Journal Logo

FRESH SCIENCE for Clinicians

News about basic science of interest and relevance for cancer clinicians

Friday, January 11, 2013

The development and widespread use of the Pap smear dramatically reduced the incidence and mortality of cervical cancer. And the relatively recent addition of HPV DNA testing to liquid Pap tests has further increased their utility.


Now, investigators at Johns Hopkins Medical Institutes and colleagues propose that the test can be adapted to diagnose endometrial cancer and some ovarian cancers.


The new work, published in the January 9 issue of Science Translational Medicine, is proof of principle that DNA assays can detect some of these tumors in liquid Pap samples. A multiplex sequencing test designed to detect 12 mutations common to these cancers accurately identified 100% of endometrial cancers (24 out of 24) and 41% of ovarian cancers (9 out of 22).


The authors, led by Isacc Kinde and colleagues, caution that there is a ways to go before the tests could be routinely used in the clinic.


Shannon Westin, from MD Anderson Cancer Center, and colleagues emphasize the same point in an accompanying editorial. They also point out that while piggybacking endometrial and ovarian cancer screening tests on the widely used Pap test is clever, it might not be as straightforward as one would hope. For example, the use of the Pap test for cervical cancer screening is being scaled back, with longer intervals recommended between tests and consideration of stopping it altogether for older women -- women who are at the highest risk of both endometrial and ovarian cancer.


Additionally, Dr. Westin and colleagues note that developing a CLIA-approved test might be technically challenging because the assay is based on multiplex sequencing. Also, even with next -generation sequencing costs dropping like a rock, the technology might exceed what a widespread test could bear.


All the cautions aside, the study does prove that ovarian and endometrial cancer cells are present for detection in the samples and that the assay can detect mutant DNA that represents less than 0.01% of the DNA copies in the sample. The test also had 100% specificity, albeit in a small sample of just 14 healthy control subjects.


Thus there are reasons for (cautious) enthusiasm: A screening test for gynecologic cancers may be possible.


One interesting aside, from my point of view, is that the paper includes whole genome sequencing data from 22 endometrial tumors and matched controls. To develop their multiplex test, the team needed to know what mutations are common in the various subtypes of ovarian and endometrial cancers. They could find data in the literature on ovarian tumor mutations (e.g., from The Cancer Genome Atlas) but little for endometrial cancer and therefore obtained their own.


Not too long ago, whole genome sequencing of that many tumors and controls would have been a paper all on its own. The fact that the researchers rolled those data into the screening study, and noted them almost in passing, shows just how far the field of cancer genomics has come in a very short time.

Friday, December 21, 2012


Although the annual meeting of the American Society for Cell Biology is unlikely to be on the travel schedule for most oncologists, researchers there presented studies that cancer scientists may find interesting.


In the first study, Gautham Venugopalan, a graduate student in the lab of Daniel Fletcher at the University of California, Berkeley, reported that a brief compressive force triggers a major morphological change in breast cancer cells (abstract 1673).


When grown in a 3-dimensional matrix under normal conditions, malignant epithelial cells from disorganized colonies. However, if the researchers subject the malignant cells to a compressive force, the cells rotate and form normal-looking acini.


Venugopalan says that about 10 minutes of force in the beginning of a 10-day experiment is enough to trigger the change. The force itself is ~5kPa (< 1psi). “That is about 5% of what you would fill up a regular party balloon with,” he said.


The investigators are not sure how the force causes the change, but they are currently looking at gene expression patterns to see if force-treated breast cancer cells return to a more normal pattern compared to untreated control cells.


When asked how these findings might translate into patient treatment in the future, Venugopalan says their work is about understanding how cells sense their environment and form organized structures. “If we know which key structures and molecules are involved in forming organized structures, we would have a new set of potential targets for cancer treatment and therapy,” he said.


In a second, unrelated abstract, Mark LaBarge, PhD, from the Lawrence Berkeley National Laboratory, and colleagues report that the relative abundance of different types of breast epithelial cells shifts as women age (abstract 1674).


Dr. LaBarge and colleagues have assembled a large collection of breast tissue samples from healthy women aged 16 to 91 years — a collection called the human mammary epithelial cell (HMEC) Aging Resource.


Comparing uncultured samples from younger and older women, the investigators found that the proportion of myoepithelial cells declines with age, whereas the proportion of luminal cells increases.


Moreover, as women age, the number of defective progenitor cells increases, giving rise to a larger number of partially differentiated luminal and myoepithelial cells. That increase, according to Dr. LaBarge, may leave older women more susceptible to breast cancer – and may partially explain why such a large proportion of breast cancer cases occur in women over the age of 50.


And for oncologists who didn’t make it the Cell Biology meeting, LaBarge and colleagues reported the full study in Cancer Research earlier this year.


Friday, October 5, 2012


SAN FRANCISCO -- Acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) are known complications associated with adjuvant chemotherapy in breast cancer patients, but the actual rate has been a matter of discussion within the oncology community recently. A new retrospective analysis, though, reported here at the Breast Cancer Symposium (Abstract 62) suggests the complication rate in a large community oncology practice is similar to what has been reported in clinical trials and is lower than that seen in recent database studies, researchers reported here at the Breast Cancer Symposium.


The meeting is co-sponsored by the American Society of Breast Disease, American Society of Breast Surgeons, American Society of Clinical Oncology, American Society for Radiation Oncology, National Consortium of Breast Centers, and Society of Surgical Oncology.

 ASCO/Todd Buchanan 2012

The researchers, led by Neelima Denduluri, MD, a medical oncologist and hematologist with Virginia Cancer Specialists and the U.S. Oncology Network, analyzed the electronic health records of 20,900 female breast cancer patients who had been treated in the U.S. Oncology system between 2007 and 2010. Patients in the study had stage I to III disease and were followed through February 2012.


With a median follow-up of 2.8 years, 11,295 women (54%) had received chemotherapy and 9,605 (46%) had not. The most striking difference between the two groups, she noted, was that the median age at diagnosis for the no-chemotherapy group was 10 years older than for the chemotherapy-treated group (64 vs. 54 years).


“Our overall event rate was very low and not statistically significant,” she said. Twelve women (0.167%) in the chemotherapy group developed AML or MDS, compared with 16 (0.106%) in the no-chemotherapy group.


In the chemotherapy-treated group, older age and anthracycline treatment were significantly associated with the risk of AML/MDS in a multivariate analysis. Pegfilgrastim use was not significantly associated with AML/MDS.


In the no-chemotherapy group, none of the variables tested were significantly associated with risk, although age and disease stage showed a trend toward significance.


During the discussion following the presentation, Clifford Hudis, MD, Chief of the Breast Cancer Medicine Service at Memorial Sloan-Kettering Cancer Center, thanked her for a clear presentation and the creative use of electronic medical records for comparative effectiveness research.


He then alluded to recent database studies that showed substantially higher rates of AML/MDS in breast cancer patients, particularly in women who had received growth factors during adjuvant treatment. “Your data are much more in line with clinical trial data,” he said. “Do you have any thoughts on that disconnect between large population-based studies like yours and huge SEER database studies?”


“I think many of the SEER database studies are using Medicare claims,” Denduluri responded. “Our population included both younger and older patients.” Clinical trials, she noted, tend to enroll younger patients, which might help explain why the rate is lower than that seen in database studies and more similar to what has been seen in large clinical trials, such as those that tested dose-dense chemotherapy.


One limitation to the study is the relatively short follow-up, she continued. “But as a comment, the vast majority of topoisomerase-induced leukemias are detected within the first three years.


“The rates of AML/MDS were low after adjuvant chemotherapy and were similar to those of non-chemotherapy treated patients. Our data confirm known risk factors of AML/MDS after chemotherapy, including increased age and anthracycline use. In our analysis, the growth factor affect was not significant.”



Monday, October 1, 2012

SAN FRANCISCO -- Women whose primary breast tumors express higher levels of RANK (receptor activator of nuclear factor kappa-B) may be at higher risk of relapse and bone metastases, compared with women whose tumors express less of the protein, researchers reported here at the Breast Cancer Symposium. Two other studies suggest that everolimus-exemestane therapy may reduce the risk of bone metastases compared with exemestane alone, and that statin use may be associated with a reduction in risk of bone metastases in breast cancer patients.


RANK Ligand as Predictor of Risk

In a study of 149 women who took part in the ISPY-1 trial (Abstract 2), Jiali Li, MD, PhD, Clinical Instructor at the University of California, San Francisco Helen Diller Family Comprehensive Cancer Center, and colleagues found that RANK expression correlated with an increased risk of relapse. Specifically, if the patients were divided between those whose primary tumors had high RANK expression (cutoff of 63 percent or greater positive cells) and those whose tumors had low expression, there was a statistically significant difference in relapse-free survival.


The observation was validated in a second, independent cohort of 425 women from an MD Anderson Cancer Center database of women with early breast cancer who had undergone neoadjuvant chemotherapy. In that database, the difference between high- and low-expressors was even more significant, with a cutoff of 61 percent cell positivity.


The investigators also found that, on average, RANK expression was significantly higher in estrogen receptor-negative tumors compared with estrogen receptor-positive lesions, and in basal subtype cancers compared with luminal A/B or HER2-positive tumors. However, RANK expression did not correlate with tumor grade or response to neoadjuvant chemotherapy.


“The most interesting finding in this study is that RANK expression in the primary tumor is associated with bone-dominant metastases,” Li said, during a presentation of the data in the general session. RANK was expressed at a significantly higher median level in the bone-dominant tumors than in the tumors that predominantly metastasize elsewhere, even after adjustment for estrogen-receptor positivity and intrinsic subtype.


“We hypothesize that RANK ligand may serve as a chemokine to attract RANK-expressing cancer cells to the bone,” she said. “Our study suggests that the RANK/OPG/RANKL pathway could be a potential target for preventing bone-dominant metastases in patients with early breast cancer.”


The team is currently testing the hypothesis in a Phase II clinical trial in which women with early breast cancer who have circulating tumor cells are treated for one year with denosumab, an antibody that binds the RANK ligand.


BOLERO Data Suggest Reduction on Bone Metastases

Meanwhile, exploratory analyses from the BOLERO-2 trial suggest that inhibiting mTOR with everolimus might also reduce the risk of bone metastases (Abstract 102). The phase III trial compared exemestane plus everolimus with exemestane plus placebo in women with metastatic or locally advanced estrogen-receptor positive, HER2-negative breast cancer refractory to non-steroidal aromatase inhibitors. The primary results showed a significant improvement in progression-free survival with the combination at a median follow-up of 18 months.


To assess the drug’s impact on bone metabolism, the researchers compared the levels of several bone markers at baseline and at six and 12 weeks. They found a decrease in all three serum markers tested in the combination-treated patients, while those in the control arm showed an increase of the markers over time.

The investigators also saw a significant reduction in disease progression in the bone with the exemestane-everolimus combination compared with the control arm. When the disease-progression analysis was restricted to just those patients with bone metastases at baseline, there was a statistically significant reduction in the risk of progression with the addition of everolimus, compared with exemestane alone.


“Everolimus does appear to reduce progression to bone, which is an important finding as bone metastasis is a considerable cause of morbidity and mortality in our patients with estrogen receptor-positive metastatic breast cancer,” said Matthew Ellis, MD, PhD, the Anheuser-Busch Professor of Medical Oncology from the Siteman Cancer Center and Washington University, St Louis, during a poster discussion session.


“How it does this is a little bit speculative, but it is thought that it may have an effect on RANK ligand production, such that when patients are on everolimus it reduces the adverse effects of exemestane on bone. This doesn’t seem to be associated with any adverse side effects we associate with that, such as osteonecrosis of the jaw,” he continued.


“This could be potentially something of tangential value when everolimus is tested in the adjuvant setting, which I believe it will be in an upcoming SWOG-NSABP study”


Fewer Bone Metastases with Statin Use

Finally, in a retrospective chart review, investigators from the Albert Einstein Medical Center in Philadelphia found that statin use was associated with a reduced likelihood of bone metastases (Abstract 40). The team examined the records of 841 patients with Stage II or III cancer treated between 1999 and 2010. Of those, 223 were identified who had taken statins for three or more months.


Eleven women (5%) taking statins developed bone metastases compared with 77 (12.5%) not taking statins, which was a statistically significant difference. The rate of metastases to other sites was not significantly different between the two groups (10.8% in the statin group vs. 18.6% in the no-statin group).


The investigators note that in-vitro studies have previously linked the mevalonic acid pathway, which statins inhibit, with promotion of a microenvironment that favors bone metastases in breast cancer. Also, prior work suggested that statins reduce neoplastic growth and osteoclast activity in vitro.


Based on these observations, the researchers suggest that statins should be examined further for the chemoprevention of bone metastases in breast cancer patients.


Thursday, August 30, 2012

It has been just over a year since Duke University officials appeared before the Institute of Medicine’s Committee to Review of Omics-Based Tests for Predicting Patient Outcomes in Clinical Trials.


At the time, Robert Califf, MD, Vice Chancellor for Clinical Research at Duke, and other officials predicted that the misconduct investigation looking into the actions of Anil Potti, MD, (discussed here and here; summarized in the IOM report here) would be completed by the end of 2011. A few months later, we were told it would be early in 2012.


Here it is the end of August and we still haven’t heard.


When I asked Duke about the delay, I received two responses. First from an official spokesperson, who provided the following statement:


Everything possible is being done to expedite this misconduct investigation despite the complexities involved and the federal requirements for such a process. According to the federal Office of Research Integrity [ORI], in 2010, the average misconduct investigation process by an institution (assessment, inquiry and investigation) was 19 months in duration with a range of 1 month to almost 4 years. ORI's portion, which follows the institutional process, averaged 7.2 months, with a range of 1 month to more than 4 years.”


Dr. Califf also responded via email to my queries. “…this is now in the hands of the misconduct committee and associated lawyers. ORI granted an extension to the committee till the end of September.”


Dr. Califf noted that ORI can accept, reject, or ask for revisions on the findings of the institutional investigation. And that it is the agency’s decision whether the results of the investigation will be made public.


Given the impact of the past events on patient care, I sincerely hope ORI will make the results known.


When asked directly, ORI said they cannot discuss the case at this time.


So here we are, waiting -- and hoping that the officials at each level do the right thing: Conduct a thorough investigation and provide answers to the many questions raised during the IOM discussions and by other researchers.


In the meantime, according to Dr. Califf, Duke is putting “resources into quality initiatives that have been very helpful.”


For Potti, the latest news, according to a report in the Grand Forks Herald, is that he is working as an oncologist at the Cancer Center of North Dakota. The Herald article quotes his new boss, William Noyes, MD, MBA, as saying that the problematic work at Duke “did not involve patient care, but clinical trials, which the Cancer Center does not do.”


Uhm, excuse me if I’m being dense, but since when did clinical trials and patient care become two exclusive domains?