Accepted Manuscript Title: Parallel screening for ALK, MET and ROS1 alterations in non-small cell lung cancer with implications for daily routine testing* Author: Philipp Jurmeister Dido Lenze Erika Berg Stefanie Mende Frank Sch¨aper Udo Kellner Hermann Herbst Christine Sers Jan Budczies Manfred Dietel Michael Hummel Maximilian von Laffert PII: DOI: Reference:
S0169-5002(14)00491-7 http://dx.doi.org/doi:10.1016/j.lungcan.2014.11.018 LUNG 4739
To appear in:
Lung Cancer
Received date: Revised date: Accepted date:
17-9-2014 8-11-2014 30-11-2014
Please cite this article as: Jurmeister P, Lenze D, Berg E, Mende S, Sch¨aper F, Kellner U, Herbst H, Sers C, Budczies J, Dietel M, Hummel M, von Laffert M, Parallel screening for ALK, MET and ROS1 alterations in non-small cell lung cancer with implications for daily routine testing*, Lung Cancer (2014), http://dx.doi.org/10.1016/j.lungcan.2014.11.018 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Parallel screening for ALK, MET and ROS1 alterations in non-small cell lung
2
cancer with implications for daily routine testing*
ip t
3
Philipp Jurmeistera, Dido Lenzea, Erika Berga, Stefanie Mendea, † Frank Schäper*b,
5
Udo Kellnerc, Hermann Herbstd, Christine Sersa, Jan Budcziesa, Manfred Dietela,
6
Michael Hummela and Maximilian von Lafferta
us
cr
4
7
* Dedicated to Frank Schäper (٭05.04.1963 - †17.01.2014)
an
8
M
9 10
a
11
Charitéplatz 1, 10117 Berlin, Germany
12
b
13
Lindenberger Weg 27, 13125 Berlin, Germany
14
c
15
32429 Minden, Germany
16
d
17
Berlin, Germany
te
d
Institute of Pathology, Campus Charité Mitte, Charité Universitätsmedizin Berlin,
Ac ce p
Pathology-Berlin, Bioptisches Institut Gemeinschaftspraxis für Pathologie,
Institute of Pathology, Johannes Wesling Klinikum Minden, Hans-Nolte-Straße 1,
Institute of Pathology, Vivantes Klinikum Berlin, Oranienburger Straße 285, 13437
18 19
Address of Correspondence
20
Maximilian von Laffert, MD
21
Institute of Pathology
22
Humboldt University Berlin, Campus Charité Mitte 1
Page 1 of 33
Charité Universitätsmedizin Berlin
24
Charitéplatz 1, 10117 Berlin, Germany
25
Tel.: +49 30 450 536116
26
Fax: +49 30 450 536943
27
E-mail: [email protected]
ip t
23
cr
28
30
us
29
Highlights
d
M
This is the first report on parallel FISH and IHC screening for ALK, MET and ROS1 Concordant results for ALK IHC and ALK FISH were seen in all cases MET FISH and IHC showed discordant results in 10.0% of cases MET overexpression can occur in ALK rearranged tumors Besides rearrangements, ROS1 amplifications can also occur in NSCLCs
te
32 33 34 35 36 37 38
an
31
Ac ce p
39
40
Abstract
41
Objectives: ALK, MET and ROS1 are prognostic and predictive markers in NSCLC,
42
which need to be implemented in daily routine. To evaluate different detection
43
approaches and scoring systems for optimal stratification of patients eligible for
44
mutation testing in the future, we screened a large and unselected cohort of NSCLCs
45
for all three alterations.
46
Material and Methods: Using tissue microarrays, 473 surgically resected NSCLCs
47
were tested for ALK and MET expression by IHC and genomic alterations in the ALK,
48
MET and ROS1 gene by FISH. For MET IHC, two different criteria (MetMab and H-
2
Page 2 of 33
score), for MET FISH, three different scoring systems (UCCC, Cappuzzo,
50
PathVysion) were investigated.
51
Results: ALK and ROS1 positivity was seen in 2.6% and 1.3% of all ADCs,
52
respectively, but not in pure SCCs. One ROS1 translocated tumor showed additional
53
ROS1 amplification. MET IHC+/FISH+ cases were found in both histological
54
subtypes (8.6% in all NSCLCs; 10.6% in ADCs; 5.0% in SCCs) and were associated
55
with pleural invasion, lymphatic vessel invasion and lymph node metastasis. MET
56
altered ADCs more frequently showed a papillary growth pattern. Whereas ALK
57
testing revealed homogenous results in IHC and FISH, we saw discordant results for
58
MET in about 10% of cases. Both MET-IHC scoring systems revealed almost
59
identical results. We did not encounter any combined FISH positivity for ALK, MET or
60
ROS1. However, three ALK positive cases harbored MET overexpression.
61
Conclusion: In daily routine, IHC could support FISH in the identification of ALK
62
altered NSCLCs. Further research is needed to assess the role of discordant MET
63
results by means of IHC and FISH as well as the relevance of tumors with an
64
increased ROS1 gene copy number.
cr
us
an
M
d
te
Ac ce p
65
ip t
49
3
Page 3 of 33
Introduction
66
In the last decade, the treatment modalities for non-small cell lung cancer (NSCLC)
67
have changed tremendously. In addition to surgery and standard chemotherapy,
68
personalized medicine such as tyrosine kinase inhibitors (TKIs) against the epidermal
69
growth factor receptor (EGFR) pose new therapeutic options with significantly
70
improved progression-free survival rates in a subset of patients [1]. Over the years,
71
additional targetable drivers have been identified in NSCLC such as the anaplastic
72
lymphoma kinase (ALK). It took only a few years from its discovery as a driver
73
mutation in NSCLC [2] to the approval of an ALK-TKI (crizotinib) in 2011. However,
74
crizotinib does not only target ALK but also MET and ROS1. This is supported by
75
recent reports that showed promising results in patients with MET amplification or
76
ROS1 translocation treated with crizotinib [3-6]. Additionally, MET was also shown to
77
be targetable by other inhibitors [7-9]. However, a clinical phase III trial was stopped
78
recently and the role of MET will need further evaluation [10]. Regardless, MET plays
79
an important role in the development of therapy resistance, especially in the context
80
of anti-EGFR-therapy [11]. Therefore, the identification of patients harboring
81
alterations in the ALK, ROS1 and still the MET gene seems of high interest.
82
In this study, we analyzed the frequency of ALK and ROS1 translocations as well as
83
MET amplifications in an unselected cohort of 473 surgically resected NSCLC
84
samples. We compared immunohistochemistry (IHC) and fluorescence in-situ
85
hybridization (FISH) as potential diagnostic assays. For MET FISH, we investigated
86
all three scoring systems so far described in literature [7,9,12-17]. Those were
87
compared among each other and correlated with two MET IHC scoring systems
88
(MetMAb and H-score). In the future, this could help to implement reliable and
89
efficient
Ac ce p
te
d
M
an
us
cr
ip t
65
screening
algorithms for
these
4
alterations
in routine
diagnostics.
Page 4 of 33
Materials and Methods
91
Patients and Samples
92
Tumor tissue of 473 NSCLC patients (289 men (61.1%) and 184 women (38.9%);
93
Table 1) who underwent pulmonary resection between 2000 and 2013 were obtained
94
from the archives of several Institutes of Pathology in Berlin (Charité University
95
Medicine Berlin, Pathology-Berlin Bioptisches Institut Gemeinschaftspraxis für
96
Pathologie, Vivantes Klinikum Berlin) and Minden (Johannes Wesling Klinikum
97
Minden). Detailed clinicopathological data was available for all patients, except for
98
two cases in which the staging was not documented. The median age was 67 years
99
(range: 29-85 years). Stage I was diagnosed in 261 (55.4%), stage II in 102 (21.7%),
100
stage III in 84 (17.8%) and stage IV in 24 (5.1%) cases. 335 (70.8%) tumors were
101
classified as adenocarcinomas (ADC). The predominant growth pattern was acinar in
102
144 (43.0%), solid in 114 (34.0%), papillary in 37 (11.0%), lepidic in 32 (9.6%) and
103
micropapillary in eight (2.4%) cases. 108 (22.8%) tumors were classified as
104
squamous cell carcinomas (SCC). The predominant growth pattern was non-
105
keratinizing in 61 (56.5%), keratinizing in 38 (35.2%) and basaloid in nine (8.3%)
106
cases. 13 (2.8%) samples were categorized as NSCLC not otherwise specified
107
(NOS), ten (2.1%) as large-cell carcinomas (LCC) and seven (1.5%) as
108
adenosquamous carcinomas (ADSC). Representative tumor areas were identified
109
and a total of ten tissue microarrays (TMA) were constructed using two cores for
110
each case (one mm diameter per core).
Ac ce p
te
d
M
an
us
cr
ip t
90
111 112
Immunohistochemistry (IHC)
113
IHC was performed as described previously on the VENTANA BenchMark XT 5
Page 5 of 33
automated slide stainer, using the anti-ALK-D5F3 Optiview kit [18] and the anti-total
115
c-MET-SP44 antibody (both Ventana Medical Systems, USA) [7]. ALK expression
116
was scored as negative (ALK IHC-) or positive (ALK IHC+); the latter was defined as
117
strong granular cytoplasmatic staining pattern [18]. MET expression was scored
118
according to the MetMAb trial criteria [7,19]: samples with no (IHC score 0) or weak
119
(IHC score 1+) membranous and/or cytoplasmatic staining in ≥50% of tumor cells
120
were considered MET IHC negative (MET IHC-). Tumors with moderate (IHC score
121
2+) or strong (IHC score 3+) membranous and/or cytoplasmatic staining in ≥50% of
122
tumor cells were considered MET IHC positive (MET IHC+). Furthermore, for MET
123
IHC we calculated a continuous H-score by multiplying the staining intensity (0: no
124
staining; 1: weak staining; 2: moderate staining; 3: strong staining) with the
125
respective percentage of tumor cells to reach a value between 0 and 300.
M
an
us
cr
ip t
114
d
126
Fluorescence in-situ hybridization (FISH)
128
FISH analysis was performed according to manufacturer’s instructions. Briefly, 4 µm
129
sections of TMA blocks were deparaffinized, dehydrated and then incubated in pre-
130
treatment solution (Dako, Denmark) for 10 minutes at 95-99°C. Samples were
131
immersed in pepsin solution for 3-6 minutes at 37°C. Following washing and
132
dehydration, slides were air dried and the respective DNA probe (ALK: Vysis LSI ALK
133
Dual Color, Abbott Molecular, USA; MET: MET/CEP7 dual-color probe, Abbott
134
Molecular, USA; ROS1: RF POSEIDON ROS1 Break probe, Kreatech, The
135
Netherlands; CEP6: CEP6 SpectrumAqua Probe, Abbott Molecular, USA) was
136
applied. The sections were sealed, denaturalized in humidified atmosphere at 82°C
137
for 5 minutes and then incubated overnight at 45°C to achieve hybridization. After
138
post-hybridization washing, slides were counterstained with 4’,6-diamidino-2-
Ac ce p
te
127
6
Page 6 of 33
139
phenylindole (DAPI).
140
Fluorescence in-situ hybridization (FISH) scoring
142
For each case, signals were enumerated in at least 50 non-overlapping tumor cells
143
using a fluorescence microscope (Axio Imager Z1, Zeiss, Germany). Computer-
144
based documentation and image analysis was performed with the ISIS imaging
145
system (MetaSystems, Germany).
146
ALK FISH positivity (ALK FISH+) was defined as the presence of split signals (SS)
147
and/or single red signals (SRS) in ≥15% of tumor cells [20-22].
148
For MET we used the three different scoring systems so far described in literature:
149
(1) UCCC (University of Colorado Cancer Center): cases with (A) MET/CEP7-ratio ≥2
150
or (B) small gene clusters (4-10 copies) or innumerable tight gene clusters in >10%
151
of tumor cells or (C) larger and brighter MET signals than CEP7 signals in >10% of
152
tumor cells or (D) >15 MET signals in >10% of tumor cells were considered as gene
153
amplification (GA). Specimens with ≥4 MET signals in ≥40% of tumor cells were
154
defined as high polysomy (HP). GA and HP were both considered to be MET FISH
155
positive (MET FISH+) [23,24]. However, as HP might be discussed as not being a
156
“true” amplification, we further discriminated between GA and HP.
157
(2) Cappuzzo: specimens with a mean gene copy number (GCN) of ≥5 MET signals
158
per tumor cell were classified as MET FISH+ [25].
159
(3) PathVysion: cases with a MET/CEP7-ratio ≥2 were defined as MET FISH+
160
[9,13,26].
161
For ROS1, tumors harboring SS and/or single green signals (SGS) in ≥15% of tumor
Ac ce p
te
d
M
an
us
cr
ip t
141
7
Page 7 of 33
162
cells were considered as ROS1 FISH positive (ROS1 FISH+) [27-30].
163
Statistical Analysis
165
Statistical analysis was performed using SPSS 21 software (IBM Corporation, USA).
166
Significance for cross tables larger than 2x2 was calculated using the statistical
167
language R. Comparison of clinicopathological data with IHC and FISH results was
168
performed using Fisher’s exact test (categorical data) and Mann-Whitney U test (age
169
and
size).
Ac ce p
te
d
M
an
tumor
us
cr
ip t
164
8
Page 8 of 33
170
Results
172
ALK IHC and ALK FISH
173
444 samples (93.9%) were evaluable by means of IHC and FISH. Concordant results
174
were seen in all 444 cases. A total of nine (2.0%) samples (2.6% of ADC) displayed
175
positivity for IHC and FISH (Figure 1). ALK rearrangements more frequently occurred
176
in younger patients with a median age of 54 years (range: 29-75 years; p=0.009). Six
177
(66.7%) of the ALK altered tumors showed lymphatic vessel invasion (p=0.023) and
178
lymph node metastasis (p=0.004). There was no significant trend (p=0.199) towards
179
advanced tumor stage (stage I: one case; stage II: three cases; stage III: two cases;
180
stage IV: one case). ALK IHC+/FISH+ tumors were mainly classified as ADC with
181
predominant acinar (five cases), solid (two cases) or papillary (one case) growth
182
pattern. One sample was classified as ADSC without keratinization in the SCC
183
component and with a solid growth pattern in the ADC component.
cr
us
an
M
d
te
Ac ce p
184
ip t
171
185
MET
186
MET IHC MetMAb criteria and MET FISH
187
440 samples (93.0%) were evaluable by means of IHC and FISH. Using the MetMAb
188
criteria for IHC and the UCCC scoring system including HP and GA for FISH,
189
homogenous results were seen in 396 cases (90.0%), with IHC-/FISH- in 358
190
(81.4%) and IHC+/FISH+ in 38 (8.6%) samples. 37 (8.4%) IHC+/FISH- and seven
191
(1.6%) IHC-/FISH+ cases were observed (Figure 2). Using the UCCC GA, Cappuzzo
9
Page 9 of 33
192
and PathVysion scoring system, the correlation was weaker with homogenous results
193
in in 383 (87,0%), 378 (85.9%) and 371 cases (84.3%), respectively (Table 3).
194
MET IHC H-score and MET FISH
196
The additionally applied MET IHC H-score showed a range between 0 and 280. A
197
good comparability of both tested IHC criteria was seen at an H-score threshold of
198
≥105, with discordant results in ten cases (2.2%; Supplemental Figure 1). Five of
199
these cases were classified as MetMAb positive and H-score negative, five as
200
MetMAb negative and H-score positive.
201
An H-score threshold of ≥180 encompassed 18 cases, all of them were considered
202
as MET FISH+ (Figure 4). Nine specimens were classified as HP and nine as GA.
203
These samples were also identified using the MetMAb criteria (IHC score 2+: 6
204
cases; IHC score 3+: 12 cases). However, this cut-off would not cover 29 (35.5%)
205
MET FISH+ samples.
cr
us
an
M
d
te
Ac ce p
206
ip t
195
207
Correlation of MET+ tumors (MetMab and UCCC) with clinicopathological data
208
MET IHC+/FISH+ samples were not associated with sex (25 men (65.8%), 13
209
women (34.2%); p=0.603) or age (median age: 67 years; range: 48-84 years;
210
p=0.694). Pleural invasion was present in in 47.2% of all IHC+/FISH+ cases
211
compared to 23.6% in tumors without MET IHC+/FISH+ results (p=0.006). MET
212
IHC+/FISH+ samples more frequently harbored lymphatic vessel invasion (60.5% vs.
213
41.0%; p=0.026) and lymph node metastasis (47.4% vs. 30.7%; p=0.048). There was
214
a non-significant trend towards advanced tumor stages (p=0.111) as alterations also
10
Page 10 of 33
occurred in early stages (stage I: 16 cases (6.8%); stage II: nine cases (9.9%); stage
216
III: eight cases (9.9%); stage IV: five cases (23.8%)). IHC+/FISH+ cases included 33
217
ADCs and five SCCs. MET alterations were more common in ADC (10.6%) than in
218
SCC (5.0%) and other NSCLC subtypes, although this difference was not significant
219
(p=0.051). We observed papillary (eleven cases), acinar (ten cases), solid (nine
220
cases) and lepidic (three cases) predominant growth patterns among the MET
221
altered ADCs with a significant trend towards papillary growth pattern (p=0.006). The
222
IHC+/FISH+ SCCs were classified as both keratinizing (one case) and non-
223
keratinizing (four cases; p=0.743).
224
Furthermore, we performed a separate analysis of the particular FISH scoring
225
systems with clinicopathological data (Supplementary Table 1). Pleural invasion and
226
advanced stage were significantly more common in the UCCC HP, UCCC GA and
227
Cappuzzo group. Blood vessel invasion was more frequent in the UCCC GA
228
(p=0.037) and Cappuzzo (p=0.005) group, while cases from the UCCC HP group
229
were significantly more likely to harbor lymphatic vessel invasion (p=0.007) and
230
lymph node metastasis (p=0.007). However, there were no significant correlations
231
with tumors that were considered positive according to the PathVysion criteria.
cr
us
an
M
d
te
Ac ce p
232
ip t
215
233
ROS1 FISH
234
ROS1 FISH was evaluable in 450 cases (94.5%). We discovered ROS1
235
rearrangements in four ADCs (0.9%), accounting for 1.3% of all ADCs. One ROS1
236
translocated tumor also showed an increased ROS1 GCN (5.4 copies per cell).
237
Another sample displayed 7.8 copies per cell but no ROS1 rearrangement (Figure 3).
11
Page 11 of 33
ROS1 translocations occurred in both men and women with a median age of 70
239
years (range 60-77 years). Three tumors showed lymphatic vessel invasion, lymph
240
node metastasis were present in two of them. Two cases were classified as stage I
241
and the other two samples as stage III. ROS1 rearranged tumors were all classified
242
as ADCs with predominantly solid (three cases) or papillary (one case) growth
243
pattern.
244
We reevaluated the two cases with an increased ROS1 GCN using an additional
245
CEP6 FISH probe. The ROS1 rearranged tumor showed amplification of the ROS1
246
gene locus with a ROS1/CEP6 ratio of 3.6, while the copy number gain in the second
247
sample was due to polysomy of chromosome 6 (Figure 3).
an
us
cr
ip t
238
M
248
Co-occurrence of ALK, MET and ROS1 alterations
250
Results for ALK, MET and ROS1 were available in 405 samples (85.6%) There was
251
no concomitant FISH positivity for any of the evaluated alterations. However, one
252
MET IHC+/FISH+ sample with gene amplification (MET/CEP7 ratio: 2.1; MetMAb
253
IHC score 2+; H-score 170) showed an increased ROS1 GCN due to polysomy of
254
chromosome 6 (Figure
255
overexpression (Figure 1). MET overexpression was more common in ALK altered
256
tumors (33% vs. 17.2%), although this was not significant (p=0.197).
257
One ALK IHC+ and MET IHC- case did not yield any evaluable FISH signals for ALK,
258
MET and ROS1, even though FISH was repeated several times, including
259
reevaluation of whole-slide tissue sections and different tumor blocks.
Ac ce p
te
d
249
3). Three ALK
IHC+/FISH+
cases showed
MET
260
12
Page 12 of 33
Discussion
261
Approximately 90% of all NSCLC diagnoses are performed on biopsy specimens
262
[31]. As the amount of tumor tissue available for molecular analysis is often limited
263
and the number of predictive targets increases steadily, the establishment of reliable
264
tests and adequate algorithms is of utmost importance. In this study we performed a
265
parallel screening for ALK, MET and ROS1 alterations and discuss the
266
consequences of our results for daily routine practice. ALK and ROS1
267
rearrangements were found in 2.6% and 1.3% of ADCs, respectively, but not in pure
268
SCCs. However, one ALK positive case showed adenosquamous histology. MET
269
IHC+/FISH+ results, on the other hand, were found in 8.6% of all NSCLCs and were
270
associated with pleural invasion. MET altered ADCs more frequently showed a
271
papillary growth pattern. Whereas ALK testing revealed reliable and concordant
272
results in IHC and FISH, the role of these two methods to identify MET amplifications
273
requires further investigation, as our data revealed discordant results in about 10.0%.
cr
us
an
M
d
te
Ac ce p
274
ip t
260
275
ALK alterations were described to be more frequent in younger patients [32-34]
276
which is consistent with the relatively low median age of 54 years in our cohort.
277
However, we encountered three ALK positive patients older than 60 years. Thus,
278
elderly patients should not generally be excluded from ALK testing [35]. For ALK, we
279
saw a clear correlation between IHC and FISH. All nine FISH+ cases were identified
280
by means of IHC and no false positive IHC results were encountered. Moreover, ALK
281
protein expression was helpful in two so-called “ALK FISH borderline” samples with
282
split signals in only 15-20% of tumor cells. These cases are particularly challenging in
283
routine diagnostics as they might be diagnosed as false negative in FISH [32,36,37].
284
Of note, we also encountered one ALK IHC+ case that was not evaluable by FISH. 13
Page 13 of 33
This case illustrates the urgent need for alternative methods to identify ALK
286
rearrangements besides FISH, as a therapy decision must be drawn under these
287
circumstances as well. In summary, our data support results of previous studies
288
[18,34,38] underlining the potential role of IHC as an alternative assay or at least as a
289
pre-screening tool to identify ALK rearrangements.
ip t
285
cr
290
Regarding MET, IHC+/FISH+ results were seen in 38 tumors (8.6%) when using the
292
MetMAb criteria for IHC and the UCCC scoring system for FISH. At a cut-off value of
293
≥105, the results from the IHC MetMAb criteria and the H-score were very similar
294
with homogenous results in 97.8% of cases. This is not surprising, as the definition of
295
a case with MetMAb IHC score 2+ (moderate staining in ≥50% of tumor cells)
296
corresponds to an H-score of at least 100. Furthermore, all tumors with an H-score
297
≥180 (n= 18) also harbored alterations of the MET gene in FISH. Although
298
statements on the predictive value of this finding is beyond the focus of this paper, an
299
IHC assay using an H-score with this threshold value could be examined as a
300
possible predictive test in future clinical trials.
an
M
d
te
Ac ce p
301
us
291
302
In accordance with literature, MET IHC+/FISH+ predominantly occurred in ADCs but
303
were also found in SCCs [17,39]. As already described by Nakamura et al. [40] we
304
observed that MET alterations were significantly more common in ADCs with
305
papillary growth pattern. Other authors described a significant correlation between
306
MET amplification and non-lepidic growth patterns [14,41]. Interestingly, activating
307
MET mutations also frequently occur in renal and thyroid carcinomas with papillary
308
growth [40]. Therefore, further research is needed to assess if MET might play an
14
Page 14 of 33
important role in the formation of papillary structures as proposed by Nakamura et al.
310
[40]. In accordance with previous studies we saw that MET alterations occurred more
311
frequently in tumors with pleural invasion, lymphatic vessel invasion and lymph node
312
metastasis [14,17]. Even though MET IHC+/FISH+ seem to be more common in
313
advanced tumors, they have also been described to occur in early stages [41].
314
Consistent with these reports, we observed MET alterations in 6.8% of all stage I
315
patients. Thus, from our data, it appears that MET testing should neither be limited to
316
a certain clinical subgroup nor to a certain histology.
317
Besides possible links with clinicopathological characteristics, the identification of the
318
most suitable method to screen for patients harboring MET amplifications is crucial
319
for routine diagnostics. The best correlation between IHC and FISH was achieved by
320
using the MetMAb criteria for IHC and the UCCC scoring system including HP for
321
FISH with concordant results in 90.0% compared to 85.9% and 84.3% when using
322
the Cappuzzo or PathVysion criteria, respectively (Table 3). The UCCC scoring
323
system was able to identify all tumors that were classified as positive according to
324
one of the other criteria. Therefore, the UCCC scoring system seems to be superior
325
to the Cappuzzo and PathVysion criteria in identifying IHC+ cases. Still, 44 samples
326
revealed contradictory results with 37 IHC+/FISH- and seven IHC-/FISH+ tumors.
327
These heterogeneous results might explain the huge differences between the
328
frequencies of MET overexpression (25-54%) and MET amplification (11.1-16.7%
329
when using the UCCC criteria) in earlier studies [7,8,12,24,39]. IHC-/FISH+ cases
330
have been reported before although they have never been described in detail [8,12].
331
It is unknown whether these tumors with a significantly increased MET GCN actually
332
do not overexpress MET or if the weak immunoreactivity in IHC was due to a specific
333
lack of antibody affinity or other mechanisms that might reduce staining intensity.
334
However, it should be noted that all MET IHC-/FISH+ specimens had been evaluable
Ac ce p
te
d
M
an
us
cr
ip t
309
15
Page 15 of 33
for other IHC markers such as TTF-1 or p63, stained in the process of the routine
336
diagnostics (data not shown). While the clinical outcome of patients with MET IHC-
337
/FISH+ tumors has not yet been investigated in detail, two recent studies showed
338
that MET IHC+/FISH- patients could also benefit from therapy with onartuzumab or
339
tivantinib [7,8]. Thus, there is a need to investigate (A) the underlying mechanisms
340
that lead to MET overexpression in FISH negative tumors; (B) potential reasons why
341
some FISH positive tumors show no or only weak protein expression in IHC; (C) the
342
clinical relevance of cases with conflicting results in IHC and FISH, especially of IHC-
343
/FISH+ patients.
an
us
cr
ip t
335
344
Recently, ROS1 has been identified as a treatable target in lung cancer [29,42,43]. In
346
our investigation, we saw an overall frequency of 0.9% (1.3% in ADCs) which is in
347
line with previous reports [29,42,43]. Recent trials and case reports discovered that
348
patients with ROS1 translocation in FISH could profit from anti-ROS1 therapy [4,5].
349
Regarding histology, we confirmed that ROS1 translocations mainly occur in ADCs
350
[42]. However, ROS1 might also be affected in SCCs and LCCs as previously
351
described [27,42,43]. Due to the low frequency of ROS1 alterations in our study, any
352
links with clinicopathological characteristics should be interpreted with care.
353
Furthermore, we encountered one case with polysomy of chromosome 6 and one
354
tumor with ROS1 amplification, the latter in addition to ROS1 translocation (Figure 3).
355
To our knowledge, “true” amplifications of the ROS1 gene locus have not been
356
described yet in NSCLC. Interestingly, tumors with an increased ROS1 GCN were
357
reported to show strong ROS1 expression in IHC [44]. These patterns need further
358
investigation as they might also benefit from targeted therapy.
Ac ce p
te
d
M
345
16
Page 16 of 33
359
Regarding concomitant alterations, we did not encounter any combined FISH+
361
results for ALK, MET or ROS1. For IHC, we observed three ALK rearranged tumors
362
showing MET overexpression. However, in contrast to recently published data
363
[17,45], MET IHC positivity was not significantly more common in ALK rearranged
364
tumors. This might be due to the lower number of ALK positive samples in our cohort.
365
Regardless, the clinical role of such cases is unknown and requires further
366
investigation (therapy resistance?).
367
While other publications showed that ROS1 alterations may overlap with KRAS,
368
EGFR and BRAF mutations [42,46], we observed no co-occurrence of ROS1
369
rearrangements with ALK or MET alterations although these results might not be
370
representative due to the low number of ROS1 translocated tumors in our study. One
371
case with an increased ROS1 GCN due to polysomy of chromosome 6 showed a
372
concomitant MET amplification (Figure 3). The clinical and biological role is unclear
373
and needs further evaluation.
cr
us
an
M
d
te
Ac ce p
374
ip t
360
375
In summary, our study is the first to report on parallel testing for ALK, MET and
376
ROS1 alterations in a large and non-preselected NSCLC-cohort. We may conclude
377
from our data that ALK and ROS1 screening should first of all be performed in ADCs
378
of the lung, whereas MET screening might be established as a further predictive
379
diagnostic option not only in ADCs but also in SCCs. In daily routine, IHC could
380
complement FISH in the diagnosis of ALK altered NSCLCs, especially with regards
381
to borderline cases or tumors without evaluable FISH signals. For MET, future
382
research is needed to evaluate the predictive value of IHC and FISH as well as the
17
Page 17 of 33
383
clinical role of patients with discordant results. Additionally, the clinical relevance of
384
tumors with increased ROS1 copy number due to polysomy of chromosome 6 or
385
amplification
the
ROS1
gene
locus
requires
further
evaluation.
Ac ce p
te
d
M
an
us
cr
ip t
of
18
Page 18 of 33
386
Acknowledgement
388
Part of this work was supported by PFIZER within the framework of the FALKE-
389
project (detection of ALK-rearrangements by FISH in NSCLC).
392
None
us
Conflict of interest
Ac ce p
te
d
M
an
391
cr
390
ip t
387
19
Page 19 of 33
393 394
References
395
[1]
Maemondo M, Inoue A, Kobayashi K, Sugawara S, Oizumi S, Isobe H, et al. Gefitinib or chemotherapy for non–small cell lung cancer with mutated EGFR.
397
N Engl J Med 2010;362:2380–8. [2]
Soda M, Choi YL, Enomoto M, Takada S, Yamashita Y, Ishikawa S, et al.
cr
398
ip t
396
Identification of the transforming EML4–ALK fusion gene in non-small-cell
400
lung cancer. Nature 2007;448:561–6.
401
[3]
us
399
Schwab R, Petak I, Kollar M, Pinter F, Varkondi E, Kohanka A, et al. Major partial response to crizotinib, a dual MET/ALK inhibitor, in a squamous cell
403
lung (SCC) carcinoma patient with de novo c-MET amplification in the
404
absence of ALK rearrangement. Lung Cancer 2013;83:109–11. [4]
Ou S-HI, Camidge R, Riely GJ, Salgia R, Shapiro G, Clark JW, et al. Efficacy
M
405
an
402
406
and safety of crizotinib in patients with advanced ROS1-rearranged non-small
407
cell lung cancer (NSCLC). Ann Oncol 2013;24:ix43–3. [5]
Bos M, Gardizi M, Schildhaus HU, Heukamp LC, Geist T, Kaminsky B, et al.
d
408
Complete metabolic response in a patient with repeatedly relapsed non-small
410
cell lung cancer harboring ROS1 gene rearrangement after treatment with
411
crizotinib. Lung Cancer 2013;81:142–3. [6]
and anaplastic lymphoma kinase (ALK) inhibitor, in a non-small cell lung
414
cancer patient with de novo MET amplification. J Thorac Oncol 2011;6:942–6.
415
[7]
Spigel DR, Ervin TJ, Ramlau RA, Daniel DB, Goldschmidt JH, Blumenschein GR, et al. Randomized Phase II Trial of Onartuzumab in Combination With
417
Erlotinib in Patients With Advanced Non-Small-Cell Lung Cancer. J Clin
418
Oncol 2013;31:4105–14.
419 420
Ou S-HI, Kwak EL, Siwak-Tapp C, Dy J, Bergethon K, Clark JW, et al. Activity of crizotinib (PF02341066), a dual mesenchymal-epithelial transition (MET)
413
416
Ac ce p
412
te
409
[8]
Rodig SJ, Sequist LV, Schiller JH, Chen Y, Halim A, Waghorne C, et al.
421
Abstract 1729: An exploratory biomarker analysis evaluating the effect of the
422
c-MET inhibitor tivantinib (ARQ 197) and erlotinib in NSCLC patients in a
423
randomized, double-blinded phase 2 study. Cancer Res 2012;72:1729.
424
[9]
Sequist LV, Sequist LV, Pawel von J, Pawel von J, Garmey EG, Garmey EG, 20
Page 20 of 33
425
et al. Randomized Phase II Study of Erlotinib Plus Tivantinib Versus Erlotinib
426
Plus Placebo in Previously Treated Non-Small-Cell Lung Cancer. J Clin
427
Oncol 2011;29:3307–15. [10]
come and gone? Clin Cancer Res 2014;20:4422–4.
429 430
Hirsch FR, Bunn PA, Herbst RS. “Companion diagnostics”: has their time
[11]
Bean J, Brennan C, Shih J-Y, Riely G, Viale A, Wang L, et al. MET
ip t
428
amplification occurs with or without T790M mutations in EGFR mutant lung
432
tumors with acquired resistance to gefitinib or erlotinib. Proc Natl Acad Sci U
433
S A 2007;104:20932–7. [12]
Dziadziuszko R, Wynes MW, Singh S, Asuncion BR, Ranger-Moore J,
us
434
cr
431
Konopa K, et al. Correlation between MET gene copy number by silver in situ
436
hybridization and protein expression by immunohistochemistry in non-small
437
cell lung cancer. J Thorac Oncol 2012;7:340–7.
438
[13]
an
435
Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski MF, Pao W, et al. Analysis of tumor specimens at the time of acquired resistance to EGFR-TKI
440
therapy in 155 patients with EGFR-mutant lung cancers. Clin Cancer Res
441
2013;19:2240–7. [14]
Tachibana K, Minami Y, Shiba-Ishii A, Kano J, Nakazato Y, Sato Y, et al.
d
442
M
439
Abnormality of the hepatocyte growth factor/MET pathway in pulmonary
444
adenocarcinogenesis. Lung Cancer 2012;75:181–8.
446 447 448 449 450 451 452
[15]
MET Increased Gene Copy Number and Primary Resistance to Gefitinib Therapy in Non-Small Cell Lung Cancer Patients. Ann Oncol 2008;20:298– 304.
[16]
Minuti G, D'Incecco A, Cappuzzo F. Targeted therapy for NSCLC with driver mutations. Expert Opin Biol Ther 2013;13:1401–12.
[17]
Tsuta K, Kozu Y, Mimae T, Yoshida A, Kohno T, Sekine I, et al. cMET/phospho-MET protein expression and MET gene copy number in nonsmall cell lung carcinomas. J Thorac Oncol 2012;7:331–9.
453 454
Cappuzzo F, Janne PA, Skokan M, Finocchiaro G, Rossi E, Ligorio C, et al.
Ac ce p
445
te
443
[18]
Nitta H, Tsuta K, Yoshida A, Ho SN, Kelly BD, Murata LB, et al. New methods
455
for ALK status diagnosis in non-small-cell lung cancer: an improved ALK
456
immunohistochemical assay and a new, Brightfield, dual ALK IHC-in situ
457
hybridization assay. J Thorac Oncol 2013;8:1019–31.
458
[19]
Spigel DR, Ervin TJ, Ramlau R, Daniel DB. Final efficacy results from 21
Page 21 of 33
459
OAM4558g, a randomized phase II study evaluating MetMAb or placebo in
460
combination with erlotinib in advanced NSCLC. J Clin Oncol 2011.
461
[20]
McLeer-Florin AA, Moro-Sibilot DD, Melis AA, Salameire DD, Lefebvre CC, Ceccaldi FF, et al. Dual IHC and FISH testing for ALK gene rearrangement in
463
lung adenocarcinomas in a routine practice: a French study. J Thorac Oncol
464
2012;7:348–54.
465
[21]
ip t
462
Thunnissen E, Bubendorf L, Dietel M, Elmberger G, Kerr K, López-Ríos F, et al. EML4-ALK testing in non-small cell carcinomas of the lung: a review with
467
recommendations. Virchows Arch 2012;461:245–57. [22]
Yi ES, Boland JM, Maleszewski JJ, Roden AC, Oliveira AM, Aubry M-C, et al.
us
468
cr
466
Correlation of IHC and FISH for ALK Gene Rearrangement in Non-small Cell
470
Lung Carcinoma IHC Score Algorithm for FISH. J Thorac Oncol 2011;6:459–
471
65.
472
[23]
an
469
Varella-Garcia M. Stratification of non-small cell lung cancer patients for therapy with epidermal growth factor receptor inhibitors: the EGFR
474
fluorescence in situ hybridization assay. Diagn Pathol 2006;1:19.
475
[24]
M
473
Go H, Jeon YK, Park HJ, Sung S-W, Seo J-W, Chung DH. High MET gene copy number leads to shorter survival in patients with non-small cell lung
477
cancer. J Thorac Oncol 2010;5:305–13.
480 481 482 483 484 485 486
te
479
[25]
Cappuzzo F, Marchetti A, Skokan M, Rossi E, Gajapathy S, Felicioni L, et al. Increased MET Gene Copy Number Negatively Affects Survival of Surgically
Ac ce p
478
d
476
Resected Non-Small Cell Lung Cancer Patients. J Clin Oncol 2009;27:1667– 74.
[26]
Tanaka A, Tanaka A, Sueoka-Aragane N, Sueoka-Aragane N, Nakamura T, Nakamura T, et al. Co-existence of positive MET FISH status with EGFR mutations signifies poor prognosis in lung adenocarcinoma patients. Lung Cancer 2012;75:89–94.
[27]
Davies KD, Davies KD, Le AT, Le AT, Theodoro MF, Theodoro MF, et al.
487
Identifying and targeting ROS1 gene fusions in non-small cell lung cancer.
488
Clin Cancer Res 2012;18:4570–9.
489
[28]
Kim MH, Shim HS, Kang DR, Jung JY, Lee CY, Kim DJ, et al. Clinical and
490
prognostic implications of ALK and ROS1 rearrangements in never-smokers
491
with surgically resected lung adenocarcinoma. Lung Cancer 2014;83:1–7.
492
[29]
Bergethon K, Bergethon K, Shaw AT, Shaw AT, Ignatius Ou SH, Ignatius Ou 22
Page 22 of 33
493
SH, et al. ROS1 Rearrangements Define a Unique Molecular Class of Lung
494
Cancers. J Clin Oncol 2012;30:863–70.
495
[30]
Yoshida A, Tsuta K, Wakai S, Arai Y, Asamura H, Shibata T, et al.
496
Immunohistochemical detection of ROS1 is useful for identifying ROS1
497
rearrangements inlung cancers. Mod Pathol 2013:1–10. Histopathology 2009;54:12–27.
499 500
Kerr KM. Pulmonary adenocarcinomas: classification and reporting.
ip t
[31] [32]
Camidge DR, Theodoro M, Maxson DA, Skokan M, O'Brien T, Lu X, et al.
cr
498
Correlations between the percentage of tumor cells showing an ALK gene
502
rearrangement, ALK signal copy number, and response to crizotinib therapy
503
in ALK FISH positive non-small cell lung cancer. Cancer Res 2012;118:4486–
504
94. [33]
an
505
us
501
Camidge DR, Hirsch FRF, Varella-Garcia MM, Franklin WAW. Finding ALKpositive lung cancer: what are we really looking for? J Thorac Oncol
507
2011;6:411–3.
508
[34]
M
506
Sholl LM, Weremowicz S, Gray SW, Wong K-K, Chirieac LR, Lindeman NI, et al. Combined use of ALK immunohistochemistry and FISH for optimal
510
detection of ALK-rearranged lung adenocarcinomas. J Thorac Oncol
511
2013;8:322–8.
514 515 516 517 518 519 520 521
te
513
[35]
Lindeman NI, Cagle PT, Beasley MB, Chitale DA, Dacic S, Giaccone G, et al. Molecular testing guideline for selection of lung cancer patients for EGFR and
Ac ce p
512
d
509
ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. J Thorac Oncol 2013;8:823–59.
[36]
V Laffert M, Warth A, Penzel R, Schirmacher P, Jonigk D, Kreipe H, et al. Anaplastic lymphoma kinase (ALK) gene rearrangement in non-small cell lung cancer (NSCLC): results of a multi-centre ALK-testing. Lung Cancer 2013;81:200–6.
[37]
V Laffert M, Warth A, Penzel R, Schirmacher P, Kerr K, Elmberger G, et al.
522
Anaplastic lymphoma kinase (ALK)-detection in non-small cell lung cancer:
523
results of the first European IHC-based (D5F3-Optiview) panel test within 16
524
institutes. J Thorac Oncol 2013;8:S484.
525 526
[38]
Savic S, Bode B, Diebold J, Tosoni I, Barascud A, Baschiera B, et al. Detection of ALK-Positive Non-Small-Cell Lung Cancers on Cytological 23
Page 23 of 33
527
Specimens: High Accuracy of Immunocytochemistry with the 5A4 Clone. J
528
Thorac Oncol 2013;8:1004–11.
529
[39]
Park S, Choi Y-L, Sung CO, An J, Seo J, Ahn M-J, et al. High MET copy
530
number and MET overexpression: poor outcome in non-small cell lung cancer
531
patients. Histol Histopathol 2012;27:197–207. [40]
Nakamura Y, Niki T, Goto A, Morikawa T, Miyazawa K, Nakajima J, et al. c-
ip t
532
Met activation in lung adenocarcinoma tissues: an immunohistochemical
534
analysis. Cancer Sci 2007;98:1006–13.
535
[41]
cr
533
Jin Y, Sun P-L, Kim H, Seo AN, Jheon S, Lee C-T, et al. MET Gene Copy Number Gain is an Independent Poor Prognostic Marker in Korean Stage I
537
Lung Adenocarcinomas. Ann Surg Oncol 2013. [42]
Warth A, Muley T, Dienemann H, Goeppert B, Stenzinger A, Schnabel PA, et
an
538
us
536
539
al. ROS1 Expression and Translocations in Non-Small Cell Lung Cancer:
540
Clinicopathological Analysis of 1478 Cases. Histopathology 2014. [43]
Rimkunas VM, Crosby KE, Li D, Hu Y, Kelly ME, Gu TL, et al. Analysis of
M
541
Receptor Tyrosine Kinase ROS1-Positive Tumors in Non-Small Cell Lung
543
Cancer: Identification of a FIG-ROS1 Fusion. Clin Cancer Res 2012;18:4449–
544
57. [44]
Lee HJ, Seol HS, Kim JY, Chun S-M, Suh Y-A, Park Y-S, et al. ROS1
te
545
d
542
Receptor Tyrosine Kinase, a Druggable Target, is Frequently Overexpressed
547
in Non-Small Cell Lung Carcinomas Via Genetic and Epigenetic Mechanisms.
548 549 550 551 552 553 554
Ac ce p
546
Ann Surg Oncol 2012;20:200–8.
[45]
Feng Y, Minca EC, Lanigan C, Liu A, Zhang W, Yin L, et al. High MET receptor expression but not gene amplification in ALK 2p23 rearrangement positive non-small-cell lung cancer. J Thorac Oncol 2014;9:646–53.
[46]
Stumpfova M, Janne PA. Zeroing in on ROS1 Rearrangements in Non-Small Cell Lung Cancer. Clin Cancer Res 2012;18:4222–4.
555
24
Page 24 of 33
Titles and legends to figures
562
Figure 1 Example of an adenocarcinoma (a; hematoxylin and eosin stain) with
563
positive results in ALK IHC (b) and ALK FISH (c; red: 3’ ALK probe, green: 5’ ALK
564
probe). Additionally, this tumor showed MET overexpression (d; MetMAb IHC score
565
2+; H-score 150) but no amplification in MET FISH (e; red: MET gene locus, cyan:
566
centromere 7).
cr
ip t
561
us
567
Figure 2 Examples of tumors with equivocal and discordant results in MET IHC and
569
FISH: cases with homogenous results with IHC-/FISH- in case 1 (a-c) and
570
IHC+/FISH+ in case 2 (d-f). Cases 3 (g-I) showed MET overexpression but revealed
571
no amplification in FISH while case 4 (j-l) showed a significant MET copy number
572
gain but lacked any MET expression. Pictures a, d and h show hematoxylin and
573
eosin stains of representative tumor areas for each case.
M
d
te Ac ce p
574
an
568
575
Figure 3 Case 1 (a-c; green: 3’ ROS1 probe, red: 5’ ROS1 probe) showed ROS1
576
translocation. Picture c shows magnification of representative tumor cells from the
577
same case. Case 2 (d-g; cyan: centromere 6) revealed amplification and
578
translocation of the ROS1 gene locus with a gene-to-chromosome ratio of 3.6 and at
579
least one split signal or single green signal (e and f, white arrows indicate rearranged
580
signals) in 30% of tumor cells. Case 3 (h-k) showed ROS1 copy number gain due to
581
polysomy of chromosome 6 and had an additional MET amplification (j; red: MET
582
gene locus, cyan: centromere 7; MET/CEP7 ratio: 2.1) with MET overexpression in
583
IHC (data not shown). Pictures a, d and h show hematoxylin and eosin stains of
584
representative tumor areas for each case. 29
Page 25 of 33
585
Figure 4 Plot of MET IHC H-score and MET FISH results. The black line indicates a
587
threshold value of ≥180. All cases with an H-score ≥180 also revealed MET alteration
588
in FISH. For reasons of clarity, cases with an H-score of 0 are only shown if they
589
were classified as FISH+. The total number of cases with an H-score ≥180 was 18
590
and they were all classified as 2+ (6 cases) or 3+ (12 cases) according to the
591
MetMAb criteria.
Ac ce p
te
d
M
an
us
cr
ip t
586
30
Page 26 of 33
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 27 of 33
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 28 of 33
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 29 of 33
300 UCCC high polysomy UCCC gene amplification UCCC negative
250
≥180
us
cr
ip
t
200
100
ce pt
ed
M
an
150
Ac
H-Score
Figure 4
50 0
Page 30 of 33
Table 1
Ac ce pt e
d
M
an
us
cr
ip t
Table 1 Correlation between MET amplification, ALK and ROS1 translocation and clinicopathological characteristics. a MET FISH+/IHC+ ALK FISH+/IHC+ ROS1 FISH+ All cases Positive p-value Positive p-value Positive p-value Total 473 38 (8.6%) 9 (1.9%) 4 (0.9%) Sex Male 289 25 5 2 0.603 0.739 0.643 Female 184 13 4 2 Age (years) Median (standard deviation) 67 ± 10 66 ± 8.7 54 ± 14.2 70 ± 7.6 0.694 0.009 0.386 Range 29 – 85 48 – 84 29 – 75 60 – 77 Stage I-II 363 25 4 2 0.111 0.199 0.220 III-IV 108 13 3 2 Lymph-node metastasis Absent 324 20 1 2 0.048 0.004 0.588 Present 147 18 6 2 Lymphatic vessel invasion Absent 276 15 1 1 0.026 0.023 0.308 Present 195 23 6 3 Blood vessel invasion Absent 369 26 4 2 0.158 0.195 0.202 Present 102 12 3 2 Pleural invasion Absent 352 19 5 3 0.006 1.000 1.000 Present 119 17 2 1 Tumor size (centimeter) Median (standard deviation) 3.0 ± 2.6 3.5 ± 2.0 0.960 2.3 ± 3.4 0.505 2,4 ± 0.8 0.321 Range 0.8 – 17.0 1.0 – 9.0 1.2 – 9.0 1.9 – 3.7 Histology ADC 335 33 8 4 SCC 108 5 0.051 0 0.182 0 0.673 b Others 30 0 1 0 ADC predominant growth pattern ADC solid 114 9 2 3 ADC acinar 144 10 5 0 ADC papillary 37 11 0.006 1 0.816 1 0.238 ADC lepidic 32 3 0 0 ADC micropapillary 8 0 0 0 SCC growth pattern Keratinizing 38 1 0 0 Non-Keratinizing 61 4 0.743 0 0 Basaloid 9 0 0 0 ALK, anaplastic lymphoma kinase; MET, mesenchymal-epithelial transition factor; ROS1, repressor of silencing 1; IHC, immunohistochemistry; FISH, fluorescence in-situ hybridization; ADC, adenocarcinoma; SCC, squamous cell carcinoma a Using the University of Colorado Cancer center (UCCC) scoring system for FISH and the MetMAb trial criteria for IHC b 13 non-small cell lung cancers not otherwise specified, ten large cell carcinomas, seven adenosquamous carcinomas
Page 31 of 33
Table 2
Ac ce pt e
d
M
an
us
cr
ip t
Table 2 Molecular pathological characteristics of patients harboring ALK or ROS1 translocation. Case # Sex Age Histology, FISH pattern ALK IHC MET MET IHC predominant (% of altered FISH MetMAb cells) criteria ALK 1 M 62 ADC, solid SS, SRS (69%) Positive Negative Positive (2+) ALK 2 M 69 ADC, acinar SRS (70%) Positive Negative Negative (0) ALK 3 F 39 ADSC, solid SS, SRS (18%) Positive Negative Negative (0) ALK 4 F 50 ADC, solid SS (20%) Positive Negative Negative (0) ALK 5 M 75 ADC, acinar SS (40%) Positive Negative Negative (1+) ALK 6 M 54 ADC, acinar SS (40%) Positive Negative Negative (0) ALK 7 M 41 ADC, papillary SS, SRS (62%) Positive Negative Positive (2+) ALK 8 F 58 ADC, acinar SS (70%) Positive Negative Negative (0) ALK 9 F 29 ADC, acinar SS, SRS (70%) Positive Negative Positive (2+) ROS1 M 60 ADC, acinar SS (71%) Negative Negative Negative (1+) 1 ROS1 F 77 ADC, solid SS (40%) Negative Negative Negative (1+) 2 ROS1 M 67 ADC, solid SS, SGS Negative Negative Negative (0) 3 (30%), A ROS1 F 74 ADC, solid SS, SGS (90%) Negative Negative Negative (1+) 4 ALK, anaplastic lymphoma kinase, ROS1, repressor of silencing 1; MET, mesenchymal-epithelial transition factor; FISH, fluorescence in-situ hybridization; IHC, immunohistochemistry; M, male; F, female; ADC, adenocarcinoma; ADSC, adenosquamous carcinoma; SS, split signals; SRS, single red signals; SGS, single green signals; A, amplification
Page 32 of 33
Table 3
Ac ce pt e
d
M
an
us
cr
ip t
Table 3 Correlation of different MET FISH amplification scoring systems and MET IHC scores. MET FISH MET IHC negative MET IHC positive All MET IHC 0 MET IHC 1+ MET IHC 2+ MET IHC 3+ a UCCC negative 395 250 108 37 0 a UCCC positive 45 5 2 26 12 UCCC HP negative 417 251 109 50 7 UCCC HP positive 23 4 1 13 5 UCCC GA negative 418 254 109 50 5 UCCC GA positive 22 1 13 7 Cappuzzo negative 423 254 109 54 6 Cappuzzo positive 17 1 1 9 6 PathVysion negative 436 255 110 59 10 PathVysion positive 6 0 0 4 2 MET, mesenchymal-epithelial transition; FISH, Fluorescence in-situ hybridization; IHC, immunohistochemistry; UCCC, University of Colorado Cancer Center; HP, high polysomy; GA, gene amplification a : Including high polysomy and gene amplification
Page 33 of 33
S0169-5002(14)00491-7 http://dx.doi.org/doi:10.1016/j.lungcan.2014.11.018 LUNG 4739
To appear in:
Lung Cancer
Received date: Revised date: Accepted date:
17-9-2014 8-11-2014 30-11-2014
Please cite this article as: Jurmeister P, Lenze D, Berg E, Mende S, Sch¨aper F, Kellner U, Herbst H, Sers C, Budczies J, Dietel M, Hummel M, von Laffert M, Parallel screening for ALK, MET and ROS1 alterations in non-small cell lung cancer with implications for daily routine testing*, Lung Cancer (2014), http://dx.doi.org/10.1016/j.lungcan.2014.11.018 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Parallel screening for ALK, MET and ROS1 alterations in non-small cell lung
2
cancer with implications for daily routine testing*
ip t
3
Philipp Jurmeistera, Dido Lenzea, Erika Berga, Stefanie Mendea, † Frank Schäper*b,
5
Udo Kellnerc, Hermann Herbstd, Christine Sersa, Jan Budcziesa, Manfred Dietela,
6
Michael Hummela and Maximilian von Lafferta
us
cr
4
7
* Dedicated to Frank Schäper (٭05.04.1963 - †17.01.2014)
an
8
M
9 10
a
11
Charitéplatz 1, 10117 Berlin, Germany
12
b
13
Lindenberger Weg 27, 13125 Berlin, Germany
14
c
15
32429 Minden, Germany
16
d
17
Berlin, Germany
te
d
Institute of Pathology, Campus Charité Mitte, Charité Universitätsmedizin Berlin,
Ac ce p
Pathology-Berlin, Bioptisches Institut Gemeinschaftspraxis für Pathologie,
Institute of Pathology, Johannes Wesling Klinikum Minden, Hans-Nolte-Straße 1,
Institute of Pathology, Vivantes Klinikum Berlin, Oranienburger Straße 285, 13437
18 19
Address of Correspondence
20
Maximilian von Laffert, MD
21
Institute of Pathology
22
Humboldt University Berlin, Campus Charité Mitte 1
Page 1 of 33
Charité Universitätsmedizin Berlin
24
Charitéplatz 1, 10117 Berlin, Germany
25
Tel.: +49 30 450 536116
26
Fax: +49 30 450 536943
27
E-mail: [email protected]
ip t
23
cr
28
30
us
29
Highlights
d
M
This is the first report on parallel FISH and IHC screening for ALK, MET and ROS1 Concordant results for ALK IHC and ALK FISH were seen in all cases MET FISH and IHC showed discordant results in 10.0% of cases MET overexpression can occur in ALK rearranged tumors Besides rearrangements, ROS1 amplifications can also occur in NSCLCs
te
32 33 34 35 36 37 38
an
31
Ac ce p
39
40
Abstract
41
Objectives: ALK, MET and ROS1 are prognostic and predictive markers in NSCLC,
42
which need to be implemented in daily routine. To evaluate different detection
43
approaches and scoring systems for optimal stratification of patients eligible for
44
mutation testing in the future, we screened a large and unselected cohort of NSCLCs
45
for all three alterations.
46
Material and Methods: Using tissue microarrays, 473 surgically resected NSCLCs
47
were tested for ALK and MET expression by IHC and genomic alterations in the ALK,
48
MET and ROS1 gene by FISH. For MET IHC, two different criteria (MetMab and H-
2
Page 2 of 33
score), for MET FISH, three different scoring systems (UCCC, Cappuzzo,
50
PathVysion) were investigated.
51
Results: ALK and ROS1 positivity was seen in 2.6% and 1.3% of all ADCs,
52
respectively, but not in pure SCCs. One ROS1 translocated tumor showed additional
53
ROS1 amplification. MET IHC+/FISH+ cases were found in both histological
54
subtypes (8.6% in all NSCLCs; 10.6% in ADCs; 5.0% in SCCs) and were associated
55
with pleural invasion, lymphatic vessel invasion and lymph node metastasis. MET
56
altered ADCs more frequently showed a papillary growth pattern. Whereas ALK
57
testing revealed homogenous results in IHC and FISH, we saw discordant results for
58
MET in about 10% of cases. Both MET-IHC scoring systems revealed almost
59
identical results. We did not encounter any combined FISH positivity for ALK, MET or
60
ROS1. However, three ALK positive cases harbored MET overexpression.
61
Conclusion: In daily routine, IHC could support FISH in the identification of ALK
62
altered NSCLCs. Further research is needed to assess the role of discordant MET
63
results by means of IHC and FISH as well as the relevance of tumors with an
64
increased ROS1 gene copy number.
cr
us
an
M
d
te
Ac ce p
65
ip t
49
3
Page 3 of 33
Introduction
66
In the last decade, the treatment modalities for non-small cell lung cancer (NSCLC)
67
have changed tremendously. In addition to surgery and standard chemotherapy,
68
personalized medicine such as tyrosine kinase inhibitors (TKIs) against the epidermal
69
growth factor receptor (EGFR) pose new therapeutic options with significantly
70
improved progression-free survival rates in a subset of patients [1]. Over the years,
71
additional targetable drivers have been identified in NSCLC such as the anaplastic
72
lymphoma kinase (ALK). It took only a few years from its discovery as a driver
73
mutation in NSCLC [2] to the approval of an ALK-TKI (crizotinib) in 2011. However,
74
crizotinib does not only target ALK but also MET and ROS1. This is supported by
75
recent reports that showed promising results in patients with MET amplification or
76
ROS1 translocation treated with crizotinib [3-6]. Additionally, MET was also shown to
77
be targetable by other inhibitors [7-9]. However, a clinical phase III trial was stopped
78
recently and the role of MET will need further evaluation [10]. Regardless, MET plays
79
an important role in the development of therapy resistance, especially in the context
80
of anti-EGFR-therapy [11]. Therefore, the identification of patients harboring
81
alterations in the ALK, ROS1 and still the MET gene seems of high interest.
82
In this study, we analyzed the frequency of ALK and ROS1 translocations as well as
83
MET amplifications in an unselected cohort of 473 surgically resected NSCLC
84
samples. We compared immunohistochemistry (IHC) and fluorescence in-situ
85
hybridization (FISH) as potential diagnostic assays. For MET FISH, we investigated
86
all three scoring systems so far described in literature [7,9,12-17]. Those were
87
compared among each other and correlated with two MET IHC scoring systems
88
(MetMAb and H-score). In the future, this could help to implement reliable and
89
efficient
Ac ce p
te
d
M
an
us
cr
ip t
65
screening
algorithms for
these
4
alterations
in routine
diagnostics.
Page 4 of 33
Materials and Methods
91
Patients and Samples
92
Tumor tissue of 473 NSCLC patients (289 men (61.1%) and 184 women (38.9%);
93
Table 1) who underwent pulmonary resection between 2000 and 2013 were obtained
94
from the archives of several Institutes of Pathology in Berlin (Charité University
95
Medicine Berlin, Pathology-Berlin Bioptisches Institut Gemeinschaftspraxis für
96
Pathologie, Vivantes Klinikum Berlin) and Minden (Johannes Wesling Klinikum
97
Minden). Detailed clinicopathological data was available for all patients, except for
98
two cases in which the staging was not documented. The median age was 67 years
99
(range: 29-85 years). Stage I was diagnosed in 261 (55.4%), stage II in 102 (21.7%),
100
stage III in 84 (17.8%) and stage IV in 24 (5.1%) cases. 335 (70.8%) tumors were
101
classified as adenocarcinomas (ADC). The predominant growth pattern was acinar in
102
144 (43.0%), solid in 114 (34.0%), papillary in 37 (11.0%), lepidic in 32 (9.6%) and
103
micropapillary in eight (2.4%) cases. 108 (22.8%) tumors were classified as
104
squamous cell carcinomas (SCC). The predominant growth pattern was non-
105
keratinizing in 61 (56.5%), keratinizing in 38 (35.2%) and basaloid in nine (8.3%)
106
cases. 13 (2.8%) samples were categorized as NSCLC not otherwise specified
107
(NOS), ten (2.1%) as large-cell carcinomas (LCC) and seven (1.5%) as
108
adenosquamous carcinomas (ADSC). Representative tumor areas were identified
109
and a total of ten tissue microarrays (TMA) were constructed using two cores for
110
each case (one mm diameter per core).
Ac ce p
te
d
M
an
us
cr
ip t
90
111 112
Immunohistochemistry (IHC)
113
IHC was performed as described previously on the VENTANA BenchMark XT 5
Page 5 of 33
automated slide stainer, using the anti-ALK-D5F3 Optiview kit [18] and the anti-total
115
c-MET-SP44 antibody (both Ventana Medical Systems, USA) [7]. ALK expression
116
was scored as negative (ALK IHC-) or positive (ALK IHC+); the latter was defined as
117
strong granular cytoplasmatic staining pattern [18]. MET expression was scored
118
according to the MetMAb trial criteria [7,19]: samples with no (IHC score 0) or weak
119
(IHC score 1+) membranous and/or cytoplasmatic staining in ≥50% of tumor cells
120
were considered MET IHC negative (MET IHC-). Tumors with moderate (IHC score
121
2+) or strong (IHC score 3+) membranous and/or cytoplasmatic staining in ≥50% of
122
tumor cells were considered MET IHC positive (MET IHC+). Furthermore, for MET
123
IHC we calculated a continuous H-score by multiplying the staining intensity (0: no
124
staining; 1: weak staining; 2: moderate staining; 3: strong staining) with the
125
respective percentage of tumor cells to reach a value between 0 and 300.
M
an
us
cr
ip t
114
d
126
Fluorescence in-situ hybridization (FISH)
128
FISH analysis was performed according to manufacturer’s instructions. Briefly, 4 µm
129
sections of TMA blocks were deparaffinized, dehydrated and then incubated in pre-
130
treatment solution (Dako, Denmark) for 10 minutes at 95-99°C. Samples were
131
immersed in pepsin solution for 3-6 minutes at 37°C. Following washing and
132
dehydration, slides were air dried and the respective DNA probe (ALK: Vysis LSI ALK
133
Dual Color, Abbott Molecular, USA; MET: MET/CEP7 dual-color probe, Abbott
134
Molecular, USA; ROS1: RF POSEIDON ROS1 Break probe, Kreatech, The
135
Netherlands; CEP6: CEP6 SpectrumAqua Probe, Abbott Molecular, USA) was
136
applied. The sections were sealed, denaturalized in humidified atmosphere at 82°C
137
for 5 minutes and then incubated overnight at 45°C to achieve hybridization. After
138
post-hybridization washing, slides were counterstained with 4’,6-diamidino-2-
Ac ce p
te
127
6
Page 6 of 33
139
phenylindole (DAPI).
140
Fluorescence in-situ hybridization (FISH) scoring
142
For each case, signals were enumerated in at least 50 non-overlapping tumor cells
143
using a fluorescence microscope (Axio Imager Z1, Zeiss, Germany). Computer-
144
based documentation and image analysis was performed with the ISIS imaging
145
system (MetaSystems, Germany).
146
ALK FISH positivity (ALK FISH+) was defined as the presence of split signals (SS)
147
and/or single red signals (SRS) in ≥15% of tumor cells [20-22].
148
For MET we used the three different scoring systems so far described in literature:
149
(1) UCCC (University of Colorado Cancer Center): cases with (A) MET/CEP7-ratio ≥2
150
or (B) small gene clusters (4-10 copies) or innumerable tight gene clusters in >10%
151
of tumor cells or (C) larger and brighter MET signals than CEP7 signals in >10% of
152
tumor cells or (D) >15 MET signals in >10% of tumor cells were considered as gene
153
amplification (GA). Specimens with ≥4 MET signals in ≥40% of tumor cells were
154
defined as high polysomy (HP). GA and HP were both considered to be MET FISH
155
positive (MET FISH+) [23,24]. However, as HP might be discussed as not being a
156
“true” amplification, we further discriminated between GA and HP.
157
(2) Cappuzzo: specimens with a mean gene copy number (GCN) of ≥5 MET signals
158
per tumor cell were classified as MET FISH+ [25].
159
(3) PathVysion: cases with a MET/CEP7-ratio ≥2 were defined as MET FISH+
160
[9,13,26].
161
For ROS1, tumors harboring SS and/or single green signals (SGS) in ≥15% of tumor
Ac ce p
te
d
M
an
us
cr
ip t
141
7
Page 7 of 33
162
cells were considered as ROS1 FISH positive (ROS1 FISH+) [27-30].
163
Statistical Analysis
165
Statistical analysis was performed using SPSS 21 software (IBM Corporation, USA).
166
Significance for cross tables larger than 2x2 was calculated using the statistical
167
language R. Comparison of clinicopathological data with IHC and FISH results was
168
performed using Fisher’s exact test (categorical data) and Mann-Whitney U test (age
169
and
size).
Ac ce p
te
d
M
an
tumor
us
cr
ip t
164
8
Page 8 of 33
170
Results
172
ALK IHC and ALK FISH
173
444 samples (93.9%) were evaluable by means of IHC and FISH. Concordant results
174
were seen in all 444 cases. A total of nine (2.0%) samples (2.6% of ADC) displayed
175
positivity for IHC and FISH (Figure 1). ALK rearrangements more frequently occurred
176
in younger patients with a median age of 54 years (range: 29-75 years; p=0.009). Six
177
(66.7%) of the ALK altered tumors showed lymphatic vessel invasion (p=0.023) and
178
lymph node metastasis (p=0.004). There was no significant trend (p=0.199) towards
179
advanced tumor stage (stage I: one case; stage II: three cases; stage III: two cases;
180
stage IV: one case). ALK IHC+/FISH+ tumors were mainly classified as ADC with
181
predominant acinar (five cases), solid (two cases) or papillary (one case) growth
182
pattern. One sample was classified as ADSC without keratinization in the SCC
183
component and with a solid growth pattern in the ADC component.
cr
us
an
M
d
te
Ac ce p
184
ip t
171
185
MET
186
MET IHC MetMAb criteria and MET FISH
187
440 samples (93.0%) were evaluable by means of IHC and FISH. Using the MetMAb
188
criteria for IHC and the UCCC scoring system including HP and GA for FISH,
189
homogenous results were seen in 396 cases (90.0%), with IHC-/FISH- in 358
190
(81.4%) and IHC+/FISH+ in 38 (8.6%) samples. 37 (8.4%) IHC+/FISH- and seven
191
(1.6%) IHC-/FISH+ cases were observed (Figure 2). Using the UCCC GA, Cappuzzo
9
Page 9 of 33
192
and PathVysion scoring system, the correlation was weaker with homogenous results
193
in in 383 (87,0%), 378 (85.9%) and 371 cases (84.3%), respectively (Table 3).
194
MET IHC H-score and MET FISH
196
The additionally applied MET IHC H-score showed a range between 0 and 280. A
197
good comparability of both tested IHC criteria was seen at an H-score threshold of
198
≥105, with discordant results in ten cases (2.2%; Supplemental Figure 1). Five of
199
these cases were classified as MetMAb positive and H-score negative, five as
200
MetMAb negative and H-score positive.
201
An H-score threshold of ≥180 encompassed 18 cases, all of them were considered
202
as MET FISH+ (Figure 4). Nine specimens were classified as HP and nine as GA.
203
These samples were also identified using the MetMAb criteria (IHC score 2+: 6
204
cases; IHC score 3+: 12 cases). However, this cut-off would not cover 29 (35.5%)
205
MET FISH+ samples.
cr
us
an
M
d
te
Ac ce p
206
ip t
195
207
Correlation of MET+ tumors (MetMab and UCCC) with clinicopathological data
208
MET IHC+/FISH+ samples were not associated with sex (25 men (65.8%), 13
209
women (34.2%); p=0.603) or age (median age: 67 years; range: 48-84 years;
210
p=0.694). Pleural invasion was present in in 47.2% of all IHC+/FISH+ cases
211
compared to 23.6% in tumors without MET IHC+/FISH+ results (p=0.006). MET
212
IHC+/FISH+ samples more frequently harbored lymphatic vessel invasion (60.5% vs.
213
41.0%; p=0.026) and lymph node metastasis (47.4% vs. 30.7%; p=0.048). There was
214
a non-significant trend towards advanced tumor stages (p=0.111) as alterations also
10
Page 10 of 33
occurred in early stages (stage I: 16 cases (6.8%); stage II: nine cases (9.9%); stage
216
III: eight cases (9.9%); stage IV: five cases (23.8%)). IHC+/FISH+ cases included 33
217
ADCs and five SCCs. MET alterations were more common in ADC (10.6%) than in
218
SCC (5.0%) and other NSCLC subtypes, although this difference was not significant
219
(p=0.051). We observed papillary (eleven cases), acinar (ten cases), solid (nine
220
cases) and lepidic (three cases) predominant growth patterns among the MET
221
altered ADCs with a significant trend towards papillary growth pattern (p=0.006). The
222
IHC+/FISH+ SCCs were classified as both keratinizing (one case) and non-
223
keratinizing (four cases; p=0.743).
224
Furthermore, we performed a separate analysis of the particular FISH scoring
225
systems with clinicopathological data (Supplementary Table 1). Pleural invasion and
226
advanced stage were significantly more common in the UCCC HP, UCCC GA and
227
Cappuzzo group. Blood vessel invasion was more frequent in the UCCC GA
228
(p=0.037) and Cappuzzo (p=0.005) group, while cases from the UCCC HP group
229
were significantly more likely to harbor lymphatic vessel invasion (p=0.007) and
230
lymph node metastasis (p=0.007). However, there were no significant correlations
231
with tumors that were considered positive according to the PathVysion criteria.
cr
us
an
M
d
te
Ac ce p
232
ip t
215
233
ROS1 FISH
234
ROS1 FISH was evaluable in 450 cases (94.5%). We discovered ROS1
235
rearrangements in four ADCs (0.9%), accounting for 1.3% of all ADCs. One ROS1
236
translocated tumor also showed an increased ROS1 GCN (5.4 copies per cell).
237
Another sample displayed 7.8 copies per cell but no ROS1 rearrangement (Figure 3).
11
Page 11 of 33
ROS1 translocations occurred in both men and women with a median age of 70
239
years (range 60-77 years). Three tumors showed lymphatic vessel invasion, lymph
240
node metastasis were present in two of them. Two cases were classified as stage I
241
and the other two samples as stage III. ROS1 rearranged tumors were all classified
242
as ADCs with predominantly solid (three cases) or papillary (one case) growth
243
pattern.
244
We reevaluated the two cases with an increased ROS1 GCN using an additional
245
CEP6 FISH probe. The ROS1 rearranged tumor showed amplification of the ROS1
246
gene locus with a ROS1/CEP6 ratio of 3.6, while the copy number gain in the second
247
sample was due to polysomy of chromosome 6 (Figure 3).
an
us
cr
ip t
238
M
248
Co-occurrence of ALK, MET and ROS1 alterations
250
Results for ALK, MET and ROS1 were available in 405 samples (85.6%) There was
251
no concomitant FISH positivity for any of the evaluated alterations. However, one
252
MET IHC+/FISH+ sample with gene amplification (MET/CEP7 ratio: 2.1; MetMAb
253
IHC score 2+; H-score 170) showed an increased ROS1 GCN due to polysomy of
254
chromosome 6 (Figure
255
overexpression (Figure 1). MET overexpression was more common in ALK altered
256
tumors (33% vs. 17.2%), although this was not significant (p=0.197).
257
One ALK IHC+ and MET IHC- case did not yield any evaluable FISH signals for ALK,
258
MET and ROS1, even though FISH was repeated several times, including
259
reevaluation of whole-slide tissue sections and different tumor blocks.
Ac ce p
te
d
249
3). Three ALK
IHC+/FISH+
cases showed
MET
260
12
Page 12 of 33
Discussion
261
Approximately 90% of all NSCLC diagnoses are performed on biopsy specimens
262
[31]. As the amount of tumor tissue available for molecular analysis is often limited
263
and the number of predictive targets increases steadily, the establishment of reliable
264
tests and adequate algorithms is of utmost importance. In this study we performed a
265
parallel screening for ALK, MET and ROS1 alterations and discuss the
266
consequences of our results for daily routine practice. ALK and ROS1
267
rearrangements were found in 2.6% and 1.3% of ADCs, respectively, but not in pure
268
SCCs. However, one ALK positive case showed adenosquamous histology. MET
269
IHC+/FISH+ results, on the other hand, were found in 8.6% of all NSCLCs and were
270
associated with pleural invasion. MET altered ADCs more frequently showed a
271
papillary growth pattern. Whereas ALK testing revealed reliable and concordant
272
results in IHC and FISH, the role of these two methods to identify MET amplifications
273
requires further investigation, as our data revealed discordant results in about 10.0%.
cr
us
an
M
d
te
Ac ce p
274
ip t
260
275
ALK alterations were described to be more frequent in younger patients [32-34]
276
which is consistent with the relatively low median age of 54 years in our cohort.
277
However, we encountered three ALK positive patients older than 60 years. Thus,
278
elderly patients should not generally be excluded from ALK testing [35]. For ALK, we
279
saw a clear correlation between IHC and FISH. All nine FISH+ cases were identified
280
by means of IHC and no false positive IHC results were encountered. Moreover, ALK
281
protein expression was helpful in two so-called “ALK FISH borderline” samples with
282
split signals in only 15-20% of tumor cells. These cases are particularly challenging in
283
routine diagnostics as they might be diagnosed as false negative in FISH [32,36,37].
284
Of note, we also encountered one ALK IHC+ case that was not evaluable by FISH. 13
Page 13 of 33
This case illustrates the urgent need for alternative methods to identify ALK
286
rearrangements besides FISH, as a therapy decision must be drawn under these
287
circumstances as well. In summary, our data support results of previous studies
288
[18,34,38] underlining the potential role of IHC as an alternative assay or at least as a
289
pre-screening tool to identify ALK rearrangements.
ip t
285
cr
290
Regarding MET, IHC+/FISH+ results were seen in 38 tumors (8.6%) when using the
292
MetMAb criteria for IHC and the UCCC scoring system for FISH. At a cut-off value of
293
≥105, the results from the IHC MetMAb criteria and the H-score were very similar
294
with homogenous results in 97.8% of cases. This is not surprising, as the definition of
295
a case with MetMAb IHC score 2+ (moderate staining in ≥50% of tumor cells)
296
corresponds to an H-score of at least 100. Furthermore, all tumors with an H-score
297
≥180 (n= 18) also harbored alterations of the MET gene in FISH. Although
298
statements on the predictive value of this finding is beyond the focus of this paper, an
299
IHC assay using an H-score with this threshold value could be examined as a
300
possible predictive test in future clinical trials.
an
M
d
te
Ac ce p
301
us
291
302
In accordance with literature, MET IHC+/FISH+ predominantly occurred in ADCs but
303
were also found in SCCs [17,39]. As already described by Nakamura et al. [40] we
304
observed that MET alterations were significantly more common in ADCs with
305
papillary growth pattern. Other authors described a significant correlation between
306
MET amplification and non-lepidic growth patterns [14,41]. Interestingly, activating
307
MET mutations also frequently occur in renal and thyroid carcinomas with papillary
308
growth [40]. Therefore, further research is needed to assess if MET might play an
14
Page 14 of 33
important role in the formation of papillary structures as proposed by Nakamura et al.
310
[40]. In accordance with previous studies we saw that MET alterations occurred more
311
frequently in tumors with pleural invasion, lymphatic vessel invasion and lymph node
312
metastasis [14,17]. Even though MET IHC+/FISH+ seem to be more common in
313
advanced tumors, they have also been described to occur in early stages [41].
314
Consistent with these reports, we observed MET alterations in 6.8% of all stage I
315
patients. Thus, from our data, it appears that MET testing should neither be limited to
316
a certain clinical subgroup nor to a certain histology.
317
Besides possible links with clinicopathological characteristics, the identification of the
318
most suitable method to screen for patients harboring MET amplifications is crucial
319
for routine diagnostics. The best correlation between IHC and FISH was achieved by
320
using the MetMAb criteria for IHC and the UCCC scoring system including HP for
321
FISH with concordant results in 90.0% compared to 85.9% and 84.3% when using
322
the Cappuzzo or PathVysion criteria, respectively (Table 3). The UCCC scoring
323
system was able to identify all tumors that were classified as positive according to
324
one of the other criteria. Therefore, the UCCC scoring system seems to be superior
325
to the Cappuzzo and PathVysion criteria in identifying IHC+ cases. Still, 44 samples
326
revealed contradictory results with 37 IHC+/FISH- and seven IHC-/FISH+ tumors.
327
These heterogeneous results might explain the huge differences between the
328
frequencies of MET overexpression (25-54%) and MET amplification (11.1-16.7%
329
when using the UCCC criteria) in earlier studies [7,8,12,24,39]. IHC-/FISH+ cases
330
have been reported before although they have never been described in detail [8,12].
331
It is unknown whether these tumors with a significantly increased MET GCN actually
332
do not overexpress MET or if the weak immunoreactivity in IHC was due to a specific
333
lack of antibody affinity or other mechanisms that might reduce staining intensity.
334
However, it should be noted that all MET IHC-/FISH+ specimens had been evaluable
Ac ce p
te
d
M
an
us
cr
ip t
309
15
Page 15 of 33
for other IHC markers such as TTF-1 or p63, stained in the process of the routine
336
diagnostics (data not shown). While the clinical outcome of patients with MET IHC-
337
/FISH+ tumors has not yet been investigated in detail, two recent studies showed
338
that MET IHC+/FISH- patients could also benefit from therapy with onartuzumab or
339
tivantinib [7,8]. Thus, there is a need to investigate (A) the underlying mechanisms
340
that lead to MET overexpression in FISH negative tumors; (B) potential reasons why
341
some FISH positive tumors show no or only weak protein expression in IHC; (C) the
342
clinical relevance of cases with conflicting results in IHC and FISH, especially of IHC-
343
/FISH+ patients.
an
us
cr
ip t
335
344
Recently, ROS1 has been identified as a treatable target in lung cancer [29,42,43]. In
346
our investigation, we saw an overall frequency of 0.9% (1.3% in ADCs) which is in
347
line with previous reports [29,42,43]. Recent trials and case reports discovered that
348
patients with ROS1 translocation in FISH could profit from anti-ROS1 therapy [4,5].
349
Regarding histology, we confirmed that ROS1 translocations mainly occur in ADCs
350
[42]. However, ROS1 might also be affected in SCCs and LCCs as previously
351
described [27,42,43]. Due to the low frequency of ROS1 alterations in our study, any
352
links with clinicopathological characteristics should be interpreted with care.
353
Furthermore, we encountered one case with polysomy of chromosome 6 and one
354
tumor with ROS1 amplification, the latter in addition to ROS1 translocation (Figure 3).
355
To our knowledge, “true” amplifications of the ROS1 gene locus have not been
356
described yet in NSCLC. Interestingly, tumors with an increased ROS1 GCN were
357
reported to show strong ROS1 expression in IHC [44]. These patterns need further
358
investigation as they might also benefit from targeted therapy.
Ac ce p
te
d
M
345
16
Page 16 of 33
359
Regarding concomitant alterations, we did not encounter any combined FISH+
361
results for ALK, MET or ROS1. For IHC, we observed three ALK rearranged tumors
362
showing MET overexpression. However, in contrast to recently published data
363
[17,45], MET IHC positivity was not significantly more common in ALK rearranged
364
tumors. This might be due to the lower number of ALK positive samples in our cohort.
365
Regardless, the clinical role of such cases is unknown and requires further
366
investigation (therapy resistance?).
367
While other publications showed that ROS1 alterations may overlap with KRAS,
368
EGFR and BRAF mutations [42,46], we observed no co-occurrence of ROS1
369
rearrangements with ALK or MET alterations although these results might not be
370
representative due to the low number of ROS1 translocated tumors in our study. One
371
case with an increased ROS1 GCN due to polysomy of chromosome 6 showed a
372
concomitant MET amplification (Figure 3). The clinical and biological role is unclear
373
and needs further evaluation.
cr
us
an
M
d
te
Ac ce p
374
ip t
360
375
In summary, our study is the first to report on parallel testing for ALK, MET and
376
ROS1 alterations in a large and non-preselected NSCLC-cohort. We may conclude
377
from our data that ALK and ROS1 screening should first of all be performed in ADCs
378
of the lung, whereas MET screening might be established as a further predictive
379
diagnostic option not only in ADCs but also in SCCs. In daily routine, IHC could
380
complement FISH in the diagnosis of ALK altered NSCLCs, especially with regards
381
to borderline cases or tumors without evaluable FISH signals. For MET, future
382
research is needed to evaluate the predictive value of IHC and FISH as well as the
17
Page 17 of 33
383
clinical role of patients with discordant results. Additionally, the clinical relevance of
384
tumors with increased ROS1 copy number due to polysomy of chromosome 6 or
385
amplification
the
ROS1
gene
locus
requires
further
evaluation.
Ac ce p
te
d
M
an
us
cr
ip t
of
18
Page 18 of 33
386
Acknowledgement
388
Part of this work was supported by PFIZER within the framework of the FALKE-
389
project (detection of ALK-rearrangements by FISH in NSCLC).
392
None
us
Conflict of interest
Ac ce p
te
d
M
an
391
cr
390
ip t
387
19
Page 19 of 33
393 394
References
395
[1]
Maemondo M, Inoue A, Kobayashi K, Sugawara S, Oizumi S, Isobe H, et al. Gefitinib or chemotherapy for non–small cell lung cancer with mutated EGFR.
397
N Engl J Med 2010;362:2380–8. [2]
Soda M, Choi YL, Enomoto M, Takada S, Yamashita Y, Ishikawa S, et al.
cr
398
ip t
396
Identification of the transforming EML4–ALK fusion gene in non-small-cell
400
lung cancer. Nature 2007;448:561–6.
401
[3]
us
399
Schwab R, Petak I, Kollar M, Pinter F, Varkondi E, Kohanka A, et al. Major partial response to crizotinib, a dual MET/ALK inhibitor, in a squamous cell
403
lung (SCC) carcinoma patient with de novo c-MET amplification in the
404
absence of ALK rearrangement. Lung Cancer 2013;83:109–11. [4]
Ou S-HI, Camidge R, Riely GJ, Salgia R, Shapiro G, Clark JW, et al. Efficacy
M
405
an
402
406
and safety of crizotinib in patients with advanced ROS1-rearranged non-small
407
cell lung cancer (NSCLC). Ann Oncol 2013;24:ix43–3. [5]
Bos M, Gardizi M, Schildhaus HU, Heukamp LC, Geist T, Kaminsky B, et al.
d
408
Complete metabolic response in a patient with repeatedly relapsed non-small
410
cell lung cancer harboring ROS1 gene rearrangement after treatment with
411
crizotinib. Lung Cancer 2013;81:142–3. [6]
and anaplastic lymphoma kinase (ALK) inhibitor, in a non-small cell lung
414
cancer patient with de novo MET amplification. J Thorac Oncol 2011;6:942–6.
415
[7]
Spigel DR, Ervin TJ, Ramlau RA, Daniel DB, Goldschmidt JH, Blumenschein GR, et al. Randomized Phase II Trial of Onartuzumab in Combination With
417
Erlotinib in Patients With Advanced Non-Small-Cell Lung Cancer. J Clin
418
Oncol 2013;31:4105–14.
419 420
Ou S-HI, Kwak EL, Siwak-Tapp C, Dy J, Bergethon K, Clark JW, et al. Activity of crizotinib (PF02341066), a dual mesenchymal-epithelial transition (MET)
413
416
Ac ce p
412
te
409
[8]
Rodig SJ, Sequist LV, Schiller JH, Chen Y, Halim A, Waghorne C, et al.
421
Abstract 1729: An exploratory biomarker analysis evaluating the effect of the
422
c-MET inhibitor tivantinib (ARQ 197) and erlotinib in NSCLC patients in a
423
randomized, double-blinded phase 2 study. Cancer Res 2012;72:1729.
424
[9]
Sequist LV, Sequist LV, Pawel von J, Pawel von J, Garmey EG, Garmey EG, 20
Page 20 of 33
425
et al. Randomized Phase II Study of Erlotinib Plus Tivantinib Versus Erlotinib
426
Plus Placebo in Previously Treated Non-Small-Cell Lung Cancer. J Clin
427
Oncol 2011;29:3307–15. [10]
come and gone? Clin Cancer Res 2014;20:4422–4.
429 430
Hirsch FR, Bunn PA, Herbst RS. “Companion diagnostics”: has their time
[11]
Bean J, Brennan C, Shih J-Y, Riely G, Viale A, Wang L, et al. MET
ip t
428
amplification occurs with or without T790M mutations in EGFR mutant lung
432
tumors with acquired resistance to gefitinib or erlotinib. Proc Natl Acad Sci U
433
S A 2007;104:20932–7. [12]
Dziadziuszko R, Wynes MW, Singh S, Asuncion BR, Ranger-Moore J,
us
434
cr
431
Konopa K, et al. Correlation between MET gene copy number by silver in situ
436
hybridization and protein expression by immunohistochemistry in non-small
437
cell lung cancer. J Thorac Oncol 2012;7:340–7.
438
[13]
an
435
Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski MF, Pao W, et al. Analysis of tumor specimens at the time of acquired resistance to EGFR-TKI
440
therapy in 155 patients with EGFR-mutant lung cancers. Clin Cancer Res
441
2013;19:2240–7. [14]
Tachibana K, Minami Y, Shiba-Ishii A, Kano J, Nakazato Y, Sato Y, et al.
d
442
M
439
Abnormality of the hepatocyte growth factor/MET pathway in pulmonary
444
adenocarcinogenesis. Lung Cancer 2012;75:181–8.
446 447 448 449 450 451 452
[15]
MET Increased Gene Copy Number and Primary Resistance to Gefitinib Therapy in Non-Small Cell Lung Cancer Patients. Ann Oncol 2008;20:298– 304.
[16]
Minuti G, D'Incecco A, Cappuzzo F. Targeted therapy for NSCLC with driver mutations. Expert Opin Biol Ther 2013;13:1401–12.
[17]
Tsuta K, Kozu Y, Mimae T, Yoshida A, Kohno T, Sekine I, et al. cMET/phospho-MET protein expression and MET gene copy number in nonsmall cell lung carcinomas. J Thorac Oncol 2012;7:331–9.
453 454
Cappuzzo F, Janne PA, Skokan M, Finocchiaro G, Rossi E, Ligorio C, et al.
Ac ce p
445
te
443
[18]
Nitta H, Tsuta K, Yoshida A, Ho SN, Kelly BD, Murata LB, et al. New methods
455
for ALK status diagnosis in non-small-cell lung cancer: an improved ALK
456
immunohistochemical assay and a new, Brightfield, dual ALK IHC-in situ
457
hybridization assay. J Thorac Oncol 2013;8:1019–31.
458
[19]
Spigel DR, Ervin TJ, Ramlau R, Daniel DB. Final efficacy results from 21
Page 21 of 33
459
OAM4558g, a randomized phase II study evaluating MetMAb or placebo in
460
combination with erlotinib in advanced NSCLC. J Clin Oncol 2011.
461
[20]
McLeer-Florin AA, Moro-Sibilot DD, Melis AA, Salameire DD, Lefebvre CC, Ceccaldi FF, et al. Dual IHC and FISH testing for ALK gene rearrangement in
463
lung adenocarcinomas in a routine practice: a French study. J Thorac Oncol
464
2012;7:348–54.
465
[21]
ip t
462
Thunnissen E, Bubendorf L, Dietel M, Elmberger G, Kerr K, López-Ríos F, et al. EML4-ALK testing in non-small cell carcinomas of the lung: a review with
467
recommendations. Virchows Arch 2012;461:245–57. [22]
Yi ES, Boland JM, Maleszewski JJ, Roden AC, Oliveira AM, Aubry M-C, et al.
us
468
cr
466
Correlation of IHC and FISH for ALK Gene Rearrangement in Non-small Cell
470
Lung Carcinoma IHC Score Algorithm for FISH. J Thorac Oncol 2011;6:459–
471
65.
472
[23]
an
469
Varella-Garcia M. Stratification of non-small cell lung cancer patients for therapy with epidermal growth factor receptor inhibitors: the EGFR
474
fluorescence in situ hybridization assay. Diagn Pathol 2006;1:19.
475
[24]
M
473
Go H, Jeon YK, Park HJ, Sung S-W, Seo J-W, Chung DH. High MET gene copy number leads to shorter survival in patients with non-small cell lung
477
cancer. J Thorac Oncol 2010;5:305–13.
480 481 482 483 484 485 486
te
479
[25]
Cappuzzo F, Marchetti A, Skokan M, Rossi E, Gajapathy S, Felicioni L, et al. Increased MET Gene Copy Number Negatively Affects Survival of Surgically
Ac ce p
478
d
476
Resected Non-Small Cell Lung Cancer Patients. J Clin Oncol 2009;27:1667– 74.
[26]
Tanaka A, Tanaka A, Sueoka-Aragane N, Sueoka-Aragane N, Nakamura T, Nakamura T, et al. Co-existence of positive MET FISH status with EGFR mutations signifies poor prognosis in lung adenocarcinoma patients. Lung Cancer 2012;75:89–94.
[27]
Davies KD, Davies KD, Le AT, Le AT, Theodoro MF, Theodoro MF, et al.
487
Identifying and targeting ROS1 gene fusions in non-small cell lung cancer.
488
Clin Cancer Res 2012;18:4570–9.
489
[28]
Kim MH, Shim HS, Kang DR, Jung JY, Lee CY, Kim DJ, et al. Clinical and
490
prognostic implications of ALK and ROS1 rearrangements in never-smokers
491
with surgically resected lung adenocarcinoma. Lung Cancer 2014;83:1–7.
492
[29]
Bergethon K, Bergethon K, Shaw AT, Shaw AT, Ignatius Ou SH, Ignatius Ou 22
Page 22 of 33
493
SH, et al. ROS1 Rearrangements Define a Unique Molecular Class of Lung
494
Cancers. J Clin Oncol 2012;30:863–70.
495
[30]
Yoshida A, Tsuta K, Wakai S, Arai Y, Asamura H, Shibata T, et al.
496
Immunohistochemical detection of ROS1 is useful for identifying ROS1
497
rearrangements inlung cancers. Mod Pathol 2013:1–10. Histopathology 2009;54:12–27.
499 500
Kerr KM. Pulmonary adenocarcinomas: classification and reporting.
ip t
[31] [32]
Camidge DR, Theodoro M, Maxson DA, Skokan M, O'Brien T, Lu X, et al.
cr
498
Correlations between the percentage of tumor cells showing an ALK gene
502
rearrangement, ALK signal copy number, and response to crizotinib therapy
503
in ALK FISH positive non-small cell lung cancer. Cancer Res 2012;118:4486–
504
94. [33]
an
505
us
501
Camidge DR, Hirsch FRF, Varella-Garcia MM, Franklin WAW. Finding ALKpositive lung cancer: what are we really looking for? J Thorac Oncol
507
2011;6:411–3.
508
[34]
M
506
Sholl LM, Weremowicz S, Gray SW, Wong K-K, Chirieac LR, Lindeman NI, et al. Combined use of ALK immunohistochemistry and FISH for optimal
510
detection of ALK-rearranged lung adenocarcinomas. J Thorac Oncol
511
2013;8:322–8.
514 515 516 517 518 519 520 521
te
513
[35]
Lindeman NI, Cagle PT, Beasley MB, Chitale DA, Dacic S, Giaccone G, et al. Molecular testing guideline for selection of lung cancer patients for EGFR and
Ac ce p
512
d
509
ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. J Thorac Oncol 2013;8:823–59.
[36]
V Laffert M, Warth A, Penzel R, Schirmacher P, Jonigk D, Kreipe H, et al. Anaplastic lymphoma kinase (ALK) gene rearrangement in non-small cell lung cancer (NSCLC): results of a multi-centre ALK-testing. Lung Cancer 2013;81:200–6.
[37]
V Laffert M, Warth A, Penzel R, Schirmacher P, Kerr K, Elmberger G, et al.
522
Anaplastic lymphoma kinase (ALK)-detection in non-small cell lung cancer:
523
results of the first European IHC-based (D5F3-Optiview) panel test within 16
524
institutes. J Thorac Oncol 2013;8:S484.
525 526
[38]
Savic S, Bode B, Diebold J, Tosoni I, Barascud A, Baschiera B, et al. Detection of ALK-Positive Non-Small-Cell Lung Cancers on Cytological 23
Page 23 of 33
527
Specimens: High Accuracy of Immunocytochemistry with the 5A4 Clone. J
528
Thorac Oncol 2013;8:1004–11.
529
[39]
Park S, Choi Y-L, Sung CO, An J, Seo J, Ahn M-J, et al. High MET copy
530
number and MET overexpression: poor outcome in non-small cell lung cancer
531
patients. Histol Histopathol 2012;27:197–207. [40]
Nakamura Y, Niki T, Goto A, Morikawa T, Miyazawa K, Nakajima J, et al. c-
ip t
532
Met activation in lung adenocarcinoma tissues: an immunohistochemical
534
analysis. Cancer Sci 2007;98:1006–13.
535
[41]
cr
533
Jin Y, Sun P-L, Kim H, Seo AN, Jheon S, Lee C-T, et al. MET Gene Copy Number Gain is an Independent Poor Prognostic Marker in Korean Stage I
537
Lung Adenocarcinomas. Ann Surg Oncol 2013. [42]
Warth A, Muley T, Dienemann H, Goeppert B, Stenzinger A, Schnabel PA, et
an
538
us
536
539
al. ROS1 Expression and Translocations in Non-Small Cell Lung Cancer:
540
Clinicopathological Analysis of 1478 Cases. Histopathology 2014. [43]
Rimkunas VM, Crosby KE, Li D, Hu Y, Kelly ME, Gu TL, et al. Analysis of
M
541
Receptor Tyrosine Kinase ROS1-Positive Tumors in Non-Small Cell Lung
543
Cancer: Identification of a FIG-ROS1 Fusion. Clin Cancer Res 2012;18:4449–
544
57. [44]
Lee HJ, Seol HS, Kim JY, Chun S-M, Suh Y-A, Park Y-S, et al. ROS1
te
545
d
542
Receptor Tyrosine Kinase, a Druggable Target, is Frequently Overexpressed
547
in Non-Small Cell Lung Carcinomas Via Genetic and Epigenetic Mechanisms.
548 549 550 551 552 553 554
Ac ce p
546
Ann Surg Oncol 2012;20:200–8.
[45]
Feng Y, Minca EC, Lanigan C, Liu A, Zhang W, Yin L, et al. High MET receptor expression but not gene amplification in ALK 2p23 rearrangement positive non-small-cell lung cancer. J Thorac Oncol 2014;9:646–53.
[46]
Stumpfova M, Janne PA. Zeroing in on ROS1 Rearrangements in Non-Small Cell Lung Cancer. Clin Cancer Res 2012;18:4222–4.
555
24
Page 24 of 33
Titles and legends to figures
562
Figure 1 Example of an adenocarcinoma (a; hematoxylin and eosin stain) with
563
positive results in ALK IHC (b) and ALK FISH (c; red: 3’ ALK probe, green: 5’ ALK
564
probe). Additionally, this tumor showed MET overexpression (d; MetMAb IHC score
565
2+; H-score 150) but no amplification in MET FISH (e; red: MET gene locus, cyan:
566
centromere 7).
cr
ip t
561
us
567
Figure 2 Examples of tumors with equivocal and discordant results in MET IHC and
569
FISH: cases with homogenous results with IHC-/FISH- in case 1 (a-c) and
570
IHC+/FISH+ in case 2 (d-f). Cases 3 (g-I) showed MET overexpression but revealed
571
no amplification in FISH while case 4 (j-l) showed a significant MET copy number
572
gain but lacked any MET expression. Pictures a, d and h show hematoxylin and
573
eosin stains of representative tumor areas for each case.
M
d
te Ac ce p
574
an
568
575
Figure 3 Case 1 (a-c; green: 3’ ROS1 probe, red: 5’ ROS1 probe) showed ROS1
576
translocation. Picture c shows magnification of representative tumor cells from the
577
same case. Case 2 (d-g; cyan: centromere 6) revealed amplification and
578
translocation of the ROS1 gene locus with a gene-to-chromosome ratio of 3.6 and at
579
least one split signal or single green signal (e and f, white arrows indicate rearranged
580
signals) in 30% of tumor cells. Case 3 (h-k) showed ROS1 copy number gain due to
581
polysomy of chromosome 6 and had an additional MET amplification (j; red: MET
582
gene locus, cyan: centromere 7; MET/CEP7 ratio: 2.1) with MET overexpression in
583
IHC (data not shown). Pictures a, d and h show hematoxylin and eosin stains of
584
representative tumor areas for each case. 29
Page 25 of 33
585
Figure 4 Plot of MET IHC H-score and MET FISH results. The black line indicates a
587
threshold value of ≥180. All cases with an H-score ≥180 also revealed MET alteration
588
in FISH. For reasons of clarity, cases with an H-score of 0 are only shown if they
589
were classified as FISH+. The total number of cases with an H-score ≥180 was 18
590
and they were all classified as 2+ (6 cases) or 3+ (12 cases) according to the
591
MetMAb criteria.
Ac ce p
te
d
M
an
us
cr
ip t
586
30
Page 26 of 33
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 27 of 33
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 28 of 33
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 29 of 33
300 UCCC high polysomy UCCC gene amplification UCCC negative
250
≥180
us
cr
ip
t
200
100
ce pt
ed
M
an
150
Ac
H-Score
Figure 4
50 0
Page 30 of 33
Table 1
Ac ce pt e
d
M
an
us
cr
ip t
Table 1 Correlation between MET amplification, ALK and ROS1 translocation and clinicopathological characteristics. a MET FISH+/IHC+ ALK FISH+/IHC+ ROS1 FISH+ All cases Positive p-value Positive p-value Positive p-value Total 473 38 (8.6%) 9 (1.9%) 4 (0.9%) Sex Male 289 25 5 2 0.603 0.739 0.643 Female 184 13 4 2 Age (years) Median (standard deviation) 67 ± 10 66 ± 8.7 54 ± 14.2 70 ± 7.6 0.694 0.009 0.386 Range 29 – 85 48 – 84 29 – 75 60 – 77 Stage I-II 363 25 4 2 0.111 0.199 0.220 III-IV 108 13 3 2 Lymph-node metastasis Absent 324 20 1 2 0.048 0.004 0.588 Present 147 18 6 2 Lymphatic vessel invasion Absent 276 15 1 1 0.026 0.023 0.308 Present 195 23 6 3 Blood vessel invasion Absent 369 26 4 2 0.158 0.195 0.202 Present 102 12 3 2 Pleural invasion Absent 352 19 5 3 0.006 1.000 1.000 Present 119 17 2 1 Tumor size (centimeter) Median (standard deviation) 3.0 ± 2.6 3.5 ± 2.0 0.960 2.3 ± 3.4 0.505 2,4 ± 0.8 0.321 Range 0.8 – 17.0 1.0 – 9.0 1.2 – 9.0 1.9 – 3.7 Histology ADC 335 33 8 4 SCC 108 5 0.051 0 0.182 0 0.673 b Others 30 0 1 0 ADC predominant growth pattern ADC solid 114 9 2 3 ADC acinar 144 10 5 0 ADC papillary 37 11 0.006 1 0.816 1 0.238 ADC lepidic 32 3 0 0 ADC micropapillary 8 0 0 0 SCC growth pattern Keratinizing 38 1 0 0 Non-Keratinizing 61 4 0.743 0 0 Basaloid 9 0 0 0 ALK, anaplastic lymphoma kinase; MET, mesenchymal-epithelial transition factor; ROS1, repressor of silencing 1; IHC, immunohistochemistry; FISH, fluorescence in-situ hybridization; ADC, adenocarcinoma; SCC, squamous cell carcinoma a Using the University of Colorado Cancer center (UCCC) scoring system for FISH and the MetMAb trial criteria for IHC b 13 non-small cell lung cancers not otherwise specified, ten large cell carcinomas, seven adenosquamous carcinomas
Page 31 of 33
Table 2
Ac ce pt e
d
M
an
us
cr
ip t
Table 2 Molecular pathological characteristics of patients harboring ALK or ROS1 translocation. Case # Sex Age Histology, FISH pattern ALK IHC MET MET IHC predominant (% of altered FISH MetMAb cells) criteria ALK 1 M 62 ADC, solid SS, SRS (69%) Positive Negative Positive (2+) ALK 2 M 69 ADC, acinar SRS (70%) Positive Negative Negative (0) ALK 3 F 39 ADSC, solid SS, SRS (18%) Positive Negative Negative (0) ALK 4 F 50 ADC, solid SS (20%) Positive Negative Negative (0) ALK 5 M 75 ADC, acinar SS (40%) Positive Negative Negative (1+) ALK 6 M 54 ADC, acinar SS (40%) Positive Negative Negative (0) ALK 7 M 41 ADC, papillary SS, SRS (62%) Positive Negative Positive (2+) ALK 8 F 58 ADC, acinar SS (70%) Positive Negative Negative (0) ALK 9 F 29 ADC, acinar SS, SRS (70%) Positive Negative Positive (2+) ROS1 M 60 ADC, acinar SS (71%) Negative Negative Negative (1+) 1 ROS1 F 77 ADC, solid SS (40%) Negative Negative Negative (1+) 2 ROS1 M 67 ADC, solid SS, SGS Negative Negative Negative (0) 3 (30%), A ROS1 F 74 ADC, solid SS, SGS (90%) Negative Negative Negative (1+) 4 ALK, anaplastic lymphoma kinase, ROS1, repressor of silencing 1; MET, mesenchymal-epithelial transition factor; FISH, fluorescence in-situ hybridization; IHC, immunohistochemistry; M, male; F, female; ADC, adenocarcinoma; ADSC, adenosquamous carcinoma; SS, split signals; SRS, single red signals; SGS, single green signals; A, amplification
Page 32 of 33
Table 3
Ac ce pt e
d
M
an
us
cr
ip t
Table 3 Correlation of different MET FISH amplification scoring systems and MET IHC scores. MET FISH MET IHC negative MET IHC positive All MET IHC 0 MET IHC 1+ MET IHC 2+ MET IHC 3+ a UCCC negative 395 250 108 37 0 a UCCC positive 45 5 2 26 12 UCCC HP negative 417 251 109 50 7 UCCC HP positive 23 4 1 13 5 UCCC GA negative 418 254 109 50 5 UCCC GA positive 22 1 13 7 Cappuzzo negative 423 254 109 54 6 Cappuzzo positive 17 1 1 9 6 PathVysion negative 436 255 110 59 10 PathVysion positive 6 0 0 4 2 MET, mesenchymal-epithelial transition; FISH, Fluorescence in-situ hybridization; IHC, immunohistochemistry; UCCC, University of Colorado Cancer Center; HP, high polysomy; GA, gene amplification a : Including high polysomy and gene amplification
Page 33 of 33
