Abstract
Background
Previously reported superior tumor response of pemetrexed in the second-line setting for advanced non-squamous non-small cell lung cancer (advNS-NSCLC) has never been confirmed in real-world studies. Platinum-based doublet is frequently used in the second-line setting for advanced NSCLC in China.
Methods
A retrospective cohort study was conducted including patients receiving pemetrexed or docetaxel-based chemotherapy in the second-line setting for advNS-NSCLC in four Chinese tertiary care hospitals. Propensity score matched treatment groups were created for head-to-head comparisons on best tumor response and clinical toxicity. Multiple regression analyses were performed to rank the impact of the four regimens on the risks of tumor progression and hematological adverse events.
Results
Three hundred and eighty-four patients were included for creating matched treatment groups for pemetrexed versus platinum/pemetrexed (33 pairs), docetaxel (17 pairs), and platinum/docetaxel (29 pairs), respectively. No significant differences were identified for best tumor response between pemetrexed and the other three regimens. However, pemetrexed was associated with significantly fewer patients experiencing anemia (39.4% vs. 69.7%, P = 0.004) and neutropenia (6.1% vs. 30.3%, P = 0.021) than platinum/pemetrexed. Multiple regression analyses indicated that pemetrexed was associated with significantly slower tumor progression (hazard ratio 0.628, P = 0.040) and a significantly lower risk of neutropenia (odds ratio 0.132, P = 0.019) than docetaxel.
Conclusions
Pemetrexed was associated with significantly postponed tumor progression and significantly less hematological toxicity than docetaxel in the real-world second-line setting for advNS-NSCLC in Chinese patients. Pemetrexed monotherapy had comparable tumor response, but significantly less hematological toxicity than pemetrexed-based doublet.
Keywords: Adenocarcinoma, docetaxel, pemetrexed, platinum-based doublet, previously treated
Introduction
Lung cancer continues to be the leading cause of cancer-related death in China and the number of lung cancer patients are expected to reach one million by 20251 as a result of uncontrolled cigarette smoking,2 the aging population,3 and worsening air pollution4 associated with industrialization. Nearly half of Chinese lung cancer patients are diagnosed at advanced stage5 because of poor awareness of lung cancer,6 unspecific symptoms associated with early tumor stage,7 and the limited capacity of X-rays for early tumor detection.8 Thus, systemic chemotherapy is the main therapeutic option to improve survival and quality of life for Chinese lung cancer patients.
Pemetrexed is the second cytotoxic agent approved in the second-line setting for advanced non-small cell lung cancer (NSCLC) because it has a superior safety profile than docetaxel. Several post-hoc analyses on randomized trials,9–12 including one trial from China,12 suggested that pemetrexed was more effective in patients with non-squamous cell histology because of the lower expression of thymidylate synthase (TS),13,14 one of the three pemetrexed-targeting enzymes in the folate pathway.15 However, no real-world studies have been performed to confirm those post-hoc analyses, which could be associated with publication bias by reporting significant results only.16 In addition, a national survey on chemotherapy pattern reported that half of the patients received doublet treatment in the second-line setting for advanced NSCLC in China.17 However, randomized trials only observed marginal clinical benefits on tumor progression, but substantial toxicity associated with doublet treatments in the second-line setting.18 Thus, the main purpose of this study was to generate real-world clinical evidence that could confirm the superior tumor response and safety profile associated with pemetrexed and resolve disputes on the use of doublet treatments in the second-line setting for advanced non-squamous NSCLC (advNS-NSCLC).
Methods
Chemotherapy for advanced NSCLC is often conducted within hospital settings in order to monitor and manage life-threatening adverse events (AEs) caused by chemotherapy. Thus, well documented tumor response and clinical toxicity associated with chemotherapy in Chinese tertiary care hospital settings allowed us to design a retrospective cohort study to assess pemetrexed or docetaxel-based monotherapy or doublet treatments in the second-line setting for advNS-NSCLC. In order to create a study cohort representing current referral and social economic patterns of lung cancer patients in China, our study selected one tumor specialized hospital and one general hospital from the national capital city (Beijing) and a provincial capital city (Changsha, Hunan province), respectively, to identify eligible patients and collect data. The study protocol was reviewed and approved by the four selected hospitals (Chinese Academy of Medical Sciences Tumor Hospital [CAMSTH], Xuanwu Hospital [XWH], Hunan Provincial Tumor Hospital [HNPTH], and Xiangya Hospital [XYH]).
Patient identification
The hospital admission registry databases in the four hospitals were used to find patients who were hospitalized for lung cancer. The searching period was three years for XYH (1 January 2010 to 31 December 2012), four years for CAMSTH (1 January 2009 to 31 December 2012), and five years for HNPTH and XWH (1 January 2007 to 31 December 2012), respectively. The key words used for searching included “lung cancer,” “NSCLC,” “non-squamous NSCLC,” “adenocarcinoma lung cancer,” and “large-cell lung cancer.” The identified patients with a primary diagnosis of lung cancer were further linked with hospital medical records to extract information on tumor stage, tumor histology, and treatment history for eligibility assessment. Our study included patients who had stage IIIB or IV with biopsy-proven non-squamous NSCLC and received subsequent monotherapy or platinum-based doublet with pemetrexed or docetaxel after the failure of first-line chemotherapy and/or maintenance chemotherapy. Our study excluded patients who had non-squamous histology mixed with small cell and/or squamous cell. Our study also excluded patients who received tyrosine-kinase inhibitors (TKIs) or epidermal growth factor receptor (EGFR) monoclonal antibodies in the first-line or second-line setting, to control possible selection bias associated with EGFR gene mutation.
Data collection
Our study defined the follow-up time as the treatment period of second-line chemotherapy from the admission date of hospitalization with the first administration of chemotherapy regimens to the discharge date of hospitalization with the last administration of chemotherapy regimens prior to the first tumor response assessment indicating disease progression or permanent treatment discontinuation, whichever occurred first. Hospital records of included patients during the defined follow-up time were reviewed to collect patients' baseline characteristics (age, gender, body mass index [BMI], residence area, public health insurance schedule, smoking status, Eastern Cooperative Oncology Group [ECOG] performance status, baseline marrow function, tumor stage, histology subtype, and metastasis status) at the initial chemotherapy; chemotherapy pattern (chemical name, dosage, administration frequency, total completed treatment cycles); admission and discharge dates of hospitalizations; tumor response assessment based on the Response Evaluation Criteria In Solid Tumors (RECIST), which classified tumor response as complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD) respectively; and AE assessment according to Common Terminology Criteria for Adverse Events (CTCAE) with modification on anemia grading for Chinese patients.19 Our study also extracted hemoglobin, neutrophilic granulocyte count, white blood cell (WBC) count, and platelet count from each laboratory blood test ordered during the follow-up as supplemental information for hematological AE assessment.
Outcome measures
The primary outcome measures in our study included time to progression (TTP) and occurrences of hematological and non-hematological AEs. TTP was defined as the time period from the onset of second-line chemotherapy to the first tumor response assessment reporting PD or permanent treatment discontinuation. Neutropenia, leukopenia, anemia, and thrombocytopenia recorded in clinical notes and identified from laboratory blood tests were taken into account to determine the occurrences of hematological AEs. The occurrences of non-hematological AEs were fully based on hospital medical records. The second outcome measures in our study included best tumor response according to the results of identified tumor response assessments and the occurrences of grade 3 or 4 hematological AEs based on CTCAE. Our study also measured treatment cycles, hospital episodes per treatment cycle, and length of hospital stay per treatment cycle to indicate the patterns of care associated with studied monotherapy and doublet treatments.
Data analyses
The analysis of variance (ANOVA) test for continuous outcomes and the chi square test for dichotomous outcomes were used to demonstrate the differences in patient baseline characteristics associated with the studied pemetrexed or docetaxel-based regimens. Our study performed multiple logistic regression analyses to identify any baseline characteristics independently and significantly associated with the use of doublet treatments in our study cohort. In order to adjust possible confounding effects associated with patient baseline characteristics, our study used both propensity score methods and conventional regression analyses for the assessment of tumor response and clinical toxicity associated with the four studied chemotherapy regimens. Propensity score methods were used to create matched treatment groups for pemetrexed versus docetaxel, platinum/pemetrexed, and platinum/docetaxel, respectively, using a greedy approach for 1:1 matching with a propensity score difference of less than 0.01 between the matched pair.20 The average of P-values associated with comparisons on baseline characteristics after propensity score matching were compared to that prior to matching to demonstrate improvement on the balance of patient baseline characteristics. A paired t-test for continuous outcomes and McNemar's test for dichotomous outcomes were used in the adjusted head-to-head comparisons on patient baseline characteristics, distribution of best tumor response, and occurrences of hematological and non-hematological AEs between propensity score matched treatment groups. Kaplan-Meier (KM) survival analyses were performed to compare survival curves associated with propensity score matched treatment groups for time to progression (TTP). Patients who had no tumor response assessment reporting PD were censored at the date of hospital discharge after the last administration of chemotherapy regimens. A paired log rank test was used to examine the differences associated with KM survival curves. Our study also performed conventional regression analyses to confirm the comparisons between the studied chemotherapy regimens in propensity score matched patients. Multiple Cox survival regression analysis with adjustment of patient baseline characteristics was performed to rank the hazard ratios (HRs) associated with the four studied regimens relative to docetaxel for the risk of PD during treatment. The odds ratios (ORs) associated with the four studied regimens relative to docetaxel for the risks of hematological AEs were also ranked respectively using multiple logistic regression analyses with adjustment of patient baseline characteristics. SAS 9.2 was used to perform the data analyses described above. Statistical significance in our study was defined as a two-sided P-value less than 0.05.
Results
Initial screening of hospital admission registry databases in the four hospitals identified 9270 patients hospitalized for lung cancer. Of these, 8886 patients were excluded because of: no pathological information (134); small cell lung cancer (181); squamous NSCLC (458); mixed with squamous cell histology (314); tumor stage less than IIIB (279); no second-line chemotherapy (7018); treatment with TKI or EGFR monoclonal antibodies in first-line (76) or second-line (211) settings; or treatment regimens that did not contain pemetrexed or docetaxel (215). Thus, 384 patients receiving pemetrexed (46), platinum/pemetrexed (161), docetaxel (61), or platinum/docetaxel (116) were included to create the study cohort. The flow charts of identifying eligible patients in the four hospitals are illustrated in Figure 1.
Figure 1.
Flow chart to identify patients receiving pemetrexed or docetaxel-based chemotherapy regimens in the second-line setting for advanced non-squamous non-small cell lung cancer (NSCLC) in the four selected tertiary hospitals. CAMSTH, Chinese Academy of Medical Sciences Tumor Hospital; EGFR, epidermal growth factor recptor; HNPTH, Hunan Provincial Tumor Hospital; TKI, tyrosine-kinase inhibitors; XYH, Xiangya Hospital; XWH, Xuanwu Hospital.
Patient baseline characteristics and predictors for doublet treatment
The four treatment groups shared a number of similar baseline characteristics, except mean age, that ranged from 53.4 years to 57.9 years (P = 0.004), prevalence of an ECOG performance status of 0 that ranged from 29.5% to 51.6% (P = 0.003), baseline neutrophilic granulocyte count that ranged from 4.1 to 5.2 × 109/L (P = 0.015), and baseline WBC count that ranged from 6.4 to 7.7 × 109/L (P = 0.015) (Table 1). Of these 384 patients, 72.1% patients received pemetrexed or docetaxel-based doublet treatment. Among the collected baseline characteristics, younger age (OR 0.964, 95% confidence interval [CI] 0.940 to 0.989, P = 0.005), higher baseline hemoglobin (OR 1.014, 95% CI 1.001 to 1.028, P = 0.037), and absence of pleural metastasis (OR 0.590, 95% CI: 0.352 to 0.988, P = 0.045) were independently and significantly associated with doublet treatment in our study cohort.
Table 1.
Summary of patient baseline characteristics associated with the studied four chemotherapy regiments used in the second-line setting for advanced non-squamous non-small cell lung cancer
| Chemotherapy regimen | Pemetrexed | Docetaxel | Platinum/demetrexed | Platinum/docetaxel | Overall | ||||
|---|---|---|---|---|---|---|---|---|---|
| Sample size | 46 | 61 | 161 | 116 | 384 | ||||
| Variables | Mean/% | SD | Mean/% | SD | Mean/% | SD | Mean/% | SD | Mean/% SD |
| Demography | |||||||||
| Age (years) | 57.8 | 12.1 | 57.9 | 8.9 | 53.9 | 9.7 | 53.4 | 10.2 | 54.8 |
| Male (%) | 60.9 | 60.7 | 62.1 | 62.9 | 62.0 | ||||
| BMI (kg/m2) | 23.9 | 3.3 | 23.4 | 4.1 | 23.6 | 3.0 | 24.1 | 3.2 | 23.7 |
| BSA (m2) | 2.0 | 0.2 | 1.9 | 0.4 | 1.9 | 0.2 | 1.9 | 0.3 | 1.9 |
| Non-smoking (%) | 39.1 | 36.1 | 41.0 | 44.8 | 41.1 | ||||
| Resident area (%) | |||||||||
| Beijing | 37.0 | 34.4 | 24.8 | 25.9 | 28.1 | ||||
| Changsha | 4.4 | 6.6 | 3.1 | 9.5 | 5.7 | ||||
| Other urban cities | 34.8 | 39.3 | 51.6 | 47.4 | 46.4 | ||||
| Rural area | 23.9 | 19.7 | 20.5 | 17.2 | 19.8 | ||||
| Public insurance type (%) | |||||||||
| Urban | 58.7 | 65.6 | 55.3 | 56.0 | 57.6 | ||||
| Rural | 17.4 | 14.8 | 14.3 | 19.0 | 16.1 | ||||
| ECOG performance status (%) | |||||||||
| 0 | 37.0 | 29.5 | 51.6 | 32.8 | 40.6 | ||||
| 1 | 54.4 | 67.2 | 44.7 | 62.9 | 54.9 | ||||
| 2 | 8.7 | 3.3 | 3.7 | 4.3 | 4.4 | ||||
| Marrow function | |||||||||
| Hemoglobin (g/l) | 118.6 | 20.9 | 118.7 | 19.3 | 124.6 | 18.1 | 122.7 | 18.0 | 122.3 |
| Neutrophilic granulocyte count (×109/l) | 5.2 | 2.6 | 4.1 | 1.8 | 4.5 | 2.1 | 4.0 | 1.9 | 4.4 |
| White cell count (×109/l) | 7.8 | 2.8 | 6.4 | 2.3 | 6.8 | 2.5 | 6.4 | 2.1 | 6.7 |
| Platelet count (×1010/l) | 24.8 | 8.7 | 24.4 | 10.7 | 23.5 | 8.5 | 24.9 | 9.0 | 24.2 |
| Tumor stage and histology (%) | |||||||||
| Stage 4 | 91.3 | 91.8 | 95.7 | 91.4 | 93.2 | ||||
| Adenocarcinoma | 95.7 | 96.7 | 98.1 | 98.3 | 97.7 | ||||
| Number of metastasis sites (%) | |||||||||
| 1 | 21.7 | 34.4 | 31.7 | 36.2 | 32.3 | ||||
| 2 | 28.3 | 34.4 | 21.7 | 29.3 | 26.8 | ||||
| 3 and above | 28.3 | 23.0 | 30.4 | 22.4 | 26.6 | ||||
| Site of metastasis (%) | |||||||||
| Brain | 15.2 | 24.6 | 25.5 | 20.7 | 22.7 | ||||
| Bone | 50.0 | 54.1 | 46.0 | 47.4 | 48.2 | ||||
| Liver | 6.5 | 6.6 | 14.3 | 8.6 | 10.4 | ||||
| Pleural | 32.6 | 44.3 | 26.7 | 31.9 | 31.8 | ||||
BMI, body mass index; BSA, body surface area; ECOG, Eastern Cooperative Oncology Group; SD, standard deviation.
Creating propensity score matched treatment groups
Propensity score methods successfully matched 17 pairs, 33 pairs, and 29 pairs for pemetrexed versus docetaxel, platinum/pemetrexed, and platinum/docetaxel doublet, respectively. The average of P-values associated with comparisons on patient baseline characteristics between pemetrexed and platinum/pemetrexed was increased by 81.8% after propensity score matching (from 0.391 to 0.711). The averages of P-values associated with the comparisons for pemetrexed versus docetaxel and platinum/docetaxel on patient baseline characteristics were increased by 10.7% (from 0.616 to 0.682) and 9.4% (from 0.573 to 0.627), respectively. No significant differences were identified between the matched treatment groups for the comparisons on patient baseline characteristics.
Head-to-head comparisons between propensity score matched treatment groups
Tumor response
The studied four regimens were administered every three weeks. The dosages of pemetrexed per treatment cycle in the monotherapy and doublet treatments were almost identical (480 ± 95 mg/m2 vs. 477 ± 114 mg/m2, P = 0.914). The docetaxel-based monotherapy and doublet regimens also had similar dosages of docetaxel (67 ± 15 mg/m2 vs. 65 ± 16 mg/m2). Pemetrexed monotherapy was associated with a similar number of treatment cycles as the two doublets, but a non-significant trend of having more treatment cycles than docetaxel monotherapy (3.9 ± 7.4 cycles vs. 1.5 ± 0.6 cycles, P = 0.193). Pemetrexed also had similar hospital episodes (ranged from 0.9 to 1.3 episodes) and length of hospital stay (ranged from 9.7 to 15.0 days) per treatment cycle as the other three regimens. Tumor response assessment was lacking in about half of the included patients and the observed distribution of best tumor response associated with pemetrexed was similar to what was observed in the matched patients receiving the other three regimens. Pemetrexed also had similar TTP as the two doublet treatments, but likely had delayed tumor progression with a considerable trend towards statistical significance in comparison with docetaxel (median TTP 31 days vs. 24 days, P = 0.104). The results of head-to-head comparisons on patterns of care and tumor response associated with propensity score matched treatment groups are summarized in Table 2.
Table 2.
Summarized results of head-to-head comparisons on patterns of care and tumor response between pemetrexed and the other three studied regimens in the propensity score matched patients
| Comparison | Pemetrexed vs. docetaxel | Pemetrexed vs. platinum/pemetrexed | Pemetrexed vs. platinum/docetaxel | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Matched pairs | 17 | 33 | 29 | ||||||
| Treatment | Pemetrexed | Docetaxel | P value | Pemetrexed | Platinum/ pemetrexed | P value | Pemetrexed | Platinum/ Docetaxel | P value. |
| Treatment pattern (mean ± SD) | |||||||||
| Cytotoxic agent (mg/m2 per cycle) | 463 ± 46 | 67 ± 15 | NA | 480 ± 95 | 477 ± 114 | 0.914 | 478 ± 101 | 65 ± 16 | NA |
| Completed treatment cycles | 3.9 ± 7.4 | 1.5 ± 0.6 | 0.193 | 3.3 ± 5.5 | 2.9 ± 2.2 | 0.721 | 3.5 ± 5.8 | 2.6 ± 2.0 | 0.433 |
| Hospital stay length per cycle (days) | 12.8 ± 15.8 | 15.0 ± 12.5 | 0.672 | 11.6 ± 12.5 | 9.7 ± 6.0 | 0.432 | 12.3 ± 14.4 | 11.5 ± 6.3 | 0.778 |
| Hospital admissions per cycle | 1.0 ± 0.1 | 1.1 ± 0.5 | 0.541 | 1.3 ± 1.6 | 1.0 ± 0.1 | 0.296 | 1.0 ± 0.4 | 0.9 ± 0.2 | 0.221 |
| Best tumor response (n, %) | |||||||||
| Partial response | 2 (11.8%) | 1 (5.9%) | 0.564 | 4 (12.1%) | 4 (12.1%) | 1.000 | 4 (13.8%) | 3 (10.4%) | 0.706 |
| Stable disease | 2 (11.8%) | 1 (5.9%) | 0.564 | 4 (12.1%) | 4 (12.1%) | 1.000 | 3 (10.3%) | 5 (17.2%) | 0.480 |
| Progressive disease | 4 (23.5%) | 4 (23.5%) | 1.000 | 9 (27.3) | 6 (18.2%) | 0.439 | 7 (24.1%) | 3 (10.3%) | 0.157 |
| Unknown | 9 (52.9%) | 11 (64.7%) | 0.480 | 16 (48.5%) | 19 (57.6%) | 0.467 | 15 (51.7%) | 18 (62.1%) | 0.366 |
| TTP (median days, 95% Cl) | 31 (13 to 79) | 24 (15 to 44) | 0.104 | 28 (23 to 56) | 43 (23 to 61) | 0.971 | 31 (25 to 49) | 30 (20 to 57) | 0.587 |
CI, confidence interval; TTP, time to progressive disease.
Clinical toxicity
Our study compared pemetrexed with the other three regimens on the occurrences of hematological and non-hematological AEs in propensity score matched patients. Pemetrexed was associated with significantly fewer patients experiencing anemia (39.4% vs. 69.7%, P = 0.004), neutropenia (6.1% vs. 30.3%, P = 0.021), and any hematological AE (42.4% vs. 75.8%, P = 0.005) than platinum/pemetrexed, and significantly fewer patients experiencing neutropenia (3.5% vs. 24.1%, P = 0.034) and leukopenia (3.5% vs. 34.5%, P = 0.007) than platinum/docetaxel. However, our study did not observe any significant differences in the occurrences of hematological AEs between pemetrexed and docetaxel in the matched patients. In addition, our study did not observe any significant differences in the occurrences of grade 3 or 4 hematological AEs between pemetrexed and the other three regimens. Further adjusted comparisons on non-hematological AEs indicated that pemetrexed was associated with significantly lower rates of nausea (42.4% vs. 66.7%, P = 0.021) and vomiting (30.3% vs. 60.6%, P = 0.008) than platinum/pemetrexed and significantly lower rates of alopecia (0% vs. 31.0%, P = 0.004), diarrhea (0% vs. 24.1%, P = 0.016), nausea (34.5% vs. 65.5%, P = 0.029), peripheral neuropathy (0% vs. 24.1%, P = 0.016), pneumonia (0% vs. 20.7%, P = 0.031), stomatitis (0% vs. 27.6%, P = 0.008), and vomiting (24.1% vs. 62.1%, P = 0.008) than platinum/docetaxel. No significant differences were observed for the occurrences of non-hematological AEs between pemetrexed and docetaxel. The results of head-to-head comparisons on the occurrences of hematological and non-hematological AEs between the propensity score matched treatment groups are summarized in Table 3.
Table 3.
Summarized results of head-to-head comparisons on the occurrences of hematological and non-hematological adverse events between pemetrexed and the other three studied regimens in the propensity score matched patients
| Comparison | Pemetrexed vs. docetaxel | Pemetrexed vs. platinum/pemetrexed | Pemetrexed vs. platinum/docetaxel | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Matched pairs | 17 | 33 | 29 | ||||||
| Treatment | Pemetrexed | Docetaxel | P value | Pemetrexed | Platinum/pemetrexed | P value | Pemetrexed | Platinum/docetaxel | P value |
| Hematolgoical AE (n, %) | |||||||||
| Anemia | 6 (35.3%) | 5 (29.4%) | 0.541 | 13 (39.4%) | 23 (69.7%) | 0.004 | 11 (37.9%) | 13 (44.8%) | 0.637 |
| Neutropenia | 2 (11.8%) | 4 (23.5%) | 0.414 | 2 (6.1%) | 10 (30.3%) | 0.021 | 1 (3.5%) | 7 (24.1%) | 0.034 |
| Leukopenia | 2 (11.8%) | 4 (23.5%) | 0.414 | 3 (9.1%) | 7 (21.2%) | 0.103 | 1 (3.5%) | 10 (34.5%) | 0.007 |
| Thrombocytopenia | 4 (23.5%) | 1 (5.9%) | 0.180 | 5 (15.2%) | 7 (21.2%) | 0.480 | 6 (20.7%) | 8 (27.6%) | 0.564 |
| Any hematologica AE | 8 (47.1%) | 9 (52.9%) | 0.416 | 14 (42.4%) | 25 (75.8%) | 0.005 | 13 (44.8%) | 20 (69.0%) | 0.071 |
| Grade 3 or 4 hematological AE (n, %) | |||||||||
| Anemia | 0 (0%) | 0 (0%) | — | 1 (3.0%) | 0 (0%) | 1.000 | 1 (3.5%) | 1 (3.5%) | 1.000 |
| Neutropenia | 2 (11.8%) | 2 (11.8%) | 1.0 | 2 (6.1%) | 2 (6.1%) | 1.000 | 1 (3.5%) | 2 (6.9%) | 0.564 |
| Leukopenia | 1 (5.9%) | 0 (0%) | 1.0 | 1 (3.0%) | 0 (0%) | 1.000 | 0 (0%) | 0 (0%) | — |
| Thrombocytopenia | 1 (5.9%) | 0 (0%) | 1.0 | 1 (3.0%) | 2 (6.1%) | 0.564 | 1 (3.5%) | 0 (0%) | 1.000 |
| Non-hematological AE (n, %) | |||||||||
| Alopecia | 0 (0%) | 1 (5.9%) | 1.000 | 0 (0%) | 1 (3.0%) | 1.000 | 0 (0%) | 9 (31.0%) | 0.004 |
| Athralgia | 0 (0%) | 1 (5.9%) | 1.000 | 0 (0%) | 3 (9.1%) | 0.250 | 0 (0%) | 4 (13.8%) | 0.125 |
| Cough | 1 (5.9%) | 2 (11.8%) | 0.317 | 2 (6.1%) | 3 (9.1%) | 0.655 | 2 (6.9%) | 6 (20.7%) | 0.103 |
| Dermatitis | 0 (0%) | 0 (0%) | — | 0 (0%) | 0 (0%) | — | 0 (0%) | 3 (10.3%) | 0.250 |
| Diarrhea | 0 (0%) | 2 (11.8%) | 0.500 | 0 (0%) | 3 (9.1%) | 0.250 | 0 (0%) | 7 (24.1%) | 0.016 |
| Dyspnea | 0 (0%) | 0 (0%) | — | 0 (0%) | 2 (6.1%) | 0.500 | 0 (0%) | 2 (6.9%) | 0.500 |
| Edema | 0 (0%) | 1 (5.9%) | 1.000 | 0 (0%) | 0 (0%) | — | 0 (0%) | 2 (6.9%) | 0.500 |
| Fatigue | 3 (17.7%) | 2 (11.8%) | 0.655 | 6 (18.2%) | 5 (15.2%) | 0.763 | 5 (17.2%) | 12 (41.4%) | 0.052 |
| Nausea | 6 (35.3%) | 6 (35.3%) | 1.000 | 14 (42.4%) | 22 (66.7%) | 0.021 | 10 (34.5%) | 19 (65.5%) | 0.029 |
| Peripheral neuropathy | 0 (0%) | 1 (5.9%) | 1.000 | 0 (0%) | 1 (3.0%) | 1.000 | 0 (0%) | 7 (24.1%) | 0.016 |
| Pneumonia | 0 (0%) | 0 (0%) | — | 0 (0%) | 1 (3.0%) | 1.000 | 0 (0%) | 6 (20.7%) | 0.031 |
| Rash | 0 (0%) | 1 (5.9%) | 1.000 | 0 (0%) | 2 (6.1%) | 0.500 | 0 (0%) | 2 (6.9%) | 0.500 |
| Stomatitis | 0 (0%) | 1 (5.9%) | 1.000 | 0 (0%) | 1 (3.0%) | 1.000 | 0 (0%) | 8 (27.6%) | 0.008 |
| Vomiting | 4 (23.5%) | 5 (29.4%) | 0.706 | 10 (30.3%) | 20 (60.6%) | 0.008 | 7 (24.1%) | 18 (62.1%) | 0.008 |
| Weight loss | 0 (0%) | 0 (0%) | — | 0 (0%) | 0 (0%) | — | 0 (0%) | 1 (3.5%) | 1.000 |
AE, adverse event.
Conventional regression analyses assessing the risks of progressive disease and hematological adverse events (AEs) associated with the studied regimens
The risks of PD associated with the studied four regimens during the follow-up were ranked by generated HRs relative to monotherapy treatment with docetaxel. The risk of PD associated with platinum/pemetrexed doublet treatment and pemetrexed monotherapy treatment were significantly reduced by 43.3% (HR 0.566, P = 0.002) and 37.2% (HR 0.628, P = 0.040), respectively. The risk of progressive disease associated with platinum/docetaxel was also reduced, but statistical significance was not reached (HR 0.798, P = 0.204). The risks of hematological AEs associated with the four studied regimens were also ranked respectively by generated ORs relative to monotherapy treatment with docetaxel. The risks of neutropenia and leukopenia associated with pemetrexed were ranked the lowest. However, statistical significance was only reached for the reduced risk of neutropenia (OR 0.132, P = 0.019), and not for leukopenia (OR 0.293, P = 0.111). The risk of anemia associated with platinum/pemetrexed was ranked as the highest with a strong trend towards statistical significance (OR 2.392, P = 0.068). The risk of thrombocytopenia associated with platinum/docetaxel was ranked the highest without reaching statistical significance (OR 2.445, P = 0.134). The ranked risks of progressive disease and hematological AEs associated with the four studied regimens are illustrated in Figures 2 and 3, respectively.
Figure 2.
Hazard ratios (HR) associated with the studied regimens for the risk of progressive disease when compared to docetaxel monotherapy in the multiple Cox survival regression analysis with adjustment of patient baseline characteristics. CI, confidence interval; ECOG, Eastern Cooperative Oncology Group.
Figure 3.
Odds ratios (ORs) associated with the studied regimens for the risk of hematological adverse events (AEs) when compared to docetaxel monotherapy in the multiple logistic regression analysis with adjustment of patient baseline characteristics. OR, odds ratio.
Discussion
Our study may be the first study demonstrating that monotherapy pemetrexed and doublet pemetrexed are likely associated with a lower risk of tumor progression than docetaxel in the second-line setting for advNS-NSCLC in a real-world Chinese patient cohort. Over 70% of patients in our study cohort were treated by doublet treatments, which is consistent with a recent national survey reporting a high prevalence of doublet therapy in the second-line setting for advanced NSCLC.17 However, our study observed that the doublet treatment with platinum/pemetrexed was not associated with significantly stronger tumor response, but more hematological and non-hematological toxicity than the monotherapy treatment with pemetrexed.
When compared to docetaxel, the risk of tumor progression associated with pemetrexed in our study was reduced by 37.5%, which was greater than the reduced risk of PD associated with pemetrexed in the adjusted comparison with docetaxel in a previous post hoc analysis reporting the treatment-by-histology interaction between pemetrexed and non-squamous histology.10 The stronger tumor response associated with pemetrexed in our study was likely a result of the prevalence of adenocarcinoma histology (97.7%) in our study cohort. Adenocarcinoma is associated with a lower expression of TS, which has been proven a key predictor for the tumor response of pemetrexed.21 We also suspect that the post hoc analysis,10 which was based on the phase III study that likely had a higher prevalence of smoker patients than our study,22 as a result could underestimate the treatment-by-histology interaction of pemetrexed because of no adjustment on smoking, a strong prognostic factor for poorer prognosis.23 The Chinese ethnicity of our study cohort might be another factor increasing the tumor response associated with pemetrexed because East Asian ethnicity is associated with a favourable response to chemotherapy treatments and is also a favourable prognostic factor.24,25 Thus, the observed lower risk of tumor progression associated with pemetrexed treatment in our Chinese patient cohort may warrant further investigation on the possible roles of ethnicity in the treatment-by-histology interaction associated with pemetrexed and further guide individualized care in patients with advanced NSCLC.
The prevalence of using doublet treatment in a second-line setting for advanced NSCLC in real-world patients was higher than what was estimated by a national survey to physicians.17 We suspect that there is a perception that “more is better” and the marginal benefits on progression-free survival associated with doublet treatments in second-line setting reported by randomized trials likely led to the increased use of doublet treatments in the second-line setting, especially in those patients with better health conditions. However, comparing doublet treatment with platinum/pemetrexed versus monotherapy treatment with pemetrexed in our patient cohort and a meta-analysis of six randomized trials comparing doublet treatments versus monotherapy treatments in second-line setting for advanced NSCLC only observed very limited improvement on tumor response,18 which was about 5% further reduction on the risk of PD in our study. Furthermore, a recent study observed no prolonged overall survival associated with doublet treatment with carboplatin plus pemetrexed when compared to pemetrexed alone in the second-line setting for advNS-NSCLC.26 When taking into account the possible impact of markedly increased hematological and non-hematological toxicities associated with doublet treatment on quality of life27 and health resource utilization,28 future clinical practice guidelines should discourage the use of doublet treatment in the second-line setting for advNS-NSCLC.
Our study has several significant implications on clinical practice and future research. Ours is the first real-world study that has demonstrated the superior tumor response associated with pemetrexed monotherapy in the second-line setting for advNS-NSCLC and further supported clinical guideline recommendations on the role of tumor histology guiding chemotherapy for advanced NSCLC. In addition, the tumor response associated with pemetrexed in our study cohort with Chinese ethnicity was likely stronger than that based on mainly Caucasian patients. Future studies are needed to confirm our hypothesis in order to maximize health outcomes of increasing lung cancer populations in China and other Eastern Asian countries. Our study observed a comparable tumor response between pemetrexed monotherapy and the two platinum-based doublets, but pemetrexed monotherapy was associated with significantly less clinical toxicity than the two doublets. This evidence can be used to correct highly prevalent, but inappropriate, use of platinum-based doublets in the second-line setting in China and avoid possible damage to quality of life and additional health resource utilization because of increased clinical toxicity of platinum-based doublets. To further address the inappropriate use of platinum-based doublets, future studies investigating the differences in quality of life and health resource utilization between monotherapy and doublet treatments in the second-line setting are needed.
The real-world evidence generated in our study should be carefully interpreted because of the small sample size, high early treatment discontinuation rate, incomplete data, and some uncontrolled confounders that could substantially bias study results. The small number of propensity score matched pairs for pemetrexed versus docetaxel in our study failed, demonstrating significant differences in tumor response and hematological toxicity between the two monotherapy treatments. Nearly half of the patients in our study cohort discontinued treatment after the completion of one or two treatment cycles and the cause for treatment discontinuation is unknown. Thus, the median TTPs associated with the four chemotherapy regimens may be substantially underestimated by being two months less than what was reported in randomized trials.12,22,26 Our study also observed much less toxicity associated with chemotherapy when compared to the toxicity reported by randomized clinical trials. We suspect that prophylactic use of granulocyte colony-stimulating factors and antibiotics in the real-world hospital setting may have contributed to this difference.29 Future real-world studies assessing toxicities associated with chemotherapy should include utilization of health resources related to AE management to identify possible undocumented AEs. Finally, our study did not have survival data to help with interpreting the impact of the reduced risk of PD associated with pemetrexed on either the or overall survival.
Conclusion
In summary, conventional regression methods in this real-world study indicated that pemetrexed monotherapy was associated with significantly reduced risks of tumor progression and neutropenia when compared to docetaxel in the second-line setting for advNS-NSCLC in Chinese patients. Comparing pemetrexed monotherapy versus platinum/pemetrexed doublet in propensity score matched patients suggested that monotherapy and doublet treatments had comparable tumor response, but monotherapy treatment was associated with significantly less hematological and non-hematological toxicity. The limitations of this study clearly demonstrate the need for more studies with larger sample sizes and more complete data on tumor response and survival for definitive conclusions.
Acknowledgments
This study was funded and monitored by Eli Lilly China. Dr. Chengping Hu was supported by China National Key Scientific & Technology Support Program: Collaborative innovation of Clinical Research for chronic obstructive pulmonary disease and lung cancer (#2013BAI09B09). We thank Lisa M. Hess, principal research scientist of Eli Lilly & Co, for her valuable comments and editorial support. We also thank Zhuanzhuan Luo, Yi Chen, Tao Peng, Chanjue Yang, Dong Wang, and the staff at the Changsha Office of Normin Health, for their logistical support of this study.
Disclosure
This article was presented in poster format at the 15th World Conference on Lung Cancer 26-30 October 2013, Sydney, Australia. Dr Yicheng Yang, Mr Narayan Rajan, and Dr Manny Papadimitropoulos are health outcome scientists of Eli Lilly & Co. Miss. Qiong Xiao and Miss Huan Zhan are employees of Normin Health. Dr Wendong Chen is the founder of Normin Health. All other authors report no conflict of interest.
References
- She J, Yang P, Hong Q, Bai C. Lung cancer in China: challenges and interventions. Chest. 2013;143:1117–1126. doi: 10.1378/chest.11-2948. [DOI] [PubMed] [Google Scholar]
- Zhang J, Ou JX, Bai CX. Tobacco smoking in China: prevalence, disease burden, challenges and future strategies. Respirology. 2011;16:1165–1172. doi: 10.1111/j.1440-1843.2011.02062.x. [DOI] [PubMed] [Google Scholar]
- Yang L, Li L, Chen Y, Parkin DM. Mortality time trends and the incidence and mortality estimation and projection for lung cancer in China. Zhongguo Fei Ai Za Zhi. 2005;8:274–278. doi: 10.3779/j.issn.1009-3419.2005.04.05. (In Chinese.) [DOI] [PubMed] [Google Scholar]
- He Q, Wu SL, Xu F, The relationship between potential years of life lost of lung cancer and air pollution in Wuhan. Zhonghua Liu Xing Bing Xue Za Zhi. 2007;28:1175–1178. Duan JJ, Nie SF, Cao YG. (In Chinese) [PubMed] [Google Scholar]
- Chen Y, Han S, Zheng MJ, Clinical characteristics of 274 non-small cell lung cancer patients in China. Onkologie. 2013;36:248–254. doi: 10.1159/000350301. Xue Y, Liu WC. [DOI] [PubMed] [Google Scholar]
- Yang Y, Wang JJ, Wang CX, Awareness of tobacco-related health hazards among adults in China. Biomed Environ Sci. 2010;23:437–444. doi: 10.1016/S0895-3988(11)60004-4. Li Q, Yang GH. [DOI] [PubMed] [Google Scholar]
- Sun C, Yu J, Liu Z. Clinical analysis of misdiagnosis for lung cancer: a report of 31 cases. Zhongguo Fei Ai Za Zhi. 2005;8:139–140. doi: 10.3779/j.issn.1009-3419.2005.02.15. (In Chinese.) [DOI] [PubMed] [Google Scholar]
- Fan Y, Jiang Y, Chang R, Retrospective analysis of screening results of lung cancer cases among occupational population at high risk of lung cancer. Zhongguo Fei Ai Za Zhi. 2007;10:102–106. doi: 10.3779/j.issn.1009-3419.2007.02.05. Yao S, Hu P, Qiao Y. (In Chinese.) [DOI] [PubMed] [Google Scholar]
- Chattopadhyay S, Moran RG, Goldman ID. Pemetrexed: biochemical and cellular pharmacology, mechanisms, and clinical applications. Mol Cancer Ther. 2007;6:404–417. doi: 10.1158/1535-7163.MCT-06-0343. [DOI] [PubMed] [Google Scholar]
- Scagliotti G, Brodowicz T, Shepherd FA, et al. Treatment-by-histology interaction analyses in three phase III trials show superiority of pemetrexed in nonsquamous non-small cell lung cancer. J Thorac Oncol. 2011;6:64–70. doi: 10.1097/JTO.0b013e3181f7c6d4. [DOI] [PubMed] [Google Scholar]
- Al-Saleh K, Quinton C, Ellis PM. Role of pemetrexed in advanced non-small-cell lung cancer: meta-analysis of randomized controlled trials, with histology subgroup analysis. Curr Oncol. 2012;19:e9–e15. doi: 10.3747/co.19.891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sun Y, Wu YL, Zhou CC, et al. Second-line pemetrexed versus docetaxel in Chinese patients with locally advanced or metastatic non-small cell lung cancer: a randomized, open-label study. Lung Cancer. 2013;79:143–150. doi: 10.1016/j.lungcan.2012.10.015. [DOI] [PubMed] [Google Scholar]
- Nakagawa T, Otake Y, Yanagihara K, et al. Expression of thymidylate synthase is correlated with proliferative activity in non-small cell lung cancer (NSCLC) Lung Cancer. 2004;43:145–149. doi: 10.1016/j.lungcan.2003.09.004. [DOI] [PubMed] [Google Scholar]
- Giovannetti E, Mey V, Nannizzi S, et al. Cellular and pharmacogenetics foundation of synergistic interaction of pemetrexed and gemcitabine in human non-small-cell lung cancer cells. Mol Pharmacol. 2005;68:110–118. doi: 10.1124/mol.104.009373. [DOI] [PubMed] [Google Scholar]
- Scagliotti G, Hanna N, Fossella F, et al. The differential efficacy of pemetrexed according to NSCLC histology: a review of two Phase III studies. Oncologist. 2009;14:253–263. doi: 10.1634/theoncologist.2008-0232. [DOI] [PubMed] [Google Scholar]
- Delgado-Rodriguez M, Llorca J. Bias. J Epidemiol Community Health. 2004;58:635–641. doi: 10.1136/jech.2003.008466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xue C, Hu Z, Jiang W, et al. National survey of the medical treatment status for non-small cell lung cancer (NSCLC) in China. Lung Cancer. 2012;77:371–375. doi: 10.1016/j.lungcan.2012.04.014. [DOI] [PubMed] [Google Scholar]
- Di Maio M, Chiodini P, Georgoulias V, et al. Meta-analysis of single-agent chemotherapy compared with combination chemotherapy as second-line treatment of advanced non-small-cell lung cancer. J Clin Oncol. 2009;27:1836–1843. doi: 10.1200/JCO.2008.17.5844. [DOI] [PubMed] [Google Scholar]
- Cheng K, Zhao F, Gao F, et al. Factors potentially associated with chemotherapy-induced anemia in patients with solid cancers. Asian Pac J Cancer Prev. 2012;13:5057–5061. doi: 10.7314/apjcp.2012.13.10.5057. [DOI] [PubMed] [Google Scholar]
- Chen W, Wong T, Tomlinson G, Prevalence and predictors of obesity among individuals with positive hepatitis C antibody in a tertiary referral clinic. J Hepatol. 2008;49:711–717. doi: 10.1016/j.jhep.2008.05.013. Krahn M, Heathcote EJ. [DOI] [PubMed] [Google Scholar]
- Christoph DC, Asuncion BR, Hassan B, et al. Significance of folate receptor alpha and thymidylate synthase protein expression in patients with non-small-cell lung cancer treated with pemetrexed. J Thorac Oncol. 2013;8:19–30. doi: 10.1097/JTO.0b013e31827628ff. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hanna N, Shepherd FA, Fossella FV, et al. Randomized phase III trial of pemetrexed versus docetaxel in patients with non-small-cell lung cancer previously treated with chemotherapy. J Clin Oncol. 2004;22:1589–1597. doi: 10.1200/JCO.2004.08.163. [DOI] [PubMed] [Google Scholar]
- Pirker R, Pereira JR, Szczesna A, et al. Prognostic factors in patients with advanced non-small cell lung cancer: data from the phase III FLEX study. Lung Cancer. 2012;77:376–382. doi: 10.1016/j.lungcan.2012.03.010. [DOI] [PubMed] [Google Scholar]
- O'Donnell PH, Dolan ME. Cancer pharmacoethnicity: ethnic differences in susceptibility to the effects of chemotherapy. Clin Cancer Res. 2009;15:4806–4814. doi: 10.1158/1078-0432.CCR-09-0344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ou SH, Ziogas A, Zell JA. Asian ethnicity is a favorable prognostic factor for overall survival in non-small cell lung cancer (NSCLC) and is independent of smoking status. J Thorac Oncol. 2009;4:1083–1093. doi: 10.1097/JTO.0b013e3181b27b15. [DOI] [PubMed] [Google Scholar]
- Ardizzoni A, Tiseo M, Boni L, et al. Pemetrexed versus pemetrexed and carboplatin as second-line chemotherapy in advanced non-small-cell lung cancer: results of the GOIRC 02-2006 randomized phase II study and pooled analysis with the NVALT7 trial. J Clin Oncol. 2012;30:4501–4507. doi: 10.1200/JCO.2012.43.6758. [DOI] [PubMed] [Google Scholar]
- Huschka MM, Mandrekar SJ, Schaefer PL, Jett JR, Sloan JA. A pooled analysis of quality of life measures and adverse events data in north central cancer treatment group lung cancer clinical trials. Cancer. 2007;109:787–795. doi: 10.1002/cncr.22444. [DOI] [PubMed] [Google Scholar]
- Isla D, González-Rojas N, Nieves D, Treatment patterns, use of resources, and costs of advanced non-small-cell lung cancer patients in Spain: results from a Delphi panel. Clin Transl Oncol. 2011;13:460–471. doi: 10.1007/s12094-011-0683-0. Brosa M, Finnern HW. [DOI] [PubMed] [Google Scholar]
- Renner P, Milazzo S, Liu JP, Primary prophylactic colony-stimulating factors for the prevention of chemotherapy-induced febrile neutropenia in breast cancer patients. Cochrane Database Syst Rev. 2012;(10) doi: 10.1002/14651858.CD007913.pub2. Zwahlen M, Birkmann J, Horneber M. CD007913. [DOI] [PMC free article] [PubMed] [Google Scholar]