In their excellent review of firstline chemotherapy for patients with advanced non-small-cell lung cancer (NSCLC), Laskin and Sandler are clear, direct, and positive. Chemotherapy works in stage IV disease:

• Survival is improved.
• There is clinical benefit.
• Toxicities can be mitigated.
• Older (> 70, > 80) and poorer performance status (PS 2) patients can benefit.
• Treatment does not need to be given "ad infinitum."
• In addition, second- and thirdline therapies add further to the improved survival duration. But the message is appropriately tempered. There is a chemotherapy efficacy plateau:
• The flow of new cytotoxics that characterized the 1990s has slowed.
• Mixing and matching agents into two-drug combinations hasn't given us a clearly "best treatment."
Three cytotoxic drug regimens add to toxicity (and perhaps response) and do not improve survival.
• We even eschew the "more effectiveplatinum," choosing carboplatin(Drug information on carboplatin) over cisplatin(Drug information on cisplatin) for the stage IV setting based on perceived tolerability.
• Developing molecular characteristics or gene profiles to select one chemotherapy regimen over another remains firmly in the investigative realm

Targeted Therapies
So where is therapy for patients with stage IIIB (wet) and IV NSCLC heading? There are some new, traditional cytotoxics currently under evaluation that may broaden first-line options. But they are unlikely to move us significantly off the efficacy plateau. Instead we have pinned our hopes mostly, if not exclusively, on work product from the molecular revolution, namely targeted therapies. And why not? The track record for targeted therapy in other disease settings has varied from beneficial (eg, bevacizumab(Drug information on bevacizumab) [Avastin] added to chemotherapy in colon cancer) to stunning (imatinib [Gleevec] for chronic myeloid leukemia and gastrointestinal stromal tumors; trastuzumab(Drug information on trastuzumab) [Herceptin] as a component of adjuvant therapy in breast cancer), while the cytotoxic pipeline has flowed more and more slowly. The oral small-molecule epidermal growth factor receptor tyrosine kinase inhibitors have demonstrated activity in NSCLC patients as second- and third-line therapy.[1] Erlotinib (Tarceva) is approved for both indications. Retrospective data on molecular predictors of the likelihood of response, lengthened progression-free survival, and perhaps even overall survival are encouraging, but need prospective testing.[ 2] In phenotypically selected subsets, response rates can exceed 50% (eg, Korean, female, never-smokers with adenocarcinoma) as first-line therapy.[ 3] Yet for the large population of patients with advanced NSCLC, firstline use of erlotinib or gefitinib(Drug information on gefitinib) (Iressa) in combination with chemotherapy has failed to improve outcome over chemotherapy alone.[4-7] As noted by Laskin and Sandler, perhaps using the cytotoxic and targeted agents all together on day 1 was wrong (remember the reduced efficacy of adjuvant tamoxifen(Drug information on tamoxifen) plus chemotherapy followed by more tamoxifen compared to adjuvant chemotherapy followed by tamoxifen for patients with breast cancer[8]). New integrated schedules, such as the day 1 chemotherapy and then day 2 through 16 erlotinib regimen of Davies et al,[9] have preclinical appeal and are in clinical development. But it is likely to be a good while before a company or cooperative group commits the resources for a large-scale phase III trial of this approach in the first-line setting. Bevacizumab Plus Paclitaxel(Drug information on paclitaxel)/Carboplatin: The ECOG 4599 Trial
Real energy has returned to frontline therapy with the recent reporting by Sandler and colleagues of the preliminary findings of the Eastern Cooperative Oncology Group trial ECOG 4599.[10] Use of the three-drug regimen of bevacizumab plus paclitaxel and carboplatin for six cycles followed by additional bevacizumab significantly improves overall survival compared to the chemotherapy alone. The median survival time of > 12 months is a landmark result for multicenter phase III trials (see Table 1 of the Laskin/Sandler article). But even this result requires cautious application. Based on concerns about risks of tracheobronchial hemorrhage or central nervous system bleeding, patients with hemoptysis, with squamous cell carcinoma, with brain metastases, or on chronic anticoagulant or nonsteroidal anti-inflammatory drug therapy were excluded. Initial radiation or chemoradiation (to the chest and/or brain depending on the distribution of disease) followed by chemotherapy and bevacizumab may mitigate bleeding risks. Squamous cancers that are not large and centralmay perhaps be able to receive bevacizumab plus chemotherapy without increased risk. Additional studies are in progress or under development to evaluate the safety and efficacy of bevacizumab plus chemotherapy among expanded groups of patients. But until clear and convincing data are in hand, these patients should remain off limits. Since the only regimen tested was paclitaxel and carboplatin, some wonder whether the same results would occur with one of the other first-line chemotherapy regimens discussed by Laskin and Sandler. This too is under current test in a sponsored European trial of gemcitabine(Drug information on gemcitabine) (Gemzar) and cisplatin alone or with the addition of bevacizumab at doses of 7.5 or 15 mg/kg once every 3 weeks. Given the ECOG 4599 data and similar or more mature data in metastatic breast (Miller et al[11]) and colon cancer (Hurwitz et al[12]), it seems likely that bevacizumab plus chemotherapy does represent another incremental advance in firstline therapy for patients with NSCLC. I expect it is not specific to combination with paclitaxel/carboplatin, but again we need data to be sure. In ECOG 4599 the bevacizumab was started day 1 cycle 1 of chemotherapy and continued after chemotherapy until disease progression. Do we need to give it that way? Are both the concurrent and continuation phases integral to generating the survival benefit? A relatively simple phase III trial could address this question. But is this a key question to generate additional forward motion? Probably not. What is ongoing (the ATLAS trial) is atest of paclitaxel and carboplatin plus bevacizumab followed by bevacizumab plus placebo or bevacizumab plus erlotinib. This will not address the duration of bevacizumab therapy, but rather assess another possible role for oral receptor tyrosine kinase inhibitors in first-line NSCLC therapy. Conclusions
Systemic therapy is now the appropriate, life-prolonging approach to recommend to the large majority of patients with advanced NSCLC. For first-line therapy it should most often be a doublet, usually platinum based. The myriad choices are fully reviewed by Laskin and Sandler. Bevacizumab can and probably should be added in patients who fit the ECOG 4599 entry criteria. Many confirmatory and clarifying studies are ongoing. Nextgeneration targeted agents and targeted therapy combinations are being tested with almost breakneck speed. It is an exciting time to share in the delivery route from bench to bedside.