The OBR Blog

June 02, 2018 - 10:06 pm Posted in ASCO Conference Coverage comments0 Comments

By Christina Bennett, MS

The press conference this morning featured several studies from the 2018 American Society of Clinical Oncology (ASCO) Annual Meeting that attendees won’t see until Sunday or Monday. Two in particular were the SANDPIPER trial for advanced breast cancer and IMpower131 trial for advanced lung cancer. Also, Silicon Valley startup GRAIL, Inc., showcased its cancer detection prototype by presenting data from their lung cancer substudy.

Cancer Detection Test, No Longer a “Pipedream” (Abstract LBA8501)

GRAIL, Inc., the biotech startup backed by billionaires Jeff Bezos and Bill Gates, is on a mission to develop a diagnostic blood test that can detect cancer early on, and according to preliminary results, the startup is on the right path.

“Two years ago, it was a pipedream,” said Geoffrey R. Oxnard, MD, Associate Professor of Medicine at Dana-Farber Cancer Institute and Harvard Medical School, who presented the results from the lung cancer substudy. “It was completely just a brainstorm and had no data supporting it, and I didn’t believe this could be done. Today, we actually have data that show that this is really feasible.”

The data come from the large prospective trial Circulating Cell-free Genome Atlas (CCGA) Study (NCT02889978). The CCGA study aims to enroll 15,000 participants, 70% of which with a cancer diagnosis and 30% without, in the United States and Canada. To date, about 12,000 participants have been enrolled across 142 sites in the United States and Canada.

From the CCGA study, several substudies have been conducted, including a breast cancer substudy featured in the poster session this morning (Abstract 536) and the lung cancer substudy presented during the press conference. Two additional studies are slated for Monday (Abstract 12021; Abstract 12003).

The findings reported for the lung cancer substudy were based on 2,800 samples in a prespecified case-control study, which were divided into a training set and a test set. A total of 118 lung cancer cases and 561 non-cancer cases were evaluable in the training set. An independent test set was used for validation.

“This is not cancer genotyping where you’re looking in the blood for a key mutation and targeting that mutation,” said Dr. Oxnard. “This is cancer detection, and it requires a different approach.”

To detect cancer, the cell-free DNA was evaluated, and three assays were used: targeted sequencing, whole-genome sequencing (WGS), and whole-genome bisulfite sequencing (WGBS).

He explained that the white blood cells are also sequenced. “The white blood cells are rich with mutations, which can pollute the data and make you think that there’s cancer present in the cell-free DNA.”

The prototype assays had a low false positive rate (2%). The targeted assay had a sensitivity of 51% for early-stage lung cancer (stage I-IIIA) and 89% for late-stage cancer (stage IIIB-IV). The WGS assay had a sensitivity of 38% for early-stage lung cancer and 87% for late-stage lung cancer. The WGBS had a sensitivity of 41% for early-stage lung cancer and 89% for late-stage lung cancer.

“The opportunity, perhaps, with a test like this is it’s very specific for finding cancer,” said Dr. Oxnard. He stressed, “This is not a blood test. This is a sequencing method that shows promise and needs to be turned into a diagnostic.”

“The next step will be to optimize those assays and machine learning algorithms and validate in larger data sets,” Charlotte Arnold, spokeswoman for GRAIL, Inc., told OBR. “We also have an ongoing study called STRIVE, which is enrolling 120,000 women at the time of mammogram screening.”

The STRIVE study (NCT03085888) is more representative of how GRAIL’s cancer detection prototype would work in a general population because, unlike the CCGA study, the participants do not know if they have cancer.

“We are hoping to complete enrollment by the end of the year, and then we need at least a year’s follow-up before we start to report results,” said Arnold.

SANDPIPER, a “Modest Step Forward” (Abstract 1006)

The phase III SANDPIPER trial in women with advanced breast cancer showed that the investigational agent taselisib slowed tumor growth and extended progression-free survival (PFS) by 2 months when combined with standard hormone therapy fulvestrant. Overall survival data have not been reported yet.

This is the first placebo-controlled, randomized trial to evaluate the efficacy and the safety of taselisib when added to fulvestrant in patients with advanced disease, asserted leady study author José Baselga, MD, PhD, Physician-in-Chief at Memorial Sloan Kettering Cancer Center.

Taselisib, a PI3K inhibitor, targets the PIK3CA mutation. Only two drugs are currently FDA approved in this class, and both are for hematological malignancies.

“This is a very appealing target,” said ASCO Expert Harold Burstein, MD, PhD. “It’s a mutation that is probably the most common in breast cancer when you do genomic sequencing, and it arises in other tumors as well.”

A total of 516 postmenopausal women with ER-positive, HER2-negative, PIK3CA-mutant locally advanced or metastatic breast cancer were enrolled in the study. Participants were randomized to receive taselisib plus fulvestrant or fulvestrant plus placebo.

The trial met its primary endpoint: The median PFS was 2 months longer for the investigational agent arm (7.4 months) vs the placebo arm (5.4 months) (HR=0.70; P=.0037). The overall response rate was higher for the investigational agent arm (28%) than the placebo arm (11.9%) (P=.0002). Participants in North America and Europe who received taselisib had an even longer median PFS, 7.9 months compared to 4.5 months in the placebo arm.

For participants in the investigational agent arm, the most common severe side effects were diarrhea, colitis, and high blood sugar. Participants in the investigational arm were also more likely to discontinue treatment as a result of side effects. Seventeen percent of participants who received taselisib stopped treatment compared to only 2% in the placebo arm.

“It’s a modest step forward, but it’s an important step forward because it does suggest that we can effectively target the pathway and hopefully this gives us something we can build on so that we can use a targeted and precision approach in this very common type of breast cancer subtype,” said Dr. Burstein.

IMpower131 Met Co-Primary Endpoint, Technically (Abstract LBA9000)

Initial findings from the IMpower131 trial show the addition of PD-L1 inhibitor atezolizumab to chemotherapy extended median PFS by about three weeks for participants with advanced squamous non–small-cell lung cancer (NSCLC).

“Squamous non–small-cell lung cancer remains a very difficult to treat disease and there have been very limited new treatment options presented to us over the last few decades,” said lead study author Robert Jotte, MD, PhD, Medical Director and Co-Chair, USON Thoracic Committee, Rocky Mountain Cancer Centers. “Approximately 25 to 30 percent of patients with non–small cell lung cancer have tumors that indeed can be classified as squamous cell carcinomas.”

For this patient population, he noted, “First-line standard of care remains primarily platinum-based chemotherapy.”

The IMpower131 study had an all-comer design and enrolled 1,021 chemotherapy-naïve patients with stage IV squamous NSCLC. Any level of PD-L1 expression was accepted.

Patients were randomized to one of three arms. Arm A participants received atezolizumab plus carboplatin and paclitaxel. Arm B participants received atezolizumab plus carboplatin and nab-paclitaxel. Arm C served as the control and participants received carboplatin and nab-paclitaxel alone.

Only results for Arm B and Arm C were presented.

Median PFS was 6.3 months in Arm B (atezolizumab plus chemotherapy) and 5.6 months in Arm C (chemotherapy alone) (HR 0.71; P=0.0001).

For the interim analysis, the difference in overall survival was not statistically significant (14 months for Arm B vs 13.5 months for Arm C).

“The overall survival results are immature,” commented ASCO Expert David Graham, MD. “We hope that the results from progression-free survival translate to improved overall survival. If and when that’s shown, I think we’ll clearly have a new standard of care for the frontline treatment of squamous cell non-small cell lung cancers.”

Tomorrow morning we’ll learn the preliminary results for a trial very similar to IMpower131, KEYNOTE-407. The difference is KEYNOTE-407 is evaluating pembrolizumab plus chemotherapy.


May 16, 2018 - 10:05 pm Posted in ASCO Conference Coverage comments0 Comments

With about two weeks to go until the 2018 ASCO Annual Meeting, a pre-meeting press cast featured abstracts of interest that are “just the tip of the iceberg,” said ASCO Chief Medical Officer Richard L. Schilsky, MD. Topics covered included a practice-changing regimen for T-cell malignancies, shorter course trastuzumab for early HER2-positive breast cancer, cost-effectiveness of next-generation sequencing, mobile sensor technology for reducing symptom severity in head and neck cancer survivors, and alarmingly low rates of lung cancer screening in heavy smokers and former heavy smokers. Here are recaps of the major findings.

T-Cell ALL and T-Cell-LL (Abstract 10500)

The largest randomized study ever conducted in T-cell acute lymphoblastic leukemia (T-ALL) and T-cell lymphoblastic leukemia (T-LL) showed the highest overall survival (OS) and disease-free survival (DFS) rates reported to date in 1,545 newly diagnosed children and young adults with these relatively uncommon malignancies. Using standard chemotherapy (a regimen called COG-augmented Berlin-Frankfurt-Munster [aBFM] chemotherapy) plus escalated dose methotrexate or high dose methotrexate, 4-year OS was 90.2% and 4-year event-free survival (EFS) was 84.1%.

In a group of 659 patients at moderate to high risk of recurrence, the addition of nelarabine to standard aBFM chemotherapy achieved 4-year disease-free survival (DFS) of 88% versus 83% in patients not randomized to nelarabine (P=.0450). Patients also received cranial radiation prophylactically or therapeutically.

Patients with T-LL (6% of the total patient population) did not benefit from nelarabine’s addition, but DFS was more than 85% at 4 years in this group of patients.

Patients randomized to receive aBFM plus escalating doses of methotrexate fared better than those treated with aBFM plus high-dose methotrexate: 4-year DFS 89.8% versus 78%, respectively. The arm with the best outcome was nelarabine plus escalated methotrexate, with a 4-year DFS of 91%.

In a small group of 43 patients who failed induction chemotherapy and were non-randomly assigned to high-dose methotrexate plus nelarabine, 4-year DFS beat historical controls: 54.3% were alive with no signs of disease in this study versus 20% of historical controls.

“These outstanding results suggest that nelarabine should become a new standard of care for newly diagnosed T-ALL and T-LL,” said lead author Kimberly Dunsmore, MD, Virginia Tech Carillon School of Medicine, Roanoke, VA.

These results apply to nelarabine when used with aBFM, which is commonly used at pediatric oncology centers in the U.S. They are not generalizable to other regimens, noted experts.

Shorter Course of Trastuzumab (Abstract 506)

A 6-month course of trastuzumab was non-inferior to the current standard of 12 months in the phase III randomized PERSEPHONE trial that included 4,088 women with early-stage, HER-2-positive breast cancer. Four-year DFS was 89.4% in women who received 6 months of trastuzumab versus 89.8% with 12 months of therapy, an absolute difference of 0.4%. Shorter-course trastuzumab appeared to improve cardiac safety. Four percent of women treated with 6 months of therapy stopped treatment due to cardiac events, compared with 8% of the 12-month arm (P<.0001).

“The Persephone trial’s researchers worked closely with patient advocates. Everyone involved in the study is very excited by these results,” said lead author Helena Earl, MD, professor at the University of Cambridge, U.K. “We are confident that this will mark the first steps towards a reduction of the duration of trastuzumab in many women with HER2-positive breast cancer.”

Longer follow-up is needed to establish survival for both arms before 6 months of trastuzumab can be considered standard of care.

PERSEPHONE, funded by the National Institute for Health Research in the U.K., is the largest trial to date to analyze the impact of shorter course trastuzumab in this setting.

Women enrolled in PERSEPHONE were also treated with chemotherapy during the trial. They were followed for a median of 5+ years. Next steps are to determine the impact of treatment length on quality of life, and cost-effectiveness of the two schedules.

“This study establishes the non-inferiority of 6 months of trastuzumab. A shorter course of therapy can reduce the number of patients who stop treatment for cardiotoxicity by half and will reduce cost as well,” said ASCO President Bruce E. Johnson, MD, who moderated the press cast.

Next-Generation Sequencing More Cost-Effective (Abstract 9031)

Genomic testing is now standard of care for non-small cell lung cancer (NSCLC), but what is the best way to do it?

Next-generation sequencing (NGS) of patients with metastatic non-small cell lung cancer (NSCLC) to test for all known cancer-related genes at diagnosis was more cost-effective and provided faster results compared with testing for one or a limited number of genes at a time, according to an economic model based on Medicare and commercial health plans with 1 million hypothetical members.

In the model, when compared with three other testing options, NGS saved between $1.4 million and $2.1 million for Medicare and between $127,402 and $250,842 for commercial insurance providers.

“The field of lung cancer treatment is moving at a rapid pace, and we need to fully characterize genomic changes to determine the best treatment for patients shortly after they are diagnosed. Today, many treatment decisions are guided by the presence or absence of certain genetic changes in a patient’s tumor, and I expect that several more genes will be identified in the near future. Therefore, it becomes even more imperative to find a cost-effective gene test that can quickly identify a large number of gene mutations that can be targeted by treatments,” stated lead author Nathan A. Pennell, MD, PhD, co-director of the Cleveland Clinic Lung Cancer Program.

Currently, there is no accepted standard for genetic testing in lung cancer as well as the timing of such tests. The model was designed to determine which gene testing approach is most cost-effective and time-efficient.

Known gene alterations in NSCLC that are targetable include EGFR, ALK, ROS1, BRAF, MET, HER2, RET, and NTRK1. EGFR, ALK, ROS1, BRAF are targetable with approved therapies. Investigational agents are targeted to the other generic changes, and newer tests also look at PD-L1 expression to predict whether a tumor will respond to immunotherapy.

The model compared one of four different approaches: upfront NGS (testing for all NSCLC-related genes and KRAS); sequential tests for one gene at a time; exclusionary KRAS test, followed by sequential testing for changes in other genes if KRAS was not mutated and no other tests if KRAS was mutated; a “hotspot” panel test for EGFR, ALK, ROS1, and BRAF, followed by either single-gene testing or NGS testing.

NGS and Hotspot panel had faster turnaround times, enabling patients to initiate appropriate therapy 2.8 and 2.7 weeks earlier, respectively. Additionally, NGS identified a higher percentage of patients with targetable alterations compared with the other three strategies.

Remote Technology to Monitor Head and Neck Cancer Symptoms (Abstract 6063)

Head and neck cancer patients treated with radiation have a high symptom burden and are at increased risk for dehydration. Using a mobile and sensor technology called CYCORE, to monitor patients’ symptoms remotely, reduced the severity of treatment- and cancer-related symptoms when compared with usual care (weekly visits to the radiation oncologist), according to a federally-funded, randomized clinical trial that included 357 people with head and neck cancer treated with radiation.

CYCORE includes a Bluetooth-enabled weight scale, Bluetooth-enabled blood pressure cuff, and mobile tablet with a symptom-tracking app that sent information directly to the physician’s office Monday through Friday. Patients who used this technology had lower symptom severity compared with patients who had standard weekly visits with their radiation oncologist.

Daily monitoring of wellbeing enabled physicians to detect symptoms earlier and respond more rapidly compared with usual care.

“Our study generated evidence on how newer technologies can be integrated into cancer care relatively easily and improve patient outcomes without interfering too much in a person’s daily life. This study was done during a rather intense period in the patients’ care for head and neck cancer. The system helped physicians to provide valuable support that ultimately resulted in lower symptom severity,” said lead author Susan K. Peterson, PhD, professor at the University of Texas MD Anderson Cancer Center, Houston, TX.

Patients were randomized 1:1 to CYCORE or standard weekly visits to their radiation oncologist. At the start of radiation therapy, self-reported health severity scores were similar between the two groups. The MD Anderson Symptom Inventory was used to track symptom severity, and patients were weighed and had blood pressure monitoring daily.

After completion of radiation therapy, CYCORE participants reported lower mean scores for general symptoms versus usual care participants (2.9 versus 3.4, respectively) on a zero to 10 scoring system for symptom severity (i.e., 10=most severe). CYCORE participants also had lower mean scores for specific head and neck cancer symptoms (4.2 versus 4.8, respectively).

Six to 8 weeks after completion of radiation therapy, these benefits persisted in favor of CYCORE monitoring. Mean age of patients was 60 years, with a range of 25 to 86 years. Adherence to daily monitoring was 80%.

“CYCORE was feasible in older patients and we had good adherence,” Dr. Peterson said. “This type of monitoring can provide timely information for clinical decision-making and can be expanded to community cancer centers.”

“This study adds to the growing body of evidence that integrating patient-reported outcomes can lead to reduced complications in a wide variety of participants,” said ASCO President, Dr. Johnson.


Lung Cancer Screening Rates Abysmal (Abstract 6504)

Only about 1.9% of almost 8 million current and former heavy smokers underwent low-dose computed tomography (LDCT) screening screened for lung cancer during 2016, according to an analysis of 1,800 lung cancer screening sites. These rates show that the 2015 screening recommendations from the U.S. Preventive Services Task Force (USPSTF) recommending LDCT for current and heavy smokers aged 55 to 80 years are largely being ignored by providers and patients.

Nationwide, a total of 1,796 accredited screening centers could have screened 7,612,975 eligible current and heavy smokers, but only 141,260 people received LDCT screenings (1.9%). By contrast, about 65% of women age 40 or older had a mammography over the same time period.

Screening rates were lowest in the West (1%), followed by the South (1.6%). The Northeast had the highest screening rate (3.5%) and the Midwest the second highest (1.9%).

“Lung cancer screening rates are much lower than screening rates for breast and colorectal cancers, which is unfortunate. It is unclear if the screening deficit is due to low provider referral or perhaps patient psychological barriers from fear of diagnosis. Lung cancer is unique in that there may be stigma associated with screening, as some smokers think that if cancer is detected, it would confirm that they have made a bad lifestyle choice,” said lead author Danh Pham, MD, medical oncologist at the James Graham Brown Cancer Center, University of Louisville, KY.

“Effective screening can prevent 12,000 premature lung cancer deaths each year,” Dr. Pham stated.

John McCleery



April 17, 2018 - 11:04 am Posted in AACR Conference Coverage comments0 Comments

Facts first, then comments.

Keynote -189: This phase III trial compared pembrolizumab (pembro), or placebo, plus pemetrexed and carboplatin as first-line therapy for metastatic NSCLC.

  • N=616; with a 2:1 randomization. Patients were stratified based on PD-L1 tumor proportion score (<1% or >1%)
  • After a median follow-up of 10.5 months, the median overall survival was not reached in the pembro arm, vs. 11.3 months in the control arm.
  • As compared with patients in the control arm, patients receiving pembro were 51% less likely to die overall, and those in the high PD-L1 score group were 58% less likely to die (for both, p<0.0001).
  • The median time to disease progression was 4.9 months for the control arm, and 8.8 for patients on pembro/chemo (P<0.0001).

Presenter of Keynote–189, Leena Gandhi, MD, PhD, Director of Thoracic Medical Oncology at the Laura and Isaac Perlmutter Cancer Center, NYU Langone Health: The rationale of the study was not to see what IO might add to chemo, but rather to see if chemotherapy – which can be immunogenic – can add or expand the benefit of IO monotherapy in patients that might not have high PD-L1 expression; and this is indeed what we are seeing.  

Alice T. Shaw, MD, PhD, Director of the Center for Thoracic Cancers at Massachusetts General Hospital Cancer Center: “This landmark study establishes the triplet as a new standard of care for advance non-squamous lung cancer – the improvement in OS with the addition of pembro to standard chemo across all patients, regardless of PD-L1 status, is really striking.”

Checkmate 227: Nivolumab (nivo) plus ipilimumab (ipi) vs. platinum-doublet chemotherapy as first-line treatment for advanced NSCLC

Note: This study was amended to include a co-primary endpoint based on a proposed new biomarker – tumor mutational burden (TMB). High TMB was defined as greater than 10 mutations per megabase (10mut/Mb). This interim report is related to those patients for whom TMB status was available; N=193.

  • Randomized, phase III trial.
  • There were 299 patients identified as high TMB; 139 received nivo/ipi, 160 received standard chemo.
  • At a minimum follow-up of 11.5 months, IO/IO patients with high TMB were 42% less likely to have their disease progress as compared to chemo patients (p<0.0002).
  • Patients with a TMB<10mut/Mb showed no difference in benefit between treatment arms.
  • The duration of response at one year was 68% vs. 25%, favoring the IO/IO treatment group (p value not given).

Presenter of Checkmate -227, Matthew Hellmann, MD, Memorial Sloan Kettering Cancer Center: “Overall, my sense is that this study has the opportunity to introduce two new standards of care for patients with NSCLC, the first is that the study demonstrates the clinical value of nivo plus ipi as a treatment option for patients that are TMB high, with durable benefit, while sparing 1st-line chemo and thereby preserving effective 2nd-line options.  Second, the study validates TMB as an important and independent biomarker to be routinely tested in treatment naïve advanced NSCLC patients.”

Texas Talk

Commenting at an investor event, sponsored by Bloomberg Intelligence, just a few days prior to the AACR conference, John Heymach, Chair of the Thoracic Cancer Program at MD Anderson Cancer Center, speculated on the outcomes from these two trials.

Regarding Keynote 189: “This is already a widely used regimen, because we have the Keynote – 21G result, which reported a really surprising 0.53 hazard ratio for PFS. So for the 189 results, we’re not looking for a practice changing result, we’re looking for practice confirming results (21G was not powered for survival).”

“For 189, if it confirms a PFS of 0.6 or below – that would solidify this regimen as front line. (The HR reported today was 0.49.)”

Regarding when to use front-line in PD-L1 high expressers: “(Community) oncologists have less experience with nivo/ipi outside of melanoma. Yes, it’s manageable – there’s no chemotherapy side effects, so there’s a good chance patients will prefer IO first and delay chemo until later. But there is more unease with IO/IO (as in Checkmate) versus pembro by itself – it’s easy to give. The question is, if you are a high PD-L1 expresser and you qualify for both IO/IO, and pembro monotherapy, what are people going to give? That’s where there will be a battle.”

Regarding the use of TMB in treatment management decisions:“Cost is a major issue. If (the use of TMB) is approved, it will be interesting to see how. A lot of different assays can produce TMB. Many centers are doing smaller panels – ours included – we’re typically running 100-plus genes, maybe 150, but that isn’t enough right now to get good TMB data. So does that mean all the centers are going change what they do and use Foundation Medicine? This is a significant potential barrier to adoption.”

A View from The Street

A quick Q&A with Patrick Rivers, Vice President of Research, Aquillo Capital

OBR: Of all the results out today, what was the most striking finding?

Rivers: The HR of 0.49 for KN-189 is really impressive. The strength of the phase II data in a much smaller group of patients was enough to get this regimen approved for first-line non-squamous NSCLC, but these results substantiate this as a go-to regimen in a broad number of patients. I was surprised that they were able to demonstrate benefit in EGFR and ALK mutant patients as well.

OBR: The results today don’t’ really suggest a “winner”.  What would you need to see – as yet unseen – that would be a true differentiator between the two PD-1 drugs?

Rivers: I don’t believe that nivolumab and pembrolizumab are appreciably different in terms of the actual drug properties. They have been tested in different clinical settings with different patient selection strategies, and that has likely given rise to the different results that we have observed. The only way to truly differentiate them would be to test them head-to-head, and we all know that isn’t going to happen. One question that hasn’t been answered, which may be critical moving forward, is how much benefit is observed for PD-1 plus chemo when you look at TMB stratification. We only have this for ipi plus nivo vs chemo.

OBR: Given your science background (Rivers came from the bench) – TMB vs. PD-L1 staining – any opinion on which has the most utility?

Rivers: They are independent variables, so they both play an important role and may represent different biological principles – PD-L1 being a marker of inflammation and TMB being a surrogate for neoantigen burden. I am surprised that they show no correlation with one another, but because they independently predict likelihood of response in different populations, they will both have to be examined. The key distinction now is that PD-L1 staining is cheaper, faster and more well-integrated into routine pathology so it currently has more utility. It will be several years before next-gen sequencing panels that give you a TMB read-out are as routine in clinical practice.

OBR: Do these data swamp out any conceivable me-too?  Is there a reason to pursue a next-gen anti PD-1?

Rivers: I don’t believe there is any reason to pursue any additional generations of PD-1 monovalent antibodies. What we do still need to explore further is additional combinations that reveal new elements of biology. There are many other environmental factors in the tumor that are not being properly addressed by PD-1, CTLA-4 and chemotherapy alone. Bi-specific antibodies that bind PD-1/L1 with one arm and another target with another arm may also yet reveal interesting new approaches that do not overlap with current PD-1 use cases.

A final word on Pepsi vs. Coke from Dr. Hellman:

“The results of these two studies are not zero sum. This is a huge step forward for patients – anyway you look at it. Two positive phase III studies, with different but complimentary messages about the use of anti PD-1 therapy in the first-line setting is a huge advance. Although now there are new questions to be answered about how we make the best decisions in the first-line setting, but these are great questions to have to grapple with. So I’m really pleased that we have all this data together.”

By Neil Canavan, Solebury Trout

March 16, 2018 - 04:03 pm Posted in AACR Conference Coverage comments0 Comments

A pre-meeting webinar provided a glimpse of some of the key presentations at the upcoming annual meeting of the American Association for Cancer Research (AACR). Four studies were selected to highlight the themes of immunotherapy, precision medicine, health disparities, and prevention.


Why isn’t CAR-T cell therapy as successful in solid tumors as it is in leukemia and other hematologic malignancies? Why is it more difficult to manufacture CAR-T cells from some pediatric leukemia patients than others?

According to an abstract from a team that pioneered use of CAR-T cells in children with leukemia, poor quality T cells may be the answer to both of these questions.

Research by David M. Barrett, MD, PhD, assistant professor of pediatrics at Children’s Hospital of Philadelphia, and colleagues found that metabolic pathways were associated with the quality of T cells. For example, T cells that use glutamine and fatty acid pathways as fuel sources had excellent CAR-T potential while those that depended on glycolysis, another fuel source, were poorly equipped to undergo the CAR-T cell manufacturing process.

This finding was based on analysis of peripheral blood samples from 157 pediatric patients with hematologic cancers and solid tumors; analysis was performed at diagnosis and after each cycle of chemotherapy.

The CAR-T potential of the T cells was poor in most tumor types prior to chemotherapy, with the exception of acute lymphoblastic leukemia and Wilms tumor (there is not yet a CAR-T product for Wilms tumor), and cumulative chemotherapy was associated with decline in CAR-T potential for all tumor types.

Cells with poor CAR-T potential tended to be those that used glycolysis as their fuel source rather than fatty acids. Certain types of chemotherapy (i.e., cyclophosphamide- and doxorubicin-containing regimens) were especially harmful to CAR-T cell potential.

“We have gotten CAR-T cells to work for leukemia but not yet been very successful in solid tumors. …Our data suggest that poor T-cell starting material may be a key first problem. The T cells from solid tumor patients may need different manufacturing protocols to be successful,” said Dr. Barrett.

Preliminary studies by Dr. Barrett’s group suggest that it is possible to force T cells to use fatty acids for fuel rather than glycolysis, and studies are ongoing to see if this will improve the CAR-T manufacturing process.

Other metabolic pathways may be responsible for poor quality T cells. The research reported today is a first step in elucidating which pathways may be implicated and how to reverse them.

“This study is a great example of the intersection between immunotherapy and precision medicine,” said AACR President Michael Caligiuri, MD, President of City of Hope National Medical Center, Duarte, CA. “Gene expression identifies the energy pathways that T cells use and predict who will do well and who won’t. Then we can try to develop alternative chemotherapies and also try to reverse the metabolic pathways involved.”

 Precision Medicine

HER2 mutations acquired during metastasis appear to confer resistance to hormone therapy in some patients with estrogen receptor (ER)-positive metastatic breast cancer, and the strategy of dual treatment with fulvestrant (a hormone therapy) plus the irreversible HER2 inhibitor neratinib appears to overcome resistance in such cases.

“Resistant ER-positive metastatic breast cancer remains the most common cause of breast cancer death. Although ER-directed therapies are highly effective, most patients invariably develop resistance and stop responding to these drugs,” said Utthara Nayar, PhD, co-lead author this abstract and research fellow at Dana-Farber Cancer Institute Harvard Medical School, Boston.

The researchers conducted whole-exome sequencing of metastatic tumor biopsies from 168 patients with ER-positive metastatic breast cancer that developed resistance to hormone therapies such as tamoxifen, fulvestrant, and palbociclib. HER2 mutations were found in 12 of 168 patients; of these, 8 had mutations previously identified as activating. Further study of 5 available biopsies from these 8 patients revealed that 4 of 5 patients with activating mutations had no evidence of pre-existing HER2 mutations, suggesting that these were acquired under selective pressure of ER-directed therapy.

“It was surprising to discover that HER2 mutations can be acquired in the metastatic setting, suggesting that these tumors evolve, and these mutations seem to be a mechanism of resistance to therapies that target the estrogen receptor….,” said senior author Nikhil Wagle, MD, deputy director of the Center for Precision Medicine at Dana-Farber and assistant professor of medicine at Harvard Medical School.

Laboratory studies showed that HER2-mutation-mediated resistance to hormone therapy could be overcome using the combination of fulvestrant plus neratinib. This strategy was effective in one patient. These early results suggest that dual treatment with these drugs may be a novel strategy for breast cancer patients resistant to ER-directed therapies with acquired HER2 mutations.

Further study is needed to determine the percentage of resistant patients who have acquired or pre-existing HER2 mutations.

Dr. Nayar mentioned that ER mutations, present in 20%-25% of patients with metastatic breast cancer, are known to be responsible for resistance to aromatase inhibitors. Prior to this study, other mechanisms of resistance to ER-directed therapy were unknown in 70%-75% of patients.

“Unlike ER mutations, HER2 mutations conferred resistance to other anti-ER agents,” she said, adding, “This study underlines the importance of profiling resistant metastatic tumors with repeat biopsies. Currently only the initial tumor is profiled in most cases.”

“Using new technologies to examine tumors pre- and post-development of resistance identifies a new mechanism for escaping ER-directed therapy. Excitingly, neratinib can overcome HER2-mediated resistance,” said AACR Program Chair Elaine Mardis, PhD, co-executive director of the Institute for Genomic Medicine at Nationwide Children’s Hospital, Columbus, OH.

“These findings play into the concept of liquid biopsies, because it is difficult to get sufficient metastatic tissue for biopsies. If we know what to look for, we can use liquid biopsies and develop strategies to prolong disease-free survival using next-generation sequencing combined with targeted therapies,” said Dr. Mardis.

Health Disparities

Pelvic inflammatory disease (PID) is associated with increased risk of ovarian cancer, and chlamydia is the leading cause of PID in the developed world. Chlamydia infection increases the risk of developing ovarian cancer two-fold, according to data from two independent studies.

In both study populations, women who had antibodies against pGP3, a protein that is an accurate marker of active or prior chlamydia infection, were about twice as likely to be diagnosed with ovarian cancer as controls. No such association with ovarian cancer risk was found with antibodies related to other infections, including human papillomavirus, herpes simplex virus, hepatitis B, and hepatitis C.

“The fact that there were no associations with antibodies against other infectious agents really supports the specificity of the association of chlamydia infection with ovarian cancer,” said lead author Britton Trabert, PhD, MS, Earl Stadtman Investigator in the Division of Cancer Epidemiology and Genetics at NCI, Bethesda, MD.

The two studies included a Polish population-based case-control study of 278 women diagnosed with ovarian cancer between 2000 and 2003 and 556 matched controls, and the NCI-sponsored Prostate, Lung, Colorectal, and Ovarian Cancer Screening trial, a nested case-control prospective trial with blood drawn prior to diagnosis, which included 160 women who developed ovarian cancer during follow-up and 159 matched controls.

The researchers evaluated antibodies to chlamydia and to other sexually-transmitted and non-sexually transmitted infectious diseases. The antibody to chlamydia was the only one found to be significantly associated with ovarian cancer.

Ovarian cancer is typically silent until it is more advanced and, therefore, is associated with a poor prognosis. Identifying a risk factor such as PID or chlamydia infection could enable diagnosis and treatment at earlier and potentially curable stages. The authors plan to confirm their findings in a larger population and determine whether chlamydia infection has a stronger association with specific subtypes of ovarian cancer.

“This study is notable for being an international collaboration for obtaining tissue samples, and it suggests that we can improve detection of ovarian cancer by routine screening for PID and chlamydia,” said Dr. Mardis.


Vulnerable populations are under-represented in clinical trials and biobanking research, hampering research on healthcare disparities. A pilot study of 78 patients and 25 health care providers from Louisiana communities suggests that both patients and providers are open to participation in clinical trials and biobanking but are not well informed about these efforts and are not invited to participate in them. Further, study subjects, including healthcare providers, did not know where to get more information.

According to study participants, the most useful information would be from a “trusted” healthcare provider. Healthcare providers wanted more information about clinical trials and biobanking, and suggested brief, plain handouts with talking points they could share with patients plus a contact person for more information as effective ways to increase participation. Few participants in the trial said that they would look to the internet or social media for information about clinical trials and biobanking.

The study was based on 14 focus groups and 7 individual interviews from January 2017 to May 2017 in urban and rural communities in Louisiana. Among the 78 patient and community participants, 78% were African American, 24% were from rural communities, and 70% reported low income. Among the 25 safety-net health care providers who participated. 10 were physicians, 7 clinical research associates, 5 nurse practitioners, and 3 behavioral health professionals.

“These results highlight a huge communication gap between the cancer research community and potential participants in cancer genomic studies as well as their providers,” said Terry C. Davis, PhD, professor at Louisiana State University Health Sciences Center in Shreveport and at the Feist-Weiller Cancer Center, and director of the Health Literacy Core of the Louisiana Clinical & Translational Science Center. “This issue must be addressed if we are to ensure that new treatments are available and effective for all patients in all segments of the population.”

“There is a communication problem. Less than 10% of the audience at two medical schools where I spoke recently raised their hands to indicate that they were aware of clinical trials and biobanking,” she noted.

Improved relationships with community providers and jointly developed materials on clinical trials and biobanking could help address this gap, Dr. Davis noted. For one thing, language has to be adjusted. In her study, words like “clinical trial,” “biobanking,” and “genomics,” were not

Well understood. “We need a new vocabulary,” she stated.

Dr. Davis said that this was a pilot study in one state, and a larger study including people from even more diverse communities would be needed to strengthen these conclusions.

“This is a simple issue, but the barriers are complex,” said Dr. Caligiuri. “A shocking statistic is that the death rate of all cancers combined is 25% higher for African-Americans than whites. This study points to some of the problems involved in this disparity.”

By John McCleery, Managing Editor, OBR

article register

Recent Posts

Recent Comments