Southwest Oncology Group Study Finds "Unprecedented" Results for Docetaxel in Non-Small Cell Lung Cancer

David R. Gandara, MD, University of California Cancer Center, Davis and the Southwest Oncology Group

At the 2000 meeting of the American Society of Clinical Oncologists (ASCO), investigators from the Southwest Oncology Group (SWOG) announced Phase II trial results they labeled as "unprecedented" in the treatment of non-small cell lung cancer (NSCLC). The study, SWOG 9504, added docetaxel (Taxotere®) sequencing after initial treatment with the combined modality of radiotherapy plus cisplatin (Platinol®)/etoposide (VePesid®).

According to Dr. David Gandara, the results of SWOG 9504 are "unprecedented in the literature . . . and [that is] the reason I am here in front of you today. This [study concentrates on] patients in whom surgery or radiotherapy alone gives an expected survival of less than ten percent. More recently, combinations of chemotherapy and radiotherapy have improved on this somewhat. But the results we achieved, with a two-year survival of 50 percent, have never been reported before."

One-year survival was 78 percent, median survival was 22 months, and median, progression-free survival was 15 months. The overall response rate was 63 percent, with four percent complete responses and 59 percent partial responses. Stable disease was noted in 28 percent and disease progression in nine percent (seven patients were not assessed), he reported.

The study included 83 patients with stage IIIB status documented through biopsy of the primary tumor or of the lymph nodes. Patients were treated with cisplatin 50 mg/m2 on days one, eight, 29, and 36; with etoposide 50 mg/m2 on days one-five, and days 29 through 33; and with concurrent radiotherpay [RT (1.8-2.0 Gy/day)] starting on day one (61 Gy total). This was followed by consolidation docetaxel 75-100 mg/m2 every 21 days for three cycles.

Study Rationale

Previously, SWOG 9019 utilized a similar regimen but treated patients after the chemoradiotherapy with additional cycles of cisplatin/etoposide. One-year survival with that regimen was 58 percent, two-year survival was 34 percent, and median survival was 15 months. "At the time, we thought this was rather good," Dr. Gandara commented.

But the rationale to substitute docetaxel for the additional cisplatin/etoposide was based on the drug's efficacy as second-line therapy. "Docetaxel . . . is a very active new agent in this disease, and it appears to have the very best activity response in lung cancer once patients are refractory to prior treatment," Dr. Gandara explained.

Furthermore, he added, docetaxel's potential molecular mechanism of action favors sequencing, specifically the fact that docetaxel does not require the p53 for best treatment effect (p53 is overexpressed as part of a germline mutation and is believed to confer susceptibility to malignancy).

Acceptable Toxicity

Docetaxel treatment was relatively well-tolerated. Grade 3-4 neutropenia of brief duration was seen in 56 percent of the patients, and only six of 83 patients developed febrile neutropenia. The chemoradiation phase caused esophagitis in 11 percent of patients. There were five cases of pneumonitis, one infection-related death, and three deaths from pulmonary complications. Dr. Gandara emphasized that the ease of administering docetaxel, and the fact that only three doses were required, probably contributed to the success of this treatment approach. "I think patients were more accepting of [this] sort of regimen," he stated.

"We are extremely encouraged by our findings, which are unprecedented in terms of survival," he said. But because this is a Phase II trial, he expressed the need to view the results with some caution. However, Dr. Gandara predicted the results would hold up since treatment and patient-related factors were well-matched and other relevant factors were controlled. "Now that patients are getting out to around three years [post-treatment], essentially none have died," he noted. "That makes us think that we are going to have a plateau, which is very encouraging."

At the moderated poster session, Katherine Pisters, MD, of MD Anderson Cancer Center, Houston, TX, commented, "I think this is a very impressive survival rate, though the pulmonary toxicity is a little worrisome. But it's an interesting concept-immediate sequencing to docetaxel as an adjuvant drug or a second-line drug. There will be roles for newer agents, such as matrix metalloproteinase inhibitors or anti-angiogenesis agents, but we need to remember that this cytotoxic drug has done a very good job in this setting."

Benefits of Chemotherapy in Advanced Lung Cancer Confirmed In Eastern Cooperative Oncology Group, Project #1594

Joan Schiller, MD, University of Wisconsin Medical School, Madison, and the Eastern Cooperative Oncology Group (ECOG)

The benefits of chemotherapy in the treatment of advanced non small-cell lung cancer (NSCLC) have been confirmed by the largest clinical trial ever conducted in this disease. Results from this study were reported at the plenary session of the American Society of Clinical Oncologists.

Current combinations have been yielding one-year survival rates of about 35-40 percent, up from 20-25 percent achieved only five years ago. But until ECOG 1594, the newer combination regimens had never been directly compared. The results from this study did not demonstrate the superiority of one single regimen over another.

"There was no statistically significant difference in survival in any of the three comparator arms when compared to cisplatin (Platinol®) and paclitaxel (Taxol®) [the reference arm]," reported the principal investigator of the study, Joan H. Schiller, MD, Professor of Medicine at the University of Wisconsin Medical School in Madison.

The randomized, multicenter Phase III trial, funded by the National Institutes of Health (NIH), enrolled 1,207 treatment-naive patients with inoperable stage IIIB NSCLC with pleural or pericardial effusion, or stage IV disease. Patients were randomized to one of four different treatment regimens: the reference arm combining paclitaxel and cisplatin (A), or "comparator" arms featuring gemcitabine (Gemzar®) plus cisplatin (B), docetaxel plus cisplatin (C), and carboplatin (Paraplatin®) plus paclitaxel (D). The objectives of the study were to compare overall survival in each of the three experimental arms with the reference arm. Secondary objectives were to determine response rate, time to progression, and toxicities. Median follow-up as of May 1, 2000, was 7.7 months.

Results based on 1,163 patients eligible for analysis showed no statistically significant differences between the four arms in terms of overall response rate (18.7 percent), overall one-year survival (33.5 percent), and overall two-year survival (12 percent). The gemcitabine/cisplatin arm, however, gave a statistically significant one-month advantage in time to progression, (4.5 months versus 3.5; p = 0.02) but was more toxic than the other arms, possibly due to the heavier doses of cisplatin and the more frequent dosing (see Table 1, page 3), Dr. Schiller reported.

The design of ECOG 1594 was based on results of an earlier ECOG trial (Bonomi, P. et al., J Clin Oncol 2000;18:623-631) that compared two dose levels of paclitaxel and cisplatin to cisplatin and etoposide. The lower-dose paclitaxel arm produced a higher response rate, better survival, and fewer side effects than the other two arms, and thus served as the reference arm for this ECOG trial.

Uniform Outcomes

The outcomes were remarkably uniform across all drug combinations and patient groups (Table 1 below). There was no difference in the primary endpoint of survival in any of the three comparator arms, nor for stage IIIB patients compared to stage IV patients across all treatment arms. However, performance status did impact survival. Patients with the best performance status (PS 0) achieved a median survival of 10.7 months compared to three months for the PS 2 patients. (The trial initially enrolled PS 0-2 patients, but this was amended to allow only PS 0-1 patients, due to the toxicities observed in the PS 2 group.)

Carboplatin/paclitaxel produced a lower response rate and more neuropathy, however, its overall side effect profile was the most favorable (nine percent versus 24 percent in the reference arm). The gemcitabine/cisplatin arm achieved the longest time to progression but was associated with more renal toxicity and other side effects, and it contained the highest percentage of patients who terminated their participation early due to complications (28 percent versus 15 percent in the reference arm-a statistically significant difference). Docetaxel/cisplatin was associated with more hypersensitivity reactions. Dr. Schiller remarked that the results give reassurance to clinicians that they have a number of good options in the treatment of advanced NSCLC, and their selection of agents can be based on considerations other than efficacy. This study raises some additional questions regarding each regimen's impact on the quality of life and the cost-effectiveness of each arm, which was not included in the analysis, she stated.

In addition, she said, "Future studies may need to control for impact by second-line therapies." These were not recorded in the current trial, and there is speculation that undocumented treatment with docetaxel as a second-line agent may have influenced the overall positive results.

Formal Critique of Trial Results

Francis Shepherd, MD, Professor of Medicine at the University of Toronto, Canada, also remarked about the second-line therapy issue and its potential impact on the trial's results. "Is it possible that a benefit in overall survival was not seen in the ECOG trial because many, if not all patients, crossed over to second-line therapy, thus, in terms of total therapy many patients may have received similar treatment? This will remain unanswered, since the details of second-line therapy were not recorded."

Expanding upon the potential influence of second-line therapy, Dr. Shepherd noted that an earlier trial that randomized patients to second-line therapy with docetaxel 75 mg/m2 versus best supportive care found a significantly longer median survival, 4.6 months versus 7.5, and one-year survival of 37 percent versus 12 percent for the docetaxel arm. She also commented that the docetaxel/cisplatin arm contained PS 2 patients who had "particularly poor outcomes. Any evaluation of this combination must await the randomized trials," she said. Docetaxel, one of the newest agents to join NSCLC treatment regimens, is being paired in an upcoming trial with cisplatin and with carboplatin, and these two arms will be compared to vinorelbine (Navelbine®)/cisplatin.

Dr. Shepherd said that "photo finish" results of ECOG 1594 were "somewhat of a disappointment to many of us." In addition to the lack of superiority of one regimen over another, the regimens produced lower response and survival rates than a number of other trials that have evaluated the same drugs, she pointed out.

The discrepancies may be due in part to the poor results achieved in PS 2 patients. She also suggested that the higher proportion of stage IIIB patients relative to stage IV patients in this trial might have contributed to the discrepancies.

While no new "gold standard" in care can be extrapolated from the results of ECOG 1594, Dr. Shepherd suggested that several options exist for first-line therapy. For second-line therapy, she suggested moving to docetaxel. "If I had recurrent disease," she said, "I would pick [docetaxel] Taxotere® for further therapy."

Phase III International Trial Shows Trend Favoring Use of Docetaxel as a Single-Agent, Neoadjuvant Therapy with Radical Treatment of Stage III Disease

Karin Verena Mattson, MD, Helsinki University Centre Hospital, Finland

A large international study has shown a survival advantage for neoadjuvant chemotherapy in selected patients with stage III NSCLC, according to Dr. Karin Verena Mattson. While the study of 277 patients did not show a difference according to the intent-to-treat (ITT) analysis, in an analysis of the fully treated population a statistically significant difference emerged for overall survival and median survival, Dr. Mattson said.

The current study confirms results from smaller studies by Roth (Roth, J.A. et al., J Natl Cancer Inst 1994;86:673-680) and Rosell (Rosell, R. et al., N Engl J Med 1994;330:153-158), indicating that chemotherapy should be included in the standard treatment of patients with locally advanced NSCLC, she said. Appropriate local treatment was determined (radical radiotherapy or surgery). Patients were then randomized to one of two study arms: a neoadjuvant treatment arm consisting of docetaxel (Taxotere®) (100 mg/m2, given over one hour by IV every three weeks for three consecutive cycles) followed by LT; or an immediate LT treatment arm consisting of radical radiotherapy or surgery. The mode of local treatment was predominantly radiotherapy (80 percent) versus surgery (20 percent). Complete surgical resections were possible in 71 percent. Response rate to docetaxel was 26 percent (nine percent complete responses and 17 percent partial responses).

Analysis was done on both the ITT population and the full treatment population, which included patients who received at least two cycles of docetaxel and either radical radiotherapy or surgery. In the randomized ITT population (n = 274) followed for a median of 14 months, 77 patients were still alive at analysis: 42 in the neoadjuvant arm, and 35 in the local treatment arm. One-year survival was approximately 60 percent in each arm, and median survival was 15 and 14 months for the docetaxel and local treatment arms, respectively.

In the full treatment analysis (n = 258), however, neoadjuvant chemotherapy was associated with a highly significant benefit. Patients receiving docetaxel had an increase of seven months in median survival: 21 months versus 14 months for local treatment only. One-year survival increased from 56 percent with local treatment to 75 percent with neoadjuvant therapy, Dr. Mattson reported (both p = 0.0014).

Neoadjuvant treatment also delayed time to progression significantly, according to the full treatment analysis. Median time to progression was 15 months in the neoadjuvant arm versus eight months in the LT arm (p = 0.005).

The benefit of neoadjuvant docetaxel also appeared in the various stage subgroups, she added, but patient numbers per subgroup were too small to permit statistical conclusions.

The addition of docetaxel produced modest toxicity, including febrile neutropenia or infection (without prophylaxis) in 15 patients. There were two treatment-related deaths. Cumulative nonhematologic toxicities were not noteworthy, since patients received only three cycles of docetaxel.

"Overall, the docetaxel treatment did not hinder our ability to deliver [docetaxel] Taxotere® and subsequent local therapy," Dr. Mattson remarked, adding that the majority of patients received all three planned cycles.

Dr. Mattson said that the survival analysis is still preliminary since 70 patients are still alive, and she is hoping that statistically significant differences will emerge in the ITT group, as were demonstrated in the full-treatment group.

"Since 1995, our understanding of the biology of stage III subgroups has improved and we would not use the same design for all subgroups today," she conceded. Nonetheless, she maintained, the current results "strongly support" the general statement that all stage III patients should be considered for combined modality treatment.

"Taxotere® [docetaxel] as a single agent is a good non-platinum option as a neoadjuvant treatment in this disease and should be studied further in chemotherapy combinations and in concurrent radiation strategies," she concluded.

Docetaxel Plus Cisplatin Versus Docetaxel Plus Irinotecan in Advanced NSCLC: A West Japan Thoracic Oncology Group Phase II Study

Koji Takeda, MD, for the WJTOG Investigators, Osaka City General Hospital, Japan

A randomized Phase II Japanese cooperative study in 108 patients has found the combination of docetaxel (Taxotere®) plus irinotecan (Camptosar®) to be a promising non-platinum regimen in advanced NSCLC. It had comparable activity and toxicity to docetaxel plus cisplatin in this large randomized Phase II study by Japanese investigators.

Dr. Takeda reported the experimental regimen had a comparable activity and toxicity profile to the more standard docetaxel plus cisplatin (Platinol®) regimen. Relative response rate was 31.6 percent for docetaxel/irinotecan (DI) and 37.3 percent for docetaxel/cisplatin (DC), which included 18 and 19 partial responses, respectively, and no complete responses in either arm. Stable disease was noted in 25 patients treated with DI and 22 treated with DC. Progressive disease occurred in 14 and six patients, respectively, Dr. Takeda reported.

The DI arm received docetaxel 60 mg/m2 on day eight and irinotecan 60 mg/m2 on days one and eight. The DC arm received docetaxel 60 mg/m2 on day one and cisplatin 80 mg/m2 on day one. The regimens were repeated every three weeks.

At the time of analysis, median time to progression was 18.1 weeks for DI patients and 21.3 for DC patients, with a median survival of 45.6 weeks for DI-treated patients and 50.3 weeks for the DC arm.

Toxicity in the DI and DC arms, respectively, was as follows: grade 3-4 anemia, 8.8 percent and 13.7 percent; grade 3-4 neutropenia, 82.5 percent and 74.5 percent; grade 3-4 thrombocytopenia, 1.8 percent and 2.0 percent; grade 2-4 serum creatinine, 15.8 percent and 17.6 percent; grade 2-4 vomiting, 52.6 percent and 86 percent; grade 2-4 diarrhea, 43.9 percent and 22 percent. Three treatment-related deaths occurred in the DC arm and none in the DI arm.

Second-line Weekly Paclitaxel in Patients with Advanced NSCLC Failing First-Line Carboplatin/Paclitaxel

Merrill S. Kies, MD, Anderson Cancer Center, Houston, Texas

In this multi-institutional trial, low-dose weekly paclitaxel (Taxol®) was given to advanced NSCLC patients failing first-line therapy with paclitaxel/carboplatin (Paraplatin®) as a means of extending the efficacy of the taxane while minimizing toxicity. The study found improved survival and a good toxicity profile with a weekly single-agent paclitaxel dose of 80 mg/m2 in 44 patients, reported Dr. Kies.

After receiving first-line therapy with paclitaxel/carboplatin, patients were divided into three groups. Group A included refractory patients who progressed after two cycles of first-line chemotherapy. Group B received four cycles of first-line chemotherapy and then was observed until progression. Group C patients received continuous paclitaxel/carboplatin until progression. The overall response rate to first-line chemotherapy was 26.6 percent.

A total of 45 patients (44 evaluable) received second-line paclitaxel after disease progression, for a median number of eight weekly treatments. Median survival from the start of second-line therapy was 5.9 months for all patients, with no differences seen among the groups.

The overall response rate to second-line weekly paclitaxel was 6.8 percent. Stable disease was seen in 38.6 percent of patients at eight weeks. The majority of the responders were patients who had received four cycles of first-line chemotherapy and were off of treatment for a median of four months before starting with weekly paclitaxel. Patients who progressed after two cycles or continuous first-line treatment did not respond to the weekly approach, Dr. Kies reported.

One-year survival was 39 percent, two-year survival was 23 percent, and median survival was almost 11 months for the group as a whole. By treatment subgroup, median survival was 6.3 months for Group A; 12.5 months for Group B; and 11.7 months for Group C. Median two-year survival by group was, respectively, 12 percent, 54 percent, and 45 percent.

Grades 2 and 3 neuropathy occurred in seven patients, and there were no grade 4 toxicities. In 26 patients assessed for quality of life, there were no differences between the three groups.

"The overall survival in this study looks promising in the early returns," Dr. Kies said at the poster session. "This is not a randomized prospective trial, but response and survival were better in the group of patients who had interrupted treatment.... This is consistent with the general impression that patients who have had a treatment-free interval are more likely to respond to a second-line program."

Second-Line Therapy with Weekly Gemcitabine and Monthly Docetaxel in NSCLC

C. Harris Spiridonidis, MD, Columbus Community Clinical Oncology Program, Ohio

By combining gemcitabine (Gemzar®) and docetaxel (Taxotere®), this Phase II study of second-line therapy in NSCLC essentially doubled the antitumor effects previously reported with each drug alone. In addition, durable responses were observed at all disease sites and in tumors refractory to prior platinum therapy.

"This resulted in one of the highest response rates ever seen in this group of relapsed lung cancer patients," commented Dr. Spiridonidis, principal investigator of the study.

The study from the Columbus Community Clinical Oncology Program enrolled 40 patients with biopsy-proven locally advanced or metastatic NSCLC who had previously been treated with a single chemotherapy regimen, excluding gemcitabine or a taxane. Most patients had received prior platinum-based chemotherapy combined with either vinorelbine (Navelbine®) or etoposide (VePesid®), and four patients had received prior single-agent therapy without platinum.

All subjects received gemcitabine 800 mg/m2 by IV over 30 minutes on days one, eight, and 15 every four weeks, plus docetaxel 100 mg/m2 by IV over one hour on day one after gemcitabine, also every four weeks.

Responses were noted in four of 14 treatment-refractory patients, six of 18 treatment-resistant patients, and two of six treatment-sensitive patients, with responses seen in eight metastatic sites. Median survival was eight months, and one-year survival was 32 percent. The median time to progression for responding patients was eight months, and two responders currently remain progression-free at over 14 and 18 months, he reported.

Grade 4 neutropenia occurred in 21 patients, and febrile neutropenia occurred in four. Grade 3-4 thrombocytopenia occurred in eight patients and transfusion was necessary in nine patients. The only other notable side effect was asthenia in 13 patients. Few patients developed other toxicities, Dr. Spiridonidis reported.

Weekly Docetaxel in Second-Line NSCLC: Phase II Trial

R. Lilenbaum, MD, Mount Sinai Community Clinical Oncology Program, Miami Beach, Florida

Based on the positive effect of docetaxel (Taxotere®) on survival as second-line therapy for advanced NSCLC, and the proven reduction in toxicity with a weekly schedule, this Phase II trial was designed to test the efficacy and toxicity of weekly docetaxel in patients with recurrent or refractory NSCLC and one prior chemotherapy regimen (adjuvant or metastatic).

The study included 23 patients who received docetaxel 36 mg/m2 a week over 15 minutes for six consecutive weeks, every eight weeks. Prior therapy consisted of gemcitabine (Gemzar®)/vinorelbine (Navelbine®) in seven patients, carboplatin(Paraplatin®)/paclitaxel (Taxol®) in six patients, and a variety of single agents or combinations in the remainder. Patients received a total of 134 treatment doses.

Two partial responses were noted in the 16 evaluable patients to date, for a 12 percent response rate. Toxicity was less than with the every-three-week regimen, with no episodes of grade 4 neutropenia and only three patients (13 percent) developing grade 3 neutropenia. There were no episodes of grades 3 or 4 anemia or thrombocytopenia. Nonhematologic toxicity consisted mainly of severe diarrhea, which was dose-limiting in six patients (27 percent). Two patients had grade 2 peripheral neuropathy and one had fluid retention after two full cycles. The study is still accruing patients.

Future Directions for New Cytotoxic Agents in Advanced Stage NSCLC: A Summary

Karen Kelly, MD, University of Colorado Cancer Center, Denver

Several cytotoxic agents have demonstrated the ability to prolong survival in NSCLC, with less toxicity than older agents, but a number of questions remain regarding their optimal use-for instance, their role as single agents and as members of two-drug and three-drug regimens, and their relative superiority to each other and to platinum compounds. Dr. Kelly addressed these issues at an educational symposium.

In the early 90s, six new chemotherapy agents were found to have activity in lung cancer: the taxanes (docetaxel [Taxotere®] and paclitaxel [Taxol®]), vinorelbine (Navelbine®), gemcitabine (Gemzar®), and the topoisomerase I inhibitors (irinotecan [Camptosar®] and topotecan [Hycamtin®]). Several randomized trials showed the advantage of combining these new agents with cisplatin (Platinol®) versus cisplatin alone, thus, the new standard of treatment in NSCLC became a novel agent plus a platinum.

The next issue then became the selection of one novel agent over the other. The Southwest Oncology Group conducted the first completed trial comparing such new regimens in patients receiving vinorelbine/cisplatin or paclitaxel/carboplatin (Paraplatin®). There was no significant difference in response (28 percent versus 25 percent), median survival time (eight months in both arms), and one-year survival (36 percent versus 38 percent), respectively. There were modest differences in toxicity-basically swapping one toxicity for another-so SWOG recommended either regimen be used in advanced NSCLC.

While ECOG 1594 (described on page 2) compared four different treatment combinations with the intention of determining the superior approach, outcomes in all arms were essentially equivalent, therefore, the question of which combination is optimal remains unanswered.

In the continuing effort to improve survival, several avenues are being explored, among them:

Non-platinum doublets. The combination of vinorelbine and gemcitabine given weekly has been extensively studied. Response rates have ranged from 19 percent to 73 percent, and median survival from nine to 12 months, with good tolerability. Other combinations of agents are now being evaluated, including paclitaxel plus gemcitabine and docetaxel and gemcitabine, which show response rates ranging from 33 percent to 38 percent with good tolerability. A randomized Phase II trial by the Cancer and Leukemia Group B (CALGB) will compare docetaxel plus irinotecan versus docetaxel plus gemcitabine.

Addition of a second new agent to the active regimens. This has not yet proved to have additional benefit over two agents, but the milder toxicity of the newer agents may make a difference. Such combinations include docetaxel/carboplatin/gemcitabine, paclitaxel/carboplatin/gemcitabine, and vinorelbine/cisplatin/gemcitabine, among others. Several trials have shown improvements in survival by the addition of gemcitabine, with median survival time up to almost ten months and one-year survival rate up to 45 percent. Sequential administration of agents. This allows efficacious doses to be given without overlapping toxicities. Sequencing with non-cross-resistant agents might also overcome drug resistance. [The report at ASCO by Gandara (page 1) suggests the potential benefit of this approach when combined with radiotherapy.]

Single-agent therapy. Platinum-containing regimens have been the mainstay of therapy in lung cancer for many years, therefore, investigators have been hesitant to exclude them from treatment regimens. However, it is possible they are less effective than the newer agents and are only adding to toxicity, therefore, a number of trials are evaluating single-agent therapy, often as a weekly regimen. Weekly taxane therapy in advanced NSCLC has proved to be an active approach that is better tolerated than a three-weekly schedule. While most studies have included weekly paclitaxel, newer studies are suggesting that docetaxel is also efficacious and less toxic than docetaxel given on a three-weekly schedule.

New chemotherapy agents. Developing newer, more active cytotoxic agents is a constant pursuit. Among these are tirapazamine, multitargeted antifolate (MTA), oxaliplatin, tegafur, and uracil (UFT). The hope is that these and other novel agents will continue to further extend survival in advanced lung cancer.