Lung Cancer 147 (2020) 130–136

Contents lists available at ScienceDirect

Lung Cancer
journal homepage: www.elsevier.com/locate/lungcan

Effectiveness and prognostic factors of first-line crizotinib treatment in T

patients with ROS1-rearranged non-small cell lung cancer: A multicenter
retrospective study
Jing Zhenga,1, He Caoa,1, Yuping Lib, Chuangzhou Raoc, Te Zhangd, Jiayou Luod, Dongqing Lve,
Yanping Zhua, Jianya Zhoua,*, Jianying Zhoua
a
Department of Respiratory Disease, Thoracic Disease Center, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, 310003 China
b
Department of Respiratory Disease, The First Affiliated Hospital of Wenzhou Medical University, 325000, China
c
Department of Respiratory Disease, Hwa Mei Hospital of the University of the Chinese Academy of Science, 325000, China
d
Department of Respiratory Disease, Huzhou Central Hospital of Zhejiang Province, Huzhou, 313003, China
e
Department of Respiratory Disease, Taizhou Hospital of Zhejiang Province, Taizhou, 318001, China

A R T I C LE I N FO A B S T R A C T

Keywords: Introduction: ROS1 rearrangements are seen in 1–2 % of non-small cell lung cancer (NSCLC) patients. The aim of
Non-small cell lung cancer this study was to determine the effectiveness of crizotinib as first-line treatment in patients with ROS1 rear-
Crizotinib ranged NSCLC.
Effectiveness Methods: The data of 56 patients with ROS1-rearranged advanced NSCLC who received first-line crizotinib
First-line treatment
treatment from 5 hospitals in China were retrospectively reviewed. The clinical characteristics, outcomes of first-
Prognostic factors
line crizotinib therapy and prognostic factors were evaluated. In addition, mechanisms of resistance to crizotinib
Resistance mechanisms
were analyzed in a cohort of crizotinib-resistant patients.
Results: The median age of the cohort was 53.0 years and most patients (91.1 %) had adenocarcinomas. The
median progression free survival (mPFS) after first-line crizotinib therapy was 23.0 months, and the median
overall survival (mOS) was 60.0 months. In the univariate analysis, mPFS was significantly shorter in female
patients compared to males (12.0 months versus 24.0 months; p = 0.015) and in patients with > 2 baseline
metastatic organ involvement compared to those with ≤2 baseline metastatic organ involvement (4.0 months vs
24.0 months; p < 0.001).In addition, the mOS was significantly shorter in patients with > 2 baseline metastatic
organ involvement relative to that in patients with ≤2 baseline metastatic organ involvement (6.0 months vs
60.0 months; p < 0.001). Multivariable analysis further showed that > 2 baseline metastatic organ involvement
was the only independent prognostic factor of PFS (p = 0.008). Four out of 8 patients (50 %) with crizotinib
resistance harbored a G2032R mutation in the ROS1 kinase domain.
Conclusions: First-line crizotinib treatment is beneficial in Chinese patients with ROS1-rearranged advanced
NSCLC. Resistance to crizotinib correlated with the G2032R mutation in the ROS1 kinase domain.

1. Introduction
Lung cancer is commonly diagnosed malignancy worldwide and a
leading cause of cancer-related deaths [1]. Non-small cell lung cancer
(NSCLC) accounts for more than 80 % of all lung cancer cases, and
identification of oncogenic mutations in the epithelial growth factor
receptor gene (EGFR) of NSCLC patients led to the development of
tyrosine kinase inhibitors (TKIs) for targeted therapy [2]. TKIs specific
for mutated EGFR and anaplastic lymphoma kinase (ALK) have dis-
tinctively improved treatment outcomes in NSCLC [2,3].
ROS1 (c-ros oncogene 1) rearrangement was identified in 2007 as a
new oncogenic receptor tyrosine kinase that activates a large number of
human tumors, including lung cancer, gastric cancer, and glioblastoma
[4–7]. The prevalence of ROS1 rearrangement in NSCLC patients is 1–2
%, and ROS1 fusion proteins with CD74, SLC34A2, and TPM3 due to
gene translocation have been detected in NSCLC [8].

⁎
Corresponding author at: Department of Respiratory Disease, Thoracic Disease Center, The First Affiliated Hospital, College of Medicine, Zhejiang University, No.
79, Qingchun Road, Shangcheng District, Hangzhou, Zhejiang, 310003, China.
E-mail address: zhoujy@zju.edu.cn (J. Zhou).
1
The two authors contribute equally to the work.

https://doi.org/10.1016/j.lungcan.2020.07.016
Received 23 April 2020; Received in revised form 12 July 2020; Accepted 14 July 2020
0169-5002/ © 2020 Elsevier B.V. All rights reserved.

DOI 10.1016/j.lungcan.2020.07.016
J. Zheng, et al.
ROS1 and ALK belong to the insulin receptor tyrosine kinases [9],
with more than 64 % sequence homology in the kinase domain and
∼84 % homology within the ATP binding site [10]. This raises the
possibility of using ALK inhibitors in ROS1-rearranged NSCLC as well
[11]. In fact, the first-generation ALK inhibitor crizotinib increased the
median progression-free survival (mPFS) of ROS1-rearranged NSCLC
patients to 19.2 months in the phase I PROFILE 1001 clinical trial
[11,12]. In a subsequent phase II study, crizotinib achieved an mPFS of
15.9 months in 127 East Asian patients with ROS1 fusion NSCLC [13].
Based on these findings, crizotinib was approved by the Chinese FDA in
2017 for treating ROS1-rearranged NSCLC. The effectiveness and safety
of crizotinib for ROS1-rearranged NSCLC have also been validated in
several small real-world cohorts [14,15]. All the aforementioned stu-
dies included only a small proportion of patients who received crizo-
tinib as the first-line treatment; for example, only 7 and 24 treatment-
naïve patients were respectively included in the phase I and phase II
trials [11].
Notwithstanding the excellent responses to crizotinib, acquired re-
sistance in ROS1 rearranged NSCLC is a frequently occurrence and as-
sociated with point mutations like G2032R, D2033 N, L2155S, S1986
F/Y, and L2026 M within the ROS1 kinase domain [16]. Furthermore,
-activation of bypass-signaling pathways can also obviate the inhibitory
effect of crizotinib and other TKIs on ROS1 rearrangement. However,
these mechanisms have been reported in only small patient cohorts or
in cell lines. The third generation ALK inhibitor lorlatinib and the ROS1
inhibitor repotrectinib have been effective in in patients that are non-
responsive to crizotinib. A phase I-II trial reported an the objective
response rate (ORR) of 35 % in lorlatinib-treated ROS1-rearranged
NSCLC patients after the crizonitib regimen [17]. The TRIDENT-1 study
found that the ORR of repotrectinib was 57 % in patitens after cirzo-
tinib resistance [18]. However, both loratinib and repotrectinib were
not effective in crizotinib-pretreated patients with ROS1 G2302R mu-
tation.
In this study, we retrospectively reviewed the data of 56 advanced
NSCLC patients with ROS1 rearrangements from 5 different hospitals in
China who received crizotinib as first-line treatment in order to de-
termine the PFS and overall survival (OS) and identify the prognostic
factors and potential drug resistance mechanisms.
2. Materials and methods
2.1. Study design and patients
This study was approved by the Institutional Review Boards of the
First Affiliated Hospital of Zhejiang University, the First Affiliated
Hospital of Wenzhou Medical University, Hwa Mei Hospital of the
University of the Chinese Academy of Science, Huzhou Central Hospital
of Zhejiang Province, and Taizhou Hospital of Zhejiang Province.
Patients from the above five hospitals in eastern China were included.
Informed consent was not required due to the retrospective study de-
sign.
Data on a total of 56 patients were analyzed in our study. All pa-
tients had been diagnosed and treated for ROS1-rearranged advanced
NSCLC in the above centers between January 2014 and December
2019. The patients all met the following criteria: (1) ROS1 rearrange-
ment in the tumor tissues was confirmed by the amplification refractory
mutation system (ARMS), fluorescence in situ hybridization (FISH), or
next-generation sequencing (NGS), (2) tumor-node-metastasis (TNM)
staging of the patients was determined as IIIB or IV according to the 7th
edition of the Lung Cancer TNM classification at diagnosis or relapse
after surgery, and (3) all patients received crizotinib as the first-line
treatment and were regularly evaluated at the treating physician’s
discretion in the above hospitals about every two months. Two in-
vestigators retrospectively reviewed the computed tomography (CT)
chest scans, brain scans and bone scans of all patients, and evaluated
the treatment outcomes of crizotinib. The time from the initiation of
131
Lung Cancer 147 (2020) 130–136
crizotinib treatment to the study cut-off date (1 January 2020) was
more than 6 months.
2.2. Tumor assessment
All images including chest CT scans, brain scans, and bone scans,
before and during the treatment of patients were obtained from the five
participating hospitals. Tumor assessment was performed retro-
spectively according to the Response Evaluation Criteria in Solid
Tumors (RECIST), version 1.1. The tumor responses to crizotinib in-
cluded 4 outcomes: complete response (CR), partial response (PR),
progressive disease (PD), and stable disease (SD). The proportion of
patients with CR or PR was defined as the objective response rate
(ORR), and the proportion that achieved CR, PR or SD was defined as
the disease control rate (DCR). PFS was defined as the time from the
start of crizotinib treatment to PD according to RECIST 1.1 or death. OS
was measured from the initiation of treatment until the date of death.
2.3. Data collection
The following patient characteristics were collected and analyzed
retrospectively in our study from medical records: sex, age, smoking
history, histologic subtype, TNM staging, baseline ECOG performance
status score, baseline metastases (including contralateral lung, pleural,
intracranial, liver, bone, adrenal glands and other organs) and con-
comitant diagnosis of VTE [including deep vein thrombosis (DVT) and
pulmonary thromboembolism (PE)]during the diagnosis of advanced
NSCLC to death or the cutoff date.
2.4. Capture-based targeted DNA sequencing analysis
Formalin-fixed paraffin-embedded (FFPE) tumor samples were col-
lected from 5 patients for capture-based targeted DNA sequencing.
Next-generation sequencing (NGS) reports were retrospectively col-
lected from 3 patients, including two who had peripheral blood se-
quencing and one who had cerebrospinal fluid sequencing. All the
tumor samples and peripheral blood samples were obtained at the time
the patients had progressed on first line crizotinib. A commercial cap-
ture-based NGS panel of 168 genes (Burning Rock Biotech, Guangzhou,
China) were used (See Supplementary Table S1 1 for gene list).
2.5. Statistical analysis
The median PFS and OS were estimated using the Kaplan–Meier
method and differences due to different variates were compared using
the log-rank test. Univariate analysis was performed on sex, age (at a
cut-off of 65 years old), smoking history, histologic subtype, TNM sta-
ging, baseline ECOG performance status score, baseline intracranial
metastases, number of baseline metastatic organ involvement and
concomitant VTE. Variates with p ≤ 0.10 were then included for
multivariable analysis using the Cox proportional hazards model. A 2-
sided p < 0.05 was considered statistically significant in the multi-
variable analyses. Statistical analysis was performed using SPSS™ soft-
ware version 22.0 (SPSS Inc., Chicago, IL).
3. Results
3.1. Clinical characteristics of the patients with ROS1 -rearranged NSCLC
Among the 56 patients with ROS1 arrangement, 53 patients were
confirmed as having ROS1 rearrangement by the use of reverse tran-
scriptase-polymerase chain reaction (RT-PCR) assays. Two patients
were diagnosed by NGS and one was diagnosed as ROS1 rearrangement
via FISH. Five of the 56 patients (8.9 %) had participated in a phase II
clinical study of crizotinib [13]. Detailed clinical and demographic
characteristics of the 56 patients are shown in Table 1.
J. Zheng, et al. Lung Cancer 147 (2020) 130–136

Table 1 3.2. Effectiveness of first-line crizotinib therapy in patients with ROS1

Characteristics of the Patients with ROS1-rearranged NSCLC. -rearranged NSCLC
Characteristics ROS1-positive
Until the cut-off date, the median follow-up time was 29.0 months
patients (n = 56) (95 % confidence interval [CI], 22.4–33.6). The best response to first-
line crizotinib treatment was evaluable in 51 patients as 5 patients had
n %
Age, years:
no measurable target lesions according to RECIST1.1. CR, PR, SD and
Median 53 PD were achieved in 1, 32, 15 and 3 patients respectively. The ORR was
Range 24−72 64.7 % (33/51) and DCR was 94.1 % (48/51).
≤65 44 78.6 Swimmer plots of PFS and OS are shown in Fig. 1. 31 patients had
> 60 12 21.4
progressive disease at the cut-off date, of which 16eventually died and2
Sex:
Male 25 44.6 lost contact with their physicians after disease progression. The mPFS
Female 31 55.4 was 23.0 months (95 % CI, 22.4–33.6) and the mOS was 60.0 months
Smoking status: (95 % CI, 40.7–79.3) [Fig. 2A–B].
Never 37 66.1
Smoker 19 33.9
Histologic classification:
3.3. The disease progression pattern in ROS1-rearranged NSCLC patients
Adenocarcinoma 51 91.1 after first-line crizotinib therapy and post-crizotinib treatments
Adenosquamous carcinoma 5 8.9
Disease stage: The disease progression pattern in 31 progressed patients was
IIIB 10 17.9
shown in Fig. 3. Intracranial progression occurred in 9 (9/31, 29.0 %)
IV or relapsed 46 82.1
ECOG score patients, whereas the remaining 22 (22/31, 71.0 %) patients showed
0 12 21.4 systemic progression. The disease progression rate was similar in pa-
1 or 2 44 78.6 tients with or without baseline (pre-crizotinib) intracranial metastases
baseline distant metastases: (63.6 % (7/11) vs. 53.3 % (24/45), P = 0.259). Eighteen patients
bone metastases 18 32.1
Contralateral pulmonary metastases 15 26.8
continued the crizotinib treatment after progression, of which 8 pa-
Intracranial metastases 11 19.6 tients received crizotinib alone and 10 received additional local radia-
Pleural metastases 9 16.1 tion therapy, including brain SRS (5), stereotactic body radiation
liver metastases 4 7.1 therapy (SBRT) of adrenal gland (1), WBRT (2) and pleural che-
adrenal glands metastases 3 5.4
motherapy (2). The systemic regimen was altered in 11 patients, of
other organs metastases* 4 7.1
Number of baseline metastatic organs: which 6 received other ALK inhibitors like ceritinib (3), loratinib (2)
≤2 52 92.9 and both the two drugs (1), 4 received chemotherapy, and 1 received
＞2 4 7.1 both chemotherapy and ceritinib. The data of the two patients that lost
Thromboembolic events: contact were not included.
Yes 8 14.3
No 48 85.7
3.4. Univariate and multivariable prognostic factors analyses of mPFS/mOS
ECOG, Eastern Cooperative Oncology Group. times with crizotinib treatment
*: other organs: 2 (3.6 %) pericardium metastases, 1 (1.8 %) pancreas metas-
tases and 1 (1.8 %) spleen metastases. As shown in Table 2, mPFS was significantly shorter in female pa-
tients than in males (12.0 months vs 24.0 months; p = 0.015), and
The median age at diagnosis was 53 years [range 24–82 years], and associated with the number of baseline metastatic organ involvement (p
most patients (44/56; 78.6 %) were younger than 65 years. Fifty-one = 0.002; Fig. 2C). Furthermore, the mPFS in patients with > 2 baseline
patients (91.1 %) had adenocarcinomas, and the remaining 5 had metastatic organ involvement was significant shorter than those with ≤
adenosquamous carcinomas. In addition, 55.4 % of the patients (31/56) 2 baseline metastatic organ involvement (4.0 months vs 24.0 months;
were female, and 66.1 % (37/56) were never smokers. At diagnosis, p < 0.001). However, the mPFS was not affected by baseline in-
73.2 % (41/56) had stage IV disease according to the 7th edition of tracranial metastases (p = 0.462).
Lung Cancer TNM classification, 17.9 % (10/56) had stage IIIB disease The number of baseline metastatic organ involvement was the only
and 8.9 % (5/56) had relapsed due to distant metastases after surgery. factor associated with the mOS (p = 0.004; Fig. 2D). The mOS in pa-
None of the 10 stage IIIB patients had received concurrent chemor- tients with > 2 baseline metastatic organ involvement was also sig-
adiation therapy. Three patients received first-line crizotinib in the nificantly shorter than patients with ≤2 baseline metastatic organ in-
phase II clinical trial (meet all the inclusion criteria and not meet all the volvement (6.0 months vs 60.0 months; p < 0.001). In addition, the
exclusion criteria) [13], 3 did not tolerate to concurrent chemoradia- mOS was similar in patients that continued crizotinib treatment or were
tion at the diagnosis (2 due to pneumonia and 1 for leucopenia) and 4 switched to other systemic therapy due to crzotinib resistance (47.0
patients refused concurrent chemoradiation and asked for oral targeted months vs 26.0 months, P=0.695).
therapy. Furthermore,46 (82.1 %) patients had metastasis, including 11 Sex, smoking history (P = 0.086) and number of organs involved in
(19.6 %) intracranial metastases and 18 (32.1 %) bone metastases, baseline metastases (≤2 versus > 2) were included in the multivariable
shown in Table 1. 6 of the 36 patients had 2 or more distant organ analysis on PFS and > 2 baseline metastatic organ involvement was the
metastases. Among those with baseline intracranial metastases, 6 pa- only independent prognostic factor of PFS (p = 0.008; Table 3).
tients had received stereotactic radiosurgery (SRS in 4 patients) or
whole brain radiotherapy (WBRT in 2 patients) along with crizotinib. 3.5. Mutation profile of crizotinib resistant ROS1-rearranged NSCLC
VTE occurred in 8 patients (14.3 %) during the disease, including 3 patients
with DVT, 3\ with PE and 2 patients with both DVT and PE. The ECOG
performance score was 0 in 12 patients (21.4 %), 1 in 42 patients (75.0 To elucidate the comprehensive mutation profile in the 8 patients
%), and 2 in 2 patients (3.6 %). with crizotinib-resistant ROS1 rearranged NSCLC, targeted sequencing
was performed on peripheral blood or tissue samples obtained when
resistance to first-line crizotinib therapy was detected, using a panel
consisting of 168 cancer-related genes (See Supplementary Fig. S1 for

132
J. Zheng, et al. Lung Cancer 147 (2020) 130–136

Fig. 1. Swimming plot of overall survival of the first-line crizotinib treatment on 56 ROS1-rearranged NSCLC patients.

the mutation profile).
ROS1 rearrangement were detected from 6 of the 8 patients (75 %)
and the ROS1 fusion partners were different, including 2 CD74-ROS1
fusions, 2 SDC4-ROS1 fusions, 1 GOPC-ROS1 fusion, and 1 LRIG3-ROS1
fusion. Furthermore, point mutation G2032R which is known to be a
resistance mechanism, was detected in 4 patients. A novel point mu-
tation within the ROS1 kinase domain, L2026 V mutation was detected
in one patient. A total of 24 concurrent mutations were detected in 16
different genes from the 8 samples. Supplementary Table S2 sum-
marizes the all the mutations detected in the cohort.
4. Discussion
This is the first study to evaluate the clinical characteristics and
prognosis of a large multicenter cohort of patients with ROS1-rear-
ranged NSCLC following first-line crizotinib therapy. In addition, re-
sistance mechanisms were analyzed in a cohort of crizotinib-resistant
patients.
Previous studies [19–23] have shown that ROS1-rearranged NSCLC
patients tend to be younger compared to those without ROS1 re-
arrangement, and that the latter correlates with female sex and non-
smokers [22,23]. Although most ROS1-rearranged NSCLC patients in
our cohort were younger than 65 years (78.6 %) and did not have a
history of smoking (66.1 %), the number of male and female patients
were similar (44.6 % vs. 55.4 %). Adenocarcinoma was the pre-
dominant histological diagnosis in our study (91.1 %), which is con-
sistent with previous reports [20,21]. Furthermore, ROS1 rearrange-
ments in NSCLC have been associated with increased rates of extra-
thoracic [21], extrapulmonary [24] and brain metastases at the time of
diagnosis [21]. Consistent with this, 19.6 % of our patients showed
baseline intracranial lesions and 44.6 % had metastases to other ex-
trapulmonary organ.
A previous retrospective study reported deep vein thrombosis in
13.6 % NSCLC patients [24], while another study [25] showed that
peri-diagnostic venous and arterial thromboembolic events were
significantly more common in patients with advanced ROS1 rearranged
NSCLC (34.7 %) in comparison with EGFR-positive (13.7 %) and KRAS-
positive (18.4 %) patients. Furthermore, Lee et al. [26] reported two
cases of ROS1-rearranged NSCLC with recurrent VTE, which was likely
associated with factor V Leiden heterozygosity [26]. VTE occurred in 8
patients (14.3 %) in our study, of which 3 were asymptomatic and di-
agnosed with VTE by ultrasound imaging. Hence, we think some
asymptomatic thromboembolic events were possibly missed in our
study as most patients did not receive a screening test for throm-
boembolic events.
The mPFS following first-line crizotinib therapy was 23 months in
our cohort, of which 5 patients had an mPFS longer than 50 months.
This is considerably longer compared to the 19.3 months reported in
the phase I PROFILE 1001 trial [12], and 15.9 months in the phase II
trial on East Asian patients with ROS1 rearranged NSCLC [13]. In other
clinical studies, the mPFS varied from 10.0 months to 17.6 months
[14,15,21]. The longer PFS in our study can be attributed to various
factors. First, all patients in our study received first-line crizotinib
therapy whereas very few treatment-naïve patients were enrolled in the
previous studies [11]. Two recent retrospective studies evaluated the
first-line use of crizotinib [27,28], of which one reported a PFS of 18.4
months and another of 20.1 months in patients with CD74-ROS1 fusion.
Thus, long-term survival of patients with ROS1- rearranged NSCLC
might be achievable with first-line use of crizotinib. Secondly, high
proportion of patients with stage IIIB NSCLC (10/56,17.9 %) were en-
rolled in our study. Concurrent chemoradiation is a standard care for
stage IIIB NSCLC patients, however, several designed clinical trials and
real-world studies have been confirmed that some of them received
crizotinib treatment for a variety of reasons. It’s worth noting that the
mOS of stage IIIB NSCLC patients with crizotinib (57.7 months) in our
study was longer than that with concomitant radiochemotherapy (28.7
months) in the Pacific study [29]. Therefore, the first-line treatment of
locally advanced ROS1 rearranged NSCLC deserves further study.
Thirdly, systemic imaging was less frequent in our study compared to
controlled clinical trials, especially for patients with long-term benefits.

133
J. Zheng, et al. Lung Cancer 147 (2020) 130–136

Fig. 2. (A) Progression-free survival of first-line crizotinib treatment in patients with ROS1-rearranged Non-small cell Lung Cancer (NSCLC); (B) Overall survival of
first-line crizotinib treatment in patients with ROS1-rearranged Non-small cell Lung Cancer; (C) Comparison of progression-free survival in patients with different
numbers of baseline metastatic organs; (D) Comparison of overall survival in patients with different numbers of baseline metastatic organs. mFPS, median pro-
gression-free survival; mOS, median overall survival; NR, not reached.

Although all patient underwent chest/brain/bone scans at diagnosis for
accurate TNM staging, chest CT scan interval in these patients might
therefore be prolonged from 2 months to about 3 months after 2 years
of follow-up, and the brain or bone scans were only performed when
patients complained of symptoms. Finally, given the different efficacy
of crizotinib in intracranial versus non-intracranial lesions [30], prior
intracranial metastases may have affected the outcome. However, since
most patients with brain metastases in our study received SRS or WBRT,
their mPFS duration was similar to patients without brain metastases.
A subgroup of patients did not respond to crizotinib or progressed
after a short period of tumor control. The number of organs involved in
baseline metastases was identified as an independent predictor for PFS.
This is not unexpected since increased metastasis is indicative of a
higher tumor burden, which may be less sensitive to targeted therapy.
One case report showed immune checkpoint inhibitors were effective in
ROS1 rearranged NSCLC [31]. However, the CheckMate 370 study re-
ported unacceptable levels of hepatotoxicity in ALK-rearranged NSCLC
patients that received a combination of crizotinib and nivolumab.
Nevertheless, sequential targeted/chemo-therapy and immunotherapy
is worth exploring for NSCLC patients with ROS1 rearrangements.
The G2032R mutation in ROS1 kinase domain is associated with
crizotinib resistance [16,32] and was detected in 50 % of the resistant
patients in our study, which is similar to the previously reported in-
cidence [33]. In addition, we identified a novel L2026 V mutation in
one crizotinib-resistant patient but could not study it further due to lack
of comparison with the baseline tumor sample. In addition to mutations

134
J. Zheng, et al. Lung Cancer 147 (2020) 130–136

Fig. 3. The disease progression pattern for all patients received first-line crizotinib treatment.

in the ROS1 kinase domain, concurrent mutations were also detected in Table 3
our crizotinib-resistant cohort. From 1 patient, we detected activating Multivariable Prognostic Factors Analyses of mPFS with Crizotinib Treatment.
mutations of BRAF (D594 N) and ERBB2 (E717 K). And from 4 of 8 Prognostic factors Multivariate analysis
patients, gene mutations in cell cycle pathway key components were
detected, including CDKN2A (cyclin-dependent kinase inhibitor 2A), HR 95 % CI P-value
CDK4 (cyclin-dependent kinase 4), CCDN1 (G1/S-specific cyclin-D1),
Sex (male vs female) 0.383 0.088−1.658 0.199
which may activate cell cycle pathway and lead to drug resistance. Smoking history (smoker vs never smoker) 1.238 0.263−5.838 0.787
However, we couldn’t determine the precise underlying mechanism of Number of baseline metastatic organs (≤2 vs 4.762 1.515−14.961 0.008
these concurrent mutations due to the absence of paired baseline DNA > 2)
sequencing results. Further prospective studies should analyze the ge-
netic/molecular alterations in both baseline and progressed tumor tis- PFS, progression-free survival.
sues to determine the resistance mechanisms.
In conclusion, first-line crizotinib treatment is beneficial to Chinese Author contributions
patients with ROS1- rearranged advanced NSCLC. Involvement of more
than two metastatic organs was associated with a significantly shorter Jing Zheng, Jianya Zhou, and Jianying Zhou conceptualized the
PFS after crizotinib therapy and the G2032R mutation in ROS1 kinase study. Yuping Li, Chuangzhou Rao, Te Zhang, Jiayou Luo and Dongqing
domain was the most common mechanism of resistance to crizotinib. Lv followed-up the patients and collected the clinical data. Jing Zheng
and Jianya Zhou reviewed the tumor assessments and wrote the
manuscript. He Cao analyzed the data and Jianying Zhou critically

Table 2
Univriate Analysis of the Influence of Clinical Characteristics on mPFS/mOS Times with First-line Crizotinib Treatment in Patients with ROS1- rearranged NSCLC.
Characteristics PFS(m) 95 %CI p-value OS(m) 95 %CI p-value

Age,years: 0.777 0.350

≤65 years 20.0 6.8−33.1 60.0 NE
> 65 years 43.0 NE NE NE
Sex: 0.015 0.381
Male 44.0 12.1−20.3 60.0 NE
Female 12.0 8.9−15.1 50.0 NE
Smoking status: 0.086 0.658
Never 12.0 6.4−17.6 NE NE
Smoker 43.0 22.2−63.8 60.0 11.6−108.4
Histologic classification: 0.732 0.987
Adenocarcinoma 23.0 12.4−33.6 60.0 40.6−79.4
Adenosquamous carcinoma NE NE NE NE
Disease stage: 0.168 0.594
IIIB NE NE NE NE
IV or relapsed 15.0 4.2−25.8 60.0 45.5−74.5
ECOG score: 0.300 0.782
0 24.0 7.0−41.0 NE NE
1 or 2 15.0 4.6−25.4 60.0 40.4−79.6
baseline intracranial metastases: 0.462 0.533
Yes 12.0 10.1−13.9 50.0 4.6−95.4
No 24.0 1.0−47.0 60.0 NE
Number of baseline metastatic organs: < 0.001 < 0.001
≤2 24.0 6.1−48.7 60.0 NE
> 2 4.0 1.1−6.9 6.0 0−15.1
Thromboembolic events: 0.599 0.576
Yes 13.0 6.1−19.9 26.0 13.5−38.5
No 24.0 0−50.7 60.0 42.8−77.2

ECOG, Eastern Cooperative Oncology Group; NE, not evaluable; OS, overall survival; PFS, progression-free survival.

135
J. Zheng, et al.
reviewed the manuscript. All authors have read and approved the final
version of the manuscript for journal submission.
Funding
This work was supported by grants from the National Natural
Science Foundation of China (Grant No: 81972179 to Jianya Zhou), the
Zhejiang Provincial Natural Science Foundation (Grant No:
LGF19H160031 to Jing Zheng) and the Key Research and Development
Program of Zhejiang Province (grant number: 2019C03042 to Jianying
Zhou).
Declaration of Competing Interest
The authors declare no potential conflicts of interest.
Acknowledgement
Editorial assistance with the manuscript was provided by Content
Ed Net Co, Shanghai. The authors thank all the patients who partici-
pated in this study and their families.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.lungcan.2020.07.016.
References
[1] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2019, CA Cancer J. Clin. 69 (1)
(2019) 7–34.
[2] T.J. Lynch, D.W. Bell, R. Sordella, S. Gurubhagavatula, R.A. Okimoto,
B.W. Brannigan, P.L. Harris, S.M. Haserlat, J.G. Supko, F.G. Haluska, D.N. Louis,
D.C. Christiani, J. Settleman, D.A. Haber, Activating mutations in the epidermal
growth factor receptor underlying responsiveness of non-small-cell lung cancer to
gefitinib, N. Engl. J. Med. 350 (21) (2004) 2129–2139.
[3] B.J. Solomon, T. Mok, D.W. Kim, Y.L. Wu, K. Nakagawa, T. Mekhail, E. Felip,
F. Cappuzzo, J. Paolini, T. Usari, S. Iyer, A. Reisman, K.D. Wilner, J. Tursi,
F. Blackhall, P. Investigators, First-line crizotinib versus chemotherapy in ALK-po-
sitive lung cancer, N. Engl. J. Med. 371 (23) (2014) 2167–2177.
[4] K. Rikova, A. Guo, Q. Zeng, A. Possemato, J. Yu, H. Haack, J. Nardone, K. Lee,
C. Reeves, Y. Li, Y. Hu, Z. Tan, M. Stokes, L. Sullivan, J. Mitchell, R. Wetzel,
J. Macneill, J.M. Ren, J. Yuan, C.E. Bakalarski, J. Villen, J.M. Kornhauser, B. Smith,
D. Li, X. Zhou, S.P. Gygi, T.L. Gu, R.D. Polakiewicz, J. Rush, M.J. Comb, Global
survey of phosphotyrosine signaling identifies oncogenic kinases in lung cancer,
Cell 131 (6) (2007) 1190–1203.
[5] A. Charest, K. Lane, K. McMahon, J. Park, E. Preisinger, H. Conroy, D. Housman,
Fusion of FIG to the receptor tyrosine kinase ROS in a glioblastoma with an in-
terstitial del(6)(q21q21), Genes Chromosomes Cancer 37 (1) (2003) 58–71.
[6] J. Lee, S.E. Lee, S.Y. Kang, I.G. Do, S. Lee, S.Y. Ha, J. Cho, W.K. Kang, J. Jang,
S.H. Ou, K.M. Kim, Identification of ROS1 rearrangement in gastric adenocarci-
noma, Cancer 119 (9) (2013) 1627–1635.
[7] K.D. Davies, R.C. Doebele, Molecular pathways: ROS1 fusion proteins in cancer,
Clin. Cancer Res. 19 (15) (2013) 4040–4045.
[8] J.F. Gainor, A.T. Shaw, Novel targets in non-small cell lung cancer: ROS1 and RET
fusions, Oncologist 18 (7) (2013) 865–875.
[9] J. Acquaviva, R. Wong, A. Charest, The multifaceted roles of the receptor tyrosine
kinase ROS in development and cancer, Biochim. Biophys. Acta 1795 (1) (2009)
37–52.
[10] M.A. Davare, N.A. Vellore, J.P. Wagner, C.A. Eide, J.R. Goodman, A. Drilon,
M.W. Deininger, T. O’Hare, B.J. Druker, Structural insight into selectivity and re-
sistance profiles of ROS1 tyrosine kinase inhibitors, Proc. Natl. Acad. Sci. U. S. A.
112 (39) (2015) E5381–90.
[11] A.T. Shaw, S.H. Ou, Y.J. Bang, D.R. Camidge, B.J. Solomon, R. Salgia, G.J. Riely,
M. Varella-Garcia, G.I. Shapiro, D.B. Costa, R.C. Doebele, L.P. Le, Z. Zheng, W. Tan,
P. Stephenson, S.M. Shreeve, L.M. Tye, J.G. Christensen, K.D. Wilner, J.W. Clark,
A.J. Iafrate, Crizotinib in ROS1-rearranged non-small-cell lung cancer, N. Engl. J.
Med. 371 (21) (2014) 1963–1971.
[12] A.T. Shaw, G.J. Riely, Y.J. Bang, D.W. Kim, D.R. Camidge, B.J. Solomon, M. Varella-
Garcia, A.J. Iafrate, G.I. Shapiro, T. Usari, S.C. Wang, K.D. Wilner, J.W. Clark,
S.I. Ou, Crizotinib in ROS1-rearranged advanced non-small-cell lung cancer
(NSCLC): updated results, including overall survival, from PROFILE 1001, Ann.
Oncol. 30 (7) (2019) 1121–1126.
[13] Y.L. Wu, J.C. Yang, D.W. Kim, S. Lu, J. Zhou, T. Seto, J.J. Yang, N. Yamamoto,
136
Lung Cancer 147 (2020) 130–136
M.J. Ahn, T. Takahashi, T. Yamanaka, A. Kemner, D. Roychowdhury, J. Paolini,
T. Usari, K.D. Wilner, K. Goto, Phase II study of crizotinib in east asian patients with
ROS1-positive advanced non-small-cell lung cancer, J. Clin. Oncol. 36 (14) (2018)
1405–1411.
[14] Z.M. Li, L. Shen, D. Ding, J. Huang, J. Zhang, Z.W. Chen, S. Lu, Efficacy of crizotinib
among different types of ROS1 fusion partners in patients with ROS1-rearranged
non-small cell lung cancer, J. Thorac. Oncol. 13 (7) (2018) 987–995.
[15] Y.M. He, W. Sheng, W.G. Hu, J. Lin, J.J. Liu, B. Yu, X.R. Mao, L. Zhang, J. Huang,
G.S. Wang, Different types of ROS1 fusion partners yield comparable efficacy to
crizotinib, Oncol. Res. 27 (8) (2019) 901–910.
[16] A. Roys, X. Chang, Y. Liu, X.B. Xu, Y.L. Wu, D.Y. Zuo, Resistance mechanisms and
potent-targeted therapies of ROS1-positive lung cancer, Cancer Chemother. Pharm.
84 (4) (2019) 679–688.
[17] A.T. Shaw, B.J. Solomon, R. Chiari, G.J. Riely, B. Besse, R.A. Soo, S. Kao, C.C. Lin,
T.M. Bauer, J.S. Clancy, H. Thurm, J.F. Martini, G. Peltz, A. Abbattista, S. Li, S.I. Ou,
Lorlatinib in advanced ROS1-positive non-small-cell lung cancer: a multicentre,
open-label, single-arm, phase 1-2 trial, Lancet Oncol. 20 (12) (2019) 1691–1701.
[18] B.C. Cho, A.E. Drilon, R.C. Doebele, D.W. Kim, J.J. Lin, J. Lee, M.J. Ahn, V.W. Zhu,
S. Ejadi, D.R. Camidge, Y.J. Liu, S. Stopatschinskaja, J.J. Cui, D.M. Hyman,
S.H.I. Ou, A.T. Shaw, Safety and preliminary clinical activity of repotrectinib in
patients with advanced ROS1 fusion-positive non-small cell lung cancer (TRIDENT-
1 study), J. Clin. Oncol. 37 (15) (2019).
[19] T. Fukui, Y. Yatabe, Y. Kobayashi, K. Tomizawa, S. Ito, S. Hatooka, K. Matsuo,
T. Mitsudomi, Clinicoradiologic characteristics of patients with lung adenocarci-
noma harboring EML4-ALK fusion oncogene, Lung Cancer (Amsterdam,
Netherlands) 77 (2) (2012) 319–325.
[20] H. Go, D.W. Kim, D. Kim, B. Keam, T.M. Kim, S.H. Lee, D.S. Heo, Y.J. Bang,
D.H. Chung, Clinicopathologic analysis of ROS1-rearranged non-small-cell lung
cancer and proposal of a diagnostic algorithm, J. Thorac. Oncol. 8 (11) (2013)
1445–1450.
[21] S. Park, B.C. Ahn, S.W. Lim, J.M. Sun, H.R. Kim, M.H. Hong, S.H. Lee, J.S. Ahn,
K. Park, Y. La Choi, B.C. Cho, M.J. Ahn, Characteristics and outcome of ROS1-
positive non-small cell lung cancer patients in routine clinical practice, J. Thorac.
Oncol. 13 (9) (2018) 1373–1382.
[22] K. Bergethon, A.T. Shaw, S.H.I. Ou, R. Katayama, C.M. Lovly, N.T. McDonald,
P.P. Massion, C. Siwak-Tapp, A. Gonzalez, R. Fang, E.J. Mark, J.M. Batten,
H.Q. Chen, K.D. Wilner, E.L. Kwak, J.W. Clark, D.P. Carbone, H.B. Ji,
J.A. Engelman, M. Mino-Kenudson, W. Pao, A.J. Iafrate, ROS1 rearrangements
define a unique molecular class of lung cancers, J. Clin. Oncol. 30 (8) (2012)
863–870.
[23] Y.F. Chen, M.S. Hsieh, S.G. Wu, Y.L. Chang, J.Y. Shih, Y.N. Liu, M.F. Tsai, T.H. Tsai,
C.J. Yu, J.C. Yang, P.C. Yang, Clinical and the prognostic characteristics of lung
adenocarcinoma patients with ROS1 fusion in comparison with other driver mu-
tations in East Asian populations, J. Thoracic Oncol. 9 (8) (2014) 1171–1179.
[24] V. Tagalakis, D. Levi, J.S. Agulnik, V. Cohen, G. Kasymjanova, D. Small, High risk of
deep vein thrombosis in patients with non-small cell lung cancer: a cohort study of
493 patients, J. Thoracic Oncol. 2 (8) (2007) 729–734.
[25] T.L. Ng, D.E. Smith, R. Mushtaq, T. Patil, A. Dimou, S. Yang, Q. Liu, X. Li, C. Zhou,
R.T. Jones, M.M. Tu, F. Yan, I.A. Bowman, S.V. Liu, S. Newkirk, J. Bauml,
R.C. Doebele, D.L. Aisner, D. Gao, S. Ren, D.R. Camidge, ROS1 gene rearrangements
are associated with an elevated risk of peridiagnosis thromboembolic events, J.
Thoracic Oncol. 14 (4) (2019) 596–605.
[26] A. Lee, V.M. Howell, M. Itchins, H.R. Wheeler, N. Pavlakis, ROS1-rearranged non-
small-cell lung cancer, factor V leiden, and recurrent venous thromboses, Clin. Lung
Cancer 19 (5) (2018) 457–459.
[27] H. Xu, Q. Zhang, L. Liang, J. Li, Z. Liu, W. Li, L. Yang, G. Yang, F. Xu, J. Ying,
S. Zhang, Y. Wang, Crizotinib vs platinum-based chemotherapy as first-line treat-
ment for advanced non-small cell lung cancer with different ROS1 fusion variants,
Cancer Med. 9 (10) (2020) 3328–3336.
[28] L. Shen, T. Qiang, Z. Li, D. Ding, Y. Yu, S. Lu, First-line crizotinib versus platinum-
pemetrexed chemotherapy in patients with advanced ROS1-rearranged non-small-
cell lung cancer, Cancer Med. 9 (10) (2020) 3310–3318.
[29] S.J. Antonia, A. Villegas, D. Daniel, D. Vicente, S. Murakami, R. Hui, T. Kurata,
A. Chiappori, K.H. Lee, M. de Wit, B.C. Cho, M. Bourhaba, X. Quantin, T. Tokito,
T. Mekhail, D. Planchard, Y.C. Kim, C.S. Karapetis, S. Hiret, G. Ostoros, K. Kubota,
J.E. Gray, L. Paz-Ares, J. de Castro Carpeno, C. Faivre-Finn, M. Reck,
J. Vansteenkiste, D.R. Spigel, C. Wadsworth, G. Melillo, M. Taboada, P.A. Dennis,
M. Ozguroglu, P. Investigators, Overall survival with durvalumab after chemor-
adiotherapy in stage III NSCLC, N. Engl. J. Med. 379 (24) (2018) 2342–2350.
[30] X. Du, Y. Shao, H.F. Qin, Y.H. Tai, H.J. Gao, ALK-rearrangement in non-small-cell
lung cancer (NSCLC), Thorac. Cancer 9 (4) (2018) 423–430.
[31] A. Jain, N. Fujioka, M. Patel, Immune checkpoint inhibitors in ROS1-Rearranged
non-small cell lung cancer: a report of two cases, J. Thoracic Oncol. 14 (8) (2019)
e165–e167.
[32] M.M. Awad, R. Katayama, M. McTigue, W. Liu, Y.L. Deng, A. Brooun, L. Friboulet,
D. Huang, M.D. Falk, S. Timofeevski, K.D. Wilner, E.L. Lockerman, T.M. Khan,
S. Mahmood, J.F. Gainor, S.R. Digumarthy, J.R. Stone, M. Mino-Kenudson,
J.G. Christensen, A.J. Iafrate, J.A. Engelman, A.T. Shaw, Acquired resistance to
crizotinib from a mutation in CD74-ROS1, N. Engl. J. Med. 368 (25) (2013)
2395–2401.
[33] F. Facchinetti, G. Rossi, E. Bria, J.C. Soria, B. Besse, R. Minari, L. Friboulet,
M. Tiseo, Oncogene addiction in non-small cell lung cancer: focus on ROS1 in-
hibition, Cancer Treat. Rev. 55 (2017) 83–95.