Paclitaxel

A number of first-line chemotherapy options for patients with advanced non-small cell lung cancer (NSCLC) are advocated in treatment guidelines and/or by various clinical investigators. Platinum-based chemotherapy has clearly demonstrated efficacy in patients with advanced NSCLC and is generally recommended as first-line therapy, although there is increasing interest in the use of non-platinum chemotherapy regimens.Among the platinum-based combinations currently used in clinical practice are regimens such as cisplatin or carboplatin combined with paclitaxel, vinorelbine, gemcitabine, docetaxel or irinotecan. The particular combinations employed may vary between institutions and geographical regions.

Several pharmacoeconomic analyses have been conducted on paclitaxel in NSCLC and most have focused on its use in combination with cisplatin. In terms of clinical efficacy, paclitaxel-cisplatin combinations achieved significantly higher response rates than teniposide plus cisplatin or etoposide plus cisplatin (previously thought to be among the more effective regimens available) in two large randomised trials. One of these studies showed a survival advantage for paclitaxel plus cisplatin [with or without a granulocyte colony-stimulating factor (G-CSF)] compared with etoposide plus cisplatin.

A Canadian cost-effectiveness analysis incorporated data from one of the large randomised comparative trials and showed that the incremental cost per life-year saved for outpatient administration of paclitaxel plus cisplatin versus etoposide plus cisplatin was $US22 181 (30 619 Canadian dollars; $Can) [1997 costs]. A European analysis incorporated data from the other large randomised study and showed slightly higher costs per responder for paclitaxel plus cisplatin than for teniposide plus cisplatin in The Netherlands ($US30 769 vs $US29 592) and Spain ($US19 923 vs $US19 724) but lower costs per responder in Belgium ($US22 852 vs $US25 000) and France ($US28 080 vs $US34 747) [1995/96 costs].

In other cost-effectiveness analyses, paclitaxel plus cisplatin was associated with a cost per life-year saved relative to best supportive care of approximately $US10 000 in a US study (year of costing not reported) or $US11 200 in a Canadian analysis ($Can15 400; 1995 costs). Results were less favourable when combining paclitaxel with carboplatin instead of cisplatin and particularly when G-CSF was added to paclitaxel plus cisplatin. The Canadian study incorporated the concept of extended dominance in a threshold analysis and ranked paclitaxel plus cisplatin first among several comparator regimens (including vinorelbine plus cisplatin) when the threshold level was $Can75 000 ($US54 526) per life-year saved or per quality-adjusted life-year gained (1995 values).

Lung cancer is the leading cause of cancer-related mortality worldwide. Almost 80% of all lung cancer cases are non-small cell lung cancer (NSCLC). Prognosis for most patients with NSCLC is poor, particularly since the majority have unresectable locally advanced (stage IIIB) or metastatic (stage IV) disease at the time of diagnosis.

An estimated $US8 billion is spent on the care of patients with lung cancer in the US each year, although an earlier assessment of direct medical costs specifically for NSCLC in the US was $US4.6 billion annually (year of costing not reported for either figure). Other data indicate that the average 2-year cost per patient with NSCLC in the US was $US47 941 (1992 costs).

Paclitaxel has been used as monotherapy and in combined regimens, typically with cisplatin or carboplatin, for the treatment of patients with advanced NSCLC. As monotherapy in chemotherapy-naive patients, paclitaxel 175 to 225 mg/m² (as a 3-hour intravenous infusion) usually given every 3 weeks for 3 to 10 cycles was associated with clinical response rates of 10 to 56% in various studies, although this is not a standard chemotherapy regimen in current guidelines.

Two large randomised trials compared the efficacy of intravenous paclitaxel plus cisplatin with that of etoposide plus cisplatin [Eastern CooperativeOncology Group (ECOG) study] or teniposide plus cisplatin [European Organization for Research and Treatment of Cancer (EORTC) study] in chemotherapy-naive patients with advanced NSCLC, and data from these studies were incorporated into subsequent cost-effectiveness analyses. At the time of these studies, combined regimens of cisplatin plus etoposide or teniposide were thought to be among the more effective regimens available for NSCLC. In the ECOG study, patients were randomised to three treatment arms: paclitaxel 250 mg/m2 plus cisplatin 75 mg/m² every 3 weeks plus filgrastim 5 μg/kg subcutaneously daily from day 3 until granulocyte count improved to ≥10 000/μL; paclitaxel 135 mg/m² plus cisplatin 75 mg/m² every 3 weeks; or etoposide 100 mg/m² on days 1, 2 and 3 plus cisplatin 75 mg/m2 every 3 weeks. Paclitaxel was administered as a 24-hour infusion. There were no statistically significant differences between the two paclitaxel dosage regimens for any clinical efficacy parameter. Combined results for the two paclitaxel dosage regimens showed median survival duration was significantly longer for paclitaxel plus cisplatin than for etoposide plus cisplatin (9.9 vs 7.6 months). Survival rates at 1 year (38.9 vs 31.8%) also significantly favoured paclitaxel-containing regimens. Response rates for high- (27.7%) and low-dose paclitaxel plus cisplatin (25.3%) were significantly greater when compared with that for etoposide plus cisplatin (12.4%).

In the EORTC trial, intravenous paclitaxel 175 mg/m2 plus cisplatin 80 mg/m² every 3 weeks was compared with teniposide 100 mg/m² on days 1, 3 and 5 plus cisplatin 80 mg/m² every 3 weeks. Paclitaxel was administered as a 3-hour infusion. Although no statistically significant differences were noted between treatment groups with respect to median survival time (9.7 vs 9.9 months), paclitaxel plus cisplatin was associated with a significantly higher response rate (41 vs 28%).

Assessment of quality-of-life changes in these two trials found no sustained differences between treatment groups when assessed by the EORTC QLQ-C30, EORTC LC-13 or Functional Assessment of Cancer-Lung questionnaires. Data from the EORTC study showed statistically significant benefits in favour of paclitaxel-based treatment in various functioning scales at the 6-week, but not the 12-week, time point. Another randomised study found no significant differences between paclitaxel and best supportive care in terms of quality-of-life changes as assessed with the Rotterdam Symptom Checklist.

The main adverse events associated with paclitaxel when administered as monotherapy are reversible alopecia, arthralgia/myalgia, neutropenia, peripheral neuropathy and gastrointestinal effects. Approximately 1 to 3% of patients receiving paclitaxel and premedication with a corticosteroid and antihistamine develop major hypersensitivity reactions with symptoms of dyspnoea, bronchospasm, urticaria and/or hypertension. Other adverse events associated with the drug include adverse haematological effects (in addition to neutropenia) and transient asymptomatic bradycardia. In comparative trials, haematological toxicity was significantly more likely to occur with teniposide- than paclitaxel-containing regimens, but hypersensitivity reactions, myalgia/arthralgia and peripheral neuropathy were more common with paclitaxel-based therapy. In general, grade 4 granulocytopenia, neurological events and myalgias were more common with paclitaxel plus cisplatin than with etoposide plus cisplatin, although statistical differences were not always consistent for the high- and low-dose paclitaxel groups versus etoposide-cisplatin.

Various cost-effectiveness and cost analyses have been conducted on paclitaxel in NSCLC, all of which focus on direct medical costs only. Five cost-effectiveness analyses have been conducted with paclitaxel in NSCLC in various countries. Four of the studies were conducted from the perspective of the government healthcare payer, the other (reported as an abstract) did not provide this information. Results have been converted to US dollars for ease of comparison.

In two of the five studies, clinical outcomes data were derived from a large randomised trial directly comparing paclitaxel-based regimens with another clinically relevant combination; both cost-effectiveness analyses have been published as full papers. One was a Canadian study incorporating data from a pivotal ECOG trial demonstrating a survival advantage for paclitaxel-cisplatin regimens versus etoposide plus cisplatin. Assuming paclitaxel is administered as a 3-hour intravenous infusion on an outpatient basis (rather than as a 24-hour infusion requiring 1 day of hospitalisation as was done in the ECOG trial), the incremental cost per life-year saved for paclitaxel plus cisplatin compared with etoposide plus cisplatin was $US22 181 (30 619 Canadian dollars; $Can) [1997 costs]. The addition of G-CSF to paclitaxel (24-hour infusion) plus cisplatin resulted in an unfavourable incremental cost-effectiveness ratio of $US100 390 ($Can138 578) per life-year saved compared with etoposide plus cisplatin.

The other cost-effectiveness study compared a regimen of paclitaxel plus cisplatin with a combination of teniposide plus cisplatin and incorporated data from the EORTC trial demonstrating a statistical advantage for the paclitaxel-cisplatin regimen in terms of response rate but showing no difference between groups in terms of survival. This European analysis evaluated data for four countries and showed slightly higher costs per responder for paclitaxel plus cisplatin than for teniposide plus cisplatin in The Netherlands ($US30 769 vs $US29 592) and Spain ($US19 923 vs $US19 724) but lower costs per responder in Belgium ($US22 852 vs $US25 000) and France ($US28 080 vs $US34 747) [1995/96 costs]. Total direct medical costs were higher for the paclitaxel-cisplatin combination than for teniposide plus cisplatin in all of the countries. Among the four countries, the incremental cost-effectiveness ratio was most favourable in France ($US12 323 per responder) and least favourable in The Netherlands ($US33 550 per responder); values for Belgium ($US17 775 per responder) and Spain ($US20 394 per responder) were in between.

All three cost-effectiveness analyses incorporating indirect clinical comparisons determined the cost per life-year saved with paclitaxel-containing regimens and comparator regimens relative to best supportive care. Two of these are Canadian studies which have been published in full, the other is a US analysis published as an abstract. Paclitaxel plus cisplatin was associated with a cost per life-year saved relative to best supportive care of about $US10 000 in a US analysis (year of costing nor reported) or $US11 200 in a Canadian analysis ($Can15 400; 1995 costs). This was increased when paclitaxel was combined with carboplatin ($US28 200 per life-year saved) or when G-CSF was added to paclitaxel plus cisplatin ($US58 200 per life-year saved) [both in a US analysis]. The ratio was reduced when paclitaxel was administered as monotherapy in an earlier Canadian analysis [$US3730 ($Can4778) per life-year saved; 1993 costs] but this is not a standard chemotherapy regimen in current guidelines. In the more recent Canadian analysis, paclitaxel plus cisplatin was generally associated with a less favourable cost-effectiveness ratio (i.e. higher cost per life-year saved relative to best supportive care) than most other regimens included in the analysis. However, the study also included a decision framework that incorporated the concept of extended dominance and ranked one treatment over another if the additional cost per additional benefit achieved was below a pre-established threshold level. At a threshold level of $Can75 000 ($US54 526) per life-year saved, paclitaxel (at the dosage level of 135 mg/m² every 3 weeks) plus cisplatin rated first among the various regimens evaluated. This combination also rated first at a threshold level of $Can75 000 per quality-adjusted life-year gained (1995 values).

Among the cost analyses, one study conducted in the US showed that paclitaxel plus carboplatin was associated with a 7% lower cost than vinorelbine plus cisplatin ($US19 322 vs $US20 790 over an unspecified period) whereas another US study indicated a 21% higher cost ($US48 940 vs $US40 292 over 24months). Details about the year of costing and specific costs measured were often lacking or inconsistent across the various cost analyses, all of which have been published only as abstracts or short reports.