Farnesyl Transferase Inhibitors for Patients with Lung Cancer

The Kirsten ras gene was initially discovered as a gene coded by a retrovirus that caused sarcomas in rats (1, 2). The human homologue of this retroviral gene, K-ras, and two related human genes, H-ras and N-ras, are collectively called the ras family of genes. These genes are commonly mutated in human cancers, the most frequent being K-ras, which is mutated in 80% of pancreatic cancers. The K-ras gene is also mutated in ∼20% of non-small cell lung cancers (NSCLCs), prompting studies to determine whether this mutation leads to biochemical changes that can be targeted with anticancer agents (3, 4). It is more common in adenocarcinomas than other types of NSCLC (3, 4).

Biochemical characterization of the proteins encoded by the ras family of genes has shown that after translation of the mRNA to protein, the protein is modified by the covalent addition of a nonpolar farnesyl group to the COOH-terminal (1, 2). This enables the ras protein to translocate from the cytoplasm to the cell membrane and allows it to participate in cell signaling, giving rise to more rapid proliferation. The inhibition of the farnesylation process can prevent the translocation of the ras protein to the cell membrane and thus reduce the ras signaling that can promote the growth of the cancer cells.

Multiple farnesyl transferase inhibitors (FTIs) have been developed to prevent the covalent linkage of the farnesyl group to the ras family of proteins. In preclinical testing, FTIs have been shown to inhibit the growth of a broad spectrum of human cell cancer lines in vitro and in vivo (5, 6, 7). Despite the hypothesized mechanism of action, tumor cell lines bearing mutated K-ras may be less sensitive to the antiproliferative effects of FTIs than cell lines bearing wild-type ras or mutated H-ras genes (5, 8). Recent evidence suggests that farnesylated proteins other than ras may also be targets for FTIs that limit the growth of cancer cells. RhoB (9, 10), the centromeric CENP-E and CENP-F proteins (11), and members of the phosphatidylinositol 3′-kinase/Akt pathway may play a role in mediating the antitumor activity of FTIs (12, 13).

In preclinical studies, FTIs inhibited the growth of human small cell lung cancer (SCLC) and NSCLC cell lines in vitro and in xenograft models (5, 6, 7, 8). Therefore, it was predicted that FTIs would be effective therapeutic agents for the treatment of lung cancer. The recommended Phase II doses were determined in Phase I studies of R115777 and L-778,123 (14, 15, 16). The dose-limiting toxicities of the agents were fatigue, myelosuppression, and neurotoxicity. The recommended Phase II doses were determined to be 300–400 mg orally twice daily for 14–21 days for R115777 and 560 mg/m² i.v./day for 14 days for L-778,123. To determine whether these FTIs have any clinical activity against lung cancer, there have been three Phase II clinical trials performed in the United States for patients with lung cancer: R115777 for patients with previously untreated NSCLC; L-778,123 for patients with previously untreated NSCLC; and R115777 for patients with sensitive relapsed SCLC. The clinical results of these trials are summarized in this article.

Patients with stage IIIB or IV NSCLC with a performance status of 0–2 who had not received chemotherapy were eligible for this trial (17). Patients were not eligible if they have been treated with radiation therapy to >25% of their bone marrow. The patients were treated with R115777 300 mg p.o. twice daily for 21 days of a 28-day cycle. The clinical parameters and toxicity were monitored in each 28-day cycle, and response was assessed with radiographs every 2 cycles (8 weeks). The drug was given until disease progression or until unacceptable toxicity was encountered.

Patients with stage IIIB with a malignant pleural effusion or stage IV NSCLC with a performance status of 0–1 who had not received chemotherapy were eligible for this trial (18). The patients were treated with L-778,123 560 mg/m² i.v. daily for 14 days of a 21-day cycle. The clinical parameters and toxicity were monitored in each 21-day cycle, and response was assessed radiographically every 2 cycles (6 weeks). The drug was given until disease progression or until excessive toxicity was encountered.

Patients with sensitive relapsed SCLC with a performance status of 0–1 were eligible for this trial (19). The patients were treated with R115777 400 mg p.o. twice daily for 14 days of a 21-day cycle. The clinical parameters and toxicity were monitored each cycle, and response was assessed with radiographs every 2 cycles (6 weeks). The drug was given until disease progression or until excessive toxicity was encountered.

There were 44 eligible patients enrolled in this trial at the Mayo Clinic and the University of Chicago between 2000 and 2001 (Table 1). The patients were treated with a median of 2 cycles with a range of 1–23 cycles (17). The most common grade 3 and 4 toxicities were neutropenia, leukopenia, dyspnea, and diarrhea (Table 2). Grade 3 dyspnea developed in 3 patients and grade 3 fatigue in 2 patients. There were no objective responses. There were 7 patients who stayed on treatment for >6 months (7, 8, 9, 11, 14, 19, and 23 months). The patients on therapy 19 and 22 months remain on treatment. The median time to progression was 2.7 months (95% confidence interval: 1.9–3.7 months). The median survival was 7.7 months (95% confidence interval: 6.3–10.5 months). Seventy percent of the patients were able to go on to conventional chemotherapy after completing R115777.

There were 23 eligible patients enrolled in the trial at the Massachusetts General Hospital and the Dana-Farber Cancer Institute between 1999 and 2000 (Table 1). The patients were treated with a median of 2 cycles of L-778,123. The hematological toxicity reported in the abstract does not appear to be as prominent as with R115777 (Table 2). Thromboembolic toxicity was observed in the patients who required central venous access for the 14-day continuous infusion. There were 3 patients who developed a pulmonary embolism and 1 patient had a superior vena cava thrombosis. The other grade 3 and 4 toxicities were a grade 3 hemoptysis in 1 patient and grade 3 fatigue in 1 patient. Nineteen of the 23 patients completed at least 1 cycle of chemotherapy; no objective responses were observed. The median survival was 11.4 months, and the 1-year survival was 48%.

There were 22 eligible patients enrolled in the trial at seven different institutions (Table 1). The patients were treated with a median of 2 cycles with a range of 1–6 cycles. The most common grade 3 and 4 toxicities were neutropenia and thrombocytopenia (Table 2). There was 1 patient who developed grade 3 fatigue. There were no objective responses. The patient who remained on therapy for the longest time was treated for 4.7 months. The median time to progression was 1.4 months.

Three Phase II trials of FTIs for patients with lung cancer have been reported in article (14) or abstract form (17, 18, 20). The agents have been relatively well tolerated with infrequent and reversible grade 3 and 4 toxicities. Nonetheless, no antitumor activity was observed in any of the three trials, and the time to progression is relatively short (2 cycles or ∼6–8 weeks). This is less than the 3–4 cycles of platinum-based chemotherapy able to be administered in patients with previously untreated NSCLC and the 4 cycles able to be given to patients with sensitive-relapsed SCLC (21, 22, 23).

The study of the trial using R115777 in NSCLC documented at least partial inhibition of farnesyl transferase activity in >80% of the patients, indicating that the drug was having the intended effect in surrogate tissues, the buccal mucosa, and peripheral blood mononuclear cells (17). The two abstracts did not report assessment of farnesyl transferase inhibition (18, 19). Thus, surrogate tissues show a biological effect at the dose and schedule used in two of the trials. These results suggest that more compelling preclinical data will be needed in lung cancer before embarking on clinical trials of unselected patients. This could include a new agent from the class of FTIs with more in vitro or in vivo activity, identification of a subclass of patients whose tumors are more likely to respond, or biomarkers that suggest increased clinical benefit.

Another FTI, lonafarnib, has been combined with paclitaxel in a Phase I/II trial for patients with taxane-refractory/resistant NSCLC (24). Five of 33 (15%) patients treated had a partial response to the combination. This outcome of this trial has prompted a randomized Phase III trial for previously untreated patients with stage IIIB or IV NSCLC. The patients are treated with paclitaxel plus carboplatin with or without lonafarnib, and we await the report of the trial (search lonafarnib).¹ If the results of this trial do not show any survival advantage for the patients treated with paclitaxel and carboplatin with lonafarnib, we do not believe there is adequate clinical or preclinical evidence to perform combination studies with FTIs until there is additional research to support these trials.

FTIs have demonstrated evidence of antitumor activity with complete or partial response rates of 10–30% as a single agent in several types of malignancies. These include breast cancer (25), acute leukemias (26), and chronic myelogenous leukemias (27). The identification of the biological characteristics leading to responses in these cancers may help identify biomarkers associated with an increased likelihood of responding and prompt additional trials in patients with lung cancer.

Dr. Paul Bunn: Maybe Dr. Herbst wants to comment on SCH66336 [lonafarnib] in relapsed non-small cell.

Dr. Roy Herbst: In taxane-refractory patients, there was evidence of activity, which is what Dr. Perez-Soler was talking about yesterday, taking a resistant population and making it sensitive with the agent. These were the best data for that approach, I believe. I thought that that would be a good trial to do in a larger group, but of course, the trial being done is the frontline trial therapy.

Dr. Bunn: Six of 12 patients had partial response to lonafarnib in combination with chemotherapy in the second-line setting. There is an ongoing randomized Phase III trial, which has probably accrued half of the patients, comparing carboplatin and paclitaxel to carboplatin/ paclitaxel with lonafarnib.

Dr. Alex Adjei: They are not testing the hypothesis that the FTI would potentially reverse resistance. The positive results occurred in patients who had paclitaxel before and whether they are resistant or refractory is open to definition. Now they are moving it to first line, where they are basically combining the paclitaxel/carboplatin and hoping that to be better than paclitaxel and carboplatin. It’s a difficult thing to interpret because there are issues of sequencing. This a study that most of us suggested they not do that way, as opposed to doing a second-line study in patients who have had a taxane. We saw a preclinical synergy with cisplatin, and there were additive interactions with gemcitabine, and we did a Phase I study, where we saw some unusual responses. We come back to this patient selection issue because we have a couple of patients still on our Phase II study with single-agent FTI and 1 patient is ∼3 years. If you look at those who stayed on for a long time, it’s maybe 5 patients out of 44 who went over a year. And what is it about those patients? There’s probably something about their tumor that made them respond, and the question is what is it and whether the phenotype is frequent enough to be worth investigating.

Dr. Geoffrey Shapiro: One issue that comes up with these compounds is whether ras is indeed the only target. There are a lot of proteins that are farnesylated, including a number of the kinetochore proteins. Effects on these proteins may in part explain the synergism with taxanes.

Dr. Bruce Johnson: If ras was the target and having mutated ras made a difference, these are not particularly active in pancreatic carcinoma, which has the highest frequency of it.

Dr. Adjei: It is possible that some of the ras isoforms could be targets, but it’s clearly not a mutated ras, so it could be wild-type ras, which is activated by upstream growth factors. For instance, it’s very active in acute myelogenous leukemia and probably will be approved in acute myelogenous leukemia. In the Phase I studies in refractory acute myelogenous leukemia phase, there was a 29% response rate, and none of those tumors had ras mutations. They are doing a Phase II study now and getting about the same response rate. It’s active in breast cancer, very low ras mutations, ras, we all know, activated growth factors. It’s active in glioma, some activity in head and neck. So, you’re right, we don’t know what the target is. It could be wild-type ras, it could be some other farnesylated proteins.

Dr. Thomas Lynch: With Dr. Adjei’s study on untreated patients and the study that we did also on untreated patients, do you think it makes sense to do more studies of this approach with this drug any longer in NSCLC or do you think it’s dead?

Dr. Johnson: With Jansen’s compound, R115777, there are hints of activity. I still think this class of compounds can be studied in previously untreated patients to look for antitumor activity. I think we’ll find out from the Phase III study in the other setting if there is any hint.

Dr. Lynch: It helped Merck for us to do that trial in 22 untreated lung cancer patients. It was negative there and in pancreatic and colon cancer, and they decided not to develop the drug any further. They were able to make that decision on <100 patients because there was absolutely no single-agent activity.