- Email: [email protected]
paclitaxel with every 3-weeks carboplatin and weekly paclitaxel in advanced non-small cell lung cancer
original article
Annals of Oncology 17: 104–109, 2006 doi:10.1093/annonc/mdj016 Published online 25 October 2005
A randomized phase II trial comparing every 3-weeks carboplatin/paclitaxel with every 3-weeks carboplatin and weekly paclitaxel in advanced non-small cell lung cancer M. A. Socinski*1, A. Ivanova1, K. Bakri1, J. Wall1, M. Q. Baggstrom1, T. A. Hensing2, A. Mears1, M. Tynan1, J. Beaumont2, A. H. Peterman2 & H. B. Niell1
Received 4 May 2005; revised 11 July 2005; accepted 8 August 2005
Background: The optimal schedule of taxane administration has been an area of active interest in several recent clinical trials.
original article
Methods: To address a pure schedule question, we randomized 161 patients with advanced stage IIIB or IV non-small-cell lung cancer (NSCLC) to either paclitaxel 225 mg/m2 every 3 weeks · 4 cycles or 75 mg/m2/week · 12 (cumulative dose on each arm = 900 mg/m2). Both arms received concurrent carboplatin AUC 6 every 3 weeks · 4 cycles. Results: The two arms were well-balanced in terms of known prognostic factors. The overall response rate and survival outcomes were similar on the two arms. There was significantly more grade 3/4 thrombocytopenia and grade 2–4 anemia on the weekly arm but less severe myalgias/arthralgias and alopecia. No difference in the rates of peripheral neuropathy was observed; however, patients on the every 3 weeks arm reported significantly more taxane therapy-related side-effects on the functional assessment of cancer therapy taxane subscale. Conclusions: This randomized trial exploring schedule-related issues with carboplatin/paclitaxel confirms the versatility of this regimen. Key words: carboplatin, non-small cell lung cancer, paclitaxel
introduction Non-small-cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality in the USA [1]. In the year 2005, 170 000 new diagnoses will be made and 157 000 deaths from NSCLC will occur [1]. Approximately 30–40% of patients present with stage IIIB or IV disease [2]. While systemic therapy can impact the duration of survival, as well as enhance quality of life (QoL) and symptom control, it does not represent curative therapy. Platinum-based doublet regimens represent the standard of care in this setting [3]; however, only 30–40% of patients survive beyond 1 year from diagnosis. Carboplatin combined with paclitaxel is a commonly used regimen and one that has been tested in several randomized phase III trials, which show its activity to be comparable to a number of other platinum-based regimens in advanced *Correspondence to: Dr M. A. Socinski, Multidisciplinary Thoracic Oncology Program, Lineberger Comprehensive Cancer Center, CB# 7305, University of North Carolina, Chapel Hill, NC 27599, USA. Tel: +1-919-966-4431; Fax: +1-919-966-6735; E-mail: [email protected]
Published by Oxford University Press 2005
NSCLC [4–6]. This regimen has typically been delivered on a day 1 every 3 weeks schedule in advanced NSCLC. On this schedule, the principal toxicities seen are neutropenia, neuropathy and myalgias/arthralgias. Although classically a microtubular targeted agent, paclitaxel may have anti-angiogenic and proapoptotic properties when delivered on low-dose weekly schedules, making this approach of interest in advanced NSCLC [7, 8]. Belani et al. [9] have reported a randomized phase II trial exploring three different weekly approaches with the combination of carboplatin and paclitaxel. The regimen which administered carboplatin at an area under the concentration curve (AUC) of 6 every 4 weeks and paclitaxel at 100 mg/m2 on days 1, 8 and 15 yielded a median survival time of 47 weeks and a 1-year survival rate of 47%. Although initially felt to be promising, a recent report of a phase III trial that compared this regimen with standard every 3 weeks carboplatin and paclitaxel failed to show a survival or response benefit of the weekly approach [10]. Prior to knowledge of that phase III experience, we sought to address a pure schedule question with the carboplatin/paclitaxel regimen. We had previously reported
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
1 Multidisciplinary Thoracic Oncology Program, Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC; 2Evanston Northwestern Healthcare and Northwestern University, Chicago, IL, USA
original article
Annals of Oncology
no benefit in survival, response rates no QoL for treatment with carboplatin/paclitaxel beyond four cycles [11]. We subsequently designed a trial in which four cycles of carboplatin were delivered every 3 weeks; however, the paclitaxel schedule varied between every 3 weeks or weekly administration achieving the same cumulative dose over four cycles (900 mg/m2). This would allow a pure schedule question to be addressed with this commonly used regimen.
patients and methods patient selection
treatment plan and response assessment Patients were randomized to either arm A or arm B (Figure 1). Patients randomized to arm A received four cycles of carboplatin at AUC 6 (using the Calvert equation [12]) and paclitaxel at 225 mg/m2 over 3 h every 21 days. Patients randomized to arm B received four cycles of carboplatin at an AUC 6, but paclitaxel was delivered at 75 mg/m2/week for 12 weeks. Patients on both arms were to receive the identical cumulative dose of paclitaxel (900 mg/m2). Standard premedications for paclitaxel were administered and included dexamethasone 20 mg intravenously (i.v.), diphenhydramine 50 mg and ranitidine 50 mg i.v. or cimetidine 300 mg i.v., given 30 min before paclitaxel. On both arms, disease assessments evaluating response to treatment was performed after cycle 2 and following completion of all four cycles. RECIST criteria were used to evaluate response. Toxicity was assessed weekly on both arms. QoL questionnaires from the Functional Assessment of Chronic Illness Therapy (FACIT) measurement system were administered to patients at baseline (prior to treatment), after cycle 2 (6-week follow-up) and cycle 4 (12-week follow-up). This included a measure of general QoL, the Functional Assessment of Cancer Therapy-General (FACT-G) [13], a lung cancer symptom scale [14], a scale measuring the neurotoxic and other side-effects of taxane therapy (TAX) [15], and a measure of fatigue (FACIT-Fatigue) [16]. Analysis of these four QoL/symptom scales provides a reasonably comprehensive picture of the general and specific QoL impact of the disease and treatment.
statistical design The trial was a prospective, randomized, multicenter trial. The LCCC Protocol Office was the coordinating center for this trial. The randomization
Volume 17 | No. 1 | January 2006
Figure 1. Design of LCCC 2003.
algorithm included PS (Karnofsky 70–80% versus 90–100%), gender and stage (IIIB versus IV) as stratification factors to minimize any imbalance between the two arms of the trial. The maximal procedure [17] was used to randomize patients within each strata. Survival was estimated using the Kaplan–Meier method. Binomial outcomes were compared using Fisher’s exact test. The primary objective of this randomized phase II trial was to estimate the survival in the two treatment arms. The median survival time in each treatment group was expected to be 8 months. Eighty patients per group will allow a 95% confidence interval for the median to be within 1.75 months if the true median survival time is 8 months (assuming exponential survival model). The secondary objectives of this trial were to estimate the response rate and overall QoL on the two arms and to select the better schedule based on the median survival.
results patient demographics Table 1 details the demographics of the patients entered onto this trial. In total, 161 patients were entered from 10 sites. The median age was 61 years (range 42–81) and 67% of the patients were male. Eighty-eight per cent had stage IV NSCLC. It should be noted that 51% of the patients had a Karnofsky PS of 70–80% while 56% had experienced >5% weight loss. Minorities comprised 32% of all patients. Eighty-one and 80 patients were randomized to arm A and arm B, respectively. No significant differences were noted between the two arms regarding baseline patient characteristics, including self-reported QoL. All patients were evaluable for survival, while one patient on each arm did not have response information available. treatment administration A total of 247 and 250 cycles of therapy were delivered on arms A and B, respectively. The median number of cycles on both arms was four. The per cent of total carboplatin delivered was 0.99 on both arms. The per cent of paclitaxel delivered was 0.96 and 0.93 on arms A and B, respectively. toxicity The rates of the common hematological and non-hematological toxicities associated with carboplatin and paclitaxel on the two arms are shown in Table 2. Patients on arm B experienced significantly more grade 3/4 thrombocytopenia (15% versus 4% arms B and A, respectively; P < 0.001) and more grade 2–4 anemia (63% versus 40% arms B and A, respectively; P < 0.001). Patients on arm A experienced more severe myalgia/arthralgia (20% versus 5% arms A and B, respectively; P = 0.02) and alopecia (62% and 42% on arms A and B, respectively; P = 0.02).
doi:10.1093/annonc/mdj016 | 105
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
Patients were required to have histologic or cytologic proof of stage IIIB/IV NSCLC. In the case of stage IIIB, patients entered were those not eligible for combined modality therapy (malignant pleural and/or pericardial effusions or advanced supraclavicular adenopathy). Patients were required to have measurable disease and could not have received prior chemotherapy. Prior radiotherapy (including whole-brain radiotherapy) was allowed; however, patients had to be at least 1 week from completion of radiotherapy. Patients were also required to have a Karnofsky performance status (PS) of ‡70% and adequate end-organ function defined as an absolute neutrophil count ‡1500/mm3, platelets ‡100 000/mm3, creatinine £1.6 mg/dl or a calculated creatinine clearance of >30 ml/min, total bilirubin £1.5· the institutional upper normal limits (IUNL), and both aspartate aminotransferase and alanine aminotransferase £2.5· IUNL. All patients were required to participate in the QoL component of the trial and were required to give written informed consent. This trial was reviewed by the Protocol Review Committee of the Lineberger Comprehensive Cancer Center (LCCC) and the Institutional Revue Board of The University of North Carolina School of Medicine (as well as all participating centers), and labeled LCCC 2003 (Figure 1).
R A Arm A-Carboplatin AUC 6 every 3 wks x 4 Paclitaxel 225 mg/m2 every 3wksx4 Stage IIIB/IV NSCLC N Stratified by: D O -Stage (IIIB vs. IV) -KPS (70-80 vs. 90-100) M Arm B-Carboplatin AUC 6 every 3/wksx 4 Paclitaxel 75 mg/m2/wk x 12 -Gender I Z E
original article
Annals of Oncology
No difference in the rates of peripheral neuropathy between the two arms was noted.
Table 1. Patient characteristics
Median age, years (range) Gender, % male Stage (%) IV IIIB Karnofsky PS (%) 90–100% 70–80% Weight loss >5% (%) Race Caucasian (%) African American (%) Others (%)
Arm A (n = 81)
Arm B (n = 80)
All patients (n = 161)
61 (42–81) 67
60 (38–85) 62
61 (38–85) 65
88 12
88 12
88 12
49 51 56
48 52 49
48 52 53
71 20 9
65 29 6
68 24 8
PS, performance status.
quality of life Compliance with the QoL administration schedule was fair, with 80 of 81 patients on arm A and 76 of 80 patients on arm B completing the baseline QoL questionnaires. At 6 weeks, 84% and 87% of eligible patients completed the QoL questionnaires, while 74% and 64% of eligible patients completed them at T3 (12 weeks). Patients were considered to be eligible for QoL questionnaire administration if they were still receiving treatment on study. Sixteen per cent of arm A patients and 12.5% of arm B patients did not complete QoL questionnaires at 6 weeks for reasons of disease progression, withdrawal from the study or death. Comparable percentages at 12 weeks were 38% and 30%. Because only 42% of patients provided complete QoL data at all three assessment points, baseline QoL scores were compared between those who did and who did not complete all QoL assessments. There was a trend for those without complete data to have lower (worse) scores on the TAX subscale (P = 0.05). Baseline values were controlled for in the subsequent QoL analyses. Table 4 contains the general and disease- and treatmentspecific QoL results. In order to utilize all available data and summarize the overall on-study experience, we opted for an approach of averaging the QoL results from the 6- and 12-week assessments, controlling for baseline QoL, then comparing the QoL between arms. Patients in arm A reported significantly more taxane therapy side-effects than arm B on the TAX subscale (P < 0.001). Patients in both arms reported increased fatigue (P values <0.006), but there was no difference in the fatigue experience of arm A versus arm B (P = 0.7). Both study arms reported stable lung cancer symptoms, with no significant within group change (P = 0.13 and P = 0.93) or between group difference evident (P = 0.5). Interestingly, arm A had poorer general QoL on the FACT-G than arm B (P = 0.044).
Table 2. Rates of grade 2–4 toxicity arm LCCC 2003 Arm A (n = 78) Hematological Neutropenia Neutropenia with fever Thrombocytopenia Anemia Non-hematological Nausea Vomiting Myalgia/arthralgia Neuropathy Alopecia
Arm B (n = 77)
P value
G2 (%)
G3 (%)
G4 (%)
G2 (%)
G3 (%)
G4 (%)
5 0 6 36
27 3 4 3
21 3 0 1
21 0 12 51
32 3 14 12
6 3 1 0
0.56 0.98 <0.001 <0.001
9 4 20 15 62
4 3 – 4 –
0 1 – 0 –
12 9 5 10 42
5 4 – 3 –
0 1 – 0 –
0.42 0.24 0.02 0.27 0.02
P values for myalgia/arthralgia and alopecia are calculated using Fisher’s exact test. P values for other types of toxicities reflect the comparison of arm A and B based on all four categories of toxicities: grade £1, grade 2, grade 3 and grade 4. P values are obtained using the linear rank test. G, grade.
106 | Socinski et al.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
response and survival rates Table 3 shows the response and survival rates for the two arms of the study. There were no differences in the response or stable disease rates between the two arms. The median survival of all patients entered was 7.7 months with a 1-year survival rate of 22%. As this was a randomized phase II trial, no formal comparison of the survival experiences of the two arms was performed. However, the survival experience seemed more favorable for the patients on arm B compared with arm A (median survival time and 1-year survival rates were 6.6 and 8.7 months and 16% and 27% on arms A and B, respectively). The survival curves for both arms are shown in Figure 2. The survival outcomes were also described for both the good (Karnofsky 90–100%) and poor (Karnofsky 70–80%) PS patients. As is shown in Table 3, the most notable survival difference was seen
in the poor PS patients with much better survival of patients receiving treatment on arm B. Regarding subsequent therapies, 34% and 41% of patients on arms A and B, respectively, received at least second-line therapy.
original article
Annals of Oncology
discussion
Table 3. Response and survival rates Arm B (n = 79)
All patients
0 32 42 10 16
1 35 41 13 10
– – – – –
6.6 16
8.7 27
7.7 22
9.2 21
8.1 20
8.7 21
5.2 11
8.7 32
5.9 22
CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; PS, performance status.
Figure 2. LCCC 2003 survival analysis by treatment. Bottom line is arm A, top line is arm B.
Although these were not planned comparisons, the QoL analyses were repeated separately for groups with poor (Karnofsky PS 70% or 80%) and good (Karnofsky PS 90% or 100%) baseline PS. The results suggest that the poor PS group may be largely responsible for the significant between-arm findings reported above. Thus, poor PS patients in arm B had significantly fewer taxane-related side-effects (P = 0.046) than poor PS patients assigned to arm A. For the good PS patient group, taxane-related side-effects were no longer significantly lower in arm B, although there was a non-significant trend in that direction (P = 0.06). Similarly, general QoL as measured by the FACT-G did not differ between arms A and B in the good PS group (P = 0.78). However, the poor PS patients on arm B did have significantly better general QoL (P = 0.01) than did patients on arm A. The average post-baseline levels of fatigue and lung cancer symptoms did not differ significantly between the two arms in either the good or poor PS groups (P values from 0.20 to 0.67).
Volume 17 | No. 1 | January 2006
Recent ASCO guidelines in advanced NSCLC recommend platinum-based doublet regimens for those patients with a good PS [3]. No one specific platinum-based doublet is recommended. The combination of carboplatin and paclitaxel represents one ‘standard care’ regimen as it has been tested in several phase III trials [4–6]. In these trials, response rates using the every 3 weeks schedule have ranged from 17–30% with median survival times and 1-year survival rates of 8–10 months and 30–40%, respectively. Another trial [11] evaluating a duration of therapy question using this regimen suggested that no clinical benefit to therapy beyond four cycles could be demonstrated. This, along with three other trials [18–20], lead the ASCO guideline committee to recommend four cycles as a standard in the first-line treatment setting in advanced NSCLC. The vast majority of phase III trials evaluating carboplatin and paclitaxel have used doses of an AUC of 6 and 225 mg/m2 over 3 h, respectively. When this trial was initially planned, no randomized data existed in NSCLC addressing the schedule of paclitaxel in combination with carboplatin. To address this issue, this trial was designed as a pure schedule question by dividing the total dose of paclitaxel/cycle into weekly infusion over the course of four cycles while maintaining the every 3 week dosing of carboplatin. As the trial was designed as a randomized phase II, direct comparison of the two arms is not appropriate. However, independent evaluation of each of the arms shows differing toxicity profiles with more hematological toxicity (thrombocytopenia and anemia) but less non-hematological toxicity (myalgia/arthralgia and alopecia) on the weekly arm compared with the every 3 weeks arm. This observation is similar to the findings seen by Belani et al. [10] on the recently reported phase III trial. Our trial, unlike that of Belani et al. [10], also included patient-reported evaluation of symptoms, side-effects and general QoL. While the increase in fatigue and the stability of lung cancer symptoms were similar between arms, patients randomized to the every 3 weeks arm reported significantly worse taxane-related side-effects. In a relatively infrequent finding, there was also a significant difference between arms in general QoL, suggesting that taxane side-effects may have had a relatively stronger effect on overall QoL than did fatigue and lung cancer symptoms. This is an interesting possibility that may deserve further evaluation. Certainly one must be cautious in the interpretation of the QoL information, as patients who ‘dropped out’ due to disease progression or toxicity may lead to an overestimation of the reported results of the QoL for the entire sample. The overall survival outcomes of patients on this trial were relatively poor even on the control arm of every 3 weeks carboplatin and paclitaxel (median survival time 7.7 months, 1-year survival rate 22%; Table 3.). This is likely explained by the characteristics of the patients entered onto this trial. As shown in Table 1, 52% of patients had a poor PS (Karnofsky 70–80%) while 53% had significant weight loss. Also, 32% of patients accrued represented minority populations. Although we did not statistically compare survival outcomes, patients on the weekly arm fared better than patients on the every 3 weeks arm,
doi:10.1093/annonc/mdj016 | 107
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
Response CR (%) PR (%) SD (%) PD (%) Non-evaluable Survival Median (months) 1 year (%) Survival by PS Karnofsky 90–100% Median (months) 1 year (%) Karnofsky 70–80% Median (months) 1 year (%)
Arm A (n = 80)
original article
Annals of Oncology
Table 4. Arm A and B differences in post baseline average quality of life controlling for baseline values Fatigue
Taxane symptoms
FACT-G
18.5 18.0 0.56
27.4 28.0 0.77
48.9 52.9 0.01
72.3 76.8 0.04
19.9 18.5 0.20
50.5 53.9 0.06
30.9 29.0 0.48
78.0 78.9 0.78
17.0 17.5 0.67
23.6 26.9 0.25
47.3 51.8 0.05
66.3 74.6 0.01
FACT-G, Functional Assessment of Cancer Therapy-General; PS, performance status.
particularly those with a poor PS (Figure 2 and Table 3). The optimal management of the poor PS patient with advanced disease remains controversial. Although ASCO guidelines recommend single-agent therapy for this group of patients, Lilenbaum et al. [21] clearly showed a survival benefit with carboplatin and paclitaxel over paclitaxel alone (median survival times and 1-year survival rates of 4.1 and 2.5 months and 18% and 10% respectively; log rank P = 0.019) in a prospectively planned subset analysis of a large phase III trial. The observation in our trial that suggested that the poor PS patients seem to have a superior survival outcome, and potentially better patient-reported taxane toxicity and general QoL, with weekly versus every 3 weeks paclitaxel should be viewed as hypothesis-generating. Given the relatively poor survival outlook and the related importance of quality of life maintenance or improvement for these patients, further trials in the poor PS exploring this issue seem warranted. As previously mentioned, Belani et al. [10] recently reported a randomized phase III trial comparing standard every 3 weeks carboplatin and paclitaxel with the best weekly regimen observed in a prior randomized phase II trial [9]. In the trial design, four cycles of therapy were delivered. There was no difference in response rates or survival times between the two groups. Toxicity patterns reported between the two arms were similar to those seen in our study. Only 12% of patients entered onto the Belani trial were PS 2 and no survival data specific to this group was reported. Both our trial and the Belani trial [10] suggest no clear advantage of the weekly approach compared with the every-3-week administration of carboplatin with paclitaxel. On the other hand, the response and survival outcomes are not inferior with the weekly approach, suggesting that the carboplatin/paclitaxel doublet has flexibility in its schedule without erosion of the efficacy seen with this regimen. This observation is important as with the standard every-3-week regimen. Certain toxicities such as myalgia/arthralgias are often dose-limiting and could be ameliorated by changing to a weekly
108 | Socinski et al.
references 1. American Cancer Society. Cancer Facts and Figures. 2004. http:// www.cancer.org 2. Socinski MA. Chemotherapy for stage IV non-small cell lung cancer. In Detterbeck FC, Rivera MP, Socinski MA, Rosenman JG (eds): Diagnosis and Treatment of Lung Cancer: An Evidence-Based Guide for the Practicing Clinician. Philadelphia, PA: WB Saunders 2001; 307–325. 3. Pfister DG, Johnson DH, Azzoli CG et al. American Society of Clinical Oncology Treatment of Unresectable Non-Small Cell Lung Cancer Guideline: Update 2003. J Clin Oncol 2004; 22: 330–353. 4. Schiller J, Harrington D, Belani C et al. Comparison of four chemotherapy regimens for advanced non-small cell lung cancer. N Engl J Med 2002; 346: 92–98. 5. Scagliotti G, De Marinis F, Rinaldi M et al. Phase III randomized trial comparing three platinum-based doublets in advanced non-small cell lung cancer. J Clin Oncol 2002; 20: 4285–4291. 6. Kelly K, Crowley J, Bunn PA Jr et al. Randomized phase III trial of paclitaxel plus carboplatin versus vinorelbine plus cisplatin in the treatment of patients with advanced non-small-cell lung cancer: a Southwest Oncology Group trial. J Clin Oncol 2001; 19: 3210–3218.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
All patients Arm A Arm B P value Karnofsky PS 90–100% Arm A Arm B P value Karnofsky PS 70–80% Arm A Arm B P value
Lung cancer symptoms
infusion schedule. There may also be a benefit in the poor PS patient where concerns of toxicity are often a reason for pursuing single-agent therapy rather than platinum-based doublets. Although our results with the weekly approach in the poorer PS group of advanced NSCLC seem intriguing, they must be viewed as speculative until tested prospectively. Neuropathy has always been a concern with the carboplatin and paclitaxel doublet. In the original phase III trials testing this regimen [4–6, 11], rates of grade ‡3 neuropathy ranged from 11% to 17%. These initial trials usually treated for six or more cycles. The neuropathy associated with this regimen is typically cumulative. In an analysis of the duration of therapy trial performed with carboplatin and paclitaxel, 19.9% of patients experienced grade ‡2 neuropathy during the first four cycles of therapy, while 43% of patients experienced grade ‡2 neuropathy by the time they had received eight cycles of therapy. Both this trial and the trial by Belani et al. [10] limited the duration of therapy with this regimen to four cycles and had rates of grade ‡3 neuropathy of 3–6%, which was not statistically different when comparing every 3 weeks versus weekly paclitaxel infusion schedules. These data strongly support the cumulative aspect of neuropathy with this regimen and show that severe neuropathy is uncommon with this regimen when appropriate durations of therapy are used in this patient population. The observation from the QoL aspect of this trial also suggest that even with limiting therapy to four cycles, taxane-related side-effects may influence an individuals QoL. In conclusion, this randomized phase II trial exploring schedule-related issues with carboplatin/paclitaxel confirms the versatility of this regimen. The weekly schedule is certainly feasible but, like the phase III trial by Belani et al., no clear efficacy advantage of the weekly schedule is evident. There are certainly differences in toxicity profiles between weekly and every 3 weeks carboplatin/paclitaxel. The observations regarding survival in the poor PS patients should be considered hypothesis-generating, but given the toxicity advantages of the weekly approach and the QoL information from this trial, the weekly approach is certainly an acceptable alternative in this population.
Annals of Oncology
Volume 17 | No. 1 | January 2006
14. Cella D, Bonomi A, Lloyd S et al. Reliability and validity of the Functional Assessment of Cancer Therapy-Lung (FACT-L) quality of life instrument. Lung Cancer 1995; 12: 199–200. 15. Cella D, Peterman A, Hudgens S et al. Measuring the side effects of taxane therapy in oncology: the functional assesment of cancer therapy-taxane (FACT-taxane). Cancer 2003; 98: 822–831. 16. Yellen SB, Cella DF, Webster K et al. Measuring fatigue and other anemia-related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage 1997; 13: 63–74. 17. Berger VW, Ivanova A, Knoll M. Minimizing predictability while retaining balance through the use of less restrictive randomization procedures. Stat Med 2003; 22: 3017–3028. 18. Smith IE, O’Brien ME, Talbot DC et al. Duration of chemotherapy in advanced non-small-cell lung cancer: a randomized trial of three versus six courses of mitomycin, vinblastine, and cisplatin. J Clin Oncol 2001; 19: 1336–1343. 19. Andresen O, Sorenson S, Bergman B et al. Duration of chemotherapy and survival in advanced non-small cell lung cancer (NSCLC). A multicenter, prospective randomised study. Lung Cancer 2003; 41: S28. 20. DePierre A, Quoix E, Mercier M et al. Maintenance chemotherapy in advanced non-small cell lung cancer (NSCLC): a randomized study of vinorelbine (V) versus observation (OB) in patients (Pts) responding to induction therapy (French Cooperative Oncology Group). Proc Am Soc Clin Oncol 2001; 20: 309a. 21. Lilenbaum RC, Herndon JE, 2nd, List MA et al. Single-agent versus combination chemotherapy in advanced non-small-cell lung cancer: the cancer and leukemia group B (study 9730). J Clin Oncol 2005; 23: 190–196.
doi:10.1093/annonc/mdj016 | 109
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
7. Reshkin SJ, Bellizzi A, Cardone RA et al. Paclitaxel induces apoptosis via protein kinase A- and p38 mitogen-activated protein-dependent inhibition of the Na(+)/H(+) exchanger (NHE) NHE isoform 1 in human breast cancer cells. Clin Cancer Res 2003; 9: 2366–2373. 8. Lanni JS, Lowe SW, Licitra EJ et al. p53-independent apoptosis induced by paclitaxel through an indirect mechanism. Proc Natl Acad Sci USA 1997; 94: 9679–9683. 9. Belani CP, Barstis J, Perry MC et al. Multicenter, randomized trial for stage IIIB or IV non-small-cell lung cancer using weekly paclitaxel and carboplatin followed by maintenance weekly paclitaxel or observation. J Clin Oncol 2003; 21: 2933–2939. 10. Belani C, Larocca R, Rinaldi D et al. A multicenter, phase III randomized trial for stage IIIB/IV NSCLC of weekly paclitaxel and carboplatin vs. standard paclitaxel and carboplatin given every three weeks, followed by weekly paclitaxel. Proc Am Soc Clin Oncol 2004; 23: 619. 11. Socinski MA, Schell MJ, Peterman A et al. Phase III trial comparing a defined duration of therapy versus continuous therapy followed by second-line therapy in advanced-stage IIIB/IV non-small-cell lung cancer. J Clin Oncol 2002; 20: 1335–1343. 12. Calvert A, Newell D, Gumbrell L et al. Carboplatin dosage: Prospective evaluation of a simple formula based on renal function. J Clin Oncol 1989; 7: 1748–1756. 13. Cella D, Tulsky D, Gray G. The functional assessment of cancer therapy scale: development and validation of the general measure. J Clin Oncol 1993; 11: 570–579.
original article
Annals of Oncology 17: 104–109, 2006 doi:10.1093/annonc/mdj016 Published online 25 October 2005
A randomized phase II trial comparing every 3-weeks carboplatin/paclitaxel with every 3-weeks carboplatin and weekly paclitaxel in advanced non-small cell lung cancer M. A. Socinski*1, A. Ivanova1, K. Bakri1, J. Wall1, M. Q. Baggstrom1, T. A. Hensing2, A. Mears1, M. Tynan1, J. Beaumont2, A. H. Peterman2 & H. B. Niell1
Received 4 May 2005; revised 11 July 2005; accepted 8 August 2005
Background: The optimal schedule of taxane administration has been an area of active interest in several recent clinical trials.
original article
Methods: To address a pure schedule question, we randomized 161 patients with advanced stage IIIB or IV non-small-cell lung cancer (NSCLC) to either paclitaxel 225 mg/m2 every 3 weeks · 4 cycles or 75 mg/m2/week · 12 (cumulative dose on each arm = 900 mg/m2). Both arms received concurrent carboplatin AUC 6 every 3 weeks · 4 cycles. Results: The two arms were well-balanced in terms of known prognostic factors. The overall response rate and survival outcomes were similar on the two arms. There was significantly more grade 3/4 thrombocytopenia and grade 2–4 anemia on the weekly arm but less severe myalgias/arthralgias and alopecia. No difference in the rates of peripheral neuropathy was observed; however, patients on the every 3 weeks arm reported significantly more taxane therapy-related side-effects on the functional assessment of cancer therapy taxane subscale. Conclusions: This randomized trial exploring schedule-related issues with carboplatin/paclitaxel confirms the versatility of this regimen. Key words: carboplatin, non-small cell lung cancer, paclitaxel
introduction Non-small-cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality in the USA [1]. In the year 2005, 170 000 new diagnoses will be made and 157 000 deaths from NSCLC will occur [1]. Approximately 30–40% of patients present with stage IIIB or IV disease [2]. While systemic therapy can impact the duration of survival, as well as enhance quality of life (QoL) and symptom control, it does not represent curative therapy. Platinum-based doublet regimens represent the standard of care in this setting [3]; however, only 30–40% of patients survive beyond 1 year from diagnosis. Carboplatin combined with paclitaxel is a commonly used regimen and one that has been tested in several randomized phase III trials, which show its activity to be comparable to a number of other platinum-based regimens in advanced *Correspondence to: Dr M. A. Socinski, Multidisciplinary Thoracic Oncology Program, Lineberger Comprehensive Cancer Center, CB# 7305, University of North Carolina, Chapel Hill, NC 27599, USA. Tel: +1-919-966-4431; Fax: +1-919-966-6735; E-mail: [email protected]
Published by Oxford University Press 2005
NSCLC [4–6]. This regimen has typically been delivered on a day 1 every 3 weeks schedule in advanced NSCLC. On this schedule, the principal toxicities seen are neutropenia, neuropathy and myalgias/arthralgias. Although classically a microtubular targeted agent, paclitaxel may have anti-angiogenic and proapoptotic properties when delivered on low-dose weekly schedules, making this approach of interest in advanced NSCLC [7, 8]. Belani et al. [9] have reported a randomized phase II trial exploring three different weekly approaches with the combination of carboplatin and paclitaxel. The regimen which administered carboplatin at an area under the concentration curve (AUC) of 6 every 4 weeks and paclitaxel at 100 mg/m2 on days 1, 8 and 15 yielded a median survival time of 47 weeks and a 1-year survival rate of 47%. Although initially felt to be promising, a recent report of a phase III trial that compared this regimen with standard every 3 weeks carboplatin and paclitaxel failed to show a survival or response benefit of the weekly approach [10]. Prior to knowledge of that phase III experience, we sought to address a pure schedule question with the carboplatin/paclitaxel regimen. We had previously reported
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
1 Multidisciplinary Thoracic Oncology Program, Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC; 2Evanston Northwestern Healthcare and Northwestern University, Chicago, IL, USA
original article
Annals of Oncology
no benefit in survival, response rates no QoL for treatment with carboplatin/paclitaxel beyond four cycles [11]. We subsequently designed a trial in which four cycles of carboplatin were delivered every 3 weeks; however, the paclitaxel schedule varied between every 3 weeks or weekly administration achieving the same cumulative dose over four cycles (900 mg/m2). This would allow a pure schedule question to be addressed with this commonly used regimen.
patients and methods patient selection
treatment plan and response assessment Patients were randomized to either arm A or arm B (Figure 1). Patients randomized to arm A received four cycles of carboplatin at AUC 6 (using the Calvert equation [12]) and paclitaxel at 225 mg/m2 over 3 h every 21 days. Patients randomized to arm B received four cycles of carboplatin at an AUC 6, but paclitaxel was delivered at 75 mg/m2/week for 12 weeks. Patients on both arms were to receive the identical cumulative dose of paclitaxel (900 mg/m2). Standard premedications for paclitaxel were administered and included dexamethasone 20 mg intravenously (i.v.), diphenhydramine 50 mg and ranitidine 50 mg i.v. or cimetidine 300 mg i.v., given 30 min before paclitaxel. On both arms, disease assessments evaluating response to treatment was performed after cycle 2 and following completion of all four cycles. RECIST criteria were used to evaluate response. Toxicity was assessed weekly on both arms. QoL questionnaires from the Functional Assessment of Chronic Illness Therapy (FACIT) measurement system were administered to patients at baseline (prior to treatment), after cycle 2 (6-week follow-up) and cycle 4 (12-week follow-up). This included a measure of general QoL, the Functional Assessment of Cancer Therapy-General (FACT-G) [13], a lung cancer symptom scale [14], a scale measuring the neurotoxic and other side-effects of taxane therapy (TAX) [15], and a measure of fatigue (FACIT-Fatigue) [16]. Analysis of these four QoL/symptom scales provides a reasonably comprehensive picture of the general and specific QoL impact of the disease and treatment.
statistical design The trial was a prospective, randomized, multicenter trial. The LCCC Protocol Office was the coordinating center for this trial. The randomization
Volume 17 | No. 1 | January 2006
Figure 1. Design of LCCC 2003.
algorithm included PS (Karnofsky 70–80% versus 90–100%), gender and stage (IIIB versus IV) as stratification factors to minimize any imbalance between the two arms of the trial. The maximal procedure [17] was used to randomize patients within each strata. Survival was estimated using the Kaplan–Meier method. Binomial outcomes were compared using Fisher’s exact test. The primary objective of this randomized phase II trial was to estimate the survival in the two treatment arms. The median survival time in each treatment group was expected to be 8 months. Eighty patients per group will allow a 95% confidence interval for the median to be within 1.75 months if the true median survival time is 8 months (assuming exponential survival model). The secondary objectives of this trial were to estimate the response rate and overall QoL on the two arms and to select the better schedule based on the median survival.
results patient demographics Table 1 details the demographics of the patients entered onto this trial. In total, 161 patients were entered from 10 sites. The median age was 61 years (range 42–81) and 67% of the patients were male. Eighty-eight per cent had stage IV NSCLC. It should be noted that 51% of the patients had a Karnofsky PS of 70–80% while 56% had experienced >5% weight loss. Minorities comprised 32% of all patients. Eighty-one and 80 patients were randomized to arm A and arm B, respectively. No significant differences were noted between the two arms regarding baseline patient characteristics, including self-reported QoL. All patients were evaluable for survival, while one patient on each arm did not have response information available. treatment administration A total of 247 and 250 cycles of therapy were delivered on arms A and B, respectively. The median number of cycles on both arms was four. The per cent of total carboplatin delivered was 0.99 on both arms. The per cent of paclitaxel delivered was 0.96 and 0.93 on arms A and B, respectively. toxicity The rates of the common hematological and non-hematological toxicities associated with carboplatin and paclitaxel on the two arms are shown in Table 2. Patients on arm B experienced significantly more grade 3/4 thrombocytopenia (15% versus 4% arms B and A, respectively; P < 0.001) and more grade 2–4 anemia (63% versus 40% arms B and A, respectively; P < 0.001). Patients on arm A experienced more severe myalgia/arthralgia (20% versus 5% arms A and B, respectively; P = 0.02) and alopecia (62% and 42% on arms A and B, respectively; P = 0.02).
doi:10.1093/annonc/mdj016 | 105
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
Patients were required to have histologic or cytologic proof of stage IIIB/IV NSCLC. In the case of stage IIIB, patients entered were those not eligible for combined modality therapy (malignant pleural and/or pericardial effusions or advanced supraclavicular adenopathy). Patients were required to have measurable disease and could not have received prior chemotherapy. Prior radiotherapy (including whole-brain radiotherapy) was allowed; however, patients had to be at least 1 week from completion of radiotherapy. Patients were also required to have a Karnofsky performance status (PS) of ‡70% and adequate end-organ function defined as an absolute neutrophil count ‡1500/mm3, platelets ‡100 000/mm3, creatinine £1.6 mg/dl or a calculated creatinine clearance of >30 ml/min, total bilirubin £1.5· the institutional upper normal limits (IUNL), and both aspartate aminotransferase and alanine aminotransferase £2.5· IUNL. All patients were required to participate in the QoL component of the trial and were required to give written informed consent. This trial was reviewed by the Protocol Review Committee of the Lineberger Comprehensive Cancer Center (LCCC) and the Institutional Revue Board of The University of North Carolina School of Medicine (as well as all participating centers), and labeled LCCC 2003 (Figure 1).
R A Arm A-Carboplatin AUC 6 every 3 wks x 4 Paclitaxel 225 mg/m2 every 3wksx4 Stage IIIB/IV NSCLC N Stratified by: D O -Stage (IIIB vs. IV) -KPS (70-80 vs. 90-100) M Arm B-Carboplatin AUC 6 every 3/wksx 4 Paclitaxel 75 mg/m2/wk x 12 -Gender I Z E
original article
Annals of Oncology
No difference in the rates of peripheral neuropathy between the two arms was noted.
Table 1. Patient characteristics
Median age, years (range) Gender, % male Stage (%) IV IIIB Karnofsky PS (%) 90–100% 70–80% Weight loss >5% (%) Race Caucasian (%) African American (%) Others (%)
Arm A (n = 81)
Arm B (n = 80)
All patients (n = 161)
61 (42–81) 67
60 (38–85) 62
61 (38–85) 65
88 12
88 12
88 12
49 51 56
48 52 49
48 52 53
71 20 9
65 29 6
68 24 8
PS, performance status.
quality of life Compliance with the QoL administration schedule was fair, with 80 of 81 patients on arm A and 76 of 80 patients on arm B completing the baseline QoL questionnaires. At 6 weeks, 84% and 87% of eligible patients completed the QoL questionnaires, while 74% and 64% of eligible patients completed them at T3 (12 weeks). Patients were considered to be eligible for QoL questionnaire administration if they were still receiving treatment on study. Sixteen per cent of arm A patients and 12.5% of arm B patients did not complete QoL questionnaires at 6 weeks for reasons of disease progression, withdrawal from the study or death. Comparable percentages at 12 weeks were 38% and 30%. Because only 42% of patients provided complete QoL data at all three assessment points, baseline QoL scores were compared between those who did and who did not complete all QoL assessments. There was a trend for those without complete data to have lower (worse) scores on the TAX subscale (P = 0.05). Baseline values were controlled for in the subsequent QoL analyses. Table 4 contains the general and disease- and treatmentspecific QoL results. In order to utilize all available data and summarize the overall on-study experience, we opted for an approach of averaging the QoL results from the 6- and 12-week assessments, controlling for baseline QoL, then comparing the QoL between arms. Patients in arm A reported significantly more taxane therapy side-effects than arm B on the TAX subscale (P < 0.001). Patients in both arms reported increased fatigue (P values <0.006), but there was no difference in the fatigue experience of arm A versus arm B (P = 0.7). Both study arms reported stable lung cancer symptoms, with no significant within group change (P = 0.13 and P = 0.93) or between group difference evident (P = 0.5). Interestingly, arm A had poorer general QoL on the FACT-G than arm B (P = 0.044).
Table 2. Rates of grade 2–4 toxicity arm LCCC 2003 Arm A (n = 78) Hematological Neutropenia Neutropenia with fever Thrombocytopenia Anemia Non-hematological Nausea Vomiting Myalgia/arthralgia Neuropathy Alopecia
Arm B (n = 77)
P value
G2 (%)
G3 (%)
G4 (%)
G2 (%)
G3 (%)
G4 (%)
5 0 6 36
27 3 4 3
21 3 0 1
21 0 12 51
32 3 14 12
6 3 1 0
0.56 0.98 <0.001 <0.001
9 4 20 15 62
4 3 – 4 –
0 1 – 0 –
12 9 5 10 42
5 4 – 3 –
0 1 – 0 –
0.42 0.24 0.02 0.27 0.02
P values for myalgia/arthralgia and alopecia are calculated using Fisher’s exact test. P values for other types of toxicities reflect the comparison of arm A and B based on all four categories of toxicities: grade £1, grade 2, grade 3 and grade 4. P values are obtained using the linear rank test. G, grade.
106 | Socinski et al.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
response and survival rates Table 3 shows the response and survival rates for the two arms of the study. There were no differences in the response or stable disease rates between the two arms. The median survival of all patients entered was 7.7 months with a 1-year survival rate of 22%. As this was a randomized phase II trial, no formal comparison of the survival experiences of the two arms was performed. However, the survival experience seemed more favorable for the patients on arm B compared with arm A (median survival time and 1-year survival rates were 6.6 and 8.7 months and 16% and 27% on arms A and B, respectively). The survival curves for both arms are shown in Figure 2. The survival outcomes were also described for both the good (Karnofsky 90–100%) and poor (Karnofsky 70–80%) PS patients. As is shown in Table 3, the most notable survival difference was seen
in the poor PS patients with much better survival of patients receiving treatment on arm B. Regarding subsequent therapies, 34% and 41% of patients on arms A and B, respectively, received at least second-line therapy.
original article
Annals of Oncology
discussion
Table 3. Response and survival rates Arm B (n = 79)
All patients
0 32 42 10 16
1 35 41 13 10
– – – – –
6.6 16
8.7 27
7.7 22
9.2 21
8.1 20
8.7 21
5.2 11
8.7 32
5.9 22
CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; PS, performance status.
Figure 2. LCCC 2003 survival analysis by treatment. Bottom line is arm A, top line is arm B.
Although these were not planned comparisons, the QoL analyses were repeated separately for groups with poor (Karnofsky PS 70% or 80%) and good (Karnofsky PS 90% or 100%) baseline PS. The results suggest that the poor PS group may be largely responsible for the significant between-arm findings reported above. Thus, poor PS patients in arm B had significantly fewer taxane-related side-effects (P = 0.046) than poor PS patients assigned to arm A. For the good PS patient group, taxane-related side-effects were no longer significantly lower in arm B, although there was a non-significant trend in that direction (P = 0.06). Similarly, general QoL as measured by the FACT-G did not differ between arms A and B in the good PS group (P = 0.78). However, the poor PS patients on arm B did have significantly better general QoL (P = 0.01) than did patients on arm A. The average post-baseline levels of fatigue and lung cancer symptoms did not differ significantly between the two arms in either the good or poor PS groups (P values from 0.20 to 0.67).
Volume 17 | No. 1 | January 2006
Recent ASCO guidelines in advanced NSCLC recommend platinum-based doublet regimens for those patients with a good PS [3]. No one specific platinum-based doublet is recommended. The combination of carboplatin and paclitaxel represents one ‘standard care’ regimen as it has been tested in several phase III trials [4–6]. In these trials, response rates using the every 3 weeks schedule have ranged from 17–30% with median survival times and 1-year survival rates of 8–10 months and 30–40%, respectively. Another trial [11] evaluating a duration of therapy question using this regimen suggested that no clinical benefit to therapy beyond four cycles could be demonstrated. This, along with three other trials [18–20], lead the ASCO guideline committee to recommend four cycles as a standard in the first-line treatment setting in advanced NSCLC. The vast majority of phase III trials evaluating carboplatin and paclitaxel have used doses of an AUC of 6 and 225 mg/m2 over 3 h, respectively. When this trial was initially planned, no randomized data existed in NSCLC addressing the schedule of paclitaxel in combination with carboplatin. To address this issue, this trial was designed as a pure schedule question by dividing the total dose of paclitaxel/cycle into weekly infusion over the course of four cycles while maintaining the every 3 week dosing of carboplatin. As the trial was designed as a randomized phase II, direct comparison of the two arms is not appropriate. However, independent evaluation of each of the arms shows differing toxicity profiles with more hematological toxicity (thrombocytopenia and anemia) but less non-hematological toxicity (myalgia/arthralgia and alopecia) on the weekly arm compared with the every 3 weeks arm. This observation is similar to the findings seen by Belani et al. [10] on the recently reported phase III trial. Our trial, unlike that of Belani et al. [10], also included patient-reported evaluation of symptoms, side-effects and general QoL. While the increase in fatigue and the stability of lung cancer symptoms were similar between arms, patients randomized to the every 3 weeks arm reported significantly worse taxane-related side-effects. In a relatively infrequent finding, there was also a significant difference between arms in general QoL, suggesting that taxane side-effects may have had a relatively stronger effect on overall QoL than did fatigue and lung cancer symptoms. This is an interesting possibility that may deserve further evaluation. Certainly one must be cautious in the interpretation of the QoL information, as patients who ‘dropped out’ due to disease progression or toxicity may lead to an overestimation of the reported results of the QoL for the entire sample. The overall survival outcomes of patients on this trial were relatively poor even on the control arm of every 3 weeks carboplatin and paclitaxel (median survival time 7.7 months, 1-year survival rate 22%; Table 3.). This is likely explained by the characteristics of the patients entered onto this trial. As shown in Table 1, 52% of patients had a poor PS (Karnofsky 70–80%) while 53% had significant weight loss. Also, 32% of patients accrued represented minority populations. Although we did not statistically compare survival outcomes, patients on the weekly arm fared better than patients on the every 3 weeks arm,
doi:10.1093/annonc/mdj016 | 107
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
Response CR (%) PR (%) SD (%) PD (%) Non-evaluable Survival Median (months) 1 year (%) Survival by PS Karnofsky 90–100% Median (months) 1 year (%) Karnofsky 70–80% Median (months) 1 year (%)
Arm A (n = 80)
original article
Annals of Oncology
Table 4. Arm A and B differences in post baseline average quality of life controlling for baseline values Fatigue
Taxane symptoms
FACT-G
18.5 18.0 0.56
27.4 28.0 0.77
48.9 52.9 0.01
72.3 76.8 0.04
19.9 18.5 0.20
50.5 53.9 0.06
30.9 29.0 0.48
78.0 78.9 0.78
17.0 17.5 0.67
23.6 26.9 0.25
47.3 51.8 0.05
66.3 74.6 0.01
FACT-G, Functional Assessment of Cancer Therapy-General; PS, performance status.
particularly those with a poor PS (Figure 2 and Table 3). The optimal management of the poor PS patient with advanced disease remains controversial. Although ASCO guidelines recommend single-agent therapy for this group of patients, Lilenbaum et al. [21] clearly showed a survival benefit with carboplatin and paclitaxel over paclitaxel alone (median survival times and 1-year survival rates of 4.1 and 2.5 months and 18% and 10% respectively; log rank P = 0.019) in a prospectively planned subset analysis of a large phase III trial. The observation in our trial that suggested that the poor PS patients seem to have a superior survival outcome, and potentially better patient-reported taxane toxicity and general QoL, with weekly versus every 3 weeks paclitaxel should be viewed as hypothesis-generating. Given the relatively poor survival outlook and the related importance of quality of life maintenance or improvement for these patients, further trials in the poor PS exploring this issue seem warranted. As previously mentioned, Belani et al. [10] recently reported a randomized phase III trial comparing standard every 3 weeks carboplatin and paclitaxel with the best weekly regimen observed in a prior randomized phase II trial [9]. In the trial design, four cycles of therapy were delivered. There was no difference in response rates or survival times between the two groups. Toxicity patterns reported between the two arms were similar to those seen in our study. Only 12% of patients entered onto the Belani trial were PS 2 and no survival data specific to this group was reported. Both our trial and the Belani trial [10] suggest no clear advantage of the weekly approach compared with the every-3-week administration of carboplatin with paclitaxel. On the other hand, the response and survival outcomes are not inferior with the weekly approach, suggesting that the carboplatin/paclitaxel doublet has flexibility in its schedule without erosion of the efficacy seen with this regimen. This observation is important as with the standard every-3-week regimen. Certain toxicities such as myalgia/arthralgias are often dose-limiting and could be ameliorated by changing to a weekly
108 | Socinski et al.
references 1. American Cancer Society. Cancer Facts and Figures. 2004. http:// www.cancer.org 2. Socinski MA. Chemotherapy for stage IV non-small cell lung cancer. In Detterbeck FC, Rivera MP, Socinski MA, Rosenman JG (eds): Diagnosis and Treatment of Lung Cancer: An Evidence-Based Guide for the Practicing Clinician. Philadelphia, PA: WB Saunders 2001; 307–325. 3. Pfister DG, Johnson DH, Azzoli CG et al. American Society of Clinical Oncology Treatment of Unresectable Non-Small Cell Lung Cancer Guideline: Update 2003. J Clin Oncol 2004; 22: 330–353. 4. Schiller J, Harrington D, Belani C et al. Comparison of four chemotherapy regimens for advanced non-small cell lung cancer. N Engl J Med 2002; 346: 92–98. 5. Scagliotti G, De Marinis F, Rinaldi M et al. Phase III randomized trial comparing three platinum-based doublets in advanced non-small cell lung cancer. J Clin Oncol 2002; 20: 4285–4291. 6. Kelly K, Crowley J, Bunn PA Jr et al. Randomized phase III trial of paclitaxel plus carboplatin versus vinorelbine plus cisplatin in the treatment of patients with advanced non-small-cell lung cancer: a Southwest Oncology Group trial. J Clin Oncol 2001; 19: 3210–3218.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
All patients Arm A Arm B P value Karnofsky PS 90–100% Arm A Arm B P value Karnofsky PS 70–80% Arm A Arm B P value
Lung cancer symptoms
infusion schedule. There may also be a benefit in the poor PS patient where concerns of toxicity are often a reason for pursuing single-agent therapy rather than platinum-based doublets. Although our results with the weekly approach in the poorer PS group of advanced NSCLC seem intriguing, they must be viewed as speculative until tested prospectively. Neuropathy has always been a concern with the carboplatin and paclitaxel doublet. In the original phase III trials testing this regimen [4–6, 11], rates of grade ‡3 neuropathy ranged from 11% to 17%. These initial trials usually treated for six or more cycles. The neuropathy associated with this regimen is typically cumulative. In an analysis of the duration of therapy trial performed with carboplatin and paclitaxel, 19.9% of patients experienced grade ‡2 neuropathy during the first four cycles of therapy, while 43% of patients experienced grade ‡2 neuropathy by the time they had received eight cycles of therapy. Both this trial and the trial by Belani et al. [10] limited the duration of therapy with this regimen to four cycles and had rates of grade ‡3 neuropathy of 3–6%, which was not statistically different when comparing every 3 weeks versus weekly paclitaxel infusion schedules. These data strongly support the cumulative aspect of neuropathy with this regimen and show that severe neuropathy is uncommon with this regimen when appropriate durations of therapy are used in this patient population. The observation from the QoL aspect of this trial also suggest that even with limiting therapy to four cycles, taxane-related side-effects may influence an individuals QoL. In conclusion, this randomized phase II trial exploring schedule-related issues with carboplatin/paclitaxel confirms the versatility of this regimen. The weekly schedule is certainly feasible but, like the phase III trial by Belani et al., no clear efficacy advantage of the weekly schedule is evident. There are certainly differences in toxicity profiles between weekly and every 3 weeks carboplatin/paclitaxel. The observations regarding survival in the poor PS patients should be considered hypothesis-generating, but given the toxicity advantages of the weekly approach and the QoL information from this trial, the weekly approach is certainly an acceptable alternative in this population.
Annals of Oncology
Volume 17 | No. 1 | January 2006
14. Cella D, Bonomi A, Lloyd S et al. Reliability and validity of the Functional Assessment of Cancer Therapy-Lung (FACT-L) quality of life instrument. Lung Cancer 1995; 12: 199–200. 15. Cella D, Peterman A, Hudgens S et al. Measuring the side effects of taxane therapy in oncology: the functional assesment of cancer therapy-taxane (FACT-taxane). Cancer 2003; 98: 822–831. 16. Yellen SB, Cella DF, Webster K et al. Measuring fatigue and other anemia-related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage 1997; 13: 63–74. 17. Berger VW, Ivanova A, Knoll M. Minimizing predictability while retaining balance through the use of less restrictive randomization procedures. Stat Med 2003; 22: 3017–3028. 18. Smith IE, O’Brien ME, Talbot DC et al. Duration of chemotherapy in advanced non-small-cell lung cancer: a randomized trial of three versus six courses of mitomycin, vinblastine, and cisplatin. J Clin Oncol 2001; 19: 1336–1343. 19. Andresen O, Sorenson S, Bergman B et al. Duration of chemotherapy and survival in advanced non-small cell lung cancer (NSCLC). A multicenter, prospective randomised study. Lung Cancer 2003; 41: S28. 20. DePierre A, Quoix E, Mercier M et al. Maintenance chemotherapy in advanced non-small cell lung cancer (NSCLC): a randomized study of vinorelbine (V) versus observation (OB) in patients (Pts) responding to induction therapy (French Cooperative Oncology Group). Proc Am Soc Clin Oncol 2001; 20: 309a. 21. Lilenbaum RC, Herndon JE, 2nd, List MA et al. Single-agent versus combination chemotherapy in advanced non-small-cell lung cancer: the cancer and leukemia group B (study 9730). J Clin Oncol 2005; 23: 190–196.
doi:10.1093/annonc/mdj016 | 109
Downloaded from http://annonc.oxfordjournals.org/ at University of Sydney Library on May 3, 2015
7. Reshkin SJ, Bellizzi A, Cardone RA et al. Paclitaxel induces apoptosis via protein kinase A- and p38 mitogen-activated protein-dependent inhibition of the Na(+)/H(+) exchanger (NHE) NHE isoform 1 in human breast cancer cells. Clin Cancer Res 2003; 9: 2366–2373. 8. Lanni JS, Lowe SW, Licitra EJ et al. p53-independent apoptosis induced by paclitaxel through an indirect mechanism. Proc Natl Acad Sci USA 1997; 94: 9679–9683. 9. Belani CP, Barstis J, Perry MC et al. Multicenter, randomized trial for stage IIIB or IV non-small-cell lung cancer using weekly paclitaxel and carboplatin followed by maintenance weekly paclitaxel or observation. J Clin Oncol 2003; 21: 2933–2939. 10. Belani C, Larocca R, Rinaldi D et al. A multicenter, phase III randomized trial for stage IIIB/IV NSCLC of weekly paclitaxel and carboplatin vs. standard paclitaxel and carboplatin given every three weeks, followed by weekly paclitaxel. Proc Am Soc Clin Oncol 2004; 23: 619. 11. Socinski MA, Schell MJ, Peterman A et al. Phase III trial comparing a defined duration of therapy versus continuous therapy followed by second-line therapy in advanced-stage IIIB/IV non-small-cell lung cancer. J Clin Oncol 2002; 20: 1335–1343. 12. Calvert A, Newell D, Gumbrell L et al. Carboplatin dosage: Prospective evaluation of a simple formula based on renal function. J Clin Oncol 1989; 7: 1748–1756. 13. Cella D, Tulsky D, Gray G. The functional assessment of cancer therapy scale: development and validation of the general measure. J Clin Oncol 1993; 11: 570–579.
original article