we move forward.

De Groot’s work continued Valter Longo’s research on fast-mimicking diets. Longo has methodically researched the effects that fasting and diets that mimic fasting have on cancer treatment. His earliest efforts to research water fasting met with poor patient compliance so he developed a fast-mimicking diet (FMD)—a low-calorie, low-protein, plant-based, intermittent meal plan that triggers similar metabolic effects as fasting. We’ve encouraged patients to use Longo’s products or imitate his eating plan for a dozen years. An article by Jennifer Couzin published in Science in August 2008 was the first we heard of Longo and his theory.³ Couzin had written about the Raffaghello study published two months prior that suggested fasting mice were more tolerant of high-dose chemo than those on a regular diet.⁴ By the time her August article appeared, our patients had begun trying Longo’s fasting at home.

According to Longo and de Groot, healthy cells switch from a proliferative state to a maintenance and repair state during fasting, while malignant cells are unable to adapt to a nutrient scarce existence. Fasting deprives proliferating cancer cells of nutrients, growth and other factors, rendering them sensitive to cancer therapy and increasing their cell death.⁵

De Groot’s data came from a multicenter, open label, phase II trial, that enrolled 129 patients from February 2014 to January 2018. They were randomly assigned to continue their regular diet or follow Longo’s fast-mimicking diet for the three days prior to and during the first day of chemo.

Thirty of the patients received six cycles of FEC-T chemotherapy (5-fluorouracil, epirubicin, cyclophosphamide, and docetaxel), while the remaining 99 received eight cycles of AC-T (doxorubicin, cyclophosphamide, docetaxel). The control group, but not the FMD group, also received dexamethasone pretreatment to minimize nausea and vomiting.

While the FMD diet is easier than fasting, it is still a challenge. Of the 65 patients randomized to the FMD, 81.5% followed the diet during their first round of chemo, around half followed the diet through two rounds, but only 20% followed it all the way through treatment.

Radiologically complete or partial responses occurred about three times more often in the FMD group than the control group in both univariate (OR 2.886) and multivariate (OR 3.168) analyses. The number of patients who had stable or progressive disease as determined by x-ray was quite a bit lower in the FMD group (11.3%) than in the control group (26.9%). The more closely a patient followed the diet, the better their response.

Toxic reactions to the chemo did not differ between groups, but recall the FMD group were not given steroids to prevent side effects. In other words, fasting worked as well as steroids at reducing chemo-induced nausea.

These results de Groot reported were satisfying but not unexpected; they support Longo’s earlier work. Tong’s results, on the other hand, left me scratching my head.

Tong ran a retrospective study analyzing data from 679 Chinese breast cancer patients, who were positive for human epidermal growth factor receptor-2 (HER-2+) and had been treated in Shanghai, China, between 2012 and 2017. Of these patients, 209 also had metabolic syndrome MetS). Overweight was defined by body mass index (BMI) ≥ 24.0 kg/m2. This is the largest study to date looking at IGF-1 and HER-2-positive breast cancer.

Tong’s researchers tracked several key measures, in particular recurrence-free survival (RFS) and overall survival (OS). Insulin-like growth factor-1 (IGF-1) was used to classify participants into high or low IGF-1 sub-groups. The usual parameters of metabolic syndrome were followed, including basal metabolic index (BMI), fasting glucose, IGF-1, IGFBP-3, insulin, C- peptide, triglycerides, TC, HDL-C, and LDL-C to see if they changed disease outcome. Tumor size, node involvement, histological grade, hormone receptor status, proliferation index, HER2-enrichment intrinsic subtype, and anti-HER2 therapy were tracked as well, as these are clear prognostic factors for HER-2+ cancers.

The theory that Valter Longo and his collaborators have bandied about since 2008 is that fasting is beneficial because it lowers IGF-1. The fast-mimicking diet used in the de Groot study was developed by trial and error to keep IGF-1 as low as possible.

After a median follow up of three years, 52 women had disease recurrence. IGF-1 levels were not associated with recurrence-free survival (RFS, P = 0.620). That was unexpected.

However, when the women were divided into two subgroups based on whether they were normal or overweight using body mass index (BMI), everything changed. Dividing the group at a BMI ≥ 24.0 kg/m2 revealed a clear association between IGF-1 and recurrence free survival (RFS). For normal-weight women, high IGF-1 was associated with a superior four-years RFS (91.1% vs. 85.0%; HR 0.53) compared with women with a low IGF-1 level. In contrast, for the overweight women, high IGF-1 was associated with an impaired four-years RFS (88.3 vs. 95.7%, HR 3.20).

It wasn’t just recurrence-free survival that varied with weight. High IGF-1 levels were independently associated with better overall survival (OS) in the whole cohort (HR 0.26 P = 0.044) as well as in the non-overweight population (HR 0.15, 95% P = 0.005). High IGF-1 was protective in non-overweight patients but appeared to be bad news for the overweight. As our colleague Dr. Ian Biers so often reminds me, progression-free survival and recurrence-free survival do not predict overall survival. We often look to these recurrence measures as they are easier for researchers to obtain than overall survival data, and many people falsely assume the two are related. They are not. Thus, these associations between IGF-1 and superior OS should be underlined.

Treatment with ‘targeted therapy’ (trastuzumab aka Herceptin) non-significantly improved OS from 96.7% to 97.7% (P = 0.149). However, significantly better four-year OS was seen in the high IGF-1 group compared to the low IGF-1 group (99.2 vs. 95.8%, P = 0.044). One might surmise having high IGF-1 was a better bet than Herceptin for staying alive. Subgroup analysis showed a modest but insignificant interaction of IGF-1 and BMI in predicting OS (P for interaction = 0.054). High IGF-1 level was associated with improved OS in normal-weight patients (4-years OS 99.4 vs. 93.7%, P = 0.005; HR 0.15), but not in overweight ones (4-years OS 98.7 vs. 98.9%, P = 0.438; HR 2.51, 95% CI 0.23–27.63, P for interaction = 0.054).

What’s all this mean?

Recall that de Groot examined only HER-2 negative cancers while Tong looked at only HER-2 positive cancers. For the moment let us assume that these findings only apply to these respective HER-2 types. The de Groot study is relatively clear. Fasting for the few days prior to chemo, as has been suggested by Longo et al over the years, seems to do what they have predicted based on animal studies; it reduces side effects of the drugs and increases odds for a good long-term outcome.

The Tong results are more difficult to describe and far more difficult to explain. For overweight HER-2+ women, lower IGF-1 levels were associated with better outcomes. We should underline “overweight” in that sentence. Fasting, because it lowers IGF-1, might be helpful for overweight women. That doesn’t change anything. It was the normal-weight women’s results that blindsided us. High IGF-1 was associated with better outcomes in these women, both short term when measured as time to progression and long term when measured as overall survival.

These findings should outweigh our general recommendations regarding fasting made in the past. In HER-2+ BC we may even want to fine-tune diet recommendations based on BMI and IGF-1 levels. Our goal in HER-2+ normal weight women should now be to increase their IGF-1 levels. High animal protein diets raise IGF-1 levels while low animal protein diets are associated with decreased IGF-1. We should be extremely cautious with fasting normal-weight women and instead probably encourage high animal protein diets to increase IGF-1. While for the overweight women they should be encouraged toward caloric restriction, fasting, and a possibly a vegan diet. Obviously, we can no longer make decisions without knowing the BMI and IGF-1 levels of these patients. And of course, what I just said applies to HER-2 positive BC.

Least we think that these data impact only a small subgroup of patients, keep in mind that in the United States, 70% of adults and 37% of adolescents are overweight or obese.⁶

Although Tong’s conclusions stand in contrast to what many of us would have predicted, our basic understanding remains that insulin-like growth factor (IGF) is critical to the growth, development, and maintenance of many tissues in the human body.⁷ IGF-1 is especially important during neonatal and pubertal growth, and acts by simulating cell proliferation and interrupting programmed cell death.⁸ IGF-1 is of particular importance in breast tissue development.⁹ Binding of IGF-1 to its receptor (IGF-1R) stimulates activation of the phosphatidylinositol 3-kinase (PI3K) and mitogen activated protein kinase (MAPK) pathways that cause cell proliferation. The bottom line remains that IGF-1 signaling is involved in 87% of invasive breast cancers.

For several years we have relied on a hypothesis that crosstalk exists between the IGF-1 pathway, insulin, and the epidermal growth factor receptor family; increased IGF signaling should lead to progression of breast cancer, metastatic invasion, and promote resistance to therapies such as chemotherapy and radiotherapy.^10,11 Elevated insulin levels bind to certain IGF-1 receptors on breast cancer cells and stimulate proliferation. This has been our rationale for encouraging these patients to reduce excess carbohydrate consumption as this might lower insulin production. An increase in IGF should lead in turn to a decrease in breast cancer survival¹² and an increased all-cause mortality in HER-2+ patients. Or at least this is what we thought.

We are left with seemingly conflicting ideas; fasting, which lowers IGF-1, seems helpful for breast cancer patients in general but, lowering IGF-1 may worsen prognosis for women with HER-2+ breast cancer.

Although De Groot’s results seem to confirm that our general theories about IGF-1 and breast cancer are on target, Tong’s data hint that other growth inducers outweigh the effect of IGF-1, at least for HER-2 + breast cancers. Biology, it turns out, is more complicated than we thought, and there are a range of chemical factors in the body which drive cancer growth beside estrogen, insulin, and IG-1. Tong’s data suggest some other factor is at play, one associated with obesity.

We should note that this isn’t the first report to differentiate the effect of IGF-1in breast cancer based on BMI. In 2013, Catherine Duggan et al reported that increased IGF-1 levels were associated with an approximate two-fold greater risk of BC-specific mortality in participants with a BMI>25 kg/m², than in lean women. On the other hand, they also found that high serum levels of IGF-1 and the IGF-1/IGFBP-3 ratio were associated with increased risk of all-cause mortality in women with breast cancer.¹³ Duggan’s study participants were not limited by HER-2 status, and their findings suggest that a similar division by BMI might apply to broad range of breast cancer patients.

We must ask why obesity has such a profound impact on cancer. Tong et al offer no theory to explain their results. The answer to the question Tong’s data elicit is probably why obesity is so closely tied to cancer in general, a phenomenon we are all aware of but tend to not pay much attention to. Probably one of those elephants in the room things.

Having unexplained and curious phenomenon in the universe, mysteries, as we might term them, is a source of delight and wonderment for many. For some people, unexplained mysteries niggle at the back of our awareness, begging for explanation. Why does obesity change cancer outcomes so much and now in particular, why this peculiar IGF-1 and obesity relationship? Let me offer up my current thinking, with the caveat that I do possess a BS degree and that by the time this is in print, the theories will have changed.

Obesity is the elephant in the cancer room; only smoking carries a greater risk of cancer. Yet surprisingly little work elucidates the actual mechanisms by which obesity promotes cancer. Perhaps this is a result of our efforts to be socially proper and not notice or make a fuss if someone is overweight? On a tissue level obesity causes a chronic inflammatory state resembling chronic injury¹⁴; but at this point, a better explanation seems necessary.

The confusing action obesity has in the Tong study is not uncommon. There are a number of cancers in which obesity increases risk, speeds progression and then, oddly provides a survival advantage. This is referred to as the “obesity paradox.”^15,16

One plausible explanation appeared in Lancet Oncology back in 2018. Jennifer McQuade at MD Anderson reported that checkpoint inhibitor drugs that target PD-1 work better in obese patients than in normal weight ones. In 330 advanced melanoma patients, those who were male and overweight lived an average of almost 27 months compared to normal weight patients who averaged only 14 months. At the start of the study her assumption had been that obesity would worsen prognosis, not improve it.¹⁷ Being fat is supposed to be bad for cancer, after all.

Tumors grow faster in obese mice and obese people, and this is apparently because their T-cells are “exhausted,” no longer capable of performing their role in limiting cancer growth. The T-cells stop proliferating and become sluggish at secreting proteins that stimulate the immune system. The T-cells also display more than average PD-1, meaning cancer cells are better at suppressing T-cell function, telling these defenders to “leave me alone.”^18,19

These effects all appear to be driven by the hormone leptin, which is closely linked to obesity. This may explain why these drugs that inhibit PD-1 work well in both obese tumor-bearing mice and overweight cancer patients. PD-1 protects cancer cells better, or more effectively, in fat bodies; it is the chronically elevated leptin levels that exhausts the T-cell so that the immune system cannot defend against cancer. Remember, high leptin increases PD-1 expression on the T-cells leaving them more vulnerable to being shut down.  PD-1 inhibitor drugs reawaken the T-cells and lead to a renewed immune defense. Thus, in the obese, PD-1 inhibitor drugs can have a greater impact.²⁰

Leptin, the hormone most linked with obesity, is associated with improved outcomes in breast cancer. In 2019 Yan Kong et al reported that in their study of 325 breast cancer patients, those who had a complete response to chemo treatment, had higher serum leptin levels and more leptin receptors than in patients whose disease progressed.²¹

Reports now link obesity with improved overall survival and progression-free survival in patients with metastatic melanoma, bladder cancer, colorectal cancer, lung cancer, and renal cell carcinoma treated with targeted or immunotherapy and high leptin seems to be the common denominator.^{22 23 24 25 26 27}

Kong’s breast cancer data wasn’t looking at patients undergoing immunotherapy treatment so we must consider and guess whether the results were related to changing leptin levels.

Which brings us back to fasting and fast-mimicking diets. Caloric restriction in humans dramatically lowers leptin. One human trial of eight weeks reported leptin dropping nearly 50%.²⁸ We may need to reinterpret our concept of fasting and caloric restriction; it does more than lower IGF-1. Caloric restriction also lowers leptin, perhaps profoundly. Fasting may also act to reawaken exhausted T-cells.

Taken together, these studies suggest that there are multiple hormones that interact to promote breast cancer growth and that we cannot expect to make sense when examining them one at a time; leptin interacts with IGF-1, IGF-binding protein-3, insulin and estrogen in breast cancer. All of these need to be examined and their interactions understood.²⁹ Our desire for simple explanations will not bring us the answers we require.

In research using human breast cancer cell lines, leptin increased HER-2 protein levels and enhanced the responsiveness of HER-2+ breast cancer cells to other growth promoting chemicals.³⁰ It is possible that leptin and IGF-1 have a greater impact on HER2+ breast cancer and this is why Tong’s data were so striking.

It seems that the interaction leptin has with other growth factors should be our new focus of attention. But should we even bother to measure leptin in our cancer patients? Or is the old BMI calculation enough to estimate it? I don’t know the answer yet. As I wrote early on, Tong’s results have left me confused, sitting at my desk scratching my head.