Palbociclib in combination with letrozole in patients with estrogen receptor–positive, human epidermal growth factor receptor 2–negative advanced breast cancer: PALOMA-2 subgroup analysis of Japanese patients

Background In PALOMA-2, palbociclib–letrozole significantly improved progression-free survival (PFS) vs placebo–letrozole in women with estrogen receptor–positive, human epidermal growth factor receptor 2–negative (ER+/HER2–) advanced breast cancer (ABC) in the first-line setting. We evaluated the efficacy, safety, and pharmacokinetics of palbociclib in Japanese women in PALOMA-2. Methods In this phase 3 study, 666 postmenopausal women with ER+/HER2– ABC were randomized 2:1 to palbociclib (125 mg/day [3 weeks on/1 week off]) plus letrozole (2.5 mg daily) or placebo plus letrozole. A prespecified, exploratory, subgroup analysis of Japanese patients (n = 46) was conducted to compare results with those of the overall population. Results At the February 26, 2016 cutoff, median PFS among the 46 Japanese patients was 22.2 months (95%CI, 13.6‒not estimable) with palbociclib–letrozole vs 13.8 months (5.6‒22.2) with placebo–letrozole (hazard ratio, 0.59 [95%CI, 0.26−1.34]). The most common adverse events (AEs) were hematologic and more frequent among Japanese patients than the overall population (neutropenia: 93.8% [87.5% grade 3/4] vs 79.5% [66.4%]; leukopenia: 62.5% [43.8%] vs 39.0% [24.8%]); no Japanese patients had febrile neutropenia. Palbociclib dose reductions due to toxicity (mainly neutropenia) were more common in Japanese patients (62.5% vs 36.0%); few permanently discontinued due to AEs. Although mean palbociclib trough concentration was higher in Japanese patients vs non-Asians (95.4 vs 61.7 ng/mL), the range of individual values of the Japanese patients was within that of non-Asians. Conclusions These results from PALOMA-2 suggest that palbociclib–letrozole merits consideration as a first-line treatment option for postmenopausal Japanese patients with ER+/HER2‒ ABC. ClinicalTrials.gov: NCT01740427. Electronic supplementary material The online version of this article (10.1007/s10147-018-1353-9) contains supplementary material, which is available to authorized users.


Introduction
Breast cancer is the most common cancer and fifth leading cause of cancer-related death in Japan [1]. Hormone receptor-positive (HR+) breast cancer accounts for approximately 71% of new cases worldwide [2]. Endocrine therapy is a mainstay of HR+ breast cancer treatment and is recommended by the Japanese Breast Cancer Society and international guidelines for initial treatment of HR+ disease [3,4]. However, hormonal blockade alone provides only modest benefit in women with advanced breast cancer (ABC), with many patients having de novo or acquired resistance to endocrine therapy [4].
Palbociclib is a selective, oral inhibitor of cyclin-dependent kinases 4 and 6 (CDK4/6) that prevents cell proliferation by blocking cell cycle progression from the G1 to the S phase [5]. In the multinational, phase 3 PALOMA-2 study, median progression-free survival (PFS) was significantly longer with palbociclib-letrozole than with placebo-letrozole as first-line treatment for postmenopausal women with estrogen receptor−positive (ER+)/human epidermal growth factor receptor 2-negative (HER2-) ABC, and toxicities were manageable [6]. Palbociclib is approved in the United States, the European Union, and Japan for the treatment of HR+/HER2-ABC in combination with endocrine therapy [7][8][9].
Limited data are available on the efficacy and safety of palbociclib-letrozole in Japanese patients. A phase 1 study in postmenopausal Japanese women with ER+/HER2-ABC determined that the optimal dosage of palbociclib is 125 mg once daily (QD) for 3 weeks on followed by 1 week off (3/1 schedule), the same as in Western patients [10]. A singlearm, phase 2 study evaluated the efficacy and safety of palbociclib-letrozole (3/1 regimen) in 42 postmenopausal Japanese patients with treatment-naive ER+/HER2-ABC; although follow-up is ongoing, initial results are encouraging with a 1-year probability of PFS of 75.0% (90%CI, 61.3%−84.4%) [11].
Data on the efficacy and safety of palbociclib-letrozole in Asian patients in PALOMA-2 have been reported [12]; however, some patients resided in Western countries, as the analysis included all patients who reported Asian ethnicity. Dietary habits and medical environments vary among Asian countries, and these factors may influence clinical outcomes [13,14]. Therefore, additional investigation in Japanese patients is warranted. This paper presents results of a prespecified exploratory analysis of the efficacy-including the association between efficacy and dose reduction-safety, and pharmacokinetics (PK) of palbociclib in Japanese women residing in Japan enrolled in PALOMA-2, as well as ad hoc analyses in the overall population of factors associated with neutropenia, the most common adverse event (AE) associated with palbociclib.

Study design and patients
The PALOMA-2 study has been described previously [6]. Briefly, postmenopausal women with histologically or cytologically confirmed ER+/HER2-ABC (locoregionally recurrent or metastatic) not suitable for resection or radiation therapy with curative intent were randomized 2:1 to receive palbociclib 125 mg QD (3/1 schedule) plus daily letrozole 2.5 mg or matching placebo plus letrozole. Dose reductions of palbociclib or placebo due to AEs were allowed; dose reductions of letrozole were not. Randomization was stratified by disease site (visceral/nonvisceral), disease-free interval from the end of (neo)adjuvant therapy (de novo metastatic, ≤ 12 months, > 12 months), and prior (neo)adjuvant hormonal therapy (yes, no). There was no stratification by country/region.
Patients included in this analysis were residing in Japan. The study was approved by an institutional review board or ethics committee at each site (Table S1), and all patients provided written informed consent. PALOMA-2 was conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonisation Good Clinical Practice Guidelines.

Outcomes and assessments
The primary endpoint was investigator-assessed PFS. Secondary endpoints included objective response (confirmed partial or complete response per RECIST v1.1), clinical benefit response (CBR; objective response or stable disease for ≥ 24 weeks per RECIST v1.1), safety, and PK. Radiological tumor assessments were performed at screening and every 12 weeks during treatment.
Laboratory analyses were performed on days 1 and 15 of cycles 1 and 2 then on day 1 of each subsequent cycle. AEs were assessed according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. Blood samples for PK analysis were obtained predose on day 15 of cycles 1 and 2. Plasma samples were analyzed using a validated high-performance liquid chromatography with tandem mass spectrometry.

Statistical analyses
Estimates of median PFS and corresponding 2-sided 95% confidence intervals were obtained using the Kaplan-Meier method. Cox proportional hazard models were used to calculate HRs [6]. Comparisons of PFS were done using 1-sided unstratified log-rank tests. A blinded independent central review (BICR) of PFS was conducted for all patients as a supportive analysis. Safety data are summarized using descriptive statistics in patients who received ≥ 1 dose of study treatment (as-treated population). Nominal P values are presented, and no adjustments were made for multiple testing. Additional details are published elsewhere [6]. Pearson correlation coefficients were calculated for the analyses evaluating the relationship between palbociclib trough concentration (C trough ) and body weight/body surface area (BSA)/body mass index (BMI) and factors associated with posttreatment neutrophil counts.
In contrast with the overall population, median duration of treatment among Japanese patients was similar with both treatments (palbociclib-letrozole, 13 (Table 3).
Using an updated data cutoff (May 31, 2017), with approximately 37-month median follow-up [15], median investigator-assessed PFS in Japanese patients was 24.9 months (95%CI, 13.6-38.6) with palbociclib-letrozole vs 13. In the evaluation of the association between efficacy and dose reduction, Japanese patients in the palbociclib-letrozole group who required dose reduction to 100 mg QD or 75 mg QD also showed long PFS (Fig. 4).  5). No apparent correlation was observed between steady state C trough and body weight (Fig. 6) or BSA/BMI (data not shown) in Japanese, Asian (excluding Japanese), and non-Asian patients.
The overall incidence of AEs associated with dose reductions was higher in the Japanese vs overall population (62.5% vs 36.0%, respectively) (Table S2). Neutropenia (31.3%) and decreased neutrophil count (28.1%) were the only AEs associated with dose reductions in > 1 Japanese patient. Five Japanese patients permanently discontinued palbociclib because of AEs (neutropenia in 1 patient, neutrophil count decreased in 2 patients, cerebral hemorrhage and pulmonary fibrosis in 1 patient each). Posttreatment (Cycle 1, Day 15) absolute neutrophil counts (ANCs) correlated with baseline ANC in Japanese patients, other Asians (excluding Japanese), and non-Asians [overall correlation    Fig. 7a]. No apparent correlation was observed between posttreatment ANC and steady state C trough (Fig. 7b), body weight (Fig. 7c), BSA/BMI (data not shown), or age (Fig. 7d).

Discussion
This prespecified, exploratory subgroup analysis of PAL-OMA-2 suggests that palbociclib-letrozole is effective for postmenopausal Japanese women with ER+/HER2-ABC  (Fig. S1). Based on the most recent data cutoff, median investigatorassessed PFS in Japanese patients was 24.9 months with palbociclib-letrozole vs 13.8 months with placebo-letrozole, while median PFS by BICR was 27.9 vs 16.6 months,  (Fig. 3). The difference between investigatorand BICR-assessed median PFS was also observed in the overall population (Figs. 1a, 2a), and was largely due to a higher censoring rate in the BICR analysis. Patients with investigator-diagnosed progressive disease were removed from study treatment and often switched to subsequent therapies with no further tumor assessment in the study. If progressive disease could not be confirmed on central review, the patient would be censored in the BICR analysis. Despite the difference in censoring rates, the results of BICR analysis showed preferable clinical benefit with palbociclib-letrozole and supported the findings of primary analysis. Consistent with the overall population, ORR was higher in Japanese patients treated with palbociclib-letrozole; however, CBR was lower with palbociclib-letrozole than with placebo-letrozole. This discrepancy reflects differences in the number of Japanese patients who experienced disease progression within 24 weeks of randomization: 5 patients in the palbociclib-letrozole group vs 1 in the placebo-letrozole group.
In PALOMA-2, the geometric mean palbociclib C trough at steady state was higher in Japanese (95.4 ng/mL) and other Asian patients (90.1 ng/mL) relative to non-Asians; however, individual values within each group were generally Progression-Free Survival (%)  + LET   24  19  19  15  13  11  11  8  14  12  8  7  5  3  3  3  2   8  3  1  0  2  2  3  within range of one another. Although Japanese patients in the palbociclib-letrozole arm had a lower body weight relative to the overall population (median, 53.9 kg vs 68.0 kg), no apparent relationships were observed between C trough and lower body weight/BSA/BMI. These results are consistent with those of a population PK analysis using data from other multinational studies of palbociclib, which showed that body weight had no clinically important effect on palbociclib PK and suggest there is no need for weight-based dosing [16]. Palbociclib is metabolized primarily by cytochrome P450 isozyme (CYP)3A and sulfotransferase (SULT) enzyme SULT2A1 [7]. There is wide interindividual variability in CYP3A metabolism, and analyses of cytochrome P450 activity among native Japanese, Chinese, Korean, and Caucasian populations indicate that CYP3A metabolism is independent of ethnicity and genotypes [17][18][19]. Although specific reasons were not identified for the observed difference in C trough between the two populations in PALOMA-2, the interindividual variability of CYP3A might be a factor.
In the aforementioned open-label phase 2 study of palbociclib-letrozole in Japanese patients, full PK analysis in a palbociclib PK profile subset (n = 6) showed a remarkable similarity to that in non-Japanese patients enrolled in PALOMA-1 [11,20]. In addition, in the global phase 3 study of palbociclib-fulvestrant (PALOMA-3), the withinpatient mean steady state palbociclib C trough in Japanese,  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  Asian (excluding Japanese), and non-Asian patients demonstrated relative consistency in the central tendency and range of observed values across groups, indicating similar palbociclib exposure in these subpopulations [21]. Considered together, these data suggest there is no clinically relevant difference in PK between Japanese and non-Japanese patients. Palbociclib-letrozole was well tolerated in Japanese patients, consistent with other palbociclib studies [6,10,11,22,23]; the most common AEs were neutropenia and leukopenia. A higher percentage of Japanese patients receiving palbociclib-letrozole experienced grade ≥ 3 neutropenia and grade ≥ 3 leukopenia compared with the overall population (87.5% vs 66.4% and 43.8% vs 24.8%, respectively); however, no febrile neutropenia was observed among Japanese patients.
Neutrophil counts vary among different ethnicities and ANC is generally lower in Asians vs non-Asians [24][25][26]. While historical comparisons must be interpreted cautiously, mean neutrophil counts reported for Japanese men (n = 3356) and women (n = 6027) were 3.8 (standard deviation, 1.3) × 10 3 /mm 3 and 3.5 (1.3) × 10 3 /mm 3 , respectively [27], which are slightly lower than neutrophil counts previously observed in a non-Hispanic white population (n = 4270; 4.35 × 10 3 /mm 3 [95%CI, 4.27−4.44]) [25]. In PALOMA-2, Japanese and most other Asians had baseline ANCs of < 6000/mm 3 , whereas many non-Asians had ANCs > 6000/mm 3 (Fig. 7a). Lower baseline neutrophil counts in Japanese and other Asians could potentially explain the higher rate of neutropenia observed in these patients. Posttreatment ANC correlated with baseline neutrophil counts in all 3 groups (R = 0.527). Together, the data suggest that the higher incidence of neutropenia among Japanese patients was not related to a higher C trough or lower body weight/BSA/BMI.
Although febrile neutropenia was not observed in Japanese patients in PALOMA-2, eight patients (1.8%, all non-Asian) in the overall population reported febrile neutropenia. Absolute neutrophil counts were assessed on day 15 of cycle 1. C trough trough concentration Baseline neutrophil counts in these patients were relatively low: median 2430/mm 3 (range, 1450-3300). These data suggest that lower baseline neutrophil counts may be a risk factor for febrile neutropenia as well as neutropenia with palbociclib, consistent with the widely reported association between lower baseline neutrophil counts and neutropenia/febrile neutropenia in patients receiving chemotherapy [28][29][30]. In vitro studies indicate that palbociclib causes reversible bone marrow suppression, clearly differentiating it from apoptotic cell death caused by cytotoxic chemotherapeutic agents [31]. This may explain the reduced frequency of febrile neutropenia seen with palbociclib vs cytotoxic chemotherapies. Although common, neutropenia was effectively managed with dose modifications, and few Japanese patients permanently discontinued palbociclib because of this AE. In addition, the duration of PFS in Japanese patients was not affected by dose reduction to 100 or 75 mg QD (Fig. 4).
Results from the open-label phase 2 study also suggest that dose reductions are unlikely to affect the efficacy of palbociclib in Japanese patients, although data from this study are immature and the impact of dose reductions on treatment response will require further evaluation [11]. Of note, dose reductions did not appear to compromise PFS in the overall populations of PALOMA-2 [32,33] or PALOMA-3 [34,35].
Stomatitis and nasopharyngitis were more commonly reported among Japanese patients receiving palbociclib-letrozole. All reported events in Japanese patients were grade 1 or 2, and no patients discontinued treatment or required dose reductions because of these AEs. The underlying cause for higher incidences of some AEs in Japanese patients is not clear. As described above, a lower body weight/BSA/BMI in Japanese patients does not necessitate dose adjustments, indicating that it is unlikely that these characteristics contributed to the higher incidence of certain AEs. Rather, differences in genetics, diet, or AE monitoring could possibly contribute to the observed differences.
This analysis suggests that the addition of palbociclib improved clinical outcomes in Japanese patients with ER+/ HER2-ABC. However, these results should be interpreted cautiously, as the small sample size lacks the power to draw definitive conclusions, particularly regarding efficacy. The safety profile of palbociclib-letrozole was consistent with those reported previously. Hematologic toxicities were more common among Japanese patients than in the overall population, but were successfully managed with dose modifications. Taken together, the results seen in the Japanese and overall populations in PALOMA-2 suggest that palbociclib-letrozole merits consideration as a first-line treatment option for postmenopausal Japanese patients with ER+/ HER2-ABC.