Ultimately, in Dr. Uzzo’s eyes, we are all managers of health care risk. We seek to “understand, predict and prevent future health care events.” As such, from the time of diagnosis, the key steps in managing a patient are: identify risk (screening, etc), utilize risk tools to risk stratify, communicate this risk to the patient, and finally, mitigate risk with intervention.
Looking at biomarkers in the kidney cancer space, he highlights the key point that biomarkers have been few and far between for RCC. At the end of the day, commonly used biomarkers such as stage, grade, and histology still remain the standards for risk stratification. While numerous biomarkers (genetic, epigenetic, etc) have been evaluated, none have been demonstrated to be superior to stage, grade and histology. As biomarkers have failed to improve upon these factors, we also looked to different models to help stratify patients. In the localized RCC disease space, these include the UISS, MSKCC, SSIGN, and Mayo clinical models (among others) to predict recurrence, but the C-index for these tests ranged between 0.76-0.89. However, all the models shared common features that are easily identified clinically – stage, grade, tumor size, performance status, presentation, age, gender, and coagulative necrosis. Models for metastatic RCC are even less capable of predicting cancer-specific mortality (C-indices ~0.6). At this point, biomarkers and models give way to common patient and pathologic characterizations for risk stratification.
Neoadjuvant Therapy (NAT)
Dr. Uzzo provided a very nice comparison of the “Halstedian” model and “Fisheresque” model of cancer progression. Dr. Halsted, a legend in oncologic surgery, believed in stepwise progression of disease from stage 1 -> stage 2 -> stage 3 -> stage 4, which supported utilization of adjuvant therapy rather than NAT. However, Dr. Fisher was a strong proponent of the idea that a subset of patients were likely metastatic at inception, which better supported the need for NAC.
When looking at NAT, there are some key questions:
1) Does it work? (does it shrink the tumor? Can it work as a “litmus test” prior to cytoreduction? Can it control distant disease?)
2) Is it safe?
3) Are there translational signals?
In terms of tumor shrinkage, based on retrospective series and phase II trials, it results in approximately 25% tumor volume reduction, with an objective response rate (ORR) in 30-40% of patients. So, if patients are referred for that indication, that is what a medical oncologist can cite to a surgeon. However, the implications of this are heavily surgeon dependent, and as Dr. Uzzo states, it is a “function of judgment and experience” – if they feel that this will allow for partial resection vs. radical nephrectomy, or make a non-operative patient operative, then it may be worthwhile to proceed. This is difficult to quantify in clinical studies, and selection bias is an unavoidable issue. In his review of NAT to facilitate partial nephrectomy (PNx), there were <200 cases amongst 7 series.1 Similarly, there have primarily only been case reports/series demonstrating tumor thrombus reduction (25-40%), but rarely does it change the level of thrombus without a concomitant risk of toxicity.
In terms of efficacy, he reviewed a few clinical trials of neoadjuvant targeted therapies, including pazopanib.2 While many of these had some tumor size reduction, they often had high rates of patients not making it to surgery due to adverse events. Importantly though, a significant portion did not make it to surgery due to progression of extrarenal disease. As such, he emphasizes that NAT may be utilized as a litmus test for patient response. Patients progressing on NAT likely wouldn’t have benefited from surgery anyway.
No biomarkers have correlated with ORR in NAT trials.
In summary, NAT is not in the guidelines, high quality guidelines are limited, and there is no long-term data. While newer therapies (cabozantinib, immune checkpoint blockage) may change management, clinical trials are the recommendation for now.
Three clinical trials in NAT space:
CARMENA – activated in 2009, still accruing but having difficulty. Testing the importance of surgery – comparing surgery + adjuvant sunitinib vs. sunitinib alone.
SURTIME – testing sequencing (sunitinib -> surgery vs. surgery -> sunitinib). While initially expecting 440 patients, they have modified study to accrue 98 patients (study closed). In data analysis phase now.
ADAPT – SUO CTC joint effort, they have accrued 713/1133 patients in 3 years. Tests sequencing, including the use of autologous dendritic cell immunotherapy and sunitinib.
Adjuvant Therapy (AT)
Recent publications on adjuvant trials have increased interest in this treatment option. However, there are still no approved ATs for RCC. Dr. Uzzo breaks down the history of AT in RCC into three time periods: the “dark ages”, the “middle ages” and “the future.”
In the “dark ages” of AT, numerous trials were done but it combined “ineffective surgery with completely ineffective systemic therapy.” None of them showed significant benefit, though many had significant flaws.3
More recently, we have come into the “middle ages”, where we utilize “ineffective surgery with more effective systemic therapy.” As is well known, S-TRAC4 and ASSURE5, presented conflicting results regarding disease-free survival outcomes in the adjuvant setting. Dr. Uzzo did highlight the key differences in the studies (only cT3-4 disease in S-TRAC, primarily clear-cell histology in S-TRAC) that may have contributed to the discrepancy. However, even when the clear-cell subset of the ASSURE cohort was analyzed, there was no DFS benefit. Two ongoing trials for whom results are pending are PROTECT (pazopanib) and SORCE (sorafenib). The PROTECT trial investigators should be presenting their results later in the meeting.
He very nicely looked at the role of adjuvant therapy in other malignancies (breast cancer, colorectal cancer, melanoma and GIST) and found DFS benefit to be 4-11% (modest), often times with significant monthly cost. As such, he makes a good point, that adjuvant therapy touted as standard of care in other malignancies doesn’t have as much of a benefit as we often put faith in.
Adjuvant therapy is only marginally effect because of
1) Poor timing and patient selection
2) Bad biology
3) Ineffective therapies
In RCC, based on prior literature regarding growth kinetics, tumor doubling time, and presentation of metastatic disease, micrometastases typically present as visible disease between 6-11 years later. Perhaps we are not giving systemic therapy at the right time?
So, while it has not been shown to be highly effective in RCC yet, he recommends:
1) Improving timing (using CTCs and biomarkers)
2) Attacking tumor stem cells (yet to be identified)
3) Attack less promiscuous upstream targets (balance toxicity for specificity)
The future is promising. The “New” Age hopes to combine “incompletely effective surgery with potentially more effective systemic therapy.” He cites two trials, the ECOG PROSPER trial (nivolumab) and the SUO-CTC INmotion trial (atezolizumab), as upcoming studies with novel therapies that may provide new standards.
Overall, in terms of perioperative systemic therapy for RCC, there are no approved options. However, clinical trials with more effective therapies and better patient selection represent the future.
Presented By: Robert G. Uzzo, MD, FACS, Fox Chase Cancer Center, Philadelphia, PA
Written By: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, Princess Margaret Cancer Centre
at the 2017 ASCO Annual Meeting - June 2 - 6, 2017 - Chicago, Illinois, USA