Selecting Therapy for Patients With Metastatic Pancreatic Cancer After Disease Progression

What are the options for patients with metastatic pancreatic cancer after disease progression? What are the key considerations that you make when creating a treatment plan for these patients?

We create a treatment plan when we first meet with each patient. The key to developing these plans is to understand potential vulnerabilities in the cancer cell based on molecular profiling and genetic testing of the tumor. It has recently become the standard of care to obtain germline genetic testing for all patients diagnosed with pancreatic cancer at any stage. The goal is to identify the minority of patients — about 10% to 15% — who carry a genetic predisposition to developing solid tumors, including pancreatic adenocarcinoma, due to mutations in BRCA1/2, PALB2, ATM CHEK2, and other genes involved in DNA repair. This is done to identify patients who would benefit from PARP inhibitors as maintenance after induction platinum-based chemotherapy.
 
Unfortunately, the majority of patients with pancreatic cancer do not carry these mutations.
 
We also test the cancer tissues genetically in search of other alterations that could be drug-amenable. For example, in the rare circumstance of the absence of KRAS mutations, pancreatic cancer cells would carry alternative mutations activating the MAP kinase pathway. These are extremely rare and typically occur in BRAF, or as the result of amplification of HER2. Without testing, you will never know, and that deprives patients of potential therapeutic options in the context of off-label use of drugs, or in the context of clinical trials.
 
Assuming that the majority of patients will fall into a general category of patients who have KRAS mutation and do not carry DNA repair pathway mutations, those patients would be given standard chemotherapy.

In your view, why are there so few clinical trials being conducted in metastatic pancreatic cancer?

There was an excellent paper in JAMA Oncology written by investigators from the Pancreatic Cancer Action Network that analyzed the unprecedented high failure rates in randomized clinical trials in pancreas cancer.¹ The reason for that sustained trend of failures is that we just simply do not understand the biology of this cancer. The major focus therefore has shifted to early-phase trials in an effort to capture signal of activity based on a solid scientific rationale, and then go to phase 3 testing.

It is thought that although tumor stroma in pancreatic cancer could obstruct adequate drug delivery, stroma also may have “protective effects in restraining tumor growth and progression in pancreatic cancer”.² What are your thoughts on this theory?

The theory that stroma plays a critical role in pancreatic cancer biology is fascinating. There is more and more evidence that this is the case. Human and mouse experiments have shown that when we try to deregulate the stroma it leads to worsening of disease.

Stroma plays a dual role and we now understand that stroma elements are extremely heterogeneous. It involves not only fibroblastic cells, which have multiple categories, but also different types of immune cells and endothelial cells that all regulate the livelihood of cancer cells. Their contribution has become more appreciated and complex. For that reason, dismantling the stroma in its entirety has not been productive. However, more nuanced approaches to stroma deregulation are underway.
 
There are ongoing clinical trials, for example, aiming to reduce activation of the fibroblasts using high-dose vitamin D. This approach is based on a paper [that] showed that high-dose vitamin D reduced activation of fibroblasts and improved the efficacy of chemotherapy.³ Now this is being tested more widely in ongoing clinical trials.
 
Another interesting advancement in the field has to do with repurposing of existing drugs. For example, there are 2 clinical trials that showed promising signals, one with the use of [the antihypertensive] medication losartan in combination with chemotherapy⁴ and another trial that used hydroxychloroquine — an autophagy inhibitor — in combination with chemotherapy.⁵ Those trials showed some signal of activity but need to be validated in larger prospective clinical trials.

At the recent ESMO World Congress on Gastrointestinal Cancer 2020 Virtual meeting, phase 1/2 clinical findings were presented that revealed in patients with newly diagnosed locally advanced or metastatic pancreatic adenocarcinoma, treatment with liposomal irinotecan plus fluorouracil (5-FU), leucovorin, and oxaliplatin (NALIRIFOX) resulted in a median progression-free survival of 9.2 months and a median overall survival of 12.6 months. These findings spurred the initiation of the NAPOLI-3 phase 3 trial (ClinicalTrials.gov Identifier: NCT04083235), which seeks to compare the safety and efficacy of liposomal irinotecan + 5-fluorouracil/leucovorin (5-FU/LV) + oxaliplatin (OX; NALIRIFOX) to gemcitabine + nab-paclitaxel in the first-line setting. Could you summarize what the key findings may be and what impact you feel this trial could have on clinical practice?

This report showed that if you incorporate liposomal irinotecan instead of a conventional irinotecan, which is part of the currently approved regimen of FOLFIRINOX, that it results in a comparable median survival and progression-free survival as what was obtained with the original FOLFIRINOX trial.⁶ Here, NALIRIFOX is given in a single-arm study.
 
The original trial of FOLFIRINOX showed a median overall survival of 11.1 months, which is comparable. This was a phase 1/2 trial and the results are usually somewhat better in an early-phase clinical trial, but it does show some tolerability and efficacy, which is great. Interestingly there has not been a study comparing liposomal and regular irinotecan head to head.
 
In NAPOLI-3, the investigators will compare NALIRIFOX to gemcitabine and nab-paclitaxel combination as the first-line chemotherapy for metastatic [pancreatic ductal adenocarcinoma (mPDAC)]. This is something that is necessary because we have never compared FOLFIRINOX vs gemcitabine/nab-paclitaxel. Both clinical trials that licensed these 2 regimens as first-line chemotherapy options for metastatic disease were compared to gemcitabine alone. For that reason, head-to-head comparison is extremely valuable.

Do you think that basket trials are the appropriate clinical trial approach to use when studying experimental therapies in patients with pancreatic cancer who have DNA mismatch repair deficiency (MMRD) or proficiency (MMRP)? Why or why not?

A basket trial is the type of experiment where you take a certain homogenous disease and stratify patients based on molecular markers. The MMRD pancreatic cancers are extremely rare. For many solid tumors, we look for MMR deficiency, or the signature of high microsatellite instability (MSI). If we find these cases, which are uncommon in pancreatic cancer, they will fall under the blanket U.S. Food and Drug Administration approval of pembrolizumab.
 
If we were to design a basket trial, it would be very difficult in pancreatic cancer because the vast majority of cases — close to 95% — will not carry this MMRD profile. Having said that, it is 1 of the biomarkers that we look for routinely. A recent large trial was done where treatments were tailored to the molecular profiling in patients with pancreatic cancer.  A few participants of Know Your Tumor registry who had MSI/MMRD cancers showed clinical responses to immune checkpoint inhibitors.⁷ This is probably the best systematic demonstration that such strategy can work if you find this molecular profile.

Please discuss frontline therapy options for mPDAC. Do the results of the phase 2 Canadian Cancer Trials Group PA.7 study of gemcitabine (GEM) plus nab-paclitaxel (Nab-P) vs GEM, Nab-P, durvalumab (D), and tremelimumab (T) convince you to change practice?⁸ Why or why not?

Essentially, this was a negative trial. Even though they tried their best to reverse immune suppression in the setting of metastatic pancreatic cancer by using dual checkpoint inhibition in combination with standard chemotherapy, the results are disappointing because any metric of the study — overall survival, progression-free survival, or response rate — were essentially indistinguishable between the 2 arms.
 
This, in a way, highlights the very different biology and mechanisms of immunosuppression in pancreatic cancer. Pancreatic cancers are very simple cancers genetically. Often, the mutation burden is vastly low, with the exception of MSI-high tumors. The rest carry a few mutations and for that reason, tumor cells do not look very different from normal cells to the immune system.
 
The fundamental problem with metastatic pancreatic cancer is that the mechanism of immunosuppression is primarily related to the overwhelming infiltration of the tumor with different types of suppressor cells, the so-called myeloid-derived suppressor cells. The tumor microenvironment in pancreatic cancer is not dominated by checkpoint proteins expressed on tumor cells; PD-1L, as an example, is always at zero. That makes this cancer somewhat unique relative to other tumor types. It is not surprising that this trial did not show any promising activity.