Targeting KRAS in Colorectal Cancer Is an Ongoing Medical Challenge

The development of therapies for colorectal cancer (CRC) with the ability to target KRAS has thus far been met by numerous barriers as most agents in development have historically been unable to demonstrate clinical activity. Due to the high prevalence of KRAS-mutant CRC and the fact that many different KRAS variants are found in these tumors, developing treatment for patients with KRAS-mutant CRC is an area of unmet medical need.¹

Afsaneh Barzi, MD, PhD, a medical oncologist and associate clinical professor, Department of Medical Oncology & Therapeutics Research at the City of Hope Comprehensive Cancer Center, explained during a conference call hosted by Cardiff Oncology, that the issue is most apparent in the second-line treatment setting where current data reveal that this patient population has poor outcomes overall, but especially in terms of survival. In addition, few studies in CRC focus specifically on patients with KRAS mutations.

“KRAS-mutated colorectal cancer is a unique entity with poorer outcomes compared to the KRAS wild-type population and is in need of drug development,” Barzi told Targeted Oncology, in an interview.

Preexisting Data from Patients With KRAS+ CRC

Barzi shared published data on patients with KRAS mutations and KRAS wild-type tumors. In the KRAS wild-type population, treatment with FOLFIRI (folinic acid, fluorouracil, and irinotecan) shows an objective response rate (ORR) of only 5%. The progression-free survival (PFS) in these patients on FOLFIRI is just 4.5 months compared with 6.4 months with FOLFIRI plus bevacizumab (Avastin) (HR, 0.61, 95% CI, 0.49-0.77; P <.0001). The overall survival (OS) in the KRAS wild-type population is 11.1 months with FOLFIRI alone versus 15.4 months with the addition of bevacizumab (HR, 0.61; 95% CI, 0.53-0.90; P = .0052).^2,3

In patients with KRAS-mutant CRC, FOLFIRI alone has a 3% ORR and a 4.1-month median PFS compared with FOLFIRI plus bevacizumab, which has a 4% ORR and a 5.5-month median PFS (HR, 0.70; 95% CI, 0.56-0.89; P = .0027). The OS with FOLFIRI in the KRAS-mutated population was 10 months compared with 10.4 months with the addition of bevacizumab (HR, 0.92; 95% CI, 0.71-1.18; P = .4969).

These data show that even with added therapy, outcomes are dismal for patients with KRAS-mutant CRC in the second-line setting, explained Barzi.

“Direct targeting of RAS is challenging and, so far, has not been fruitful. However, using other strategies, such as synthetic lethality, is an opportunity to develop agents that can effectively treat patients with KRAS-mutated tumors.”

The Search for a RAS Gene to Target

No RAS inhibitors are FDA approved for the treatment of CRC. Experts including Barzi consider many RAS proteins to be undruggable and this problem has inspired the clinical development of novel drugs.¹

One type of agent for targeted the KRAS oncogene is the KRAS G12C covalent inhibitor, which is actively being researched. Barzi noted, however, that not many of these mutations are seen in patients with KRAS-mutant CRC. While KRAS mutations are found in about 45% of CRCs, the G12C subset only represents 8% of the overall population.

Alternative strategies for targeting the KRAS oncogene include the use of PI3K/AKT/mTOR inhibitors and RAF/MEK/ERK inhibitors. However, using these targets doesn’t fill the area of unmet need considering the data so far have shown limited activity and heavy toxicity.

There is also exploration in the field around disrupting KRAS plasma membrane localization with anti-farnesyl inhibitors, targeting mutant KRAS-regulated metabolic pathways, and targeting synthetic lethal partners of mutant KRAS. What is known about these novel strategies is that no clinical efficacy has been demonstrated with anti-farnesyl inhibitors. But preclinical research has suggested promise with utilizing PLK1 inhibitors to target synthetic lethality genes. One such inhibitor that demonstrated strong cell growth inhibition and mitotic arrest inKRAS-mutant CRC cells compared with wild-type cells was onvansertib.

Onvansertib as Treatment of KRAS-Mutant CRC

During the conference call, the data for onvansertib in patients with KRAS-mutant CRC were presented by Heinz-Josef Lenz, MD, FACP, professor of medicine and associate director of Clinical Research, at the University of California Keck School of Medicine.

“We need more effective therapies for KRAS-mutant colorectal cancers; this is an unmet need. Onvansertib inhibiting PLK1 was shown to be very effective in synthetic lethality models. Preclinical models show synergism with irinotecan, 5-FU, and bevacizumab. Based on preclinical data, a phase 1/2 trial was initiated in second-line mCRC with onvansertib in combination with FOLFIRI and bevacizumab. Early data suggest that onvansertib is well tolerated and shows promising efficacy.”

In a preclinical model, the PLK1 inhibitor onvansertib demonstrated antitumor activity in a patient with KRAS-mutant CRC when administered in combination with irinotecan and 5-FU. The agent was found to have synergy with both standard-of-care chemotherapy partners. These preclinical findings serve as rationale for the launch of a phase 1b/2 clinical trial of onvansertib combined with the FOLFIRI regimen, as well as bevacizumab (NCT03829410).

The study enrolled patients with metastatic/unresectable CRC with a KRAS mutation in the primary tumor or metastasis. All patients had failed frontline treatment with 5-FU and oxaliplatin with or without bevacizumab or those who were intolerant to the same frontline therapies. All patients progressed on less than 6 months of frontline treatment.

As a primary end point in the phase 1b portion of the study, investigators assessed dose-limiting toxicities (DLTs), adverse events (AEs), and tolerability. The primary end point of the phase 2 portion was objective response rate (ORR). The secondary end point explored in the study included progression-free survival (PFS) and the reduction in KRAS allelic burden, which was determined through liquid biopsies.

In a population of 13 patients, there were 2 DLTs of grade 4 neutropenic fever at the 12 mg/m² dose and grade 4 neutropenia at the 18 mg/m² dose level. Both DLTs were attributed to the bolus treatment with 5-FU. The dose levels of onvansertib considered to be safe in the combination for patients with KRAS-mutant CRC were between 12 mg/m² and 15 mg/m^2.

Overall, onvansertib with irinotecan and 5-FU was well tolerated in patients with only 9% experiencing a grade 3 or 4 AE. The grade 3/4 AEs that were reported in more than 2 patients included neutropenia (n = 5) andabdominal pain (n = 2), but these AEs were resolved within 2.5 weeks. The most common any-grade treatment-emergent AEs were fatigue (n = 10), diarrhea (n =8), nausea (n = 8), and neutropenia (n = 8). Lenz noted that onvansertib did not cause any major toxicities in this study.

Durable response of greater than 6 months (range, 6-12) was observed in 73% of patients. There was only 1 patient who progressed in less than 6 months. Of the responses observed, 45% were partial responses (PRs), including 4 confirmed PRs and 1 unconfirmed. Notably, better response were observed in females compared with males, in patients with a higher ECOG performance status compared with 0, in those with the colon as their primary tumor site versus the rectum, in patients with metastatic disease, and in patients who had prior treatment with bevacizumab.

Liquid Biopsy Results

Serial monitoring of KRAS was performed in the study to evaluate whether liquid biopsies can predict subsequent radiographic response in patients. The monitoring identified KRAS variants in 9 out of 11 patients. After the first cycle of treatment with the onvansertib combination, all patients had a decrease in KRAS mutant allele frequency (MAF).

Patients who achieved a PR to therapy were identified as having the greatest change in KRAS MAF. Most notably, the patients who had progressive disease only had a 55% decrease in KRAS MAF.

Looking at the how PLK1 inhibition impacted pathways in this patient population, 8 pathways were found to have significantly different alterations rates in PLK1 quartile 1 compared with PLK1 quartile 4. The difference was P <.05. In addition, 7 pathways were more frequently altered in PLK1 quartile 4, including the WNT signaling pathway, TP53 pathway, chromatin remodeling pathway, lynch syndrome/MMR gene pathway, receptor tyrosine kinanses.co-factors pathway,notch signaling pathway, response to immunotherapy pathway, and the signal transduction pathway.

Lenz noted that clinical responses were seen in multiple KRAS variants, which included the 3 most common KRAS point mutations found in patients with CRC.

In terms of next steps with the study of onvansertib, Lenz stated, “if this study confirms efficacy, a next step could be a registration trial in randomized phase 2/3 settings. A more interesting project is to determine what other combination with onvansertib might be effective. There are data suggesting that PLK1 is associated with immune signaling and DNA repair indicating that combination with immune therapies and DNA damaging agents or PARP inhibitors should be tested in preclinical models.”

Furthering the Use of Promising Strategies

The next step to continue to fill the gap for effective treatment strategies in KRAS-mutant colorectal tumors is to promote wider adoption of targeted therapies. Currently, adoption differs in academic centers compared with community centers. Although FDA-approved drugs are available, there is a gap in the genetic testing of this patient population seen in the community setting.

Barzi told Targeted Oncology, “Better understanding and documentation of KRAS mutations is something that can and should be addressed in the community. For many providers who treat patients with colorectal cancer, timely identification of KRAS status and documentation of the mutation type and coexisting mutations can open the door to understanding the behavior of this mutation in patients, which can lead to novel therapies.”

Lenz commented about targeted therapy adoption in a separate interview, stating: “for the community, we need registered FDA approved drugs for patients with KRAS-mutant CRC. As soon [as a] drug is approved there is no doubt that oncologists in the community will use this strategy.”

_References

_{1. Cardiff Oncology KOL Call: KRAS in Clinical Practice in Colorectal Cancer. Conference Call. Cardiff Oncology, Inc. September 23, 2020. Accessed September 23, 2020. https://bit.ly/362qDOk​}

_{2. Hess LM, Cui ZL, Mytelka DS, et al. Treatment patterns and survival outcomes for patients receiving second-line treatment for metastatic colorectal cancer in the USA. Int J Colorectal Dis. 2019;34(4):581-588. doi:10.1007/s00384-018-03227-5}

_{3. Ahn DG, Erlander M, Ridinge M et al. Phase 1b/2 study of the polo-like kinase 1 (PLK1) inhibitor, Onvansertib, in combination with FOLFIRI and bevacizumab for second line treatment of KRAS-mutated metastatic colorectal cancer. Presented at: 2020 ESMO Virtual Congress; September 19-21, 2020; Virtual. Abstract 436P.}