Small-cell lung cancer (SCLC) is one of the solid tumors most responsive to chemotherapy. Com monly used combination chemotherapy regimens for small-cell lung cancer are usually assembled from five drugs or their analogues. Many permutations and variations of regimens containing cyclophosphamide(Drug information on cyclophosphamide) (Cytoxan) (C), doxorubicin(Drug information on doxorubicin) (Adriamycin) (A), etoposide(Drug information on etoposide), (VePesid) (E), and cisplatin(Drug information on cisplatin) (Platinol) (P) have been reported and a number of regimens have been regarded as standard. The CAV (cyclophosphamide/Adriamycin/vincristine [Oncovin]), CAVE (CAV plus etoposide), EP (etoposide/Platinol), and alternating CAV/EP have been used most consistently in North America. The standard cycle interval is 3 weeks, and the standard duration of chemotherapy is four to six cycles.
The three most common themes of small-cell lung cancer chemotherapy investigations include increased drug delivery by dose and dose-intensity escalation, enhanced drug diversity by drug addition studies or alternating combinations, and new drug trials. Because of overlapping toxicity, an increase in drug diversity is usually accompanied by a decrease in dose intensity of component drugs.
At the University of British Columbia, we attempted to simultaneously increase dose intensity and drug diversity by changing the cycle interval from 3 weeks to 1 week, alternating myelosuppressive and nonmyelosuppressive weeks of therapy, and administering supportive drugs (steroids and antibiotics). The CODE (cisplatin/Oncovin/doxorubicin/etoposide) regimen was designed to deliver as much cisplatin, more doxorubicin, more vincristine, and more etoposide in 9 weeks than the alternating CAV/EP regimen does in 18 weeks. Patients responding to CODE without residual disease outside the chest after chemotherapy received consolidative thoracic irradiation and prophylactic cranial irradiation.
The population treated had extensive small-cell lung cancer, good performance status, and was less than 66 years of age. A recent update of the data shows a complete response rate of 48% (thoracic
response assigned by chest radiograph rather than computed tomography scan), a median survival of 55 weeks, 2-year survival of 20%, and 5-year survival of 10%. The pilot study suggested that the cure barrier could be broken in extensive small-cell lung cancer, allowing a small but definite proportion of long-term survivors.
Two phase III trials in extensive small-cell lung cancer have been completed comparing CODE with alternating CAV/EP. The study by the Japanese Clinical Oncology Group (JCOG) and the Intergroup (National Cancer Institute of Canada/Southwest Oncology Group [NCIC/SWOG]) trial differed in a number of respects. Differences included age eligibility (patients more than 65 years of age allowed by JCOG but not by NCIC/SWOG), etoposide administration route (all etoposide intravenous by JCOG but oral etoposide given on days 2 and 3 in NCIC/SWOG), cyclophosphamide dose (20% lower in JCOG than in NCIC/SWOG), and radiotherapy policy (thoracic and cranial radiotherapy given to responders in NCIC/SWOG but not routinely by JCOG). An important additional variance between the trials was the type of supportive care administered. JCOG used a granulocyte colony-stimulating factor where-as NCIC/SWOG used steroids (prednisone) and prophylactic antibiotics (cotrimoxazole).
The JCOG study was presented at the 1996 American Society of Clinical Oncology meeting by Furuse et al. Notably, the median age of 64 years in the JCOG trial is older than the median age in the NCIC/SWOG trial (58 years). The granulocyte colony-stimulating factor as used by JCOG appeared to allow delivery of the CODE protocol with good fidelity, as about 85% of patients received all 9 weeks of intended treatment within 10 weeks. Although the overall response rate was somewhat higher with CODE (85.3%) than CAV/EP (76.7%), the difference was not statistically significant, and the proportion of complete responses was about 15% in both arms. The survival curves separate in favor of CODE, but the median (11.9 months for CODE and 10.6 months for CAV/EP) differs by only 6 weeks and the P value was not significant.
Based on the results of the Vancouver pilot study, the treatment-related mortality in the NCIC/SWOG trial was anticipated to be 3% to 4%. Unfortunately, after 220 patients were accrued, the toxic death rate in the CODE arm reached 10%. Clearly, the steroids and antibiotics were insufficient to protect patients from CODE toxicity in the multi-institutional setting. Based on toxicity and an interim analysis, the trial was closed in April of 1996. The median dose intensity and total dose in the CODE arm were the same as intended. The median survival of all patients in the trial was about 1 year, and approximately 20% are alive at 2 years.
Taken together, the preliminary results of these two randomized trials do not demonstrate an advantage to doubling the intensity of multiagent chemotherapy in extensive small-cell lung cancer while maintaining a similar total dose. Although the median survival outcomes for CODE in both phase III trials are similar to those in the Vancouver pilot study (about 1 year), the patient populations overall appear to have better prognostic factors than in many extensive small-cell lung cancer trials. Mature results of both JCOG and NCIC/SWOG studies will be of interest with respect to the percentage of long-term survivors.
Based on the results of randomized trials performed to date, there appears to have been little meaningful progress in improving the effectiveness of chemotherapy for extensive small-cell lung cancer over the past 20 years. The median survival time of 30 to 40 weeks and 2-year survival rates of less than 5% are similar for many combination chemotherapy regimens. The EP regimen has become increasingly popular in both extensive and limited small-cell lung cancer because it has acceptable toxicity and four cycles are considered adequate.
A major factor that has driven EP popularity in limited small-cell lung cancer is the manageable toxicity associated with administration of this regimen with concurrent thoracic irradiation. Early reports[4,5] showed that concurrent EP and thoracic irradiation had less pulmonary, cutaneous, and esophageal toxicity than other chemotherapy regimens, and importantly, both modalities could be administered without dose or treatment delay compromise.
Treatment results for limited small-cell lung cancer have improved over the past 10 years, with median survival times increasing from 14 to 16 months to 20 to 24 months and long-term survival rates increasing from 10% to 20%. This improvement is probably more closely associated with early integrated use of chemoradiation than with better efficacy of the chemotherapy or thoracic irradiation components. Although not all studies of thoracic irradiation timing conclude that early chemoradiation is superior, it is notable that no large study of limited small-cell lung cancer has achieved a state-of-the-art long-term survival of 20% without using early chemoradiation.
As shown in Table 1, many cooperative group trials from around the world were based on the belief that EP was the best chemotherapy regimen for small-cell lung cancer. The trials in Table 1 ask interesting questions, but none compares EP with other chemotherapy regimens. It is curious that, to date, the EP regimen has never been shown to be superior when compared with any other combination chemotherapy regimen in randomized trials of either extensive or limited small-cell lung cancer. It appears that a number of chemotherapy regimens are capable of controlling the chemosensitive component of small-cell lung cancer, and new drugs are needed to deal with the resistant clones that are the principal barriers to progress.
The properties that new drugs should have in order to join or displace standard therapies are considerable. Desirable characteristics of an optimal new agent are shown in Table 2. Because analogues of existing drugs have not demonstrated convincing advantages over the parent compounds in controlled trials, new chemotherapeutic agents must have a different structure with a novel mechanism of action and, more importantly, a novel mechanism of resistance.
A truly interesting drug for small-cell lung cancer should have a monotherapy response rate of ³ 40% to 50% in previously untreated patients, and be able to achieve ³ 10% complete responses. Activity in previously treated patients and evidence of non-cross-resistance with cisplatin and etoposide is crucial. Moreover, because of the increasing acceptance of integrated chemoradiation in limited small-cell lung cancer, it should be possible to administer the new agent concurrently with or adjacent to thoracic irradiation for limited small-cell lung cancer without radiotherapy-chemotherapy interactions that cause serious normal tissue toxicity.
A preferential radiosensitizing effect on the tumor would be a great advantage. Activity should be present over a range of schedules so that dose-response and intensity-response characteristics can be defined. Unless the drug has incredible monotherapy activity, it should combine with existing active drugs in an additive or, preferably, a synergistic fashion.
An obvious corollary is that the new drug must have a toxicity profile that does not preclude administration at full therapeutic doses when combined with the best existing agents or combinations. Cumulative toxicity should not decrease the likelihood of completing the entire treatment program at full dosage on time. The well-being of the increased percentage of long-term survivors must not be endangered by chronic toxicity. The ultimate criterion will be the demonstration that the program incorporating the new drug is unequivocally superior to standard therapy in reproducible randomized comparisons.