Pemetrexed (ALIMTA), A Novel Multitargeted Antineoplastic Agent

Perturbation of nucleotide metabolism for cancer therapy has primarily focused on the utilization of nucleoside and nucleic acid base analogs, which compete with their physiological counterparts for incorporation into DNA and RNA (1), as well as antifolate agents. A recent, novel approach to inhibiting purine and pyrimidine synthesis is represented by the development of pemetrexed, which inhibits tumor growth by inhibiting folate-dependent enzymes, primarily thymidylate synthase (TS; Ref. 2).

Pemetrexed enters cells via the reduced folate carrier, with transport kinetics similar to that of methotrexate, and binds to folate receptor-α with a very high affinity, similar to that of folic acid (3). Pemetrexed also appears to be a substrate for multidrug resistance protein transporters (4). Intracellularly, pemetrexed is polyglutamated to the active pentaglutamide by a reaction catalyzed by folylpolyglutamate synthase. Pemetrexed is one of the best substrates for folylpolyglutamate synthase, when compared to other antifolates such as methotrexate (5).

Pemetrexed can be considered a prodrug, because its pentaglutamate form is the predominant intracellular form, and is over 60-fold more potent in its inhibition of TS than the parent compound (2). Polyglutamation traps pemetrexed and enhances its intracellular retention. The parent drug is polyglutamated 90- to 200-fold more efficiently than methotrexate and 6- to 13-fold more efficiently than the glycinamide ribonucleotide formyltransferase inhibitor, lometrexol (6). The increased cellular retention of polyglutamated pemetrexed forms may explain the success of the 3-week administration schedule.

Pemetrexed inhibits multiple enzyme targets involved in both pyrimidine and purine synthesis. One of these primary enzyme targets is TS (7, 8). TS, a folate-dependent enzyme, catalyzes the transformation of dUMP to dTMP. Inhibition of TS results in decreased thymidine necessary for DNA synthesis (7, 9). In addition to TS, pemetrexed inhibits dihydrofolate reductase, aminoimidazole carboxamide ribonucleotide formyltransferase, as well as glycinamide ribonucleotide formyltransferase; the latter is a folate-dependent enzyme that is involved in purine synthesis (10). These targets are related to the cytotoxicity of pemetrexed, because both thymidine and hypoxanthine are required to circumvent cellular death caused by pemetrexed (11). Pemetrexed is 30–200 times more potent an inhibitor of TS than of aminoimidazole carboxamide ribonucleotide formyltransferase or glycinamide ribonucleotide formyltransferase, suggesting that its cytotoxicity may be mediated predominantly through TS inhibition. However, pemetrexed demonstrates activity against H630 colon cancer cell lines, which are resistant to raltitrexed and 5-fluorouracil because of TS amplification (9). Thus, inhibition of the other folate enzymes is important for the clinical activity of pemetrexed.

The role of folate deficiency in the toxicity of pemetrexed has been comprehensively reviewed recently (12). Severe, unpredictable, and occasionally fatal toxicities (predominantly myelosuppression and gastrointestinal) were initially seen with pemetrexed. It was hypothesized that this toxicity may be related to folate deficiency in patients. Data accumulated in the past few years suggest that plasma homocysteine is a more sensitive measure of the functional folate status of patients than RBC or serum folate (13). With this insight into cellular folate metabolism, the role of the folate status of patients in the toxicity of pemetrexed has been delineated.

Niyikiza et al. (14) used multivariate stepwise regression methods to analyze baseline physiological and demographic data from individuals enrolled in early pemetrexed trials. The objective was to identify predictive variables for severe toxicity. Variables analyzed included serum homocysteine, cystathionine, methylmalonic acid, albumin, and other hepatic enzymes. These analyses indicated that an elevated plasma homocysteine concentration was indicative of preclinical folate deficiency and resulted in a more severe toxicity profile that typically included thrombocytopenia, neutropenia, severe diarrhea, and mucositis. The threshold baseline homocysteine value of 10 μm appeared to be the cutoff for these toxicities after cycle 1 (χ² = 6.2, P = 0.01), with levels >10 μm predicting an increased rate of toxicity.

As a result of the above findings, all patients treated with pemetrexed over the past 3 years received concomitant oral folic acid tablets 350-1000 μg starting at least 5 continuous days before pemetrexed administration and continuing throughout the course of therapy, and vitamin B12 1000 μg administered i.m. with folic acid and repeated every 9 weeks while the patient is on therapy. Current data indicate that this supplemental use of vitamins can ameliorate some of the severe drug-induced toxic effects, resulting in an improved safety profile and efficacy for this drug (15). There is an ongoing Phase I study evaluating a combination of pemetrexed and folic acid, with dose escalation of pemetrexed (16).

There are no published data on the evaluation of pemetrexed in small cell lung cancer. Thus, all available clinical data for this agent are in the treatment of non-small cell lung cancer (NSCLC).

Two Phase II studies with single-agent pemetrexed in previously untreated patients with advanced NSCLC have been completed. The starting pemetrexed dose of 600 mg/m² was later reduced to 500 mg/m² because of early toxicity. Four of 8 patients with stage IIIB disease and 3 of 25 patients with stage IV disease achieved partial responses for an overall response rate of 23% (17). Median survival was 9.6 months with a median time to progression of 3.8 months. The second front-line study was conducted jointly in Australia and South Africa. The starting dose of pemetrexed was 600 mg/m². Nine objective responses were achieved in 57 available patients for a response rate of 16%. Median survival was 8.4 months with a median time to progression of 5.6 months (18). These results are comparable to those obtained for single-agent studies of agents that are active in NSCLC. There was no vitamin supplementation in these studies, and the incidence of grades 3 and 4 neutropenia was approximately 20%. Anemia, mucositis, rash, and diarrhea were also common.

In the only study of pemetrexed as second-line therapy for NSCLC, all patients received a starting dose of 500 mg/m² and were classified according to whether their prior regimen did or did not contain a platinum agent. There were 44 evaluable patients in the platinum pretreated group. Two partial responses were seen for a response rate of 5%. Median survival was 6.4 months, median time to progression was 2.3 months, and 1-year survival was 25%. In the non-platinum pretreated group, 5 objective responses (1 complete response, 4 partial responses) were seen in 35 evaluable patients for an overall response rate of 14%. The median survival time was 4 months, median time to progression was 1.6 months, and 1-year survival was 20% (19). Thus, for this entire group, the response rate was 7 of 79 or 9%, again comparing favorably with responses seen with agents that are felt to be active in the second-line setting.

On the basis of the activity seen in this second-line study, a randomized Phase III trial comparing single-agent pemetrexed with single-agent docetaxel as second-line therapy in patients with advanced NSCLC was undertaken. A total of 571 patients were randomized to pemetrexed (500 mg/m² i.v. infusion) or docetaxel (75 mg/m² i.v. infusion). The two regimens demonstrated equivalent efficacy, with a median survival of approximately 8 months. Unlike the Phase II trial, the pemetrexed arm of this study was supplemented with vitamin B12 and folic acid. In the docetaxel arm, grade 3 or 4 neutropenia was seen in 40% of patients, and 7% of patients developed neutropenic fever. In the pemetrexed arm, grade 3 or 4 neutropenia was seen in 5.3% of patients with a 2% incidence of neutropenic fever (20).

Pemetrexed has been combined with cisplatin, carboplatin and oxaliplatin in Phase II studies in NSCLC. Clinical activity has been seen with all three combinations, with modest toxicity.

Two Phase II trials evaluating the combination of pemetrexed (500 mg/m²) and cisplatin (75 mg/m²) in untreated NSCLC have been reported. There was no vitamin supplementation in these studies. In the first study, 36 patients (18 stage IIIB and 18 stage IV) were treated (21). There were 14 partial responses for an overall response rate of 39%. Median survival was 10.9 months and 1-year survival was 50%. The most common toxicity was neutropenia, with a grade 3/4 incidence of 21%. Grade 3/4 anemia occurred in 5% of patients. The most common non-hematological toxicity was fatigue (grade 2/3 in 18% of patients). In the second study, 31 patients (5 stage IIIB and 26 stage IV) were treated (22). There were 13 partial responses for a response rate of 45%. Median survival was 8.9 months, and 1-year survival was 49%. In this study, the most common toxicity was fatigue, with a grade 2/3 incidence in 21% of patients. Grade 3/4 neutropenia occurred in 11% of patients, whereasgrade 3/4 anemia occurred in 6% of patients. Thrombocytopenia was mild and infrequent in both studies.

A randomized Phase II study evaluating pemetrexed plus carboplatin and pemetrexed plus oxaliplatin has been reported. Eighty chemotherapy-naïve patients with advanced or metastatic NSCLC were randomized to receive pemetrexed (500 mg/m² with vitamin supplementation and dexamethasone) plus carboplatin (AUC 6) or oxaliplatin (120 mg/m²), with both drugs given once every 3 weeks. There was less grade 3 neutropenia (5%) in the oxaliplatin arm compared with the carboplatin (26%) arm, although there was less peripheral neuropathy in the oxaliplatin arm compared with the carboplatin arm. Confirmed responses and proportions of patients with stable disease were very similar: 33% and 41% in the pemetrexed/carboplatin and 27% and 44% in the pemetrexed/oxaliplatin arm. The survival data are not mature (23).

The regimen of vinorelbine in combination with pemetrexed has been studied in Australia in the Phase II setting. Pemetrexed (500 mg/m²; day 1) and vinorelbine (30 mg/m²; days 1 and 8) were administered as 10-min infusions every 21 days, with folic acid and vitamin B12 supplementation. The most common hematological toxicity was neutropenia. Fifteen of 35 evaluable patients developed grade 4 neutropenia. Grades 3 and 4 non-hematological toxicities included fatigue in seven patients, dyspnea in five, and nausea in three patients. There was 1 complete response and 12 partial responses for an overall objective rate of 37%, with a median time to progression of 3.8 months (24). Follow-up is brief; thus, survival data are not available at this time.

Adjei et al. (25) performed a Phase I study of pemetrexed in combination with gemcitabine. After two schedules were tested, the recommended doses and schedule for Phase II studies were gemcitabine administered on days 1 and 8 with pemetrexed administered on day 8, 90 min after gemcitabine, at doses of gemcitabine 1250 mg/m² and pemetrexed 500 mg/m². Promising clinical activity was seen in a variety of tumor types including NSCLC in 13 of 55 evaluable patients. From a sample of patients analyzed for pharmacokinetics, the disposition of pemetrexed was not altered by prior administration of gemcitabine.

On the basis of these promising results, a Phase II study has been completed. Fifty-four evaluable patients were treated with 1250 mg/m² of gemcitabine on days 1 and 8 and 500 mg/m² pemetrexed on day 8. Three patients received no vitamin supplementation; 8 patients started vitamin B12 and folic acid supplementation while on study; and 43 patients had vitamin supplementation throughout the study. There were nine partial responses for an overall response rate of 17% (95% confidence interval, 7.9–29.3). Median duration of response was 3.3 months. Twenty-nine patients had prolonged disease stabilization for 4 months or longer. Median survival was 11.3 months with a 44% 1-year survival (26). Neutropenia was the most common toxicity (29% grade 3, 34% grade 4). Grade 3 thrombocytopenia occurred in 5% of patients. There was no grade 4 thrombocytopenia. The most common non-hematological toxicity was fatigue (23% grade 3).

The sequence of drug administration for the Phase I trial of pemetrexed and gemcitabine was based on preclinical studies that revealed synergistic cytotoxicity when gemcitabine exposure preceded pemetrexed exposure in HCT-8 cultured human colon cancer cell lines (27). Another report, however, demonstrated synergistic cytotoxicity for the opposite sequence of pemetrexed exposure followed 24 h later by gemcitabine exposure in HT29 colon cancer cell lines and xenografts (28). These data suggest that patterns of interaction with these two agents may be cell-line specific. In an attempt to identify the optimal sequence of pemetrexed and gemcitabine, a three-arm randomized Phase II study of gemcitabine and pemetrexed, testing various schedules and sequences, has been completed at Mayo Clinic and the North Central Cancer Treatment Group. Results are awaited with interest.

Pemetrexed in combination with radiation therapy is an area of increasing research. A study evaluating this combination in lung cancer is currently ongoing.

Pemetrexed is a novel antimetabolite with inhibitory activity against a number of folate-dependent enzymes. This agent has demonstrated promising single-agent activity in both the first- and second-line treatment of NSCLC. The recently reported Phase III second-line study may lead to regulatory approval of pemetrexed in NSCLC in the next few months. Full-dose combinations with other cytotoxic agents have been feasible because of an acceptable toxicity profile. More importantly, a number of these combinations (gemcitabine, cisplatin, carboplatin, oxaliplatin, and vinorelbine) have demonstrated significant activity in NSCLC. Pemetrexed should be further studied in NSCLC, especially in combination with novel agents.

Dr. Roman Perez-Soler: This seems to me as good a molecularly targeted therapy as anything I have heard. In addition, it works. So, why we don’t propose doing biopsy studies, subset analysis of patients, determinants of sensitivity…

Dr. Alan Sandler: Because it works! That’s why we don’t have to do it!

Dr. Perez-Soler: Well, it would be nice to increase the response rate to 40%, too. This is the methotrexate of the new millennium. Nobody wants to do the studies, but we want to do studies for the drugs that have a 10% response rate.

Dr. Alex Adjei: A neoadjuvant breast cancer study has been done where biopsies were collected, and those data are being analyzed.

Dr. Thomas Lynch: But, it’s not totally there. People like Dr. Rosell are spending time trying to figure out why there are responses to certain platinum agents and I think they haven’t abandoned trying to enhance the response rate of traditional agents.

Dr. Adjei: In terms of what has been looked at recently, everybody was looking at thymidylate synthase, GARFT [glycinamide ribonucleotide formyltransferase], dihydrofolate reductase, because those are the known targets of the drug, but now people are looking at MTAP [methylthioadenosine phosphorylase] also.

Dr. Geoffrey Shapiro: That’s really fascinating because in many cases of mesothelioma, the p16 is deleted, so you would lose MTAP. In lung cancer p16 is usually methylated so that the adjacent MTAP may still be there, and there is a far lower rate of p16 deletion in which you might co-delete MTAP. One simple thing to do, at least among the lung cancer population, would be to see if the response rate could be associated with who deletes p16 and MTAP versus who doesn’t. That would be just one simple question that could be asked. Maybe the reason why the success in mesothelioma is better is because there’s a much higher rate of deletion.

Dr. Rafael Rosell: In the sequential study, you mentioned gemcitabine followed by pemetrexed. Is that the better combination? And what interval do you recommend?

Dr. Adjei: The data are conflicting, depending on the cell line used. It is like Dr. Shapiro’s story with flavopiridol. In the lab the interval was actually 4 hours. But in the clinic we figured that was too long, so we picked 90 minutes as something that is reasonable.

Dr. Rosell: Do you have theories as to why gemcitabine administered before is better?

Dr. Adjei: No. Actually, I would have theorized the opposite, I would have theorized that pemetrexed first should be better, so it’s difficult to understand. One last point I will make in terms of the efficacy with the added vitamins. One thing that became clear to us in our studies is that you can get late responses with this drug, and so it may be that without vitamins a number of patients have toxicity and have to go off study early. Some the effect of the vitamins on efficacy might just be because they tolerated treatment better and they get more treatment.

Dr. Roy Herbst: What struck me is how many cycles you can give and how well tolerated it is. It suggests the potential of this agent in the maintenance setting, because you could give 12 or even 15 cycles, and patients do well with a very convenient schedule. So it might be a reasonable regimen to start building on in the future. SWOG [Southwest Oncology Group] is going to be doing a study in the second line, and we have a trial of bevacizumab plus pemetrexed we are in the process of developing now.

Dr. Lynch: Dr. Scagliotti, you have studied pemetrexed with either carboplatin or oxaliplatin. Any reason to say that one combination is better than the other?

Dr. Giorgio Scagliotti: The median survival time was close to 10 months with both combinations. The most important point is not the survival but rather the lower level of hematologic and nonhematologic toxicity with these two combinations, which makes it more attractive for treatment of stage IV disease. The combination with carboplatin was slightly more toxic in terms of hematologic toxicity; the oxaliplatin combination was slightly more toxic in terms of nonhematologic toxicity, mainly grade 2 and 3 neurosensitivity. Concerning the other point about the feasibility of the combination as maintenance: two-thirds of our patients had more than 4 cycles. In the MD Anderson trial, most patients received more than 4 cycles.

Dr. Herbst: We had it written there was no limit. We had as many as 13 cycles.

Dr. Bruce Johnson: The thing that surprised me in the mesothelioma study is the impact vitamins had on platinum alone. Considering the effect of leucovorin with 5-FU [5-fluorouracil], should we reexamine the vitamin phenomenon with another combination?