Bartlett Family Professor of Gastrointestinal Oncology
Chief, Solid Tumor Services
Director, Gastrointestinal Oncology Program
Co-Director, Center for Advanced Digestive Care
Weill Cornell Medical College
New York, New York
Manish A. Shah, MD, FASCO, has disclosed that he has received funds for research support from Boston Biomedical and Roche.
In an educational activity from early 2017, I discussed the central role of cancer stem cells (CSC) in tumorigenesis, relapse, and metastasis and emerging clinical trial data on novel agents targeting CSC in gastrointestinal (GI) cancers. Since then, multiple studies have added novel information to this conversation, with even more results forthcoming. Herein, I review how recent developments are shaping our understanding of CSC-targeting agents and their potential placement in the current treatment paradigm for GI cancers.
The CSC inhibitor that is furthest along in clinical development is napabucasin, a first-in-class agent specific for STAT3 that was shown to have antitumor activity in a range of GI cancers in early-phase trials. More recently, in a prespecified biomarker analysis of the phase III CO23 study evaluating napabucasin monotherapy vs placebo in patients with advanced colorectal cancer (CRC) in whom all available standard therapies had failed, napabucasin extended OS in the pSTAT3-positive subset of patients from 3.0 months to 5.1 months (HR: 0.41; 95% CI: 0.23-0.73; P = .0025) but not in the overall unselected patient population (P = .34), suggesting that identification of the patient population most likely to benefit from these agents is going to be key to their success in the treatment of GI cancers. These results led to the phase III CanStem303C trial assessing FOLFIRI with or without napabucasin in previously treated metastatic CRC, which is projected to report in 2020 and will add valuable data to our understanding of napabucasin activity in pSTAT3-positive CRC.
In the second‑line setting for gastric and gastroesophageal junction (GEJ) cancer, results of the phase III BRIGHTER trial presented at ASCO 2018 were somewhat of a disappointment. This study showed no improvements in OS or PFS with napabucasin plus paclitaxel vs placebo plus paclitaxel in patients with pretreated advanced gastric or GEJ adenocarcinoma. One of the key questions that emerged during review of the BRIGHTER data was: Why did this combination fail? A key possibility informed by the CO23 data in CRC is that we did not select for the patient population most likely to benefit. In fact, analyses in the pSTAT3-positive patient population are currently ongoing. That said, there was a potential OS benefit with combination therapy for HER2-positive tumors (HR: 0.76; 95% CI: 0.51-1.15). This is particularly interesting because HER2 overexpression is associated with downstream activation of STAT3 and, therefore, may be a surrogate indicator that napabucasin is, indeed, hitting its target. However, we ultimately need to identify the patients for which there is definitive napabucasin activity.
Another possible explanation for the failure of BRIGHTER is that increases in toxicity with the combination led to lower paclitaxel exposures. Napabucasin itself is associated with higher rates of diarrhea and, when added to paclitaxel, led to grade ≥ 3 diarrhea in 16% of patients who received the combination vs 1% of patients in the placebo plus chemotherapy arm, with mean paclitaxel exposures of 10.7 doses vs 11.7 doses, respectively. Whether this affected efficacy of the combination we cannot say for sure. However, it could be hypothesized that slightly increased higher-grade diarrhea leading to slightly decreased delivery of cytotoxic therapy could have overwhelmed the efficacy of a marginally active drug in an unselected population. Thus, whatever benefit may have occurred in patients with the target could have been washed out by lower administration of the drug secondary to slightly higher toxicity.
ASCO 2018 also saw presentation of remarkable preliminary results from a phase I/II study of first-line napabucasin plus gemcitabine/nab‑paclitaxel in metastatic pancreatic adenocarcinoma. In 59 patients, the overall disease control rate was 78% and the ORR was 48%, with 2 patients having a CR and 26 patients having a PR. This impressive antitumor activity is being further investigated in the ongoing phase III CanStem111P trial, which is projected to report in late 2020. More refractory tumors such as pancreatic cancer, or perhaps even esophageal cancer, may be particularly good candidates for CSC-targeting agents owing to enrichment of cancer stemness properties in these cancers.
Outside of napabucasin, ongoing research continues to appraise the significance of other CSC inhibitors and stemness properties in cancer. One study published as a part of ASCO 2018 examined the association between CSC levels and CRC tumor invasion. Immunohistochemistry of tumor tissue samples from 60 patients for the CSC markers CD44 and CD133 showed that CD133 expression was generally higher in tumors from patients with more advanced cancer, including in tumors from patients with vs without distant metastases, in tumors classified as maximally vs minimally invasive, and in liver metastases compared with corresponding primary tumors. These data suggest that as cancers become more aggressive, their stemness properties increase. This is compelling evidence supporting that cancer stemness is important, particularly for patients with metastatic disease, because these tumors are the ones that are going to continue to propagate, develop resistance, and ultimately cause death.
Another interesting facet of ongoing research is whether combining CSC inhibitors with immunotherapies may augment immune responses. Data suggest that cancer stemness properties affect the tumor microenvironment and can lead to immunoresistance. The rationale for cotreatment, then, is that combining a CSC inhibitor with, say, a checkpoint inhibitor may overcome this obstacle. In support of this rationale, preclinical data from 2018 American Association for Cancer Research meeting demonstrated synergistic antitumor activity in syngeneic murine tumor models when combining Wnt antagonists vantictumab or ipafricept with an anti–CTLA-4 or anti–PD-1 checkpoint inhibitor. Such combinations decreased tumor volume, immunosuppressive myeloid cells, and Tregs while increasing cytotoxic T-cell and APC activity.
As we move forward, efforts to identify the patient populations most likely to benefit from each of the different CSC inhibitors will be paramount. Selecting patients with tumors exhibiting a higher level of stemness properties or stem cells could ensure that drugs like napabucasin can exert their maximum potential. Although we are early in the developmental process leading to practical use of these drugs, much hope exists that this therapeutic class could significantly change the course of how cancers can be treated. Although there may be activity in CSC-targeting agents as monotherapy, it appears that the more substantial activity is going to be in combination with chemotherapy or with immunotherapy in the optimal patient population. Desire to develop CSC inhibitors in the adjuvant setting also exists, although proof of principle will likely be required in more advanced patients before transitioning to other settings.
How has your perspective on CSC-targeting agents for GI cancers changed during the last 2 years? Do you consider molecular profiling a promising strategy for identification of patient populations that will benefit from these agents? Please join the conversation by sharing your feedback in the comments box below and responding to the polling question to the right of your screen.
For more information on key results in GI cancers from ASCO 2018, please see our conference coverage here.