- Email: [email protected]
Clinical benefit of continuing ALK inhibition with crizotinib beyond initial disease progression in patients with advanced ALK-positive NSCLC
original articles
Annals of Oncology
16. Motzer RJ, Escudier B, Oudard S et al. Phase 3 trial of everolimus for metastatic renal cell carcinoma: final results and analysis of prognostic factors. Cancer 2010; 116: 4256–4265. 17. Yao JC, Shah MH, Ito T et al. Everolimus for advanced pancreatic neuroendocrine tumors. N Engl J Med 2011; 364: 514–523. 18. Ferte C, Paci A, Zizi M et al. Natural history, management and pharmacokinetics of everolimus-induced-oral ulcers: insights into compliance issues. Eur J Cancer 2011; 47: 2249–2255. 19. Soria JC, Massard C, Izzedine H. From theoretical synergy to clinical supra-additive toxicity. J Clin Oncol 2009; 27: 1359–1361. 20. Negrier S, Gravis G, Perol D et al. Temsirolimus and bevacizumab, or sunitinib, or interferon alfa and bevacizumab for patients with advanced renal cell carcinoma (TORAVA): a randomised phase 2 trial. Lancet Oncol 2011; 12: 673–680. 21. Baselga J, Campone M, Piccart M et al. Everolimus in postmenopausal hormone receptor-positive advanced breast cancer. N Engl J Med 2012; 366: 520–529. 22. Morrow PK, Wulf GM, Ensor J et al. Phase I/II study of trastuzumab in combination with everolimus (RAD001) in patients with HER2-overexpressing metastatic breast cancer who progressed on trastuzumab-based therapy. J Clin Oncol 2011; 29: 3126–3132.
Annals of Oncology 25: 415–422, 2014 doi:10.1093/annonc/mdt572
Clinical benefit of continuing ALK inhibition with crizotinib beyond initial disease progression in patients with advanced ALK-positive NSCLC S.-H. I. Ou1*, †, P. A. Jänne2,†, C. H. Bartlett3, Y. Tang4, D.-W. Kim5, G. A. Otterson6, L. Crinò7, P. Selaru4, D. P. Cohen4, J. W. Clark8 & G. J. Riely9 1 Chao Family Comprehensive Cancer Center, University of California at Irvine, Irvine; 2Lowe Center for Thoracic Oncology and the Belfer Institute for Applied Cancer Science, Dana-Farber Cancer Institute, Harvard Medical School, Boston; 3Pfizer Oncology, New York; 4Pfizer Oncology, La Jolla, USA; 5Department of Internal Medicine, Seoul National University Hospital, Seoul, Korea; 6Department of Internal Medicine, The Ohio State University Comprehensive Cancer Center, Columbus, USA; 7 Department of Oncology, University Hospital of Perugia, Perugia, Italy; 8Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston; 9 Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, USA
Received 25 September 2013; revised 22 November 2013; accepted 27 November 2013
Background: Crizotinib is approved to treat advanced ALK-positive non-small-cell lung cancer (NSCLC), but most patients ultimately develop progressive disease (PD). We investigated whether continuing ALK inhibition with crizotinib beyond PD (CBPD) is clinically beneficial and attempted to identify clinicopathologic characteristics associated with patients who experience clinical benefit. Patients and methods: Patients with advanced ALK-positive NSCLC enrolled in two ongoing multicenter, single-arm trials who developed RECIST-defined PD were allowed to continue crizotinib if they were deriving ongoing clinical benefit. In the present retrospective analysis, continuation of CBPD was defined as >3 weeks of crizotinib treatment after PD documentation. Patients who had PD as best response to initial crizotinib treatment were excluded. Baseline and postprogression characteristics, sites of PD, and overall survival (OS) were compared in patients who continued CBPD versus those who did not. The impact of continuing CBPD on OS after adjusting for potential confounding factors was assessed. Results: Among 194 crizotinib-treated patients with RECIST-defined PD, 120 (62%) continued CBPD. A significantly higher proportion of patients who continued CBPD than patients who did not had an ECOG performance status (PS) of 0/1 at PD (96% versus 82%; P = 0.02). CBPD patients had significantly longer OS from the time of PD [median 16.4 versus 3.9 *Correspondence to: Dr Sai-Hong I. Ou, Chao Family Comprehensive Cancer Center, University of California Irvine School of Medicine, 101 City Drive, Bldg 56, RT81, Orange, CA 92868, USA. Tel: +1-714-456-8104; Fax: +1-714-456-2242; E-mail: [email protected] †
Both authors contributed equally.
© The Author 2014. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
10. Soria JC, Shepherd FA, Douillard JY et al. Efficacy of everolimus (RAD001) in patients with advanced NSCLC previously treated with chemotherapy alone or with chemotherapy and EGFR inhibitors. Ann Oncol 2009; 20: 1674–1681. 11. Buck E, Eyzaguirre A, Brown E et al. Rapamycin synergizes with the epidermal growth factor receptor inhibitor erlotinib in non-small-cell lung, pancreatic, colon, and breast tumors. Mol Cancer Ther 2006; 5: 2676–2684. 12. Johnson BE, Jackman D, Janne PA. Rationale for a phase I trial of erlotinib and the mammalian target of rapamycin inhibitor everolimus (RAD001) for patients with relapsed non small cell lung cancer. Clin Cancer Res 2007; 13: s4628–s4631. 13. Papadimtrakopoulou VA, Soria JC, Jappe A et al. Everolimus and erlotinib as second- or third-line therapy in patients with advanced non-small cell lung cancer. J Thorac Oncol 2012; 7: 1594–1601. 14. Price KA, Azzoli CG, Krug LM et al. Phase II trial of gefitinib and everolimus in advanced non-small cell lung cancer. J Thorac Oncol 2010; 5: 1623–1629. 15. Milton DT, Riely GJ, Azzoli CG et al. Phase 1 trial of everolimus and gefitinib in patients with advanced nonsmall-cell lung cancer. Cancer 2007; 110: 599–605.
original articles
Annals of Oncology
months; hazards ratio (HR) 0.27, 95% confidence interval (CI): 0.17–0.42; P < 0.0001] and from the time of initial crizotinib treatment (median 29.6 versus 10.8 months; HR 0.30, 95% CI: 0.19–0.46; P < 0.0001). The multiple-covariate Cox regression analysis revealed that CBPD remained significantly associated with improved OS after adjusting for relevant factors. Conclusions: Patients who continued CBPD were more likely to have good ECOG PS (0/1) at the time of PD. Continuing ALK inhibition with crizotinib after PD may provide survival benefit to patients with advanced ALK-positive NSCLC. Key words: treatment beyond disease progression, crizotinib, TKI, RECIST, oncogene addiction, ALK-positive NSCLC
introduction
patients and methods patients and procedures Patients with advanced ALK-positive NSCLC enrolled in either PROFILE 1001 expansion cohort or PROFILE 1005 (and treated by 1 February 2011
| Ou et al.
survival analysis The Kaplan−Meier method was used to estimate all time to tumor progression (TTP) and overall survival (OS) endpoints. Two-sided 95% confidence intervals (CIs) were determined using the Brookmeyer–Crowley method. The Cox proportional hazards models with multiple covariates were used to assess the impact of continuing CBPD on OS from the first dose of crizotinib and from the time of PD with adjustment for various potential confounding factors, retaining statistically significant factors after model selection (factors analyzed are listed in supplementary Data, available at Annals of Oncology online, along with additional details about the methodology). The hazard ratios (HRs) between the two groups were estimated using the Cox proportional hazard models with 95% Wald CIs. All analyses were carried out using SAS statistical software, version 9.2 (SAS Institute Inc., Cary, NC).
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Anaplastic lymphoma kinase (ALK)-positive non-small-cell lung cancer (NSCLC) is a heterogeneous molecular subtype of NSCLC with at least 27 ALK-fusion variants identified [1–3]. ALK kinase activity in ALK-positive NSCLC is aberrantly activated through dimerization of various fusion partners (EML4, KIF5B, KLC1, or TFG) in the cytoplasm. Crizotinib is a multitargeted small-molecule ALK inhibitor approved for the treatment of advanced ALK-positive NSCLC [4, 5]. In a randomized phase III trial, crizotinib demonstrated significantly improved progression-free survival (PFS) when compared with standardof-care chemotherapy for previously treated advanced ALKpositive NSCLC [6]. Most patients with advanced ALK-positive NSCLC treated with crizotinib eventually develop progressive disease (PD), as observed in patients with epidermal growth factor receptor (EGFR)-mutation-positive (EGFRm) NSCLC treated with EGFR tyrosine kinase inhibitors (TKIs). In patients with EGFRm NSCLC, EGFR TKI therapy is sometimes continued beyond the conventional radiographic definition of PD per Response Evaluation Criteria in Solid Tumors (RECIST) [7, 8]. Additionally, patients with EGFRm NSCLC may develop a disease flare manifested as either rapid radiographic PD or significant worsening of symptoms, including death, upon discontinuation of EGFR TKI therapy, indicating that EGFRm tumors continue to depend on EGFR signaling [9]. Disease flare following discontinuation of crizotinib in ALK-positive NSCLC has also been reported [10]. However, whether continuation of ALK inhibition with crizotinib beyond PD (CBPD) would also be beneficial to patients with ALK-positive NSCLC treated with crizotinib remains unknown. Two ongoing single-arm studies of crizotinib in advanced ALK-positive NSCLC, the molecularly enriched expansion cohort of a phase I trial (PROFILE 1001) and a phase II trial (PROFILE 1005), allowed the continuation of crizotinib beyond RECIST-defined PD in patients who continued to derive clinical benefit from crizotinib. Camidge et al. [11] have described the feasibility and tolerability of continuing CBPD in 39 patients with advanced ALK-positive NSCLC. In this report, we carried out a retrospective analysis to investigate whether or not continuation of ALK inhibition with crizotinib is clinically beneficial and whether there are any clinicopathologic characteristics of patients with ALK-positive NSCLC who could potentially benefit from this approach.
in the latter study) were included in this analysis (supplementary Figure S1, available at Annals of Oncology online). ALK rearrangement was detected by break-apart FISH [4]. Key inclusion and exclusion criteria for PROFILE 1001 have been previously described [4, 11]. Key inclusion and exclusion criteria for PROFILE 1005 were similar to those of PROFILE 1001 [5]. Major differences included treatment setting: patients in PROFILE 1005 must have failed at least one line of treatment, while patients in PROFILE 1001 could have been treatment-naïve. Disease assessment was carried out every 6 (PROFILE 1005) or 8 weeks (PROFILE 1001). Response confirmation was required in both studies. PROFILE 1005, but not PROFILE 1001, required baseline brain imaging, and if brain metastases were detected, follow-up brain imaging was required. Otherwise, imaging to assess brain metastases was carried out as clinically indicated in both trials. In this analysis, PD was based on overall investigator assessment of tumor data and RECIST-defined, unless specifically noted. All patients received crizotinib at a starting dose of 250 mg twice daily with appropriate dosing modification as needed. Data collected at protocol-specified intervals that were the nearest, but prior, to the time of PD included performance status (PS), adverse events (AEs), and clinical laboratory tests. For CBPD patients, the same schedule of activities as that used before PD was followed for collection of AE data. A period of 3 weeks of treatment after first documentation of PD was chosen as the cut-off for defining continuing or not continuing CBPD (>3 versus ≤3 weeks, respectively) because it was within one treatment cycle for the majority of the patients, the time frame within which investigators generally decide whether or not to continue CBPD. Patients who had received crizotinib as first-line treatment in PROFILE 1001 were excluded from this analysis to ensure that comparable populations of patients from the two trials were analyzed. Patients from either study whose best overall response to initial crizotinib treatment was PD were also excluded from this analysis since inclusion of these patients could confound the comparison of post-PD survival in patients who did or did not continue CBPD. These patients are highly unlikely to benefit from continuation of CBPD and are more likely to fall into the group that did not continue CBPD, thereby biasing the analysis. All patients provided written informed consent before initial study entry and Institutional Review Board (IRB) approved the protocols. The data cut-off date for this analysis was 2 January 2012.
original articles
Annals of Oncology
results patient characteristics at baseline and at the time of PD
responses to initial crizotinib treatment Maximum percentage reductions from baseline in target lesions among both groups of patients while on crizotinib treatment are shown in supplementary Figure S2, available at Annals of Oncology online. In all, 97% of patients in both groups had some tumor shrinkage. CBPD patients had a numerically higher objective response rate to initial crizotinib treatment (74%, 95% CI 65–82) than did patients who did not continue CBPD (55%, 95% CI 43–67), although this difference was not significant, as the 95% CIs overlapped (Table 2). Among patients who achieved RECIST-defined complete response (CR) or partial response (PR) to initial crizotinib treatment, 68% (89/130) continued CBPD. Among patients who achieved SD as best response to initial crizotinib treatment, 48% (31/64) continued CBPD. Among patients who maintained SD for ≥6 months on initial crizotinib treatment, 70% (16/23) continued CBPD. The median TTP from initial crizotinib treatment for the entire patient group was 6.9 months (95% CI 6.2–7.9). The median TTP from initial crizotinib treatment was numerically longer among patients who continued CBPD (7.3 months, 95% CI 6.8–8.9) than among those who did not continue CBPD (5.7
Volume 25 | No. 2 | February 2014
treatment duration beyond PD The median duration of crizotinib treatment post-PD (total number of active dosing days from first PD to last dose) among the 120 CBPD patients in this analysis was 19.4 weeks (95% CI 16.7–28.9). The duration of CBPD was comparable between patients who received crizotinib as second-line treatment (n = 27; 20.7 weeks, 95% CI 11.9–29.6) and those who received crizotinib as third- or later-line treatment (n = 93; 19.4 weeks, 95% CI 16.0–32.0). Among the 74 patients who did not continue CBPD, 37 (50%) received subsequent systemic treatment, and the remaining 37 (50%) received best supportive care only.
survival analysis The median OS for the overall population from the time of PD was 8.9 months (95% CI 7.2–16.4; Table 2). The median OS from the time of PD was significantly longer for CBPD patients than for patients who did not continue CBPD [16.4 months (95% CI 14.5–not reached) versus 3.9 months (95% CI 2.7–5.1); HR 0.27, 95% CI 0.17–0.42; P < 0.0001; Figure 1A]. Six- and 12month survival probabilities after PD for CBPD patients were 76.3% (95% CI 66.7–83.5) and 64.7% (95% CI 53.0–74.3), respectively, compared with 6- and 12-month survival probabilities of 31.2% (95% CI 20.0–43.0) and 23.9% (95% CI 13.3– 36.1), respectively, for patients who did not continue CBPD. A possible reason for the survival disadvantage seen among patients who did not continue CBPD was that a subset of these patients never received further anticancer therapy post-PD. To address this, we evaluated the impact of receiving or not receiving subsequent systemic anticancer treatment in this group. The median OS post-PD among patients who did not continue crizotinib but received subsequent systemic therapy (n = 37; 5.4 months, 95% CI 3.8–12.3) was longer than that of those who did not continue crizotinib and received no subsequent therapy (n = 37; 2.2 months, 95% CI 1.1–3.8; Figure 1B). Nevertheless, the median OS of patients who did not continue crizotinib but received subsequent systemic therapy remained significantly shorter than that of CBPD patients (5.4 versus 16.4 months; HR 0.38, 95% CI 0.22–0.66; P = 0.0005). The median OS for the overall population (n = 194) was 21.9 months (95% CI 16.6–not reached) from the start of initial crizotinib treatment (Table 2). The median OS from the start of initial crizotinib treatment was significantly longer for CBPD patients than for patients who did not continue crizotinib [29.6 months (95% CI: 23.1–not reached) versus 10.8 months (95% CI: 8.9–14.7); HR 0.30, 95% CI: 0.19–0.46; P < 0.0001; Figure 1C]. After adjusting for potential confounding factors (supplementary Data, available at Annals of Oncology online), continuing CBPD remained significantly associated with improved OS post-PD (HR 0.27; P < 0.0001; Table 3). Both response (CR/PR) to initial crizotinib treatment (HR 0.48; P = 0.01) and longer TTP (HR 0.78; P = 0.01) were also found to be factors associated with improved OS (Table 3). Eight patients (all did not continue CBPD) died ≤3 weeks after PD documentation. After adjusting for potential confounding factors and analyzing the 186 patients
doi:10.1093/annonc/mdt572 |
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
A total of 414 patients with advanced ALK-positive NSCLC who had been enrolled in PROFILE 1001 (expansion cohort) or PROFILE 1005 (and treated by 1 February 2011) were included in this analysis (supplementary Figure S1, available at Annals of Oncology online). A total of 194 patients who had developed PD by the time of data cut-off constituted the final study population. One hundred and twenty patients (62%) continued crizotinib for >3 weeks post-PD, while 74 patients (38%) did not continue crizotinib beyond 3 weeks post-PD at the discretion of the investigators. The median duration of follow-up post-PD for the 120 patients who continued CBPD was 25.2 weeks (range 3.7– 115.9). Baseline and post-progression characteristics of these patients are provided in Table 1. The groups were comparable in age at baseline, proportions of patients from the two trials, smoking history, histology, and prior treatment history. The only significant difference between the two groups was the PS at the time of PD (Table 1). Among patients with Eastern Cooperative Oncology Group PS (ECOG PS) of 0 or 1 at the time of PD, 65% (115/176) continued CBPD. On the other hand, among patients with ECOG PS of ≥2 at the time of PD, only 29% (4/14) continued CBPD. We carried out an analysis of the sites of PD to investigate whether or not the brain may be a common site of PD in patients who continued CBPD. In the absence of standard minimum tumor growth criteria defining individual sites of PD when there are multiple sites of progressing target lesions, we analyzed the subset of patients who had PD involving new lesions and/or non-target lesions (N = 138). Among these patients, half of the patients (51%) who continued CBPD had brain as a site of PD, while patients who did not continue CBPD had both liver (37%) and brain (28%) as a site of PD (supplementary Table S1, available at Annals of Oncology online).
months, 95% CI 5.4–6.8), although the 95% CIs overlapped (Table 2).
original articles
Annals of Oncology
Table 1. Baseline and post-progression characteristics of patients who continued crizotinib beyond progressive disease and those who did not continue crizotinib beyond progressive disease Characteristic
n (%) All patients (n = 194)
Continued CBPD (n = 120)
Did not continue CBPD (n = 74)
50 (50) 21–78
52 (54) 28–78
0.30b
103 (86) 17 (14)
68 (92) 6 (8)
0.20c
65 (54) 55 (46)
35 (47) 39 (53)
0.35c
76 (63) 39 (32) 3 (3) 2 (2)
34 (46) 34 (46) 5 (7) 1 (1)
0.06d
48 (40) 72 (60)
23 (31) 51 (69)
0.21c
83 (69) 35 (29) 2 (2)
53 (72) 19 (26) 2 (3)
0.73d
33 (28) 70 (58) 16 (13) 1 (1)
14 (19) 48 (65) 11 (15) 1 (1)
0.53d
37 (31) 78 (65) 2 (2) 2 (2) 1 (1)
18 (24) 43 (58) 8 (11) 2 (3) 3 (4)
0.02d
113 (94) 7 (6)
71 (96) 3 (4)
0.90c
27 (22) 93 (78)
15 (20) 59 (80)
0.71c
115 (96) 5 (4)
72 (97) 2 (3)
0.71d
Percentages may not total 100% due to rounding. Continued CBPD versus did not continue CBPD. b t-test of means. c 2 χ test of general association. d Fisher’s exact test of general association when sample size requirement for χ 2 test is not met. e Last ECOG PS determined before disease progression. f Includes large-cell carcinoma, squamous-cell carcinoma, and adenosquamous carcinoma. g Includes neoadjuvant, adjuvant, and metastatic use. CBPD, crizotinib beyond progressive disease; ECOG PS, Eastern Cooperative Oncology Group performance status. a
| Ou et al.
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Age (years) Mean (median) 51 (51) Range 21–78 Age category <65 years 171 (88) ≥65 years 23 (12) Sex Male 100 (52) Female 94 (48) Ethnicity White 110 (57) Asian 73 (38) African-American 8 (4) Other 3 (2) Clinical trial PROFILE 1001 71 (37) PROFILE 1005 123 (63) Smoking history Never-smoker 136 (70) Former smoker 54 (28) Current smoker 4 (2) ECOG PS at baseline 0 47 (24) 1 118 (61) 2 27 (14) 3 2 (1) ECOG PS at disease progressione 0 55 (28) 1 121 (62) 2 10 (5) ≥3 4 (2) Missing 4 (2) Histology Adenocarcinoma 184 (95) Otherf 10 (5) Prior lines of therapy for advanced or metastatic disease 1 42 (22) ≥2 152 (78) Prior platinum-based therapyg Yes 187 (96) No 7 (4)
P-valuea
original articles
Annals of Oncology
Table 2. Clinical outcomes in patients who continued CBPD versus those who did not continue CBPD Outcome
Continued CBPD (n = 120)
Did not continue CBPD (n = 74)
130 (67) 60–74 64 (33)
89 (74) 65–82 31 (26)
41 (55) 43–67 33 (45)
7±7 5–48 92 (71) 24 (18) 9 (7) 5 (4)
7±8 5–48 61 (69) 19 (21) 5 (6) 4 (4)
7±5 5–24 31 (76) 5 (12) 4 (10) 1 (2)
7 (11) 34 (53) 13 (20) 7 (11) 3 (5)
2 (6) 13 (42) 7 (23) 7 (23) 2 (6)
5 (15) 21 (64) 6 (18) 0 (0) 1 (3)
6.9 6.2–7.9 59.0 51.7–65.6 83 (43)
7.3 6.8–8.9 65.6 56.3–73.3 36 (30)
5.7 5.4–6.8 48.5 36.7–59.3 47 (64)
8.9 7.2–16.4 59.4 51.3–66.5 49.3 40.5–57.5
16.4 14.5–NR 76.3 66.7–83.5 64.7 53.0–74.3
3.9 2.7–5.1 31.2 20.0–43.0 23.9 13.3–36.1
21.9 16.6–NR 93.8 89.4–96.4 68.5 61.3–74.6
29.6 23.1–NR 97.5 92.5–99.2 81.7 73.2–87.7
10.8 8.9–14.7 87.8 77.9–93.5 46.6 34.7–57.8
a
Derived from investigator-reported tumor data in order to more accurately characterize best response. Among patients who had PR or CR on initial crizotinib treatment. c Time from first dose to disease progression among patients who had SD on initial crizotinib treatment. CBPD, crizotinib beyond progressive disease; CR, complete response; PR, partial response; ORR, objective response rate; CI, confidence interval; SD, stable disease; TTP, time to tumor progression; OS, overall survival; NR, not reached. b
who survived beyond the 3 weeks defining CBPD, continuing CBPD remained significantly associated with improved OS post-PD.
discussion Molecularly targeted therapy is increasingly being continued beyond PD, especially in cancers harboring an underlying oncogenic dependence on the molecular target, such as trastuzumab in HER2-positive breast cancer [12, 13] and EGFR TKIs in
Volume 25 | No. 2 | February 2014
EGFRm NSCLC [14]. In the present retrospective analysis, we showed that patients with advanced ALK-positive NSCLC derived clinical benefit from continued ALK inhibition with crizotinib after RECIST-defined PD. These patients were more likely to have good PS at the time of PD, had responded to and exhibited extended TTP from initial crizotinib treatment, and had a site of PD particularly amenable to local therapy (brain). Not surprisingly, these patients also had a better prognosis as demonstrated by their longer OS from the start of initial crizotinib treatment (Figure 1C). While it may seem intuitive that
doi:10.1093/annonc/mdt572 |
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Best response to initial crizotinib treatmenta CR/PR (ORR), n (%) 95% CI SD, n (%) Time to first response (weeks)b Median ± standard deviation Range 0 to <8 weeks, n (%) 8 to <16 weeks, n (%) 16 to <24 weeks, n (%) ≥24 weeks, n (%) Duration of SD,c n (%) 0 to <3 months 3 to <6 months 6 to <9 months 9 to <12 months ≥12 months TTP (months) Median 95% CI 6-month TTP probability (%) 95% CI Deaths, n (%) OS from the time of PD Median (months) 95% CI 6-month OS probability (%) 95% CI 12-month OS probability (%) 95% CI OS from first dose of crizotinib Median (months) 95% CI 6-month OS probability (%) 95% CI 12-month OS probability (%) 95% CI
All patients (n = 194)
original articles A
Annals of Oncology
100
Continued crizotinib Did not continue crizotinib
Probability of survival (%)
80
60
40
20
0
0
5
10
15
20
25
30
Time (months)
B
29 7
4 0
100
0
Continued crizotinib Did not continue crizotinib, + ST Did not continue crizotinib, – ST
Probability of survival (%)
80
60
40
20
0
0
10
5
15
30
25
20
Time (months) Number at risk Continued 120 Did not continue, + ST 37 Did not continue, – ST 37
C
29 5 2
4 0 0
Probability of survival (%)
100
0
Continued crizotinib Did not continue crizotinib
80
60
40
20
0 0
5
10
15
20
25
30
35
40
Time (months) Number at risk Continued 120 Did not continue 74
104 40
30 8
6 0
1
Figure 1. Overall survival of patients who continued crizotinib versus those who did not continue crizotinib beyond progressive disease. (A) Overall survival from the time of progressive disease. (B) Overall survival from the time of progressive disease of patients who continued crizotinib beyond progressive disease versus those who did not and who either received or did not receive subsequent systemic therapy (ST). (C) Overall survival from the time of initial crizotinib treatment. Shaded areas are 95% Hall–Wellner confidence bands.
| Ou et al.
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Number at risk Continued 120 Did not continue 74
original articles
Annals of Oncology
Table 3. The Cox proportional hazards analysis of overall survival to assess the impact of continuing crizotinib beyond progressive disease while adjusting for other patient/disease characteristics or outcomes using backward model selectiona
Crizotinib beyond PD (yes versus no) Sex (female versus male) Smoking status(never versus ex/smoker) ECOG PS by the time of PDc 1 versus 0 ≥2 versus 0 Best response (CR/PR versus SD)d TTP (1-month increments)e New lesions only at PDd (yes versus no)
OS from the time of PD (n = 194) HRb 95% CI
P-value
OS from first dose of crizotinib (n = 194) HRb 95% CI P-value
0.27 — 0.54
0.17–0.45 — 0.32–0.89
<0.0001 — 0.02
0.28 0.58 —
0.18–0.44 0.37–0.91 —
<0.0001 0.02 —
1.75 10.94 0.48 0.78 1.87
1.00–3.04 4.92–24.3 0.28–0.85 0.66–0.93 1.06–3.30
0.05 <0.0001 0.01 0.01 0.03
— — — — —
— — — — —
— — — — —
a
patients who derive clinical benefit on initial crizotinib treatment benefit from continued ALK inhibition with crizotinib, our report represents the largest systematic collection of the clinicopathologic data from these patients to date and shows, reassuringly, that continuation of ALK inhibition with CBPD was practiced by many of the investigators in the two trials. ALK-positive NSCLC is a heterogeneous disease with many EML4-ALK fusion variants that likely present a diverse clinical picture and may exhibit differential responses to crizotinib and other ALK inhibitors [15]. Several crizotinib resistance mechanisms have been reported, including generation of secondary resistance mutations, amplification of the ALK gene, and activation of other bypass signaling pathways [16–19]. However, the resistance mechanism is unknown in the majority of the patients whose disease progressed on crizotinib due to low frequency of re-biopsy, especially when the progressing metastatic lesions occur in the brain which are not easily amenable to biopsy. There are limitations to the present retrospective analysis despite efforts to adjust for confounding factors by the Cox regression analysis with multiple covariates and to exclude patients who did not respond at all to initial crizotinib treatment. While there were no gross imbalances in the clinicopathologic characteristics of the two groups of patients analyzed, there may be unknown confounding factors not considered in our analysis. Therefore, the ability to continue CBPD could simply represent a subgroup with favorable prognostic factors rather than a direct post-PD treatment effect. Data on the use of ALK inhibitors other than crizotinib, subsequent disease progression while on continued CBPD, the type of chemotherapy regimen used after PD, and the mode of any radiation therapy delivered to patients with PD on crizotinib were not captured.
Volume 25 | No. 2 | February 2014
As CBPD is not expected to provide clinical benefit indefinitely, other strategies such as the addition of local therapy [20], sequencing chemotherapy [21], or addition of HSP90 inhibitors [22] to CBPD should be evaluated prospectively in this treatment setting. Furthermore, several ALK inhibitors are in clinical development primarily for PD after crizotinib [23–25]. Given the potential survival benefit in ALK-positive NSCLC patients who continue CBPD reported herein, direct comparative studies between these newer ALK inhibitors versus CBPD plus additional treatment modalities would be needed to inform optimal treatment after PD on crizotinib.
acknowledgements We would like to thank all of the patients and their families, as well as the investigators, research nurses, study coordinators, and operations staff who participated in PROFILE 1001 and PROFILE 1005. We would also like to thank Kristen Letrent (Pfizer) for her contributions to the design of the analyses and Keith Wilner (Pfizer) for his lead role in the conduct of both studies.
funding This work was sponsored by Pfizer Inc. Editorial support was provided by Wendy Sacks at ACUMED® (New York, NY, USA) and was funded by Pfizer Inc [no grant numbers].
disclosure S-HIO has served as a consultant for and received honoraria from Pfizer. PAJ has served as a consultant for Pfizer. CHB, YT, PS, and DPC are employees of Pfizer and own Pfizer stock.
doi:10.1093/annonc/mdt572 |
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
A complete list of variables included in this analysis and a summary of the methodology are provided in supplementary Data, available at Annals of Oncology online. b For categorical variables, HR < 1 favors the first category and HR > 1 favors the second category; for continuous variables, for every unit increase in the variable, there is an increase (HR > 1) or decrease (HR < 1) in the risk of death. c Last ECOG PS determined before disease progression. d Derived from investigator-reported tumor data in order to more accurately characterize best response and site/mode of PD. e Continuous variable. OS, overall survival; PD, progressive disease; HR, hazard ratio; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; CR, complete response; PR, partial response; SD, stable disease; TTP, time to tumor progression.
original articles D-WK has served as a consultant for Pfizer, GlaxoSmithKline, and Lilly and has received honoraria from Pfizer and Lilly. GJR’s institution has received research funding from Pfizer for research carried out by GJR.
Annals of Oncology
13.
14.
references
| Ou et al.
15.
16. 17. 18.
19.
20.
21.
22.
23.
24. 25.
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
1. Soda M, Choi YL, Enomoto M et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature 2007; 448: 561–566. 2. Rikova K, Guo A, Zeng Q et al. Global survey of phosphotyrosine signaling identifies oncogenic kinases in lung cancer. Cell 2007; 131: 1190–1203. 3. Ou SH, Bartlett CH, Mino-Kenudson M et al. Crizotinib for the treatment of ALKrearranged non-small cell lung cancer: a success story to usher in the second decade of molecular targeted therapy in oncology. Oncologist 2012; 17: 1351–1375. 4. Kwak EL, Bang YJ, Camidge DR et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010; 363: 1693–1703. 5. Kim DW, Ahn MJ, Shi Y et al. Results of a global phase II study with crizotinib in ALK-positive non-small cell lung cancer NSCLC. J Clin Oncol 2012; 30(Suppl): abstr 7533. 6. Shaw AT, Kim DY, Nagakawa K et al. Crizotinib versus chemotherapy in advanced ALK-positive lung cancer. N Engl J Med 2013; 368: 2385–2394. 7. Riely GJ, Kris MG, Zhao B et al. Prospective assessment of discontinuation and reinitiation of erlotinib or gefitinib in patients with acquired resistance to erlotinib or gefitinib followed by the addition of everolimus. Clin Cancer Res 2007; 13: 5150–5155. 8. Nishie K, Kawaguchi T, Tamiya A et al. Epidermal growth factor receptor tyrosine kinase inhibitors beyond progressive disease: a retrospective analysis for Japanese patients with activating EGFR mutations. J Thorac Oncol 2012; 7: 1722–1727. 9. Chaft JE, Oxnard GR, Sima CS et al. Disease flare after tyrosine kinase inhibitor discontinuation in patients with EGFR-mutant lung cancer and acquired resistance to erlotinib or gefitinib: implications for clinical trial design. Clin Cancer Res 2011; 17: 6298–6303. 10. Pop O, Pirvu A, Toffart AC et al. Disease flare after treatment discontinuation in a patient with EML4-ALK lung cancer and acquired resistance to crizotinib. J Thorac Oncol 2012; 7: e1–e2. 11. Camidge DR, Bang YJ, Kwak EL et al. Activity and safety of crizotinib in patients with ALK-positive non-small-cell lung cancer: updated results from a phase 1 study. Lancet Oncol 2012; 13: 1011–1019. 12. von Minckwitz G, du Bois A, Schmidt M et al. Trastuzumab beyond progression in human epidermal growth factor receptor 2-positive advanced breast cancer: a
German Breast Group 26/Breast International Group 03-05 study. J Clin Oncol 2009; 27: 1999–2006. von Minckwitz G, Schwedler K, Schmidt M et al. Trastuzumab beyond progression: overall survival analysis of the GBG 26/BIG 3-05 phase III study in HER2-positive breast cancer. Eur J Cancer 2011; 47: 2273–2281. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology; Non-Small Cell Lung Cancer, v.2. 2013; June 2013, date last accessed. Heuckmann JM, Balke-Want H, Malchers F et al. Differential protein stability and ALK inhibitor sensitivity of EML4-ALK fusion variants. Clin Cancer Res 2012; 18: 4682–4690. Choi YL, Soda M, Yamashita Y et al. EML4-ALK mutations in lung cancer that confer resistance to ALK inhibitors. N Engl J Med 2010; 363: 1734–1739. Katayama R, Shaw AT, Khan TM et al. Mechanisms of acquired crizotinib resistance in ALK-rearranged lung cancers. Sci Transl Med 2012; 4: 120ra17. Sasaki T, Koivunen J, Ogino A et al. A novel ALK secondary mutation and EGFR signaling cause resistance to ALK kinase inhibitors. Cancer Res 2011; 71: 6051–6060. Doebele RC, Pilling AB, Aisner DL et al. Mechanisms of resistance to crizotinib in patients with ALK gene rearranged non-small cell lung cancer. Clin Cancer Res 2012; 18: 1472–1482. Weickhardt AJ, Scheier B, Burke JM et al. Local ablative therapy of oligoprogressive disease prolongs disease control by tyrosine kinase inhibitors in oncogene-addicted non-small-cell lung cancer. J Thorac Oncol 2012; 7: 1807–1814. Browning ET, Weickhardt AJ, Camidge DR. Response to crizotinib rechallenge after initial progression and intervening chemotherapy in ALK lung cancer. J Thorac Oncol 2013; 8: e21. Sang J, Acquaviva J, Friedland JC et al. Targeted inhibition of the molecular chaperone Hsp90 overcomes ALK inhibitor resistance in non-small cell lung cancer. Cancer Discov 2013; 3: 430–443. Camidge DR, Bazhenova L, Salgia R et al. First-in-human dose-finding study of the ALK/EGFR inhibitor AP26113 in patients with advanced malignancies: updated results. J Clin Oncol 2013; 31(Suppl): abstr 8031. Shaw AT, Mehra R, Kim DW et al. Clinical activity of the ALK inhibitor LDK378 in advanced, ALK-positive NSCLC. J Clin Oncol 2013; 31(Suppl): abstr 8010. Ou S, Gadgeel S, Chiappori A et al. Safety and efficacy analysis of RO5424802/ CH5424802 in anaplastic lymphoma kinase (ALK)-positive non-small cell lung cancer (NSCLC) patients who have failed crizotinib in a dose-finding phase I study (AF-002JG, NCT01588028). Eur J Cancer 2013; 49(Suppl 2): LBA44.
Annals of Oncology
16. Motzer RJ, Escudier B, Oudard S et al. Phase 3 trial of everolimus for metastatic renal cell carcinoma: final results and analysis of prognostic factors. Cancer 2010; 116: 4256–4265. 17. Yao JC, Shah MH, Ito T et al. Everolimus for advanced pancreatic neuroendocrine tumors. N Engl J Med 2011; 364: 514–523. 18. Ferte C, Paci A, Zizi M et al. Natural history, management and pharmacokinetics of everolimus-induced-oral ulcers: insights into compliance issues. Eur J Cancer 2011; 47: 2249–2255. 19. Soria JC, Massard C, Izzedine H. From theoretical synergy to clinical supra-additive toxicity. J Clin Oncol 2009; 27: 1359–1361. 20. Negrier S, Gravis G, Perol D et al. Temsirolimus and bevacizumab, or sunitinib, or interferon alfa and bevacizumab for patients with advanced renal cell carcinoma (TORAVA): a randomised phase 2 trial. Lancet Oncol 2011; 12: 673–680. 21. Baselga J, Campone M, Piccart M et al. Everolimus in postmenopausal hormone receptor-positive advanced breast cancer. N Engl J Med 2012; 366: 520–529. 22. Morrow PK, Wulf GM, Ensor J et al. Phase I/II study of trastuzumab in combination with everolimus (RAD001) in patients with HER2-overexpressing metastatic breast cancer who progressed on trastuzumab-based therapy. J Clin Oncol 2011; 29: 3126–3132.
Annals of Oncology 25: 415–422, 2014 doi:10.1093/annonc/mdt572
Clinical benefit of continuing ALK inhibition with crizotinib beyond initial disease progression in patients with advanced ALK-positive NSCLC S.-H. I. Ou1*, †, P. A. Jänne2,†, C. H. Bartlett3, Y. Tang4, D.-W. Kim5, G. A. Otterson6, L. Crinò7, P. Selaru4, D. P. Cohen4, J. W. Clark8 & G. J. Riely9 1 Chao Family Comprehensive Cancer Center, University of California at Irvine, Irvine; 2Lowe Center for Thoracic Oncology and the Belfer Institute for Applied Cancer Science, Dana-Farber Cancer Institute, Harvard Medical School, Boston; 3Pfizer Oncology, New York; 4Pfizer Oncology, La Jolla, USA; 5Department of Internal Medicine, Seoul National University Hospital, Seoul, Korea; 6Department of Internal Medicine, The Ohio State University Comprehensive Cancer Center, Columbus, USA; 7 Department of Oncology, University Hospital of Perugia, Perugia, Italy; 8Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston; 9 Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, USA
Received 25 September 2013; revised 22 November 2013; accepted 27 November 2013
Background: Crizotinib is approved to treat advanced ALK-positive non-small-cell lung cancer (NSCLC), but most patients ultimately develop progressive disease (PD). We investigated whether continuing ALK inhibition with crizotinib beyond PD (CBPD) is clinically beneficial and attempted to identify clinicopathologic characteristics associated with patients who experience clinical benefit. Patients and methods: Patients with advanced ALK-positive NSCLC enrolled in two ongoing multicenter, single-arm trials who developed RECIST-defined PD were allowed to continue crizotinib if they were deriving ongoing clinical benefit. In the present retrospective analysis, continuation of CBPD was defined as >3 weeks of crizotinib treatment after PD documentation. Patients who had PD as best response to initial crizotinib treatment were excluded. Baseline and postprogression characteristics, sites of PD, and overall survival (OS) were compared in patients who continued CBPD versus those who did not. The impact of continuing CBPD on OS after adjusting for potential confounding factors was assessed. Results: Among 194 crizotinib-treated patients with RECIST-defined PD, 120 (62%) continued CBPD. A significantly higher proportion of patients who continued CBPD than patients who did not had an ECOG performance status (PS) of 0/1 at PD (96% versus 82%; P = 0.02). CBPD patients had significantly longer OS from the time of PD [median 16.4 versus 3.9 *Correspondence to: Dr Sai-Hong I. Ou, Chao Family Comprehensive Cancer Center, University of California Irvine School of Medicine, 101 City Drive, Bldg 56, RT81, Orange, CA 92868, USA. Tel: +1-714-456-8104; Fax: +1-714-456-2242; E-mail: [email protected] †
Both authors contributed equally.
© The Author 2014. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
10. Soria JC, Shepherd FA, Douillard JY et al. Efficacy of everolimus (RAD001) in patients with advanced NSCLC previously treated with chemotherapy alone or with chemotherapy and EGFR inhibitors. Ann Oncol 2009; 20: 1674–1681. 11. Buck E, Eyzaguirre A, Brown E et al. Rapamycin synergizes with the epidermal growth factor receptor inhibitor erlotinib in non-small-cell lung, pancreatic, colon, and breast tumors. Mol Cancer Ther 2006; 5: 2676–2684. 12. Johnson BE, Jackman D, Janne PA. Rationale for a phase I trial of erlotinib and the mammalian target of rapamycin inhibitor everolimus (RAD001) for patients with relapsed non small cell lung cancer. Clin Cancer Res 2007; 13: s4628–s4631. 13. Papadimtrakopoulou VA, Soria JC, Jappe A et al. Everolimus and erlotinib as second- or third-line therapy in patients with advanced non-small cell lung cancer. J Thorac Oncol 2012; 7: 1594–1601. 14. Price KA, Azzoli CG, Krug LM et al. Phase II trial of gefitinib and everolimus in advanced non-small cell lung cancer. J Thorac Oncol 2010; 5: 1623–1629. 15. Milton DT, Riely GJ, Azzoli CG et al. Phase 1 trial of everolimus and gefitinib in patients with advanced nonsmall-cell lung cancer. Cancer 2007; 110: 599–605.
original articles
Annals of Oncology
months; hazards ratio (HR) 0.27, 95% confidence interval (CI): 0.17–0.42; P < 0.0001] and from the time of initial crizotinib treatment (median 29.6 versus 10.8 months; HR 0.30, 95% CI: 0.19–0.46; P < 0.0001). The multiple-covariate Cox regression analysis revealed that CBPD remained significantly associated with improved OS after adjusting for relevant factors. Conclusions: Patients who continued CBPD were more likely to have good ECOG PS (0/1) at the time of PD. Continuing ALK inhibition with crizotinib after PD may provide survival benefit to patients with advanced ALK-positive NSCLC. Key words: treatment beyond disease progression, crizotinib, TKI, RECIST, oncogene addiction, ALK-positive NSCLC
introduction
patients and methods patients and procedures Patients with advanced ALK-positive NSCLC enrolled in either PROFILE 1001 expansion cohort or PROFILE 1005 (and treated by 1 February 2011
| Ou et al.
survival analysis The Kaplan−Meier method was used to estimate all time to tumor progression (TTP) and overall survival (OS) endpoints. Two-sided 95% confidence intervals (CIs) were determined using the Brookmeyer–Crowley method. The Cox proportional hazards models with multiple covariates were used to assess the impact of continuing CBPD on OS from the first dose of crizotinib and from the time of PD with adjustment for various potential confounding factors, retaining statistically significant factors after model selection (factors analyzed are listed in supplementary Data, available at Annals of Oncology online, along with additional details about the methodology). The hazard ratios (HRs) between the two groups were estimated using the Cox proportional hazard models with 95% Wald CIs. All analyses were carried out using SAS statistical software, version 9.2 (SAS Institute Inc., Cary, NC).
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Anaplastic lymphoma kinase (ALK)-positive non-small-cell lung cancer (NSCLC) is a heterogeneous molecular subtype of NSCLC with at least 27 ALK-fusion variants identified [1–3]. ALK kinase activity in ALK-positive NSCLC is aberrantly activated through dimerization of various fusion partners (EML4, KIF5B, KLC1, or TFG) in the cytoplasm. Crizotinib is a multitargeted small-molecule ALK inhibitor approved for the treatment of advanced ALK-positive NSCLC [4, 5]. In a randomized phase III trial, crizotinib demonstrated significantly improved progression-free survival (PFS) when compared with standardof-care chemotherapy for previously treated advanced ALKpositive NSCLC [6]. Most patients with advanced ALK-positive NSCLC treated with crizotinib eventually develop progressive disease (PD), as observed in patients with epidermal growth factor receptor (EGFR)-mutation-positive (EGFRm) NSCLC treated with EGFR tyrosine kinase inhibitors (TKIs). In patients with EGFRm NSCLC, EGFR TKI therapy is sometimes continued beyond the conventional radiographic definition of PD per Response Evaluation Criteria in Solid Tumors (RECIST) [7, 8]. Additionally, patients with EGFRm NSCLC may develop a disease flare manifested as either rapid radiographic PD or significant worsening of symptoms, including death, upon discontinuation of EGFR TKI therapy, indicating that EGFRm tumors continue to depend on EGFR signaling [9]. Disease flare following discontinuation of crizotinib in ALK-positive NSCLC has also been reported [10]. However, whether continuation of ALK inhibition with crizotinib beyond PD (CBPD) would also be beneficial to patients with ALK-positive NSCLC treated with crizotinib remains unknown. Two ongoing single-arm studies of crizotinib in advanced ALK-positive NSCLC, the molecularly enriched expansion cohort of a phase I trial (PROFILE 1001) and a phase II trial (PROFILE 1005), allowed the continuation of crizotinib beyond RECIST-defined PD in patients who continued to derive clinical benefit from crizotinib. Camidge et al. [11] have described the feasibility and tolerability of continuing CBPD in 39 patients with advanced ALK-positive NSCLC. In this report, we carried out a retrospective analysis to investigate whether or not continuation of ALK inhibition with crizotinib is clinically beneficial and whether there are any clinicopathologic characteristics of patients with ALK-positive NSCLC who could potentially benefit from this approach.
in the latter study) were included in this analysis (supplementary Figure S1, available at Annals of Oncology online). ALK rearrangement was detected by break-apart FISH [4]. Key inclusion and exclusion criteria for PROFILE 1001 have been previously described [4, 11]. Key inclusion and exclusion criteria for PROFILE 1005 were similar to those of PROFILE 1001 [5]. Major differences included treatment setting: patients in PROFILE 1005 must have failed at least one line of treatment, while patients in PROFILE 1001 could have been treatment-naïve. Disease assessment was carried out every 6 (PROFILE 1005) or 8 weeks (PROFILE 1001). Response confirmation was required in both studies. PROFILE 1005, but not PROFILE 1001, required baseline brain imaging, and if brain metastases were detected, follow-up brain imaging was required. Otherwise, imaging to assess brain metastases was carried out as clinically indicated in both trials. In this analysis, PD was based on overall investigator assessment of tumor data and RECIST-defined, unless specifically noted. All patients received crizotinib at a starting dose of 250 mg twice daily with appropriate dosing modification as needed. Data collected at protocol-specified intervals that were the nearest, but prior, to the time of PD included performance status (PS), adverse events (AEs), and clinical laboratory tests. For CBPD patients, the same schedule of activities as that used before PD was followed for collection of AE data. A period of 3 weeks of treatment after first documentation of PD was chosen as the cut-off for defining continuing or not continuing CBPD (>3 versus ≤3 weeks, respectively) because it was within one treatment cycle for the majority of the patients, the time frame within which investigators generally decide whether or not to continue CBPD. Patients who had received crizotinib as first-line treatment in PROFILE 1001 were excluded from this analysis to ensure that comparable populations of patients from the two trials were analyzed. Patients from either study whose best overall response to initial crizotinib treatment was PD were also excluded from this analysis since inclusion of these patients could confound the comparison of post-PD survival in patients who did or did not continue CBPD. These patients are highly unlikely to benefit from continuation of CBPD and are more likely to fall into the group that did not continue CBPD, thereby biasing the analysis. All patients provided written informed consent before initial study entry and Institutional Review Board (IRB) approved the protocols. The data cut-off date for this analysis was 2 January 2012.
original articles
Annals of Oncology
results patient characteristics at baseline and at the time of PD
responses to initial crizotinib treatment Maximum percentage reductions from baseline in target lesions among both groups of patients while on crizotinib treatment are shown in supplementary Figure S2, available at Annals of Oncology online. In all, 97% of patients in both groups had some tumor shrinkage. CBPD patients had a numerically higher objective response rate to initial crizotinib treatment (74%, 95% CI 65–82) than did patients who did not continue CBPD (55%, 95% CI 43–67), although this difference was not significant, as the 95% CIs overlapped (Table 2). Among patients who achieved RECIST-defined complete response (CR) or partial response (PR) to initial crizotinib treatment, 68% (89/130) continued CBPD. Among patients who achieved SD as best response to initial crizotinib treatment, 48% (31/64) continued CBPD. Among patients who maintained SD for ≥6 months on initial crizotinib treatment, 70% (16/23) continued CBPD. The median TTP from initial crizotinib treatment for the entire patient group was 6.9 months (95% CI 6.2–7.9). The median TTP from initial crizotinib treatment was numerically longer among patients who continued CBPD (7.3 months, 95% CI 6.8–8.9) than among those who did not continue CBPD (5.7
Volume 25 | No. 2 | February 2014
treatment duration beyond PD The median duration of crizotinib treatment post-PD (total number of active dosing days from first PD to last dose) among the 120 CBPD patients in this analysis was 19.4 weeks (95% CI 16.7–28.9). The duration of CBPD was comparable between patients who received crizotinib as second-line treatment (n = 27; 20.7 weeks, 95% CI 11.9–29.6) and those who received crizotinib as third- or later-line treatment (n = 93; 19.4 weeks, 95% CI 16.0–32.0). Among the 74 patients who did not continue CBPD, 37 (50%) received subsequent systemic treatment, and the remaining 37 (50%) received best supportive care only.
survival analysis The median OS for the overall population from the time of PD was 8.9 months (95% CI 7.2–16.4; Table 2). The median OS from the time of PD was significantly longer for CBPD patients than for patients who did not continue CBPD [16.4 months (95% CI 14.5–not reached) versus 3.9 months (95% CI 2.7–5.1); HR 0.27, 95% CI 0.17–0.42; P < 0.0001; Figure 1A]. Six- and 12month survival probabilities after PD for CBPD patients were 76.3% (95% CI 66.7–83.5) and 64.7% (95% CI 53.0–74.3), respectively, compared with 6- and 12-month survival probabilities of 31.2% (95% CI 20.0–43.0) and 23.9% (95% CI 13.3– 36.1), respectively, for patients who did not continue CBPD. A possible reason for the survival disadvantage seen among patients who did not continue CBPD was that a subset of these patients never received further anticancer therapy post-PD. To address this, we evaluated the impact of receiving or not receiving subsequent systemic anticancer treatment in this group. The median OS post-PD among patients who did not continue crizotinib but received subsequent systemic therapy (n = 37; 5.4 months, 95% CI 3.8–12.3) was longer than that of those who did not continue crizotinib and received no subsequent therapy (n = 37; 2.2 months, 95% CI 1.1–3.8; Figure 1B). Nevertheless, the median OS of patients who did not continue crizotinib but received subsequent systemic therapy remained significantly shorter than that of CBPD patients (5.4 versus 16.4 months; HR 0.38, 95% CI 0.22–0.66; P = 0.0005). The median OS for the overall population (n = 194) was 21.9 months (95% CI 16.6–not reached) from the start of initial crizotinib treatment (Table 2). The median OS from the start of initial crizotinib treatment was significantly longer for CBPD patients than for patients who did not continue crizotinib [29.6 months (95% CI: 23.1–not reached) versus 10.8 months (95% CI: 8.9–14.7); HR 0.30, 95% CI: 0.19–0.46; P < 0.0001; Figure 1C]. After adjusting for potential confounding factors (supplementary Data, available at Annals of Oncology online), continuing CBPD remained significantly associated with improved OS post-PD (HR 0.27; P < 0.0001; Table 3). Both response (CR/PR) to initial crizotinib treatment (HR 0.48; P = 0.01) and longer TTP (HR 0.78; P = 0.01) were also found to be factors associated with improved OS (Table 3). Eight patients (all did not continue CBPD) died ≤3 weeks after PD documentation. After adjusting for potential confounding factors and analyzing the 186 patients
doi:10.1093/annonc/mdt572 |
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
A total of 414 patients with advanced ALK-positive NSCLC who had been enrolled in PROFILE 1001 (expansion cohort) or PROFILE 1005 (and treated by 1 February 2011) were included in this analysis (supplementary Figure S1, available at Annals of Oncology online). A total of 194 patients who had developed PD by the time of data cut-off constituted the final study population. One hundred and twenty patients (62%) continued crizotinib for >3 weeks post-PD, while 74 patients (38%) did not continue crizotinib beyond 3 weeks post-PD at the discretion of the investigators. The median duration of follow-up post-PD for the 120 patients who continued CBPD was 25.2 weeks (range 3.7– 115.9). Baseline and post-progression characteristics of these patients are provided in Table 1. The groups were comparable in age at baseline, proportions of patients from the two trials, smoking history, histology, and prior treatment history. The only significant difference between the two groups was the PS at the time of PD (Table 1). Among patients with Eastern Cooperative Oncology Group PS (ECOG PS) of 0 or 1 at the time of PD, 65% (115/176) continued CBPD. On the other hand, among patients with ECOG PS of ≥2 at the time of PD, only 29% (4/14) continued CBPD. We carried out an analysis of the sites of PD to investigate whether or not the brain may be a common site of PD in patients who continued CBPD. In the absence of standard minimum tumor growth criteria defining individual sites of PD when there are multiple sites of progressing target lesions, we analyzed the subset of patients who had PD involving new lesions and/or non-target lesions (N = 138). Among these patients, half of the patients (51%) who continued CBPD had brain as a site of PD, while patients who did not continue CBPD had both liver (37%) and brain (28%) as a site of PD (supplementary Table S1, available at Annals of Oncology online).
months, 95% CI 5.4–6.8), although the 95% CIs overlapped (Table 2).
original articles
Annals of Oncology
Table 1. Baseline and post-progression characteristics of patients who continued crizotinib beyond progressive disease and those who did not continue crizotinib beyond progressive disease Characteristic
n (%) All patients (n = 194)
Continued CBPD (n = 120)
Did not continue CBPD (n = 74)
50 (50) 21–78
52 (54) 28–78
0.30b
103 (86) 17 (14)
68 (92) 6 (8)
0.20c
65 (54) 55 (46)
35 (47) 39 (53)
0.35c
76 (63) 39 (32) 3 (3) 2 (2)
34 (46) 34 (46) 5 (7) 1 (1)
0.06d
48 (40) 72 (60)
23 (31) 51 (69)
0.21c
83 (69) 35 (29) 2 (2)
53 (72) 19 (26) 2 (3)
0.73d
33 (28) 70 (58) 16 (13) 1 (1)
14 (19) 48 (65) 11 (15) 1 (1)
0.53d
37 (31) 78 (65) 2 (2) 2 (2) 1 (1)
18 (24) 43 (58) 8 (11) 2 (3) 3 (4)
0.02d
113 (94) 7 (6)
71 (96) 3 (4)
0.90c
27 (22) 93 (78)
15 (20) 59 (80)
0.71c
115 (96) 5 (4)
72 (97) 2 (3)
0.71d
Percentages may not total 100% due to rounding. Continued CBPD versus did not continue CBPD. b t-test of means. c 2 χ test of general association. d Fisher’s exact test of general association when sample size requirement for χ 2 test is not met. e Last ECOG PS determined before disease progression. f Includes large-cell carcinoma, squamous-cell carcinoma, and adenosquamous carcinoma. g Includes neoadjuvant, adjuvant, and metastatic use. CBPD, crizotinib beyond progressive disease; ECOG PS, Eastern Cooperative Oncology Group performance status. a
| Ou et al.
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Age (years) Mean (median) 51 (51) Range 21–78 Age category <65 years 171 (88) ≥65 years 23 (12) Sex Male 100 (52) Female 94 (48) Ethnicity White 110 (57) Asian 73 (38) African-American 8 (4) Other 3 (2) Clinical trial PROFILE 1001 71 (37) PROFILE 1005 123 (63) Smoking history Never-smoker 136 (70) Former smoker 54 (28) Current smoker 4 (2) ECOG PS at baseline 0 47 (24) 1 118 (61) 2 27 (14) 3 2 (1) ECOG PS at disease progressione 0 55 (28) 1 121 (62) 2 10 (5) ≥3 4 (2) Missing 4 (2) Histology Adenocarcinoma 184 (95) Otherf 10 (5) Prior lines of therapy for advanced or metastatic disease 1 42 (22) ≥2 152 (78) Prior platinum-based therapyg Yes 187 (96) No 7 (4)
P-valuea
original articles
Annals of Oncology
Table 2. Clinical outcomes in patients who continued CBPD versus those who did not continue CBPD Outcome
Continued CBPD (n = 120)
Did not continue CBPD (n = 74)
130 (67) 60–74 64 (33)
89 (74) 65–82 31 (26)
41 (55) 43–67 33 (45)
7±7 5–48 92 (71) 24 (18) 9 (7) 5 (4)
7±8 5–48 61 (69) 19 (21) 5 (6) 4 (4)
7±5 5–24 31 (76) 5 (12) 4 (10) 1 (2)
7 (11) 34 (53) 13 (20) 7 (11) 3 (5)
2 (6) 13 (42) 7 (23) 7 (23) 2 (6)
5 (15) 21 (64) 6 (18) 0 (0) 1 (3)
6.9 6.2–7.9 59.0 51.7–65.6 83 (43)
7.3 6.8–8.9 65.6 56.3–73.3 36 (30)
5.7 5.4–6.8 48.5 36.7–59.3 47 (64)
8.9 7.2–16.4 59.4 51.3–66.5 49.3 40.5–57.5
16.4 14.5–NR 76.3 66.7–83.5 64.7 53.0–74.3
3.9 2.7–5.1 31.2 20.0–43.0 23.9 13.3–36.1
21.9 16.6–NR 93.8 89.4–96.4 68.5 61.3–74.6
29.6 23.1–NR 97.5 92.5–99.2 81.7 73.2–87.7
10.8 8.9–14.7 87.8 77.9–93.5 46.6 34.7–57.8
a
Derived from investigator-reported tumor data in order to more accurately characterize best response. Among patients who had PR or CR on initial crizotinib treatment. c Time from first dose to disease progression among patients who had SD on initial crizotinib treatment. CBPD, crizotinib beyond progressive disease; CR, complete response; PR, partial response; ORR, objective response rate; CI, confidence interval; SD, stable disease; TTP, time to tumor progression; OS, overall survival; NR, not reached. b
who survived beyond the 3 weeks defining CBPD, continuing CBPD remained significantly associated with improved OS post-PD.
discussion Molecularly targeted therapy is increasingly being continued beyond PD, especially in cancers harboring an underlying oncogenic dependence on the molecular target, such as trastuzumab in HER2-positive breast cancer [12, 13] and EGFR TKIs in
Volume 25 | No. 2 | February 2014
EGFRm NSCLC [14]. In the present retrospective analysis, we showed that patients with advanced ALK-positive NSCLC derived clinical benefit from continued ALK inhibition with crizotinib after RECIST-defined PD. These patients were more likely to have good PS at the time of PD, had responded to and exhibited extended TTP from initial crizotinib treatment, and had a site of PD particularly amenable to local therapy (brain). Not surprisingly, these patients also had a better prognosis as demonstrated by their longer OS from the start of initial crizotinib treatment (Figure 1C). While it may seem intuitive that
doi:10.1093/annonc/mdt572 |
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Best response to initial crizotinib treatmenta CR/PR (ORR), n (%) 95% CI SD, n (%) Time to first response (weeks)b Median ± standard deviation Range 0 to <8 weeks, n (%) 8 to <16 weeks, n (%) 16 to <24 weeks, n (%) ≥24 weeks, n (%) Duration of SD,c n (%) 0 to <3 months 3 to <6 months 6 to <9 months 9 to <12 months ≥12 months TTP (months) Median 95% CI 6-month TTP probability (%) 95% CI Deaths, n (%) OS from the time of PD Median (months) 95% CI 6-month OS probability (%) 95% CI 12-month OS probability (%) 95% CI OS from first dose of crizotinib Median (months) 95% CI 6-month OS probability (%) 95% CI 12-month OS probability (%) 95% CI
All patients (n = 194)
original articles A
Annals of Oncology
100
Continued crizotinib Did not continue crizotinib
Probability of survival (%)
80
60
40
20
0
0
5
10
15
20
25
30
Time (months)
B
29 7
4 0
100
0
Continued crizotinib Did not continue crizotinib, + ST Did not continue crizotinib, – ST
Probability of survival (%)
80
60
40
20
0
0
10
5
15
30
25
20
Time (months) Number at risk Continued 120 Did not continue, + ST 37 Did not continue, – ST 37
C
29 5 2
4 0 0
Probability of survival (%)
100
0
Continued crizotinib Did not continue crizotinib
80
60
40
20
0 0
5
10
15
20
25
30
35
40
Time (months) Number at risk Continued 120 Did not continue 74
104 40
30 8
6 0
1
Figure 1. Overall survival of patients who continued crizotinib versus those who did not continue crizotinib beyond progressive disease. (A) Overall survival from the time of progressive disease. (B) Overall survival from the time of progressive disease of patients who continued crizotinib beyond progressive disease versus those who did not and who either received or did not receive subsequent systemic therapy (ST). (C) Overall survival from the time of initial crizotinib treatment. Shaded areas are 95% Hall–Wellner confidence bands.
| Ou et al.
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
Number at risk Continued 120 Did not continue 74
original articles
Annals of Oncology
Table 3. The Cox proportional hazards analysis of overall survival to assess the impact of continuing crizotinib beyond progressive disease while adjusting for other patient/disease characteristics or outcomes using backward model selectiona
Crizotinib beyond PD (yes versus no) Sex (female versus male) Smoking status(never versus ex/smoker) ECOG PS by the time of PDc 1 versus 0 ≥2 versus 0 Best response (CR/PR versus SD)d TTP (1-month increments)e New lesions only at PDd (yes versus no)
OS from the time of PD (n = 194) HRb 95% CI
P-value
OS from first dose of crizotinib (n = 194) HRb 95% CI P-value
0.27 — 0.54
0.17–0.45 — 0.32–0.89
<0.0001 — 0.02
0.28 0.58 —
0.18–0.44 0.37–0.91 —
<0.0001 0.02 —
1.75 10.94 0.48 0.78 1.87
1.00–3.04 4.92–24.3 0.28–0.85 0.66–0.93 1.06–3.30
0.05 <0.0001 0.01 0.01 0.03
— — — — —
— — — — —
— — — — —
a
patients who derive clinical benefit on initial crizotinib treatment benefit from continued ALK inhibition with crizotinib, our report represents the largest systematic collection of the clinicopathologic data from these patients to date and shows, reassuringly, that continuation of ALK inhibition with CBPD was practiced by many of the investigators in the two trials. ALK-positive NSCLC is a heterogeneous disease with many EML4-ALK fusion variants that likely present a diverse clinical picture and may exhibit differential responses to crizotinib and other ALK inhibitors [15]. Several crizotinib resistance mechanisms have been reported, including generation of secondary resistance mutations, amplification of the ALK gene, and activation of other bypass signaling pathways [16–19]. However, the resistance mechanism is unknown in the majority of the patients whose disease progressed on crizotinib due to low frequency of re-biopsy, especially when the progressing metastatic lesions occur in the brain which are not easily amenable to biopsy. There are limitations to the present retrospective analysis despite efforts to adjust for confounding factors by the Cox regression analysis with multiple covariates and to exclude patients who did not respond at all to initial crizotinib treatment. While there were no gross imbalances in the clinicopathologic characteristics of the two groups of patients analyzed, there may be unknown confounding factors not considered in our analysis. Therefore, the ability to continue CBPD could simply represent a subgroup with favorable prognostic factors rather than a direct post-PD treatment effect. Data on the use of ALK inhibitors other than crizotinib, subsequent disease progression while on continued CBPD, the type of chemotherapy regimen used after PD, and the mode of any radiation therapy delivered to patients with PD on crizotinib were not captured.
Volume 25 | No. 2 | February 2014
As CBPD is not expected to provide clinical benefit indefinitely, other strategies such as the addition of local therapy [20], sequencing chemotherapy [21], or addition of HSP90 inhibitors [22] to CBPD should be evaluated prospectively in this treatment setting. Furthermore, several ALK inhibitors are in clinical development primarily for PD after crizotinib [23–25]. Given the potential survival benefit in ALK-positive NSCLC patients who continue CBPD reported herein, direct comparative studies between these newer ALK inhibitors versus CBPD plus additional treatment modalities would be needed to inform optimal treatment after PD on crizotinib.
acknowledgements We would like to thank all of the patients and their families, as well as the investigators, research nurses, study coordinators, and operations staff who participated in PROFILE 1001 and PROFILE 1005. We would also like to thank Kristen Letrent (Pfizer) for her contributions to the design of the analyses and Keith Wilner (Pfizer) for his lead role in the conduct of both studies.
funding This work was sponsored by Pfizer Inc. Editorial support was provided by Wendy Sacks at ACUMED® (New York, NY, USA) and was funded by Pfizer Inc [no grant numbers].
disclosure S-HIO has served as a consultant for and received honoraria from Pfizer. PAJ has served as a consultant for Pfizer. CHB, YT, PS, and DPC are employees of Pfizer and own Pfizer stock.
doi:10.1093/annonc/mdt572 |
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
A complete list of variables included in this analysis and a summary of the methodology are provided in supplementary Data, available at Annals of Oncology online. b For categorical variables, HR < 1 favors the first category and HR > 1 favors the second category; for continuous variables, for every unit increase in the variable, there is an increase (HR > 1) or decrease (HR < 1) in the risk of death. c Last ECOG PS determined before disease progression. d Derived from investigator-reported tumor data in order to more accurately characterize best response and site/mode of PD. e Continuous variable. OS, overall survival; PD, progressive disease; HR, hazard ratio; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; CR, complete response; PR, partial response; SD, stable disease; TTP, time to tumor progression.
original articles D-WK has served as a consultant for Pfizer, GlaxoSmithKline, and Lilly and has received honoraria from Pfizer and Lilly. GJR’s institution has received research funding from Pfizer for research carried out by GJR.
Annals of Oncology
13.
14.
references
| Ou et al.
15.
16. 17. 18.
19.
20.
21.
22.
23.
24. 25.
Volume 25 | No. 2 | February 2014
Downloaded from http://annonc.oxfordjournals.org/ at Belgorod State University on February 4, 2014
1. Soda M, Choi YL, Enomoto M et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature 2007; 448: 561–566. 2. Rikova K, Guo A, Zeng Q et al. Global survey of phosphotyrosine signaling identifies oncogenic kinases in lung cancer. Cell 2007; 131: 1190–1203. 3. Ou SH, Bartlett CH, Mino-Kenudson M et al. Crizotinib for the treatment of ALKrearranged non-small cell lung cancer: a success story to usher in the second decade of molecular targeted therapy in oncology. Oncologist 2012; 17: 1351–1375. 4. Kwak EL, Bang YJ, Camidge DR et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010; 363: 1693–1703. 5. Kim DW, Ahn MJ, Shi Y et al. Results of a global phase II study with crizotinib in ALK-positive non-small cell lung cancer NSCLC. J Clin Oncol 2012; 30(Suppl): abstr 7533. 6. Shaw AT, Kim DY, Nagakawa K et al. Crizotinib versus chemotherapy in advanced ALK-positive lung cancer. N Engl J Med 2013; 368: 2385–2394. 7. Riely GJ, Kris MG, Zhao B et al. Prospective assessment of discontinuation and reinitiation of erlotinib or gefitinib in patients with acquired resistance to erlotinib or gefitinib followed by the addition of everolimus. Clin Cancer Res 2007; 13: 5150–5155. 8. Nishie K, Kawaguchi T, Tamiya A et al. Epidermal growth factor receptor tyrosine kinase inhibitors beyond progressive disease: a retrospective analysis for Japanese patients with activating EGFR mutations. J Thorac Oncol 2012; 7: 1722–1727. 9. Chaft JE, Oxnard GR, Sima CS et al. Disease flare after tyrosine kinase inhibitor discontinuation in patients with EGFR-mutant lung cancer and acquired resistance to erlotinib or gefitinib: implications for clinical trial design. Clin Cancer Res 2011; 17: 6298–6303. 10. Pop O, Pirvu A, Toffart AC et al. Disease flare after treatment discontinuation in a patient with EML4-ALK lung cancer and acquired resistance to crizotinib. J Thorac Oncol 2012; 7: e1–e2. 11. Camidge DR, Bang YJ, Kwak EL et al. Activity and safety of crizotinib in patients with ALK-positive non-small-cell lung cancer: updated results from a phase 1 study. Lancet Oncol 2012; 13: 1011–1019. 12. von Minckwitz G, du Bois A, Schmidt M et al. Trastuzumab beyond progression in human epidermal growth factor receptor 2-positive advanced breast cancer: a
German Breast Group 26/Breast International Group 03-05 study. J Clin Oncol 2009; 27: 1999–2006. von Minckwitz G, Schwedler K, Schmidt M et al. Trastuzumab beyond progression: overall survival analysis of the GBG 26/BIG 3-05 phase III study in HER2-positive breast cancer. Eur J Cancer 2011; 47: 2273–2281. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology; Non-Small Cell Lung Cancer, v.2. 2013; June 2013, date last accessed. Heuckmann JM, Balke-Want H, Malchers F et al. Differential protein stability and ALK inhibitor sensitivity of EML4-ALK fusion variants. Clin Cancer Res 2012; 18: 4682–4690. Choi YL, Soda M, Yamashita Y et al. EML4-ALK mutations in lung cancer that confer resistance to ALK inhibitors. N Engl J Med 2010; 363: 1734–1739. Katayama R, Shaw AT, Khan TM et al. Mechanisms of acquired crizotinib resistance in ALK-rearranged lung cancers. Sci Transl Med 2012; 4: 120ra17. Sasaki T, Koivunen J, Ogino A et al. A novel ALK secondary mutation and EGFR signaling cause resistance to ALK kinase inhibitors. Cancer Res 2011; 71: 6051–6060. Doebele RC, Pilling AB, Aisner DL et al. Mechanisms of resistance to crizotinib in patients with ALK gene rearranged non-small cell lung cancer. Clin Cancer Res 2012; 18: 1472–1482. Weickhardt AJ, Scheier B, Burke JM et al. Local ablative therapy of oligoprogressive disease prolongs disease control by tyrosine kinase inhibitors in oncogene-addicted non-small-cell lung cancer. J Thorac Oncol 2012; 7: 1807–1814. Browning ET, Weickhardt AJ, Camidge DR. Response to crizotinib rechallenge after initial progression and intervening chemotherapy in ALK lung cancer. J Thorac Oncol 2013; 8: e21. Sang J, Acquaviva J, Friedland JC et al. Targeted inhibition of the molecular chaperone Hsp90 overcomes ALK inhibitor resistance in non-small cell lung cancer. Cancer Discov 2013; 3: 430–443. Camidge DR, Bazhenova L, Salgia R et al. First-in-human dose-finding study of the ALK/EGFR inhibitor AP26113 in patients with advanced malignancies: updated results. J Clin Oncol 2013; 31(Suppl): abstr 8031. Shaw AT, Mehra R, Kim DW et al. Clinical activity of the ALK inhibitor LDK378 in advanced, ALK-positive NSCLC. J Clin Oncol 2013; 31(Suppl): abstr 8010. Ou S, Gadgeel S, Chiappori A et al. Safety and efficacy analysis of RO5424802/ CH5424802 in anaplastic lymphoma kinase (ALK)-positive non-small cell lung cancer (NSCLC) patients who have failed crizotinib in a dose-finding phase I study (AF-002JG, NCT01588028). Eur J Cancer 2013; 49(Suppl 2): LBA44.