Targeted therapy for advanced anaplastic lymphoma kinase (ALK)‐rearranged non‐small cell lung cancer

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To evaluate the safety and efficacy of anaplastic lymphoma kinase (ALK) inhibitors given as monotherapy to treat advanced ALK‐rearranged non‐small cell lung cancer (NSCLC).

Background

Description of the condition

Lung cancer leads to the most cancer deaths worldwide of any tumour type, causing approximately one‐quarter of all cancer deaths annually (Siegel 2019).

Non‐small cell lung cancer (NSCLC) comprises approximately 85% of lung cancers overall (Novello 2016), and is further divided into squamous and non‐squamous histological subtypes, including adenocarcinoma. Typically, NSCLC is diagnosed at an advanced stage with five‐year survival for stage IV cancer estimated at 3% to 6% (Chansky 2017). Platinum doublet chemotherapies have been the standard first‐line chemotherapy and achieve modest improvements with median overall survivals between 7.9 months found in Schiller 2002 and 11.8 months in Scagliotti 2008.

Evolution in understanding lung cancer at the molecular level has led to new and tailored therapies showing significant efficacy in defined subgroups of lung cancer patients (Hirsch 2016). Potentially targetable mutations have been found in up to 64% of lung adenocarcinomas (Kris 2014), however not all of these potentially targetable mutations have clinically validated treatments.

Non‐squamous NSCLC has been further classified according to the presence of driving gene mutations and current international molecular testing guidelines recommend universal testing of patients with advanced adenocarcinoma for epidermal growth factor receptor (EGFR) mutations and rearrangements of the anaplastic lymphoma kinase (ALK) and c‐ros oncogene 1 (ROS1) genes (Lindeman 2018). This is regardless of clinical characteristics, and furthermore, extended panels are recommended in some patients to include detection of KRAS, BRAF, HER2, RET and MET mutations. These driving mutations are considered mutually exclusive, to the extent that sequential testing strategies can be used. Furthermore, as the name implies, targeted therapies are only effective against a tumour where the specific target (or gene mutation) has been found on molecular testing of biopsy material or circulating tumour DNA.

Advanced NSCLC driven by EGFR mutations were the first molecularly defined subgroup where an impact on survival was demonstrated using a targeted therapy approach. First‐line tyrosine kinase inhibitors, including erlotinib and gefitinib, have demonstrated efficacy superior to chemotherapy, but only when tailored to treat cancers driven by sensitising mutations in exons 18‐21 of the EGFR gene (Fukuoka 2011).

In 2007, ALK gene rearrangements were discovered to occur in NSCLC (Rikova 2007; Soda 2007). Patients with ALK‐rearranged advanced NSCLC have subsequently been identified as the next subgroup of lung cancer patients to gain survival benefit from targeted therapy. Approximately 5% of NSCLC is driven by the ALK oncogene (Barlesi 2016; Solomon 2009), with patients typically being younger, light or never smokers with adenocarcinoma histology (Shaw 2009).

Independent of the discovery that driving mutations respond to targeted therapy, immunotherapy has recently demonstrated efficacy in NSCLC. Checkpoint inhibitors, including PD1 and PDL1, targeting antibodies, such as pembrolizumab, atezolizumab and nivolumab have changed the landscape of advanced NSCLC treatment with improved and durable survival when compared to chemotherapy in the first‐ and second‐line setting (Kim 2018).

Subgroup analysis of second‐line immunotherapy trials have demonstrated that patients with NSCLC driven by EGFR or ALK mutations do not gain the same benefits from immunotherapy when given as a single agent (Gainor 2016). More recently, trials are underway to investigate the combination of targeted therapy (including ALK inhibitors) with immunotherapy (Moya‐Horno 2018). This strategy is not current standard practice, and in fact one trial revealed unacceptable toxicity with the combination of gefitinib and durvalumab immunotherapy (Creelan 2019).

Although first‐line EGFR‐targeted therapy combined with chemotherapy has been shown to be effective (Seike 2018), this has not been demonstrated in the ALK setting. International guidelines currently recommend ALK inhibitors to be used as monotherapy.

When thinking broadly about lung cancer treatments, clinicians now routinely consider the role of five options: surgery, radiotherapy, cytotoxic chemotherapy, targeted therapy and immunotherapy. This review focusses on the role of targeted therapy, specifically ALK inhibitors, to treat patients with NSCLC driven by the ALK gene rearrangement.

Description of the intervention

Over the last decade, multiple tyrosine kinase inhibitors have been developed to target the ALK fusion kinase. These ALK inhibitors are medications taken orally up to twice a day and continuously to maintain effect.

Crizotinib (formerly known as PF02341066) was the first‐in‐class ALK inhibitor and is a tyrosine kinase inhibitor with activity against ALK, c‐MET and ROS1 kinases. Crizotinib has been compared to chemotherapy in the treatment of ALK‐rearranged advanced NSCLC in phase III clinical trials. In both the first‐ and second‐line setting (Shaw 2013; Solomon 2014), crizotinib was the first drug shown to have significantly higher response rates and progression‐free survival than chemotherapy in this subgroup of lung cancer patients. Although these trials were not of cross‐over design, participants were allowed to receive crizotinib after progression on chemotherapy and this has resulted in a failure to show a statistically significant difference in overall survival. The impact of crossover on overall survival is also apparent in subsequent ALK‐inhibitor trials. Longer follow‐up of the initial phase III first‐line crizotinib studies now demonstrates a four‐year survival of 57%; the median overall survival is not yet reached at 46 months follow‐up (Solomon 2018). This is a significant and meaningful outcome when considering the five‐year survival more broadly for stage IV lung adenocarcinoma is estimated at 2% (Cetin 2011).

Despite the efficacy of crizotinib in treating ALK‐rearranged NSCLC, acquired resistance to crizotinib inevitably results in disease progression while on treatment. The pattern and mechanism of disease progression can guide the most appropriate next step in treatment (Lin 2017). There is a high rate of central nervous system metastases that occurs in this setting (Costa 2015), due to lower levels of crizotinib reaching the brain (Costa 2011), which may represent a pharmacokinetic failure of the tyrosine kinase inhibitor. At a molecular level, acquired resistance mechanisms can include employment of other oncogenic pathways to bypass the inhibited ALK fusion protein or alteration of the ALK target through the gain of additional mutations (Camidge 2012).

This acquired resistance was the trigger to develop more potent, central nervous system penetrant and specific next generation ALK inhibitors including ceritinib, alectinib, brigatinib, ensartinib, entrectinib and lorlatinib. Each tyrosine kinase inhibitor has a different potency (Gainor 2016a), with some also able to inhibit kinases other than the ALK fusion kinase, including ROS1, RET and NTRK. This review will be limited to efficacy against the ALK kinase. Next generation ALK inhibitors were initially tested in crizotinib‐resistant patients and have been found to be effective, including for the treatment of central nervous system metastases (Gainor 2015). Evidence has rapidly evolved to demonstrate efficacy in the first‐line setting, compared to crizotinib in Shaw 2017 and chemotherapy in Soria 2017. Current guidelines give an option of four different ALK inhibitors (Planchard 2018), all supported by phase III trial data. With an approach of sequential ALK inhibitors, survival outcomes achieved in ALK‐positive patients are unprecedented in the advanced NSCLC setting. Real world data now demonstrates a median overall survival of nearly seven years (Pacheco 2019), and five‐year overall survival of 60% (Pacheco 2019a), which is comparable to resected stage II lung cancer (Goldstraw 2016).

How the intervention might work

The chromosomal rearrangement of the ALK gene results in a constitutively active ALK‐fusion kinase protein. This kinase was demonstrated in preclinical studies to be an oncogenic driver (Soda 2008), that was sensitive to inhibition by a tyrosine kinase inhibitor (McDermott 2008). This finding was rapidly followed by early phase studies including patients with lung cancer, confirming clinical efficacy and safety of crizotinib. International guidelines responding to results of phase III clinical trials recommend first‐line targeted therapy in NSCLC patients with ALK gene rearrangements rather than cytotoxic chemotherapy or immunotherapy (Planchard 2018).

Why it is important to do this review

ALK inhibitors have already been established as the standard of care in treating advanced ALK‐positive lung cancer. The necessary development of multiple, potent next generation ALK inhibitors contributes to a rapidly evolving treatment paradigm. Our initial literature search has found a number of expert opinion publications (Cameron 2015), but only one systematic review (Barrows 2019), and no meta‐analysis looking at comparable efficacy and sequence of multiple ALK inhibitors. This paper reviewed publications up until July 2017 and already further relevant phase III trials have been published since. A systematic review and meta‐analysis of ALK‐inhibitor toxicities has been completed (Costa 2018), including studies up to July 2017.

With multiple next generation drugs being tested in first and subsequent treatment settings, determining the ideal sequence becomes more complex. Strategies have been proposed for sequence of therapies based on potential resistance mechanisms (Gainor 2016a), but with a rapidly evolving evidence base, a robust assessment of high level clinical trial outcomes is needed to inform and update the current optimal treatment algorithm for this specific group of lung cancer patients.

The issue confronting clinicians now is which ALK inhibitor should be used first and in what sequence to gain the optimal survival for ALK‐rearranged NSCLC patients (Recondo 2018).

Objectives

To evaluate the safety and efficacy of anaplastic lymphoma kinase (ALK) inhibitors given as monotherapy to treat advanced ALK‐rearranged non‐small cell lung cancer (NSCLC).

Methods

Criteria for considering studies for this review

Types of studies

We will consider all randomised trials including open‐label, single‐blind and double‐blind studies.

Types of participants

The participants will be people with advanced (stage III or IV) non‐small cell lung cancer (NSCLC) harbouring anaplastic lymphoma kinase (ALK) gene rearrangement diagnosed histologically or cytologically using immunohistochemistry or fluorescence in situ hybridisation (FISH) analysis. There will be no limitation by age, gender or demographics. We will not include people with ROS1 mutation.

Types of interventions

We will consider any administration of therapies targeting the oncogenic ALK fusion kinase (ALK inhibitors) versus chemotherapy or other ALK inhibitors, when used as a first (naive to systemic therapy) or subsequent (previously treated) line of treatment.

ALK inhibitors will include, but not be limited to, crizotinib, ceritinib, alectinib, entrectinib, lorlatinib, brigatinib and ensartinib. We will exclude studies that involve ALK inhibitors in combination with other systemic treatments. Epidermal growth factor receptor (EGFR) and ALK‐driven lung cancers can be considered different pathologies, and as such, we will not include EGFR‐targeted therapies in this review.

We plan the following comparisons.

• First‐line treatment

  • Any ALK inhibitor compared to any cytotoxic chemotherapy

  • One ALK inhibitor compared to another ALK inhibitor

• Second‐line treatment or subsequent treatment

  • Any ALK inhibitor compared to chemotherapy

  • One ALK inhibitor compared to another ALK inhibitor

Types of outcome measures

Primary outcomes

• Progression‐free survival (assessed from date of randomisation to date of objective disease progression by response evaluation criteria in solid tumours (RECIST; Eisenhauer 2009)) or death, whichever occurs first

• Adverse events, as defined by common terminology criteria for adverse events (CTCAE; CTCAE v4) grading (1‐5)

Secondary outcomes

The secondary outcomes will include the following.

• Overall survival, which is assessed from date of randomisation to date of death or study end date if the participant was alive

• Overall survival at one year

• Overall response rate by RECIST criteria (Eisenhauer 2009). We will also report partial response and complete response rates.

• Health‐related quality of life, as measured on a validated scale

Search methods for identification of studies

Electronic searches

We will search the following databases.

• Cochrane Lung Cancer Group's Specialised Register

• Cochrane Central Register of Controlled Trials (CENTRAL) (Appendix 1)

• MEDLINE (2007 to present) (Appendix 2)

• Embase (2007 to present) (Appendix 3)

The Cochrane Lung Cancer Group Information Specialists developed the search strategies for the three main databases: CENTRAL (Appendix 1), MEDLINE (Appendix 2), and Embase (Appendix 3). The search string for MEDLINE is developed according to the Cochrane Highly Sensitive Search Strategy, sensitivity maximising version (2008 version) as referenced in Chapter 6.4.11.1 and detailed in box 6.4.b of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will also conduct a search of ClinicalTrials.gov (ClinicalTrials.gov), and the World Health Organization (WHO) trials portal (www.who.int/ictrp/en/).

We will commence searches from 2007 because it is the year that ALK rearrangements were discovered in NSCLC. We will place no restrictions on language or publication type, this includes abstract format.

Searching other resources

We will also search the conference proceedings of:

• American Society for Clinical Oncology, year 2017 to present;

• European Society of Medical Oncology, year 2017 to present;

• International Association for the Study of Lung Cancer (IASLC) World Conference on Lung Cancer, year 2017 to present.

In addition, we will search the reference lists of included studies and narrative reviews.

Data collection and analysis

We will use standard methodologies of Cochrane, for data collection (Higgins 2011).

Selection of studies

All authors (excluding BS) will use the Covidence tool for screening studies identified (Covidence 2017). Initially LC will independently screen titles and abstracts to assess possible eligibility, following this LC and NH will independently assess full texts according to the criteria defined above for inclusion in the review. We will resolve disagreements by discussing with VJ or RM. We will document the reasons for exclusion according to full‐text review in the report and record the selection process in detail to complete a PRISMA flow diagram (Moher 2009). In the case of multiple publications for the same trial, we will include the most mature outcome data.

Data extraction and management

Two authors (NH and LC) will independently collect data including the following.

• Study details: study design
, study period, year of publication, number of participants randomised
, participant discontinuation rate.

• Participant characteristics: age, gender, ethnicity
, NSCLC staging
, sites of disease
, histological subtype, 
smoking status, 
performance status,
 previous treatment.

• Intervention: dose, frequency and duration of targeted therapy and comparator intervention.

• Outcomes: progression‐free survival, adverse events, overall survival, one‐year overall survival, response rate, health‐related quality of life measures.

The comparisons we plan are outlined in the Types of interventions section.

Assessment of risk of bias in included studies

Two authors (LC and NH) will be assessing risk of bias independently using the 'Risk of bias' tool, described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Disagreements will be discussed with a third review author (RM or VJ) to reach consensus. We will justify the risk of bias according to the following domains.

• Random sequence generation

• Allocation concealment

• Blinding of participants and personnel

• Blinding of outcome assessment

• Incomplete outcome data, evaluated for each outcome measure

• Selective outcome reporting

• Other bias

We will be assigning each risk of bias domain into one of the following categories, 'low risk of bias', 'high risk of bias', or 'unclear risk of bias'. We anticipate that a number of studies will not be blinded due to the different methods of drug administration. For each included study, we will describe the methods used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. We will judge studies as low risk of bias if they were blinded, or if we judge that the lack of blinding could not have affected the results, such as mortality. We will assess blinding and incomplete outcome data separately for different outcomes or classes of outcomes. We will also assess blinding of outcome assessors who will likely use RECIST criteria of tumour measurement to assess response.

Measures of treatment effect

We will report hazard ratios (HRs) and their standard errors (SEs) for time‐to‐event outcomes: progression‐free survival and overall survival. We will include these in a meta‐analysis using the generic inverse variance method. Additionally, if the data has been analysed using a Cox proportional hazards model, we will use this to produce a direct estimate of the HR and it's SE. If HRs have not been supplied, but survival curves have been given, we will use these to estimate the HRs. If a HR has been supplied with a measure of error, which is not a SE (such as a confidence interval (CI) or P value) we will use the inbuilt calculator in Review Manager 5 to convert these to a SE (Review Manager 2014).

We anticipate that participant crossover in included trials will result in overall survival differences being less apparent when presented as HRs, and so for overall survival, we will also present median length of survival time with ranges in a table.

For the continuous outcome, health‐related quality of life, we will report the results as mean differences (MDs), and calculate them with 95% CIs for studies that used the same scale. If quality of life has been measured using different scales in the different studies, we will calculate standardised mean differences (SMDs) with 95% CIs . We will assess the health‐related quality of life scale used and only include it if it is a recognised and validated scale. We will also obtain the minimally important difference for the scales used and use this to interpret our results.

For the dichotomous variables, response rate, adverse event rates and one‐year overall survival, we will present results as risk ratios (RRs) with 95% CIs. If the number of events and the total number of participants are not presented in the paper, we will use percentages to back calculate the number of events. If only a summary RR is presented, then we will use generic inverse variance to allow calculation of an overall summary statistic.

Unit of analysis issues

The unit of analysis will be an individual participant. We will exclude trials with cross‐over design. For the studies with multiple intervention groups, we will divide the analysis into pairwise comparisons (e.g. A versus placebo, B versus placebo, A versus B) in order to make sure a group of participants will not be included twice in the same meta‐analysis. When A and B must be analysed together, we will halve the placebo group to avoid double‐counting.

Dealing with missing data

We will contact authors of included studies to obtain any missing data.

Assessment of heterogeneity

We will present the included trials on forest plot graphs and visually inspect the plot for heterogeneity. We will explore potential causes of heterogeneity. For pooled analyses we will use Review Manager 5 to calculate the I² statistic (Review Manager 2014), which gives a percentage of variation across studies due to heterogeneity rather than chance. If I² is higher than 75%, we will consider not pooling the data. Where we find moderate heterogeneity, we will explore sources of clinical and statistical heterogeneity, including clinically significant subgroup analysis.

Assessment of reporting biases

If 10 or more studies are included, we will create a funnel plot to assess possible publication bias or small study effects. We will not limit the search to English studies in order to avoid language bias.

Data synthesis

We will analyse data using Review Manager 5 software (Review Manager 2014). If appropriate, we will perform meta‐analyses. We will use the fixed‐effect model if the studies have sufficiently low clinical and statistical heterogeneity. Where multiple publications refer to the same trial we will include the most mature data set for outcomes.

For the primary outcome of progression‐free survival and for overall survival, we will calculate an overall HR using the inverse variance method to combine the summary data from each of the studies. For those studies not supplying a HR, we will include a table of median survival times. We will not combine these times to provide an overall average.

For health‐related quality of life, we will combine the MDs obtained from the individual studies using either a MD if all measures were taken on the same scale or a SMD if different scales were used. If a MD is used to calculate the summary statistic, we will be able to analyse studies that report endpoints with studies that report change scores, however, if a SMD is used, we will separate the endpoint scores from the change scores and analyse these separately. For dichotomous data supplied for adverse events, overall response rate and one‐year overall survival, we will combine data to produce an overall RR using the Mantel‐Haenszel method with CIs.

We will create a 'Summary of findings' table to present the results. The outcomes on the table will include progression‐free survival, adverse events, overall survival, one‐year overall survival, overall response rate and health‐related quality of life. We will use the GRADE approach to rate the quality of evidence for the meta‐analysis and present the results on the 'Summary of findings' table (GRADEPro GDT 2014). We will assess the quality of a body of evidence using the five GRADE aspects (study limitations, indirectness, unexplained heterogeneity or inconsistency of results, imprecision of results and publication bias (Guyatt 2008). We will add comments and footnotes to the 'Summary of findings' table in order to help readers understand the results.

Subgroup analysis and investigation of heterogeneity

We plan to perform subgroup analysis by type of ALK inhibitors and types of previous treatment, with a focus on systemic therapy and disease sites, including central nervous system.

Sensitivity analysis

We will conduct sensitivity analyses to confirm the robustness of analysis results. If sufficient trials are eligible for the analysis, we will select trials with low risk of bias for the majority of domains. We will also perform sensitivity analysis by utilising a random‐effects model and comparing results to our default fixed‐effect model.

Appendices

Appendix 1. CENTRAL search strategy

#1. MeSH descriptor: [Carcinoma, Non‐Small‐Cell Lung] explode all trees #2. nsclc #3. "lung cancer*" #4. "lung carcinom*" #5. "lung neoplasm*" #6. "lung tum*" #7. "non small cell*" #8. "nonsmall cell*" #9. (#3 or #4 or #5 or #6) and (#7 or #8) #10. #1 #2 or #9 #11. "anaplastic lymphoma kinase" #12. alk #13. "anaplastic lymphoma kinase" #14. "anaplastic lymphoma receptor tyrosine kinase" #15. "CD 246" #16. MeSH descriptor: [Molecular Targeted Therapy] explode all trees #17. crizotinib #18. xalkori #19. ceritinib #20. zykadia #21. LDK378 #22. alectinib #23. entrectinib #24. brigatinib #25. AP26113 #26. lorlatinib #27. X396 OR Ensartinib #28. "X 396" #29. MeSH descriptor: [Receptor Protein‐Tyrosine Kinases] explode all trees #30. MeSH descriptor: [Protein Kinase Inhibitors] explode all trees #31. #13 or #14 or #15 #32. #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30 #33. #10 and #31 and #32

Appendix 2. MEDLINE search strategy

1	Carcinoma, Non‐Small‐Cell Lung[MeSH Terms]
2	nsclc[Title/Abstract]
3	lung cancer*[Title/Abstract]
4	lung carcinoma[Title/Abstract]
5	lung neoplasm*[Title/Abstract]
6	lung tumor*[Title/Abstract]
7	lung tumour*[Title/Abstract]
8	non small cell*[Title/Abstract]
9	nonsmall cell*[Title/Abstract]
10	(#3 OR #4 OR #5 OR #6 OR #7) AND (#8 OR #9)
11	#1 OR #2 OR #10
12	anaplastic lymphoma kinase[Supplementary Concept]
13	alk[Title/Abstract]
14	anaplastic lymphoma kinase[Title/Abstract]
15	anaplastic lymphoma receptor tyrosine kinase[Title/Abstract]
16	CD246[Title/Abstract]
17	CD 246[Title/Abstract]
18	molecular targeted therapy[MeSH Terms]
19	target*[Title/Abstract]
20	crizotinib[Supplementary Concept]
21	crizotinib[Title/Abstract]
22	PF 02341066
23	PF02341066
24	xalkori[Title/Abstract]
25	ceritinib[Supplementary Concept]
26	ceritinib[Title/Abstract]
27	zykadia[Title/Abstract]
28	LDK378
29	CH5424802[Supplementary Concept]
30	alectinib[Title/Abstract]
31	RO5424802
32	entrectinib[Supplementary Concept]
33	entrectinib[Title/Abstract]
34	RXDX‐101
35	NMS‐E628
36	brigatinib[Title/Abstract]
37	AP26113
38	7‐amino‐12‐fluoro‐2,10,16‐trimethyl‐15‐oxo‐10,15,16,17‐tetrahydro‐2H‐8,4‐(metheno)pyrazolo(4,3‐h)(2,5,11)benzoxadiazacyclotetradecine‐3‐carbonitrile
39	lorlatinib[Title/Abstract]
40	PF06463922
41	PF‐06463922
42	ensartinib[Title/Abstract]
43	X396
44	receptor protein‐tyrosine kinases[MeSH Terms]
45	protein kinase inhibitors[MeSH Terms]
54	(((AP26113[Supplementary Concept]) OR AP26113) OR brigatinib[Title/Abstract]) OR Alunbrig[Title/Abstract]
58	((((entrectinib[Supplementary Concept]) OR entrectinib[Title/Abstract]) OR RXDX‐101) OR NMS‐E628) OR N‐(5‐(3,5‐difluorobenzyl)‐1H‐indazol‐3‐yl)‐4‐(4‐methyl‐1‐piperazinyl)‐2‐(tetrahydro‐2H‐pyran‐4‐ylamino)benzamide
60	#12 OR #13 OR #14 OR #15 OR #16 OR #17
61	#18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #54 OR #58
62	#11 AND #60 AND #61
63	randomized controlled trial[Publication Type]
64	controlled clinical trial[Publication Type]
65	randomized[Title/Abstract]
66	placebo[Title/Abstract]
67	drug therapy[MeSH Subheading]
68	randomly[Title/Abstract]
69	trial[Title/Abstract]
70	groups[Title/Abstract]
71	#63 OR #64 OR #65 OR #66 OR #67 OR #68 OR #69 OR #70
72	animals [MeSH Terms] NOT humans [MeSH Terms]
73	#71 NOT #72
74	#62 AND #73

Appendix 3. Embase search strategy

#50	#11 AND #48 AND #49
#49	'crossover procedure'/exp OR 'double‐blind procedure'/exp OR 'randomized controlled trial'/exp OR 'single‐blind procedure'/exp OR random* OR factorial* OR crossover* OR (cross NEXT/1 over*) OR placebo* OR (doubl* NEAR/1 blind*) OR (singl* NEAR/1 blind*) OR assign* OR allocat* OR volunteer*
#48	#12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47
#47	'x 396':ti,ab
#46	'nms e628':ti,ab
#45	'rxdx 101':ti,ab
#44	'pf 02341066':ti,ab
#43	'pf 06463922':ti,ab
#42	ro5424802:ti,ab
#41	xalkori:ti,ab
#40	alunbrig:ti,ab
#39	lorbrena:ti,ab
#38	zykadia:ti,ab
#37	alecensa:ti,ab
#36	x396:ti,ab
#35	pf06463922:ti,ab
#34	ap26113:ti,ab
#33	rxdx101:ti,ab
#32	ch5424802:ti,ab
#31	ldk378:ti,ab
#30	pf02341066:ti,ab
#29	ensartinib:ti,ab
#28	'lorlatinib':ti,ab
#27	'lorlatinib'/exp
#26	'brigatinib':ti,ab
#25	'brigatinib'/exp
#24	'entrectinib':ti,ab
#23	'entrectinib'/exp
#22	'alectinib':ti,ab
#21	'alectinib'/exp
#20	'ceritinib':ti,ab
#19	'ceritinib'/exp
#18	'crizotinib':ti,ab
#17	'crizotinib'/exp
#16	'protein tyrosine kinase inhibitor'/exp
#15	'tyrosine kinase inhib*':ti,ab
#14	'anaplastic lymphoma kinase':ti,ab
#13	'anaplastic lymphoma kinase inhibitor'/exp
#12	alk:ti,ab
#11	#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10
#10	'non‐small cell*':ab,ti
#9	'nonsmall cell*':ab,ti
#8	'lung tumour*':ab,ti
#7	'lung tumor*':ab,ti
#6	'lung neoplasm*':ab,ti
#5	'lung cancer*':ab,ti
#4	'lung carcinom*':ab,ti
#3	'nsclc':ab,ti
#2	'lung tumor'/exp
#1	'non small cell lung cancer'/exp

Contributions of authors

Laird Cameron, Nadia Hitchen, Ben Solomon, Renée Manser and Vanessa Jordan designed the review strategy. All authors have reviewed and approved the final draft of this protocol.

Declarations of interest

LC: none known. LC is a member of the Management Advisory Committee of the Australian Lung Trials Group.

NH: none known

VJ: none known

RM: none known

BS is an author on phase III clinical studies of crizotinib, ceritinib and lorlatinib. He has served on advisory boards for Pfizer, Novartis, Roche‐Genentech, AstraZeneca, Merck, Bristol Myers Squibb, Gritstone Oncology and Loxo Oncology and he has spoken at meetings where costs were sponsored by industry.

New