Characterization of KIT mutation in melanoma - Dermatologica Sinica

DERMATOLOGICA SINICA 32 (2014) 7e12

Contents lists available at SciVerse ScienceDirect

Dermatologica Sinica journal homepage: http://www.derm-sinica.com

ORIGINAL ARTICLE

Characterization of KIT mutation in melanoma Chi-Yuan Tzen 1, 2, Yu-Hung Wu 3, Chin-Yuan Tzen 2, 4, 5, * 1

Department of Pathology, Mackay Memorial Hospital, Taipei, Taiwan National Taipei College of Nursing, Taipei, Taiwan 3 Department of Dermatology, Mackay Memorial Hospital, Taipei, Taiwan 4 Department of Pathology and Laboratory Medicine, Cathay General Hospital, Taipei, Taiwan 5 College of Medicine, Fu Jen Catholic University, Taipei, Taiwan 2

a r t i c l e i n f o

a b s t r a c t

Article history: Received: Jan 25, 2013 Revised: May 14, 2013 Accepted: May 29, 2013

Background/objectives: Recent studies have shown that the KIT mutational type appears to be a predictive marker for the efficacy of imatinib in treating melanoma. However, a wide range of KIT mutation rates was reported in different types of melanoma, suggesting that the mutation frequency of KIT may be associated with clinicopathological subsets of melanoma. Methods: To characterize their relationship, we sequenced exons 11, 13, 17, and 18 of KIT in 80 of 85 melanomas collected from two hospitals and categorized KIT mutation by tumor type, age and sex of patients, and mutation hot spots of KIT. Results: Our results showed that KIT mutation rates were 25%, 22%, and 8% in acral, mucosal, and cutaneous melanomas, respectively. Approximately 38% (5/13) of male patients with acral melanoma and 45% (5/11) of female patients with mucosal melanomas of the anorectal and genitourinary regions had a KIT mutation. Approximately 81% of KIT mutations occurred in L576P, K642E, V559A, and D820Y. Conclusion: This result shows that KIT mutation is enriched in a certain subset of melanoma patients and mutation hot spots do exist. Copyright Ó 2013, Taiwanese Dermatological Association. Published by Elsevier Taiwan LLC. All rights reserved.

Keywords: acral melanoma cutaneous melanoma KIT mutation rate mucosal melanoma

Introduction Melanoma, a malignant tumor of melanocytic origin, is a lifethreatening disease once metastasized, with a median survival of 6e10 months.1 The emerging targeted therapy is promising for these patients in that a variety of mutations have been discovered in melanoma, and small molecules specifically inhibiting some of these mutations, such as KIT and BRAF, are available. The success of targeted therapy for patients with cancer has been well exemplified by imatinib mesylate for patients with gastrointestinal stromal tumors.2 In this paradigm, imatinib demonstrates a better efficacy if the tumor harbors mutated KIT (particularly the exon 11 mutation), whereas imatinib shows minimal effect if the tumor is a wild type for KIT or PDGFRA. A similar scenario appears to happen in treating melanoma patients with imatinib because the earlier clinical trials showed disappointing Conflicts of interest: The authors declare that this study was supported in part by Cathay General Hospital and Novartis Oncology, Taiwan. * Corresponding author. Department of Pathology and Laboratory Medicine, Cathay General Hospital, Number 280, Section 4, Jen-Ai Road, Taipei, Taiwan. Tel.: þ886 2 2690 7965x2518; fax: þ886 2 2691 9800. E-mail address: [email protected] (C.-Y. Tzen).

results of imatinib therapy in unselected melanoma patients.3e5 By contrast, results of recent studies have showed clinical responses to imatinib in patients with KIT-mutated melanoma.6e9 More specifically, in a Phase II study, 43 patients with melanoma harboring mutated KIT were treated with imatinib, and the result showed an overall response rate of 23.3%.10 In addition, some patients may even have a complete response after imatinib treatment.11 If a KIT mutation is a reliable marker for selecting melanoma patients for imatinib treatment, then it is important to know how often KIT is mutated in melanoma. In this regard, earlier observations showed that KIT mutations are rare in unselected melanomas. For example, Went et al12 reported that only one of 39 in their series had a KIT mutation and Willmore-Payne et al13 reported that two of 100 cases had a KIT mutation. However, subsequent studies showed a higher mutation rate, but with variable data in different types of melanoma. In 2006, Curtin et al14 reported that KIT mutation rates in acral (11%) and mucosal (21%) melanomas were higher than that in cutaneous melanomas (8%). In 2008, Beadling et al15 confirmed this observation and noted that the KIT mutation rate was 23% in acral melanomas, 16% in mucosal melanomas, and 2% in cutaneous melanomas. Carvajal et al16 also reported that the KIT mutation rate in cutaneous melanomas (16%) was close to, but

1027-8117/$ e see front matter Copyright Ó 2013, Taiwanese Dermatological Association. Published by Elsevier Taiwan LLC. All rights reserved. http://dx.doi.org/10.1016/j.dsi.2013.05.005

8

C.-Y. Tzen et al. / Dermatologica Sinica 32 (2014) 7e12

still lower than, that in acral (21%) and mucosal (18%) melanomas. In contrast to these studies that analyze American patients, Kong et al17 analyzed 451 Chinese patients and found that KIT mutation rates were similar in acral (12%), mucosal (10%), and cutaneous (12%) melanomas. This disparity leads to an important question. Does the KIT mutation rate have ethnic differences so that the difference of mutation frequency among cutaneous, acral, and mucosal melanomas in American patients was not observed in Chinese patients? If KIT mutations are to be used as markers to select melanoma patients for imatinib treatment, there is no doubt that tumor specimens should be subjected to mutational analysis. However, practical problems will be encountered when equipment (such as an FDA-approved DNA sequencer) is not available or the budget for healthcare is limited. In this situation, one would like to know if mutation hot spots of KIT exist in melanoma so that a simpler technique can be used for mutational analysis. Furthermore, clinicians may want to know if there is a subset of melanoma patients whose tumors have a much higher chance of harboring a KIT mutation. Therefore, imatinib can be given to these patients in the absence of mutational analysis. These questions are important not only for clinicians who order the molecular test, but also for health insurance providers who cover the cost of the test. In this study, we aimed to compare the KIT mutation rate in acral, mucosal, and cutaneous melanomas in Chinese patients, to further categorize KIT mutation by various clinical parameters (such as age, sex, and anatomical location), and to examine whether there are mutation hot spots of KIT in melanoma. Materials and methods Patient series and tumor tissue samples We retrospectively collected consecutive 85 melanomas of acral, mucosal, and cutaneous types, all of which were diagnosed between 2005 and 2010 from two medical centers in Taipei. This series consisted of 52 melanomas from Mackay Memorial Hospital and 33 melanomas from Cathay General Hospital. These samples were examined by hematoxylin and eosin staining, and an immunohistochemical analysis for HMB-45 or S-100 was performed to confirm the diagnosis when necessary. Histologic slides were reviewed by two surgical pathologists (C.-Y.T. and C.-Y.T.), who confirmed the diagnosis. This study was approved by the institutional review board of both hospitals. DNA extraction DNA was isolated from formalin-fixed paraffin-embedded tumors. In brief, tissue blocks were cut at 8 mm using a clean disposable microtome blade for each block. The paraffin sections were incubated in xylene at 25 C for 5 minutes, rinsed in absolute alcohol at room temperature, air dried, and transferred into polymerase chain reaction (PCR) tubes. The pellets were then processed using a QIAamp DNA FFPE Tissue Kit (Qiagen) according to the manufacturer’s instructions. Mutational analysis For mutational analysis, exons 11, 13, 17, and 18 of KIT were analyzed. We amplified these exons by PCR in at least two separate preparations of genomic DNA. The PCR primers for exon 11 were 50 -T GT TCT CTC TCC AGA GTG CT-30 (forward) and 50 -AAA AGG GGC GCA ATT TCA CA-30 (reverse), for exon 13 were 50 -AGA TGC TCA AGC GTA AGT TC-30 (forward) and 50 -GAG AAC AAC AGT CTG GGT AA-30

(reverse), for exon 17 were 50 -GTG AAC ATC ATT CAA GGC GT-30 (forward) and 50 -TGT GAT ATC CCT AGA CAG GA-30 (reverse), and for exon 18 were 50 -CAT TTC AGC AAC AGC AGC AT-30 (forward) and 50 -C AA GGA AGC AGG ACA CCA AT-30 (reverse). The PCR was carried out in GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA, USA). The final reaction mixtures consisted of 10 ng/mL of extracted DNA, 200 pmol of each of the oligonucleotide primers, and ThermoHotStarr Mastermix 1 (JMR). The reaction protocol, performed in a microcentrifuge tube, was as follows: (1) denaturation at 95 C for 10 minutes; (2) amplification for 40 cycles at 95 C for 1 minute, 55 C for 1 minute, and 72 C for 1 minute; and (3) extension at 72 C for 10 minutes. The PCR products were sequenced by ABI PRISM BigDye Terminator v3.1 Cycle Sequencing Kit and 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). Sequencing analyses were performed using the National Center for Biotechnology Information database.

Results A total of 85 melanomas were included in this study. These cases consisted of 29 acral melanomas, 28 mucosal melanomas, and 28 cutaneous melanomas. Approximately 24% of patients were diagnosed after recurrent (n ¼ 10) or metastatic (n ¼ 10) lesions had occurred, with cutaneous melanomas occurring least frequently in this setting (Table 1). There was a slight female predominance among these 85 patients, with an overall male-to-female ratio of 1:1.23 (Table 1). The female predominance was more pronounced for patients with recurrent or metastatic diseases at diagnosis (1:1.85) and those with mucosal melanoma (1:2.5). Therefore, the female predominance was even dramatic (1:8) for patients with combinations of both, that is, mucosal melanoma with recurrent/metastatic disease at diagnosis. By contrast, there was a male predominance in cutaneous melanoma, with an overall male-to-female ratio of 1.5:1. The mean age of these 85 patients was 64 years (range: 22e101 years, median age of 62 years), with approximately 90% of patients between 40 and 89 years of age. Five patients were under the age of 40, and three patients were 90 years of age or older. In general, patients with cutaneous melanoma (mean age: 56 years) were younger than those with acral (mean age: 69 years) and mucosal melanomas (mean age: 67 years) (Table 1). However, there was no significant age difference between patients with primary tumors and those with recurrent/metastatic tumors at diagnosis. Among these 85 melanomas, 80 specimens could be analyzed for their DNA sequences. We found that 15 (17%) tumors had a KIT

Table 1 Sex and age distribution among various groups of melanoma. Melanoma

Acral

Mucosal

Cutaneous

Total

a

29 13:16 69 21 9:12 69 8 4:4 67 7 5:2 75

28 8:20 67 19 7:12 68 9 1:8 64 6 1:5 59

28 17:11 56 25 15:10 55 3 2:1 62 2 1:1 70

85 38:47 64 65 31:34 64 33 7:13 66 15 7:8 68

Total Male: female Mean age (y) Primarya,b Male: female Mean age (y) Secondarya,b Male: female Mean age (y) KIT mutanta Male: female Mean age (y)

y ¼ years. a Case number. b For recurrent or metastatic melanomas, only the site of secondary tumor is considered.


mutation including seven male patients (mean age: 70 years) and eight females (mean age: 66 years) (Table 1). KIT mutations were detected in 25% (7/28) of acral melanomas, 22% (6/27) of mucosal melanomas, and 8% (2/25) of cutaneous melanomas. KIT mutations showed a male predominance in acral melanomas (5 males and 2 females) and female predominance in mucosal melanomas (1 male and 5 females). KIT mutation in mucosal melanomas (mean age: 59 years) occurred more than a decade earlier than that in acral melanomas (mean age: 75 years) and cutaneous melanomas (mean age: 70 years). Among 27 mucosal melanomas analyzed for KIT mutation, 10 occurred in the head and neck region (including 7 tumors of the nasal cavity/sinus, 1 of the tongue, 1 of the conjunctiva, and 1 of the middle ear), six in the anorectum, five in the gastrointestinal tract other than the anorectum, five in the genitourinary tract, and one in the gallbladder. KIT mutations were detected in three of six melanomas from the anorectum, two of five from the genitourinary tract, and one of six from the gastrointestinal tract (excluding the anorectum). Of 15 KIT mutations, seven occurred in exon 11, five in exon 13, three in exon 17, and none in exon 18 (Table 2). The most common mutation was L576P (n ¼ 6) and K642E (n ¼ 6), followed by D820Y (n ¼ 2), and finally Y559D (n ¼ 1), L657F (n ¼ 1), and G827R (n ¼ 1). Although only three of 20 (15%) recurrent/metastatic melanomas had a KIT mutation, the majority of them were D820Y (2/3). By contrast, D820Y was not detected in all primary melanomas at diagnosis, which had a KIT mutation rate of 20% (12/61) and approximately 92% (11/12) occurred in exons 11 and 13. There were three cases of in situ melanoma in this study. Two of them had DNA sequence data, and were wild type for the analyzed KIT exons. Discussion By collecting unselected cases from two hospitals that are not known for being dermatology referral centers, we analyzed 85 melanomas in this study. A larger series of melanoma in Taiwan was studied by Chang et al in 2004,18 who collected 181 cases from a single institute that is known for being a referral center for dermatology patients. Both our series and theirs showed a similar median age (62 vs. 61 years), a similar male-to-female ratio (1:1.23 vs. 1:1.13), and a similar percentage of metastatic disease at the time of diagnosis (24% vs. 25%), suggesting that both were representative for melanoma patients in Taiwan. It is of interest that both cohorts pointed out that acral melanoma is a common type of melanoma in Taiwan (Chinese population) and mucosal melanoma is not uncommon either. By contrast, acral melanoma is uncommon in Caucasians, accounting for less than 10% of all melanomas, and mucosal melanoma is even rare, accounting for only 1.2% of all melanomas in Western countries.19 In this study, patients with cutaneous melanomas were about a decade younger than those with acral and mucosal melanomas,

Table 2 Exon distribution among various groups of melanoma. Melanoma

Acral

Mucosal

Cutaneous

Total

Exon 11

L576P 3 V559D 1 L657F 1 K642E 1 D820Y 1a

L576P 3

d

7

K642E 2

K642E 1

5

D820Y 1b

G827R 1c

3

Exon 13 Exon 17 a b c

Melanoma of lymph node, metastasized from thumb. Melanoma of vulva, recurrent. Cutaneous melanoma.

9

Table 3 Comparison of KIT mutation rate in various types of melanoma. Literature

Exons tested

Acral

Mucosal

Curtin et al14 (2006, USA) Beadling et al15 (2008, USA) Carvajal et al16 (2011, USA) Kong et al17 (2011, China) This study (Taiwan) Total

11, 13, 17, 18

11% (3/28)

21% (8/38)

8% (3/36)

11, 13, 17

23% (3/13)

16% (7/45)

2% (1/58)

18% (17/93)

16% (5/32)

9, 11, 13, 17, 18 21% (18/84)

Cutaneous

9, 11, 13, 17, 18 12% (23/193) 10% (16/167) 12% (11/91) 11, 13, 17, 18

25% (7/28) 22% (6/27) 16% (54/346) 15% (54/370)

8% (2/25) 9% (22/242)

which is consistent with the opinion that cutaneous melanoma is easier to be detected than other types. This is echoed by the observation that cutaneous melanomas occurred least frequently in recurrent or metastatic disease. There was no significant difference in age between patients with primary melanoma and those who presented with recurrent/metastatic melanoma at initial diagnosis, suggesting a rapid progression of this disease. In addition, onethird of patients with melanoma were diagnosed after recurrent/ metastatic disease had occurred, indicating that many patients neglected this malignant disease in the first place. To our knowledge, KIT mutation rates in cutaneous, mucosal, and acral melanomas have so far been compared in four studies (Table 3).14e17 Two of them showed that the KIT mutation rate was lower in cutaneous melanomas than in acral and mucosal melanomas in American patients. However, one of the four studies, which analyzed Chinese patients, showed no significant difference of KIT mutation rates among these three types of melanoma. These opposite results may suggest a racial difference in KIT mutation frequency among cutaneous, mucosal, and acral melanomas. However, our study on Chinese patients showed that the KIT mutation rate in mucosal and acral melanomas was threefold higher than that in cutaneous melanoma, suggesting that there is no racial difference in this issue. When putting all analyzed patients in these five studies together (Table 3), we noticed that the KIT mutation rate in cutaneous melanomas (9%) is indeed lower than that of acral (16%) and mucosal (15%) melanomas.

Table 4 Comparison of KIT mutation in acral melanoma. Exons tested

Mutation rate

Mutated exon

Curtin et al14 (2006, USA) Beadling et al15 (2008, USA) Ashida et al22 (2008, Japan) Torres-Cabala et al23 (2009, USA)

11, 13, 17, 18

11% (3/28)

11, 13, 17

23% (3/13)

exon 11 (1), exon 13 (2) d

11, 13, 17, 18

13% (2/16)

11, 13, 17

13% (5/39)

Handolias et al6 (2010, Australia) Carvajal et al16 (2011, USA) Yun et al24 (2011, Korea) Kong et al17 (2011, China) This study (Taiwan)

11, 13, 17

6% (1/16)

exon exon exon exon exon exon

9, 11, 13, 17, 18

21% (18/84)

d

11, 13, 17, 18

10% (4/40)

exon 11 (3), exon 13 (1)

9, 11, 13, 17, 18

12% (23/193)

11, 13, 17, 18

25% (7/28)

Total

14% (66/457)

exon exon exon exon exon exon

13 17 11 13 17 17

11 13 17 11 13 17

(1), (1) (3), (1), (1) (1)

(4), (2), (1) (12), (7), (4)

10


Table 5 Comparison of mucosal melanoma in terms of anatomical locations. Exons tested

Curtin et al14 (2006, USA) Antonescu et al25 (2007, USA) Rivera et al26 (2007, Asia) Beadling et al15 (2008, USA) Satzger et al20 (2008, Germany) Ashida et al22 (2008, Japan) Torres-Cabala et al23 (2009, USA) Sekine et al27 (2009, Japan) Handolias et al6 (2010, Australia) Abysheva et al28 (2011, Rossia) Carvajal et al16 (2011, USA) Yun et al24 (2011, Korea) Kong et al17 (2011, China) Ni et al29 (2012, China) This study (Taiwan) Total a

Head and necka

Total

11, 13, 17, 18 11, 13, 17 11, 13 11, 13, 17 9, 11, 13, 17, 18 11, 13, 17, 18 11, 13, 17 11, 13, 17 11, 13, 17 9, 11, 13, 17 9, 11, 13, 17, 18 11, 13, 17, 18 9, 11, 13, 17, 18 9, 11, 13, 17 11, 13, 17, 18

3/38 3/20 4/15 7/45 6/37 0/3 9/52 1/16 6/16 8/48 17/93 3/55 16/167 7/40 6/27 102/672 (15%)

Genitourinary tract

Not specified d 4/15 4/49a 2/18 Not specified Not specified d 0/3 0/7 Not specified Not specified Not specified d 0/10a 10/102 (10%)

Gastrointestinal tract

Others

Anorectum

Others

d d d 3/9

3/20 d 4/9 1/8

d d d d

d d d d

d 2/7 1/15

d 4/6 6/42

1/16 d 1/1

d d 0/1

7/40 2/5 15/76 (20%)

d 3/6 21/91 (23%)

d 1/5 3/22 (14%)

d 0/1 0/2 (0%)

Including conjunctiva.

Table 6 Comparison of type and frequency of KIT mutation. Ref

7

8

9

24

17

25

26

20

3

27

29

21

a

Total

Exon 11

K550N M552I Y553C Y553N W557R K558N K558R V559A V559D V559G V560D G562V N566D Y568C V569G T574A L576P

M1 M1 M1 A1 M1

17

R634W K642E

N655K L657F K807R D816H

1 1b

M1

1b 1b

M2

A1

A1

M2c M1 M1 M2 S1

5b

M3

M3

A1 M3

a b c

Unspecified type. Double mutation. This study.

M1

M1

M1

M1 A1 M1

M2

A3 M3

M1 M2c A1 A1 A1 M3

M1

M2

A1

A1 M1

M1

M2

M2c

M1

M1

A1

1 2 7

A1 M2 S1 A1 A1 M1

M1 S1 M1 A1 M1

D820Y

18

M1

M1

D816N D816V

N822I N822K Y823D G827R A829P I841V

A1 A1

M1

M1 M1

W577R E583V Del 13

b

A1

A1 M1

A1 M1 M1

b

S1 S1 M1

1 15

2 1 1 2 1 3

M1

M1c 1

1 1 1 1 3 1 1 6 2 2 1 2 1 1 1 1 31

5 1 1 1 1 1 1


For acral melanoma, the KIT mutation rate (25%) in our series is close to that of Beadling et al (23%)15 and Carvajal et al (21%),16 but is higher than that of Curtin et al (11%)14 and Handolias et al (6%).6 We have no explanation for this disparity. However, it should be addressed that acral melanoma is usually very small in size, thus making a false negative result difficult to avoid. To date, KIT mutations have been analyzed in 457 acral melanomas (Table 4).6,14e17,22e24 The average KIT mutation rate in these patients is approximately 14%. As for mucosal melanoma, the KIT mutation rate in our series was 22%, similar to the findings of Curtin et al,14 who showed a mutation rate of 21%. Since 2006, there have been 15 reports including ours that analyzed KIT mutations in mucosal melanomas (Table 5).6,14e17,20,22e29 From these 672 cases, we noticed that the KIT mutation rate in mucosal melanoma is approximately 15%, ranging from 5% (Korea) to 38% (Australia). In this study, we found that KIT mutations occur most commonly in the anorectum and genitourinary tract. Although our series is small in sample size, this trend is compatible with the data when we accumulated all analyzed patients from these 15 reports together (Table 5). Of interest, none of our mucosal melanomas of the head and neck had KIT mutation and eight of 11 melanomas of this category occurred in the nasal cavity/sinus. Although there was no difference in the mean age of onset between acral and mucosal melanomas, KIT mutation in mucosal melanomas occurred more than a decade earlier than that in acral melanomas. Regarding the sex factor, we noticed that KIT mutation among our few cases is more commonly seen in male patients with acral melanoma despite a male predominance in prevalence is characteristic for acral melanoma. By contrast, however, KIT mutation is more commonly seen in female patients with mucosal melanoma (a male-to-female ratio of 1:5) even though a female predominance is already noticed in mucosal melanomas (a maleto-female ratio of 1:2.5). To our knowledge, there are 66 KIT-mutated acral melanomas and 102 KIT-mutated mucosal melanomas reported in the literature, including this study (Table 6). Among the analyzed exons in these melanomas, KIT mutation was most commonly seen in exon 11, followed by exon 13, and finally exon 17. Exon 18 mutation was not found in all acral melanomas, but was present in one mucosal melanoma.20 Of these, a total of 104 different point mutations of KIT have been described (Table 6). The most common one is L576P (30%), followed by K642E (14%), then V559A (6%) and D820Y (5%). These four mutations account for 55% of all KIT mutations in the literature, suggesting that efforts should be focused on these hot spots when budget for mutational analysis is limited. Indeed, 81% of KIT mutations in our series occur in these hot spots. Because mutational analysis may not be available in some areas around the world, it will be important to know whether KIT mutation more commonly occurs in certain subsets of melanoma so that imatinib can be empirically given to this subset of patients. From this point of view, it appears that male patients with acral melanoma and female patients with mucosal melanoma of the anorectal and genitourinary regions belong to this mutation-prone subset. In our series, 10 of 13 KIT mutants belong to this subset. More specifically, five of 13 male patients with acral melanoma and five of 11 female patients with mucosal melanoma of the anorectal and genitourinary regions had KIT mutation. In short, approximately 42% (10/24) of melanomas belong to this mutation-prone subset. However, it should be addressed that mutant D820Y is reported to be resistant to imatinib.21 Therefore, one-third (8/24) of patients in this subset will be responsive to imatinib treatment. From this point of view, future clinical trials should select the patients of the mutation-prone subset if KIT mutation is not one of the

11

inclusion criteria. Ideally, those with tumors harboring D820Y should be excluded. In summary, we showed that acral and mucosal melanomas have a higher KIT mutation rate than cutaneous melanoma. Furthermore, KIT mutation is most commonly seen in male patients with acral melanoma and female patients with mucosal melanomas of the anorectal and genitourinary regions. When mutational analysis is to be performed, exons 11, 13, and 17 should be examined. Alternatively, four hot spots of KIT should be checked if laboratory analysis is limited. Although no definite conclusions can be drawn from a study with a small sample size, our study shows some characteristics of KIT mutation in melanoma, which may provide practical information in the era of targeted therapy for patients with melanoma. Acknowledgments This study was supported in part by Cathay General Hospital (CGHMR-10126) and Novartis Oncology (Taiwan). The authors thank Ms Hsiang-Hua Yang in the Department of Molecular Medicine of Cathay General Hospital for technical expertise. References 1. Manola J, Atkins M, Ibrahim J, Kirkwood J. Prognostic factors in metastatic melanoma: a pooled analysis of Eastern Cooperative Oncology Group Trials. J Clin Oncol 2000;18:3782e93. 2. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347: 472e80. 3. Ugurel S, Hildenbrand R, Zimpfer A, et al. Lack of clinical efficacy of imatinib in metastatic melanoma. Br J Cancer 2005;92:1398e405. 4. Wyman K, Atkins MB, Prieto V, et al. Multicenter Phase II trial of high-dose imatinib mesylate in metastatic melanoma: significant toxicity with no clinical efficacy. Cancer 2006;106:2005e11. 5. Kim KB, Eton O, Davis DW, et al. Phase II trial of imatinib mesylate in patients with metastatic melanoma. Br J Cancer 2008;99:734e40. 6. Handolias D, Hamilton AL, Salemi R, et al. Clinical responses observed with imatinib or sorafenib in melanoma patients expressing mutations in KIT. Br J Cancer 2010;102:1219e23. 7. McDonnell K, Betz B, Fullen D, Lao CD. V559A and N822I double KIT mutant melanoma with predictable response to imatinib? Pigment Cell Melanoma Res 2011;24:390e2. 8. Lutzky J, Bauer J, Bastian BC. Dose-dependent, complete response to imatinib of a metastatic mucosal melanoma with a K642E KIT mutation. Pigment Cell Melanoma Res 2008;21:492e3. 9. Hodi FS, Friedlander P, Corless CL, et al. Major response to imatinib mesylate in KIT-mutated melanoma. J Clin Oncol 2008;26:2046e51. 10. Guo J, Si L, Kong Y, et al. Phase II, open-label, single-arm trial of imatinib mesylate in patients with metastatic melanoma harboring c-Kit mutation or amplification. J Clin Oncol 2011;29:2904e9. 11. Brown MC, Casasola RJ. Complete response in a melanoma patient treated with imatinib. J Laryngol Otol 2012;126:638e40. 12. Went PT, Dirnhofer S, Bundi M, et al. Prevalence of KIT expression in human tumors. J Clin Oncol 2004;22:4514e22. 13. Willmore-Payne C, Holden JA, Tripp S, Layfield LJ. Human malignant melanoma: detection of BRAF- and c-kit-activating mutations by high-resolution amplicon melting analysis. Hum Pathol 2005;36:486e93. 14. Curtin JA, Busam K, Pinkel D, Bastian BC. Somatic activation of KIT in distinct subtypes of melanoma. J Clin Oncol 2006;24:4340e6. 15. Beadling C, Jacobson-Dunlop E, Hodi FS, et al. KIT gene mutations and copy number in melanoma subtypes. Clin Cancer Res 2008;14:6821e7. 16. Carvajal RD, Antonescu CR, Wolchok JD, et al. KIT as a therapeutic target in metastatic melanoma. JAMA 2011;305:2327e34. 17. Kong Y, Si L, Zhu Y, et al. Large-scale analysis of KIT aberrations in Chinese patients with melanoma. Clin Cancer Res 2011;17:1684e91. 18. Chang JW, Yeh KY, Wang CH, et al. Malignant melanoma in Taiwan: a prognostic study of 181 cases. Melanoma Res 2004;14:537e41. 19. Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report on cutaneous and noncutaneous melanoma: a summary of 84,836 cases from the past decade. The American College of Surgeons Commission on Cancer and the American Cancer Society. Cancer 1998;83:1664e78. 20. Satzger I, Schaefer T, Kuettler U, et al. Analysis of c-KIT expression and KIT gene mutation in human mucosal melanomas. Br J Cancer 2008;99:2065e9. 21. Antonescu CR, Besmer P, Guo T, et al. Acquired resistance to imatinib in gastrointestinal stromal tumor occurs through secondary gene mutation. Clin Cancer Res 2005;11:4182e90.

12


22. Ashida A, Takata M, Murata H, Kido K, Saida T. Pathological activation of KIT in metastatic tumors of acral and mucosal melanomas. Int J Cancer 2009;124:862e8. 23. Torres-Cabala CA, Wang WL, Trent J, et al. Correlation between KIT expression and KIT mutation in melanoma: a study of 173 cases with emphasis on the acral-lentiginous/mucosal type. Mod Pathol 2009;22:1446e56. 24. Yun J, Lee J, Jang J, et al. KIT amplification and gene mutations in acral/mucosal melanoma in Korea. APMIS 2011;119:330e5. 25. Antonescu CR, Busam KJ, Francone TD, et al. L576P KIT mutation in anal melanomas correlates with KIT protein expression and is sensitive to specific kinase inhibition. Int J Cancer 2007;121:257e64.

26. Rivera RS, Nagatsuka H, Gunduz M, et al. C-kit protein expression correlated with activating mutations in KIT gene in oral mucosal melanoma. Virchows Arch 2008;452:27e32. 27. Sekine S, Nakanishi Y, Ogawa R, Kouda S, Kanai Y. Esophageal melanomas harbor frequent NRAS mutations unlike melanomas of other mucosal sites. Virchows Arch 2009;454:513e7. 28. Abysheva SN, Iyevleva AG, Efimova NV, et al. KIT mutations in Russian patients with mucosal melanoma. Melanoma Res 2011;21:555e9. 29. Ni S, Huang D, Chen X, et al. c-kit gene mutation and CD117 expression in human anorectal melanomas. Hum Pathol 2012;43:801e7.