Summary Basis of Decision for Tykerb ™

3.3.1 Pharmacodynamics

A Phase I, open-label study was conducted to determine the oral dose range within which Tykerb^™ elicits a biological effect as measured by a decrease in ErbB1 and/or ErbB2 phosphorylation in cancer patients with solid tumours. Inclusion criteria for the study required that patients had a histologically confirmed diagnosis of malignancy that overexpressed ErbB1 or ErbB2.

Although the effect of Tykerb^™ on the targeted receptors (ErbB1 and ErbB2) was not demonstrated, likely due to technical difficulties, the minimal activity of the downstream signalling pathways is acceptable evidence to reflect the mechanism of action. No clear correlation between the dose administered and resulting biological activity could be determined, therefore the optimal dose could not be confirmed from this study.

3.3.2 Pharmacokinetics

Absorption

Absorption following oral administration of Tykerb^™ is incomplete and variable. Serum concentrations appear after a median lag time of 0.25 hours (range 0 to 1.5 hour). Peak plasma concentrations (C_max) of Tykerb^™ are achieved approximately 4 hours after administration. Daily dosing of 1250 mg produces steady-state geometric mean C_max values of 2.43 µg/mL and area under the curve (AUC) values of 36.2 µg•hr/mL.

Systemic exposure to Tykerb^™ is increased when administered with food. Tykerb^™ AUC values were approximately 3- and 4-fold higher (C_max approximately 2.5- and 3-fold higher) when administered with a low fat [5% fat (500 calories)] or with a high fat [50% fat (1,000 calories)] meal, respectively. To ensure a target level of exposure, Tykerb^™ should be taken at least one hour before or at least one hour after a low-fat meal.

Distribution

Tykerb^™ is highly bound (≥99%) to albumin and alpha-1 acid glycoprotein. In vitro studies indicate that Tykerb^™ is a substrate for the transporters BCRP and Pgp. In addition, Tykerb^™ has also been shown to inhibit these efflux transporters in vitro, as well as the hepatic uptake transporter OATP 1B1, at clinically relevant concentrations (IC₅₀ values were equal to 2.3 µg/mL). The clinical significance of these effects on the PKs of other drugs or the pharmacological activity of other anti-cancer agents is unknown.

Tykerb^™ undergoes extensive metabolism, primarily by CYP3A4 and CYP3A5, with minor contributions from CYP2C19 and CYP2C8, to a variety of oxidated metabolites, none of which account for more than 14% of the dose recovered in the faeces or 10% of Tykerb^™ concentration in plasma.

Tykerb^™ inhibits CYP3A4 and CYP2C8 in vitro at clinically relevant concentrations.

In healthy volunteers receiving ketoconazole, a CYP3A4 inhibitor, at 200 mg twice daily for 7 days, systemic exposure to Tykerb^™ was increased approximately 3.6-fold, and half-life increased 1.7-fold.

In healthy volunteers receiving carbamazepine, a CYP3A4 inducer, at 100 mg twice daily for 3 days and 200 mg twice daily for 17 days, systemic exposure to Tykerb^™ decreased approximately 72%.

Grapefruit juice may inhibit CYP3A4 in the gut wall and increase the bioavailability of lapatinib and should therefore be avoided during treatment with Tykerb^™.

Excretion

After single doses, the half-life of Tykerb^™ increased with increasing dose. However, daily dosing with Tykerb^™ results in achievement of steady-state within 6 to 7 days, indicating an effective half-life of 24 hours. Tykerb^™ is predominantly eliminated through metabolism by CYP3A4/5.

In a study with human volunteers, faecal excretion was the predominant route of elimination, accounting for a median of 91.8% of the recovered dose. Urinary excretion was minimal, with a median of 1.16% of the total radioactivity recovered in the urine. The median total recovery of radioactivity was 93.1% and excretion of radioactivity was completed within 168 hours.

Special Populations

Hepatic Insufficiency

The PKs of Tykerb^™ were examined in subjects with moderate (n = 8) or severe (n = 4) hepatic impairment and in healthy control subjects (n = 8). Systemic exposure (AUC) to Tykerb^™ after a single oral 100 mg dose increased approximately 56% and 85% in subjects with moderate and severe hepatic impairment, respectively. Administration of Tykerb^™ in patients with hepatic impairment should be conducted with caution due to increased exposure to the drug. There is no safety data from clinical trials on the use of Tykerb^™ in patients with severe hepatic impairment, however based on PK modeling, a dose reduction is recommended although the safety and efficacy of this dose has not been demonstrated. Patients who develop severe hepatotoxicity while taking Tykerb^™ should discontinue treatment and should not be retreated with lapatinib.

3.3.3 Clinical Efficacy

The efficacy of Tykerb^™ for the proposed indication was assessed mainly from data submitted for a single pivotal Phase III, multicentre, randomized, open-label trial, comparing Tykerb^™ plus capecitabine versus capecitabine alone in women with ErbB2 (HER2)-overexpressing advanced or metastatic breast cancer. Eligible subjects had received prior therapy with anthracyclines, taxanes, and trastuzumab. Prior treatment with capecitabine was not permitted.

Patients were randomized to receive either Tykerb^™ 1250 mg once daily (continuously) plus capecitabine (2000 mg/m²/day on days 1-14 every 21 days) (combination treatment), or to receive capecitabine alone (2500 mg/m²/day on days 1-14 every 21 days).

The primary endpoint of the study was time to progression (TTP) (defined as the interval between the date of randomization and the earliest date of disease progression or death due to breast cancer) as assessed by an independent review committee (IRC). Secondary endpoints included (but were not limited to) overall survival (OS) and progression-free survival (PFS).

The study was designed to incorporate two planned interim analyses to allow for an independent monitoring committee (IDMC) to recommend whether the trial should continue based on the safety and efficacy data. Based on the first interim analysis performed on November 15, 2005, the IDMC recommended that the study be halted due to improvement in TTP patients receiving Tykerb^™ plus capecitabine. At this time, patients receiving capecitabine alone were permitted to cross over to the treatment arm and receive Tykerb^™. An additional 75 patients were enrolled in the study between the time of the interim analysis and the updated analysis (when enrolment into the study was halted) on April 3, 2006, for a total of 399 subjects randomized to treatment (Tykerb^™ + capecitabine: 198; capecitabine: 201).

The efficacy and safety analysis for this submission included data up to the interim analysis on November 15, 2005. A supplementary results report which included data from the updated analysis (at the time of study termination April 3, 2006) was also submitted.

At the updated analysis, the IRC and investigator data demonstrated that lapatinib in combination with capecitabine significantly increased TTP compared to capecitabine alone. However, the IRC and investigator assessments of TTP were discordant (the IRC analysis of TTP was likely overestimated), thus the magnitude of improvement in TTP could not be quantified in this trial. Although investigator's results are often affected by assessment bias in unblinded studies, the investigator's results in this trial are considered a more accurate assessment of TTP.

The results of the updated analysis of the investigator's data demonstrated a statistically significant improvement in the median TTP of 5.6 weeks favouring the combination-treatment arm. One patient receiving combination treatment achieved a complete response compared to no patients in the control group. A total of 23% of patients in the combination-treatment group compared to 14% in the control group had a partial response. However, a statistically significant OS advantage, or palliation due to therapy, has not been demonstrated. The most recent analysis of OS on September 28, 2007 demonstrates an unadjusted hazard ratio of 0.90 [95% Confidence Interval (CI): 0.71, 1.12; p = 0.336]. The median OS is 65.9 weeks for capecitabine alone compared to 74.0 weeks for lapatinib + capecitabine.

On December 13, 2007, the NDS for Tykerb^™ was given an NON as a positive benefit/risk assessment could not be drawn due to multiple issues regarding the safety and efficacy of Tykerb^™. A response was filed by the sponsor on March 11, 2008. All safety issues were satisfactorily addressed by the sponsor in the Response to NON, however, the efficacy issues were still outstanding and on September 19, 2008, an NON/W was issued for Tykerb^™. The OD's pivotal concern was the lack of a convincing demonstration of efficacy of Tykerb^™ for the proposed indication. The discordance between the investigator's and the IRC results in the only pivotal trial was disconcerting in that the quality of the data could be questioned and the true magnitude of the clinical benefit could not be measured accurately. Based on the discordance in the assessed TTP benefit and without supportive studies or other measures of clinical benefit, such as evidence of palliation or OS, the OD could not recommend approval at the time of the initial review of the submission.

A Request for Reconsideration was submitted by the sponsor and on December 4, 2008, the OoS held a Reconsideration Panel (RP) hearing consisting of Canadian medical experts external to Health Canada. The OoS and the RP concluded that sufficient efficacy and clinical relevance was demonstrated in this drug submission. Although efficacy was not demonstrated as second-line therapy (as per the indication) in Phase II supportive trials, the RP was satisfied with the degree of efficacy seen in first-line Phase II trials of heavily pre-treated patients. In the one pivotal trial submitted, the RP acknowledged the discrepancy between the IRC and investigator data and agreed with the OD that the IRC's data was likely overestimated due to the imbalance in censoring between the two data sets. They also shared the opinion of the OD that the magnitude of improvement cannot be accurately measured. However, the RP felt that both the IRC and investigator's analyses consistently demonstrated an improvement in TTP, however small, and that the investigator's results, although not exempt from bias, were a more accurate analysis of TTP. The RP concluded that sufficient, although modest, efficacy was demonstrated for the intended indication, despite no other measures of clinical benefit demonstrated, such as evidence of palliation or OS.

The submission was returned to the OD for re-evaluation and the OD agreed to accept the RP's position that the data presented in the submission were sufficient to support efficacy of Tykerb^™ in the setting of ErbB2 (HER2)-overexpressing metastatic breast cancer. Although the improvement in TTP was modest, it is statistically significant and clinically relevant, and there appears to be no disadvantage in survival outcome.

3.3.4 Clinical Safety

The safety of Tykerb^™ was evaluated in the pivotal study described in section 3.3.3 Clinical Efficacy. The mean duration of treatment in the combination-treatment arm (lapatinib/capecitabine) was 21.6/20.7 weeks. The mean duration of treatment in the monotherapy (capecitabine) arm was 15.1 weeks.

The most common adverse reactions during therapy with Tykerb^™ plus capecitabine were GI (diarrhoea, nausea, and vomiting), dermatologic (palmar-plantar erythrodysaesthesia and rash), and fatigue. Diarrhoea was the most common adverse reaction (60% all grades) resulting in discontinuation of study medication (5% of patients). The most common Grade 3 and 4 adverse reactions based on the National Cancer Institute Common Toxicity Criteria Version 3.0 (NCI CTC v3.0) were diarrhoea and palmar-plantar erythrodysaesthesia.

Gastrointestinal

Diarrhoea, including severe diarrhoea, has been reported with Tykerb^™ treatment. Proactive management with anti-diarrhoea agents is important.

Hepatic/Biliary/Pancreatic

Hepatotoxicity (increased levels of alanine transaminase/aspartate transaminase, alkaline phosphatase, and bilirubin) has been observed in clinical trials (<1% of patients) and post-marketing experience. Hepatotoxicity may occur days to several months after initiation of treatment. The hepatotoxicity may be severe and deaths have been reported, although the relationship to Tykerb^™ is uncertain.

Cardiovascular

Left Ventricular Ejection Fraction and Heart Failure

Tykerb^™ has been reported to decrease left ventricular ejection fraction (LVEF). The majority (>60%) of LVEF decreases occurred within the first 9 weeks of treatment, but data on long-term exposure are limited. Rare but serious events of congestive heart failure, cardiac arrest, and sudden death have also been reported with Tykerb^™.

Due to potential cardiac toxicity with HER2 (ErbB2) inhibitors, LVEF was monitored in clinical trials at approximately 8-week intervals. Decreases in LVEF were considered a serious adverse event (SAE) if signs or symptoms of deterioration in LVEF were ≥Grade 3 according to the NCI Common Terminology Criteria for Adverse Events (CTCAE), or in cases of a ≥20% decrease in LVEF relative to baseline value and below the institution's lower limit of normal. In the pivotal study, among 177 patients who received lapatinib plus capecitabine and had screening plus at least one on-treatment LVEF measurement, 10 patients (6%) experienced a ≥20% decrease in LVEF, including 4 patients (2%) that met the SAE criteria above. Amongst 150 patients who received capecitabine monotherapy and had screening plus at least one on-treatment LVEF measurement, 9 patients (6%) experienced a ≥20% decrease in LVEF, including 4 patients (3%) that met the SAE criteria.

In 3 supportive monotherapy studies, among 338 patients who had screening plus at least one on-treatment LVEF measurement, 17 patients (5%) experienced a ≥20% decrease in LVEF, including 7 patients (2.1%) that met the SAE criteria.

Prior to commencing therapy with Tykerb^™, LVEF should be evaluated in all patients to ensure that their baseline LVEF is within normal limits. LVEF should continue to be evaluated during treatment with Tykerb^™ to ensure that it does not decline to an unacceptable level. Caution should be taken if Tykerb^™ is to be administered to patients with conditions that could impair left ventricular function.

QT Prolongation

Tykerb^™ is associated with QT/QTc interval prolongation. Many drugs that cause QT/QTc prolongation are suspected to increase the risk of torsade de pointes. If sustained, torsade de pointes can progress to ventricular fibrillation and sudden cardiac death. Events of ventricular fibrillation, cardiac arrest, and sudden death have been reported with Tykerb^™ in clinical trials.

Particular care should be exercised when administering Tykerb^™ to patients who are suspected to be at an increased risk of experiencing torsade de pointes during treatment with a QT/QTc-prolonging drug.

Respiratory

Tykerb^™ has been associated with reports of interstitial lung disease and pneumonitis. Patients should be monitored for pulmonary symptoms indicative of interstitial lung disease/pneumonitis. Tykerb^™ should be discontinued in patients who experience pulmonary symptoms indicative of interstitial lung disease/pneumonitis which are ≥Grade 3.

In addition to the AEs outlined above, a notably large number of neutropaenia SAEs were observed that were associated with clinical trials involving lapatinib. Events of GI haemorrhage, liver, and renal SAEs were also noted.