Annals of Oncology Advance Access published May 21, 2015

original article

Annals of Oncology 00: 1–6, 2015 doi:10.1093/annonc/mdv208

Biomarker testing and time to treatment decision in patients with advanced nonsmall-cell lung cancer† C. Lim1, M. S. Tsao2, L. W. Le3, F. A. Shepherd1, R. Feld1, R. L. Burkes4, G. Liu1, S. Kamel-Reid2,5, D. Hwang2, J. Tanguay2, G. da Cunha Santos2 & N. B. Leighl1* 1 Division of Medical Oncology, Princess Margaret Cancer Centre, Toronto; 2Laboratory Medicine Program, University Health Network and Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto; 3Department of Biostatistics, Princess Margaret Cancer Centre, Toronto; 4Division of Medical Oncology, Mount Sinai Hospital, Toronto; 5Molecular Diagnostics Laboratory, University Health Network, Toronto, Canada

Received 19 November 2014; revised 30 March 2015; accepted 20 April 2015

introduction The approach to treating advanced nonsmall-cell lung cancer (NSCLC) has evolved rapidly with the increasingly widespread use of molecularly targeted agents. Tyrosine kinase inhibitors (TKIs) targeting the epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) have consistently demonstrated improved clinical outcomes, including response, quality of life, favorable toxicity profile and progression-free survival compared with conventional chemotherapy for selected

*Correspondence to: Dr Natasha B. Leighl, Division of Medical Oncology, Princess Margaret Cancer Centre, University Health Network, 5-105, 610 University Avenue, Toronto, ON, Canada M5G 2M9. Tel: +1-416-946-4645; E-mail: [email protected]

Study previously presented at: American Society of Clinical Oncology 2014 Annual Scientific Meeting, Canadian Association of Medical Oncologists 2014 Annual Scientific Meeting.

patient populations [1–4]. A pooled analysis of two EGFR TKI trials has even demonstrated improved overall survival [5]. Given the evidence supporting first-line use of targeted agents, clinical guidelines have been developed with recommendations regarding biomarker testing in NSCLC [6–8]. Routine testing for EGFR mutation and ALK rearrangement in advanced nonsquamous NSCLC now has become a global standard in the management of advanced NSCLC. However, approaches to address this requirement have varied by country and region. The addition of biomarker testing adds to the complexity of lung cancer diagnostic algorithms and can affect the timeliness of treatment decisions, particularly when biopsies yield insufficient tissue for analysis. Given that first-line molecular targeted therapies yield better outcomes and less toxicity than chemotherapy, lack of biomarker results at the time of initial oncology consultation can result in missed opportunities for patient benefit. European cancer institutes and national health systems have

© The Author 2015. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]

original article

Downloaded from http://annonc.oxfordjournals.org/ at CSU Fresno on July 9, 2015

Background: Testing for EGFR mutations and ALK rearrangement has become standard in managing advanced nonsmall-cell lung cancer (NSCLC). However, many institutions in Europe, North America and other world regions continue to face a common challenge of facilitating timely molecular testing with rapid result turnaround time. We assessed the prevalence of biomarker testing for advanced NSCLC patients and whether testing affected the timeliness of treatment decisions. Methods: We conducted a retrospective chart review of a random sample of one-quarter of all patients with advanced NSCLC referred to the Princess Margaret Cancer Centre from 1 April 2010 to 31 March 2013. Results: Of 300 patients reviewed, 175 seen by medical oncology had nonsquamous NSCLC, 72% of whom had biomarker testing carried out. Patients tested for biomarkers were more likely to be female (47% versus 21%, P = 0.002), Asian (27% versus 6%, P = 0.005) and never smokers (42% versus 8%, P < 0.0001). Only 21% of patients with biomarker testing had results available at their initial oncology consultation. This group had a shorter median time from consultation to treatment decision (0 versus 22 days, P = 0.0008) and time to treatment start (16 versus 29, P = 0.004). Thirteen percent underwent repeat biopsy for molecular testing after the initial consultation. Of those with positive EGFR or ALK results, 19% started chemotherapy before biomarker results became available. Conclusions: Awaiting biomarker testing results can delay treatment decisions and treatment initiation for patients with advanced NSCLC. This may be avoided by incorporating reflex biomarker testing into diagnostic algorithms for NSCLC at the level of the pathologist, and further education of specialists involved in obtaining diagnostic cancer specimens to ensure they are sufficient for molecular testing. Key words: NSCLC, biomarkers, quality of care, molecular testing, EGFR, ALK

original article pioneered molecular testing networks to improve patient access to timely molecular analysis on a nationwide basis [9]. While this initiative may serve as a blueprint for other health jurisdictions, health care infrastructure to support molecular testing is still being developed. Many institutions around the world continue to face the challenge of facilitating timely molecular testing and results to avoid delays in starting first-line treatment. At our Canadian institution, EGFR testing became available in March 2010 and ALK testing in April 2012. For surgical resection specimens, reflex biomarker testing at the time of initial diagnosis of nonsquamous lung cancer was introduced in 2011. Testing of cytopathology specimens was initiated upon clinician request. We assessed the prevalence and timing of biomarker testing for advanced NSCLC patients and whether testing affected the timeliness of treatment decisions.

methods A retrospective chart review was conducted of patients with advanced NSCLC referred to the Princess Margaret Cancer Centre/University Health Network (UHN) from 1 April 2010 to 31 March 2013. A random sample of 300 patients with advanced NSCLC from this time period was selected for analysis using a random numbers generator, representing one-fourth of all advanced NSCLC patient referrals to the center during the study period. UHN is the lead regional testing center for molecular diagnostics in lung cancer. EGFR testing was carried out using polymerase chain reaction fragment analysis for exon 19 deletions and Sau961 restriction enzyme digestion for the L858R mutation in exon 21. Initial ALK testing was carried out using immunohistochemistry (IHC) with confirmatory fluorescence in situ hybridization (FISH) analysis if positive staining was detected [10]. The institutional research ethics board approved this study. Patients were included if they had biopsy-proven stage IV NSCLC at initial assessment between 1 April 2010 and 31 March 2013. Patients were excluded if they were not assessed by an oncologist at the institution.

outcomes The primary outcome for this study was to determine the proportion of patients, with a focus on those with nonsquamous subtype, that had biomarker testing carried out and results available at the time of initial consultation with an institutional medical oncologist. We summarized biomarker testing prevalence over time, test outcomes and additional outcomes including assessing the time interval between initial oncologist assessment and when biomarker test results became available. We examined whether this time interval affected the timeliness of starting treatment. Among patients with EGFR mutant or ALK rearranged lung cancer, we examined the timing and appropriateness of treatment decisions.

data collection Data were abstracted from patient electronic medical records at the Princess Margaret Cancer Centre, Princess Margaret Cancer Registry and UHN molecular laboratory information system (CL, NL). Demographics including age, sex, cancer stage, ethnicity, Eastern Co-operative Oncology Group (ECOG) performance status and smoking status were collected. Tumor pathology, biopsy method, tissue sample adequacy and biomarker results were determined by reviewing pathology reports. Detailed timeline information including time points for biopsy, biomarker status determination, initial oncologist consultation and treatment decision were abstracted from clinical notes, registry data and the molecular laboratory information system.

 | Lim et al.

data analysis Descriptive analysis was used to summarize patient demographics, biomarker testing prevalence over time and outcomes. The association between patient demographics and biomarker testing was examined using Fisher’s exact test and the Wilcoxon–Mann–Whitney test. The Kruskal–Wallis test was carried out to assess the impact of biomarker testing result availability at the initial medical oncologist assessment. Statistical analysis was carried out using SAS v9.3 (Cary, NC).

results all patients A random sample of 300 advanced NSCLC patients seen in consultation between April 2010 and March 2013 was evaluated (Figure 1). Among all 300 patients, the median age was 65 years (range 20–95 years), 41% were female, 15% were Asian and 21% were never smokers. There was a trend toward increasing prevalence of biomarker testing over time, rising from 46% in the first year to 60% during the last year of the study period for EGFR testing, and from 5% to 40% for ALK testing (supplementary Figure S1, available at Annals of Oncology online).

nonsquamous patients There were 258 nonsquamous patients seen by medical (n = 175) or radiation (n = 83) oncologists; 58% had biomarker testing carried out. EGFR testing was carried out for 150 patients, of whom 44 (29%) had tumor mutations. Three percent (5/150) of EGFR tests were reported as inconclusive. ALK testing was carried out for 50 patients, 3 (6%) had tumoral ALK rearrangements, and 3 (6%) had insufficient tumor cellularity for ALK analysis by IHC (and FISH). Six patients underwent multiplex biomarker testing in the context of a molecular profiling trial.

patients seen by institutional medical oncologists Of all 300 patients, 202 had a consultation with a medical oncologist at the cancer center (Figure 1). Demographic characteristics of nonsquamous NSCLC patients seen by a medical oncologist (n = 175) are listed in Table 1. Within this group, 72% (126/175) of patients had biomarker testing carried out. Those tested for biomarkers were significantly more likely to be female (47% versus 20%, P = 0.002), Asian (27% versus 6%, P = 0.005) or never smokers (42% versus 8%, P < 0.0001). Patients tested for biomarkers were also more likely to have adenocarcinoma histology (95% versus 71%, P < 0.0001). There were no significant differences in testing patterns by age or ECOG performance score. There were no appreciable changes over time in demographic characteristics of patients with biomarker testing. Of the 98 patients referred only to radiation oncologists at the center, only 24 had molecular testing results available. Further data were not available from outside institutions on subsequent choice and timing of treatment, and these patients were therefore excluded from subsequent analyses.

diagnostic sample adequacy Of 126 nonsquamous patients with biomarker testing carried out at the time of initial oncology consultation, 20 (16%) had inadequate tissue for biomarker analysis (Figure 1). Sixteen (13%)

Downloaded from http://annonc.oxfordjournals.org/ at CSU Fresno on July 9, 2015

study design and population

Annals of Oncology

original article

Annals of Oncology

All patients reviewed (n = 300)

Seen by rad onc only (n = 98)

Seen by PMH med onc (n = 202)

Non-squamous pathology (n = 175)

Adequate tissue for analysis (n = 106)

Biomarkers not tested (n = 49)

Non-squamous (n = 83)

Biomarkers tested (n = 24)

Squamous cell pathology (n = 15)

Biomarkers not tested (n = 58)

Inadequate tissue for analysis (n = 20)

Repeat biopsy performed (n = 16)

No repeat biopsy (n = 4)

Figure 1. Flow diagram of patients studied.

had repeat biopsies carried out after their initial consultation to obtain additional tissue for biomarker analysis. Most of the repeat biopsies (94%) had sufficient tissue for subsequent analysis.

days and to treatment start was 15 days (Table 2). Both the time to treatment decision (P = 0.0008) and time to treatment start (P = 0.004) were significantly shorter in the group with biomarker results available at initial oncology consultation.

turnaround time of biomarker results Only 21% (27/126) of nonsquamous NSCLC patients undergoing biomarker testing had test results available to the oncologist at the time of initial consultation. There was a trend over time toward an increasing proportion of patients with biomarker results available at initial consultation (Figure 2). For this group, biomarker results were reported a median of 25 days before the date of the initial oncology assessment. For patients without biomarker results available at initial oncology assessment, the median time to biomarker result reporting was 21 days after the initial consultation (Table 2). For patients with biomarker test results available at the time of initial oncologist assessment (n = 27), the median time from consultation to treatment decision was 0 days and to treatment start was 16 days. For patients that did not have biomarker results available at the time of initial oncologist assessment (n = 99), the median time from consultation to treatment decision was 22 days and to treatment start was 29 days. Meanwhile, for patients that did not undergo biomarker testing (n = 49), the median time from consultation to treatment decision was 10

biomarker-directed treatment EGFR mutations or ALK rearrangements were detected in 43 patient samples. The majority (74%) started treatment after biomarker results were available. Three did not receive therapy (two declined, one lost to follow-up). The remaining 19% started firstline chemotherapy before biomarker test results becoming available. Six patients had tumor EGFR mutations and two had ALK rearrangements. Once results were reported to the clinician, three patients continued to receive chemotherapy per the original treatment plan, two switched early (poor response to chemotherapy) and another three started TKI upon progression.

discussion In our institution’s experience, we found that despite the lack of reflex testing for most cases, the majority of nonsquamous NSCLC patients did have biomarker testing initiated, and that the prevalence of biomarker testing increased over time. There was selection bias based on clinical factors, with more never

doi:10.1093/annonc/mdv208 | 

Downloaded from http://annonc.oxfordjournals.org/ at CSU Fresno on July 9, 2015

Biomarkers tested (n = 126)

Squamous cell pathology (n = 27)

original article

Annals of Oncology

Table 1. Demographic features of patients with nonsquamous NSCLC seen by a medical oncologist (n = 175) No biomarker tested (n = 49)

Total (N = 175)

P value

66 (36–87)

65 (20–89)

65 (20–89)

0.40

10 (20.4%) 39 (79.6%)

59 (46.8%) 67 (53.2%)

69 (39.4%) 106 (60.6%)

0.002

35 (71.4%) 3 (6.1%) 11 (22.4%)

74 (58.7%) 34 (27.0%) 18 (14.3%)

109 (62.3%) 37 (21.1%) 29 (16.6%)

0.005

4 (8.2%) 27 (55.1%) 18 (36.7%)

50 (41.7%) 50 (41.7%) 20 (16.7%)

54 (32.0%) 77 (45.6%) 38 (22.5%)

Biomarker testing and time to treatment decision in patients with advanced nonsmall-cell lung cancer.

Testing for EGFR mutations and ALK rearrangement has become standard in managing advanced nonsmall-cell lung cancer (NSCLC). However, many institution...
172KB Sizes 2 Downloads 18 Views