The role of chemotherapy in advanced lung cancer was not proven until the mid-1990s. In 1995, a meta-analysis by the Non-small Cell Lung Cancer Collaborative Group showed that cisplatin-based chemotherapy was effective in patients with advanced non-small cell lung cancer (NSCLC) (1). More recently, in 2000, second-line chemotherapy with docetaxel was shown to prolong survival in NSCLC patients previously treated with platinum-based chemotherapy as a first-line chemotherapy treatment (2). As more novel agents have been been shown to be effective against lung cancer, such as EGFR inhibitors, third-line chemotherapy can be accepted as a reasonable therapeutic option (3,4). These types of chemotherapy were also shown to improve quality of life (QOL) (5).

The role of additional palliative chemotherapy for patients with progressive disease after third-line chemotherapy remains unclear. If additional chemotherapy improves survival or QOL of patients even after third-line chemotherapy, further treatment may be a reasonable option in patients with a favorable performance status. However, little data are available on the results of additional treatments including more than third-line chemotherapy. Therefore, we report on treatment outcomes including more than third-line chemotherapy for advanced NSCLC patients.

Eighty-two patients with inoperable advanced NSCLC were admitted for platinum-based chemotherapy as a first-line treatment at the Korea University Hospitals between March 2003 and February 2007. All 82 patients were initially treated with cisplatin-based first-line chemotherapy in combination with gemcitabine or taxanes. When disease progression was confirmed, the patients moved on to additional lines of chemotherapy with drugs such as docetaxel, gefitinib or erlotinib, vinorelbine, irinotecan and pemetrexed. Vinorelbine and irinotecan were known to be effective in combination with the platinum agent in first-line chemotherapy (6). These drugs were provided as monotherapy based on the attending physician’s decision based on patient performance status and drug toxicity. Selection of drugs with same toxicity profiles as the previous treatment was avoided to minimize cumulative toxicity. After every 2 cycles of chemotherapy, the patients were evaluated for a response. Toxicity was evaluated after every cycle of chemotherapy. The total cycles of chemotherapy and withdrawal of treatment were based on the physician’s decision, as well as the benefit of treatment and risk of toxicity. Chemotherapy was changed or withdrawn if disease progression or unacceptable toxicity was observed. In cases of grade 4 hematological toxicities and more than a grade 3 non-hematological toxicity, the chemotherapy dose was reduced by 25%.

The electronic medical records of the enrolled patients were retrospectively reviewed using the hospital’s computer network. Information on the patients including performance status, based on the Eastern Cooperative Oncology Group (ECOG); performance scale; histological type of malignancy, chemotherapy regimens; overall survival; initial response; duration of response; total chemotherapy treatment lines and number of cycles of treatment were recorded. Collected information was then analyzed and evaluated for overall survival, duration of response and the total number of chemotherapy cycles provided. Patients with an initial response of partial response or better were further classified as the response group. The subgroup analysis for overall survival, total number of chemotherapy cycles and duration of initial response was determined and compared among the different response groups. In addition, a subgroup analysis for overall survival between smokers and non-smokers was performed.

The response to chemotherapy was classified according to the Response Evaluation Criteria in Solid Tumors criteria (7). Definitions of complete response (CR), partial response (PR), stable disease (SD) and progressive disease (PD) were as follows: CR, disappearance of all target lesions; PR, at least a 30% decrease in the sum of the longest diameters of target lesions compared to the baseline; PD, at least a 20% increase in the sum of the longest diameters of the target lesions, using as a reference the smallest sum of the longest diameters recorded since onset of the treatment or the appearance of one or more new lesions and SD, neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for PD.

The survival curves were plotted according to the method of Kaplan-Meyer. Log-rank and Chi-square analyses were performed to determine statistical significance.

The median age of the enrolled patients was 65 and there were more males. At the time of diagnosis most patients had stage IV lung cancer except for three patients with stage IIIb disease. Most patients had a favorable performance status. Adenocarcinoma (47.5%) was the most common histological type of cancer, followed by squamous cell carcinoma (26.8%; Table I).

All patients started first-line chemotherapy with a regimen of cisplatin combined with either gemcitabine (56.1%) or taxanes (43.9%). Thirty-three out of 82 patients (40.2%) had a PR or better on their initial response evaluation, including two patients who were observed with a CR. The median number of chemotherapy lines and cycles received were 2 and 7, respectively. A total of 33 patients were candidate third-line chemotherapy or more. The highest number of lines and cycles of chemotherapy was eighth-line and total 33 cycles (Table I). The median overall survival of the studied patients was 15 months (Fig. 1A).

In the treatment of patients with inoperable, advanced lung cancer, first- and second-line chemotherapy have proven to be effective (1,2). Agents such as docetaxel, pemetrexed and EGFR inhibitors are widely used for second- and third-line chemotherapy (3,4,8). However, even after completion of third-line chemotherapy, some patients still exhibit a favorable performance status despite disease progression. The question is whether present with a favorable performance status after third-line chemotherapy. Chemotherapy beyond third-line treatment is supported by medical insurance in Korea. Thus, in practice, when these treatments are shown to improve QOL even after third-line chemotherapy, a physician may approve additional cyles of chemotherapy after disease progression.

In this study, the subgroup analysis showed that 49 patients were treated with less than second-line chemotherapy and 33 patients (40.2%) were treated with more than third-line chemotherapy. This result is similar to that of Murillo and Koeller, who also reported that 41% patients received more than third-line chemotherapy (9).

Vigorous efforts are ongoing to prolong survival by increasing drug exposure or duration. Sequential, alternating and maintenance therapy have been proposed in an attempt to improve survival. These treatments are appealing as they may increase the number of non-cross resistant agents and/or duration of chemotherapy. As maintenance therapy, paclitaxel has shown success (10–12), whereas agents such as docetaxel and gemcitabine have failed to show significant benefits (10,13,14). Based on the studies published thus far, however, it is unlikely that sequential and alternating treatment will prove to be superior to the current standard therapy for all advanced NSCLC patients (15). However, selective groups, especially chemo-sensitive patients, may benefit from such treatments.

In this retrospective study, the overall treatment results for patients with inoperable advanced NSCLC had a median overall survival of 15 months. This finding was not significantly different from the results of previous reports. However, in the subgroup analysis of patients with a partial response or better (response group) to first-line chemotherapy, the median overall survival increased to 24 months. Compared to the entire study group, patients of the response group in first-line chemotherapy received a median of 1 more line and 8 more cycles of chemotherapy. They were exposed to more chemotherapeutic agents than the non-responsive group. The patient, for example, who was treated with the highest number of chemotherapy lines and cycles in this study received as many as 8 lines and 33 cycles of chemotherapy. The median survival was 23 months in patients treated with more than third-line chemotherapy, compared to 7 months in patients treated with less than second-line chemotherapy.

Notably, since the patients lived longer, they could be potential candidates for additional chemotherapy. However, the increase of the median overall survival of 24 months in the response group is difficult to explain solely by additional chemotherapy. In a study that identified an objective response to chemotherapy as a marker for further survival, a median survival time of 41 and 19 weeks was reported for the responding and non-responding patients, respectively (16).

For patients with colorectal cancer, exposure to 3 active cytotoxic drugs prolonged survival regardless of whether the drugs were administered as monotherapy or in combination (17–20, Pitot HC, et al, J Clin Oncol 23: abs. 3506, 2005). It is possible that a similar mechanism applies to patients with NSCLC. Multi-drug exposure shown to be non-cross resistant may increase survival in the response group. If this is also true for NSCLC patients, additional chemotherapy after second- or third-line chemotherapy may be an effective and feasible strategy for treating patients with inoperable NSCLC. Careful patient selection is important, since many of the patients that were resistant to the initial chemotherapeutic agents appeared to be resistant to second- and third-line drugs as well.

The subgroup analysis comparing the survival of NSCLC patients who were smokers to non-smokers showed that the smokers had a significantly worse prognosis. This finding is comparable to the results of previously published studies (21). Certain investigators have suggested that NSCLC among non-smokers exhibits different clinical and pathological features compared to that of smokers (22–24).

Treatment-related toxicities are always a concern when further chemotherapy is required. Although not studied intensively, we did not encounter unmanageable toxicities associated with increased chemotherapy exposure and related cumulative toxicities. The fact that most of the additional lines were single agent lines of chemotherapy may have played a role in reducing the toxicities. In addition, chemotherapeutic agents were carefully selected in order that the major toxicities of these drugs did not overlap the major toxicities of previously used drugs.

Significant limitations of this study were that it was not a prospective randomized study, and that the sample size is small. However, the results suggested that NSCLC patients with a favorable response to initial chemotherapy benefit from further chemotherapy beyond third-line treatment by multi-drug exposure. Chemotherapy may, therefore, benefit NSCLC patients who remain fit for treatment following the completion of third-line cycles of chemotherapy. Prospective large scale studies are needed to confirm these results.