SEARCH

SEARCH BY CITATION

Keywords:

  • ovarian cancer;
  • BRAF;
  • outcomes;
  • low-grade serous ovarian cancer;
  • early stage disease

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

BACKGROUND:

Low-grade serous (LGS) ovarian cancer is a chemoresistant disease that accounts for 10% of serous ovarian cancers. Prior studies have reported that 28% to 35% of serous borderline (SB)/LGS ovarian tumors harbor a BRAF mutation, suggesting that BRAF inhibitors may be a rational therapeutic approach for this disease. In the current study, the authors sought to determine whether BRAF or KRAS mutation status was associated with disease stage and/or histology in patients with SB and LGS ovarian cancer.

METHODS:

Genetic profiles were constructed for 75 SB and LGS ovarian tumors to determine BRAF and KRAS mutation status. The incidence and identity of BRAF and KRAS mutations were defined, and the results were correlated with disease stage, response to treatment, and overall survival.

RESULTS:

Of 75 samples examined, 56 tumors (75%) had SB histology, and 19 tumors (25%) had LGS histology. Fifty-seven percent of tumors harbored either a KRAS mutation (n = 17) or a BRAF mutation (a valine-to-glutamate substitution at residue 600 [V600E]; n = 26). The BRAF V600E mutation was associated significantly with early disease stage (stage I/II; P < .001) and SB histology (P = .002). KRAS mutations were not associated significantly with disease stage or histology. Of the 22 patients (29%) who required chemotherapy, 20 had tumors with wild-type KRAS/BRAF, 2 had KRAS mutant tumors, and none had tumors that harbored a BRAF mutation. All patients with BRAF tumors remained alive at a median follow-up of 3.6 years (range, 1.9–129.3 months).

CONCLUSIONS:

V600E BRAF mutations were present in 35% of patients who had SB/LGS ovarian cancers. The presence of the BRAF V600E mutation in SB/LGS ovarian cancer was associated with early stage disease and improved prognosis. The authors concluded that patients with SB/LGS ovarian cancer who require systemic therapy are unlikely to have BRAF mutant tumors. Cancer 2013. © 2012 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Mutations in BRAF and KRAS, components of the mitogen-activated protein kinase (MAPK) cascade, are common in low-grade serous and serous borderline ovarian tumors. In contrast, they are present in <1% of high-grade serous ovarian tumors.1-6 Conversely, tumor protein 53 (p53) mutations are present in 96% of high-grade serous ovarian tumors but are rare in low-grade serous tumors.1,7

On the basis of these clinical and genetic differences between low-grade and high-grade serous ovarian cancer, a 2-tiered grading system has been created.8-11 This dualistic model of ovarian carcinogenesis proposes that high-grade disease arises de novo from distal fallopian tube epithelium; whereas low-grade serous ovarian tumors evolve through a step-wise progression from a benign serous cystadenoma, to a serous borderline neoplasm, to an invasive low-grade serous carcinoma (Fig. 1).12 Serous borderline neoplasms can be subclassified further into atypical proliferative serous tumors and noninvasive micropapillary serous carcinomas. When micropapillary serous carcinomas develop invasion, they become synonymous with low-grade serous carcinomas.13,14 The presence of micropapillary features in serous borderline neoplasms is associated with an increased frequency of bilateral ovarian disease, peritoneal implants, and recurrent disease compared with serous borderline neoplasms without micropapillary features.15,16 In contrast to patients with high-grade disease, patients with low-grade serous ovarian cancer present at a younger age (ages 45–57 years), and their tumors typically display a low mitotic index and largely are resistant to chemotherapy (Table 1)17-20 Contrary to previous findings, it was reported recently that BRAF mutation is rare in patients with advanced-stage, low-grade serous ovarian cancer and that patients with BRAF or KRAS mutations may have an improved clinical outcome.7

thumbnail image

Figure 1. (A) A serous borderline (SB) tumor is observed with papillary architecture and abundant micropapillae growing on the surface of the large, papillary fronds. Like low-grade serous (LGS) tumors, SB tumors display low nuclear grade and a low mitotic index but lack stromal invasion. (B) An LGS carcinoma is observed with the typical invasion pattern of micropapillae embedded in stroma surrounded by an artifactual cleft. LGS carcinomas possess low nuclear grade and a low mitotic index compared with high-grade serous (HGS) carcinomas. (C) An HGS carcinoma is observed with high nuclear grade and focal anaplasia, a markedly elevated mitotic index, and glandular architecture with abundant tufting and budding of cells.

Download figure to PowerPoint

The primary objective of the current study was to determine whether BRAF or KRAS mutation status is associated with disease stage and/or histology in patients with low-grade serous and serous borderline ovarian cancer. To achieve this objective, we retrospectively analyzed tumor samples and associated clinical data from all patients with low-grade serous or serous borderline ovarian cancer who underwent surgery at Memorial Sloan-Kettering Cancer Center between 2000 and 2010.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Patient Samples

After Institutional Review Board approval, clinical data were collected on all patients with a diagnosis of low-grade serous or serous borderline ovarian cancer who underwent surgery at Memorial Sloan-Kettering Cancer Center between 2000 and 2010. All included patients were required to have formalin-fixed, paraffin-embedded tissue available from at least 1 prior staging or debulking operation.

The original pathology reports were reviewed to determine stage based on the 2010 American Joint Committee on Cancer (AJCC) staging system (seventh edition) for ovarian and primary peritoneal cancer. Patients' records also were reviewed to determine clinical status, date of diagnosis, date of last follow-up, date of recurrence(s), and treatment history. For all specimens, tumor histology was reviewed and confirmed by a reference specialty pathologist (K.G. or D.D.).

Tissue Analysis

Formalin-fixed, paraffin-embedded tissue samples were macrodissected to remove stromal contamination and to ensure tumor cellularity of ≥80%. Tumor DNA was extracted using the DNeasy tissue kit according to the manufacturer's instructions (Qiagen, Valencia, Calif). Each specimen was analyzed using a custom iPLEX assay (Sequenom, Inc., San Diego, Calif) to detect KRAS and BRAF hotspot mutations.21 Each variant detected was manually reviewed. Tumors that harbored a mutation and had sufficient DNA underwent confirmation of mutation status with an orthogonal method. All primer sequences are available upon request.

Statistical Analysis

This was a single-institution, retrospective analysis of archived tumor tissues and associated clinical variables. For the purposes of this analysis, patients were grouped into those with early stage (stage I or II disease at presentation) or advanced stage (stage III or IV disease at presentation) based on AJCC seventh edition staging. Fisher exact tests were used to determine the association between mutation status (KRAS mutant, KRAS wild type (WT), BRAF mutant, BRAF WT, or KRAS/BRAF WT) and disease stage at presentation and the association between mutation status and histology (serous borderline or low-grade serous). Overall survival intervals were calculated from the date of diagnosis to the date of death or last follow-up and were estimated using the Kaplan-Meier method.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Twelve patients were excluded because of an inadequate quantity of tumor DNA. Seventy-five patients were included in the final analysis. Eighty percent of samples (n = 60) were collected from a primary staging or debulking operation, 16% (n = 12) were collected from a secondary debulking, 3% (n = 2) were collected from a tertiary debulking, and 1% (n = 1) were collected from a quaternary debulking. Seven patients had died and 68 remained alive at the time of this report. The median follow-up for the living patients was 35.9 months (range, 0.8–129.3 months) (Table 2).

Table 1. Clinical Features of Low-Grade Versus High-Grade Serous Ovarian Cancer
  Serous Ovarian Cancer
Clinical FeatureReferenceLow-GradeHigh-Grade
  1. Abbreviations: HPFs, high-power fields. y, years.

Median age, y 45–5755–65
Patients with serous ovarian cancer, %Schmeler, 200819∼10∼90
Growth patternVang, 200910Micropapillary-richLarge papillae, solid and glandular growth
Nuclear atypiaBodurka, 20128Mild-to-moderate atypiaMarked atypia
Mitotic rate, mitoses per 10 HPFsBodurka, 20128≤12>12
Five-y survival rate for stage >1, %Vang, 20091040–569–34
Response rate to platinum-based neoadjuvant chemotherapy, %Schmeler, 200819480
Table 2. Population Characteristics
CharacteristicNo. of Patients (%)
Total75
Age at diagnosis: Median/mean [range], y47/46.3 [15–79]
Disease stage 
 I/II45 (60)
 III/IV30 (40)
Histology 
 Serous borderline56 (74.7)
 Low-grade serous19 (25.3)
Mutation 
 KRAS G12D/G12V17 (22.7)
 BRAF V600E26 (34.7)
 WT for KRAS and BRAF32 (42.7)
Clinical status 
 Alive at last follow-up68 (90.7)
 Dead7 (9.3)

Fifty-seven percent of tumors harbored either a BRAF mutation (n = 26) or a KRAS mutation (n = 17). Figure 2 displays a representative mass spectrometry trace of a BRAF valine-to-glutamate substitution at residue 600 (V600E), a KRAS glycine-to-aspartic acid substitution at codon 12 (G12D), and a KRAS glycine-to-valine substitution at codon 12 (G12V) from three separate tumors. The same mutations were detected using Sanger sequencing. BRAF and KRAS mutations were mutually exclusive. All BRAF mutations were the V600E substitution. KRAS mutations were either G12D (n = 11) or G12V (n = 6) (Fig. 3, top). All 11 KRAS G12D mutations were confirmed by Sanger sequencing or by a second Sequenom assay. Two of 6 KRAS G12V mutations and 7 of 26 BRAF V600E mutations could not be confirmed by Sanger sequencing or by a second Sequenom assay because of insufficient amounts of DNA. The BRAF V600E mutation (n = 26) was associated significantly with early stage disease (P < .001) and serous borderline histology (P = .002) compared with BRAF WT tumors (n = 49) (Table 3). The presence of the BRAF V600E mutation also was associated significantly with early stage disease (P < .001) and serous borderline histology (P < .001) compared with KRAS/BRAF WT tumors (n = 32). In contrast, tumors with KRAS mutations (n = 17) were not associated significantly with disease stage or histology compared with KRAS WT tumors (n = 58) or KRAS/BRAF WT tumors (n = 32). Eleven of 56 serous borderline tumors (20%) had micropapillary features. Two of those 11 tumors (18%) with micropapillary features harbored BRAF mutations, whereas 23 of 45 tumors (51%) without micropapillary features had a BRAF mutation identified.

thumbnail image

Figure 2. Representative mass spectrometry traces and corresponding Sanger traces are shown for 3 tumors that harbored (A) a (B) G indicates guanine; T, thymine; C, cytosine; A, adenine.

Download figure to PowerPoint

thumbnail image

Figure 3. (Top) Of 75 patients with low-grade serous or borderline serous ovarian cancer, 17 harbored either a KRAS G12D (n = 11) or a KRAS G12V (n = 6) mutation, 26 harbored a BRAF V00E mutation, and 32 had WT KRAS and BRAF. (Bottom) Overall, 62.5% of patients who had WT KRAS and BRAF received chemotherapy in either the adjuvant setting or the recurrent setting compared with 11.8 % of patients with KRAS mutations and 0% of patients with BRAF mutations.

Download figure to PowerPoint

Table 3. BRAF Mutation Is Associated Significantly With Stage and Histology in Patients With Ovarian Cancer
  No. of Patients (%) No. of Patients (%) 
MutationTotal No.Stage I/IIStage III/IVPSBLGSP
  1. Abbreviations: +, positive.

BRAF V600E+2624 (53.3)2 (6.7)<.00125 (44.6)1 (5.3).002
BRAF WT4921 (46.7)28 (93.3) 31 (55.4)18 (94.6) 
KRAS G12D+/G12V+1710 (22.2)7 (23.3)1.00014 (25)3 (15.8).534
KRAS WT5835 (77.8)23 (76.7) 42 (75)16 (84.2) 

Twenty-two patients (29%) received chemotherapy (4 serous borderline tumors, 18 low-grade serous tumors) in either the adjuvant or the recurrent setting; 2 patients had KRAS mutant tumors, 0 had BRAF mutant tumors, and 20 had KRAS/BRAF WT tumors (Fig. 3, bottom). The need for chemotherapy was inversely associated with BRAF mutation status (P < .001). All patients with BRAF mutant tumors remained alive at a median follow-up of 43.3 months (range, 1.9–129.3 months), with the suggestion of improved overall survival compared with patients who had KRAS mutant or KRAS/BRAF WT tumors. Median survival has not yet been reached in any of the mutation cohorts (Fig. 4).

thumbnail image

Figure 4. (Top) This Kaplan-Meier curve illustrates overall survival (OS) for all 75 patients from the current study. Seven patients died. (Bottom) This Kaplan-Meier curve illustrates the proportion of surviving patients according to mutation type. The median OS had not been reached for any group at the time of the current report.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Our current findings indicate that, within the disease continuum of serous borderline and low-grade serous ovarian cancer, the V600E BRAF mutation is associated with early stage at presentation, serous borderline histology, and improved outcome. We postulate that the presence of a BRAF mutation in patients with serous borderline disease prevents progression to more aggressive disease. This is in contrast to papillary thyroid cancer, in which the presence of the V600E BRAF mutation is associated with advanced stage and a poor prognosis, and metastatic colon cancer, in which the V600E BRAF mutation also is associated with a poor prognosis, indicating that BRAF mutation status has tumor lineage-specific prognostic implications.22-27

Serous borderline and low-grade serous ovarian cancers typically are chemotherapy-resistant diseases, and the reported response rates to cytotoxic chemotherapy are 4% in the neoadjuvant setting and 2.1% to 4.9% in the recurrent setting.18,19 Given the high prevalence of BRAF and KRAS mutations in serous borderline and low-grade serous ovarian cancers, there has been recent interest in testing inhibitors targeting the MAPK pathway in patients with advanced disease. To date, 29% of patients in the current study have received at least 1 line of systemic chemotherapy; however, none of the patients who required systemic therapy had a BRAF mutant tumor. These findings suggest that patients with aggressive, low-grade serous ovarian cancers—the population most in need of novel, effective, systemic therapies—are unlikely to harbor BRAF mutant tumors. Because the selective RAF inhibitor vemurafenib lacks activity in KRAS mutant and BRAF/RAS WT tumors, our data suggest that vemurafenib will be of limited utility in this disease.28

Notably, a recently completed phase 2 trial of the MEK inhibitor AZD6244 in women with recurrent, low-grade serous carcinoma of the ovary or peritoneum reported a radiographic response rate of 15.4%. Of the 34 patients in that study who had sufficient DNA for mutation analysis, only 2 tumors harbored BRAF mutations, supporting our finding that BRAF mutation is rare in those patients who require systemic therapy. A correlation between RAS or BRAF mutation status and disease response was not observed.29 This lack of correlation between BRAF and RAS mutation status and response to AZD6244 may have been a result of the presence of occult genomic events in the BRAF and RAS WT cohort, which phenocopy the effects of RAS mutation, such as neurofibromin 1 (NF1) mutation or loss. It is noteworthy, however, that the 15.4% response rate observed after treatment with AZD6244 was markedly higher than what was reported previously with cytotoxic chemotherapies in patients with low-grade serous ovarian cancer. These results indicate that a subset of low-grade serious ovarian cancers depend on MEK activity and that further evaluation of MEK inhibitors is warranted in this disease.

In summary, low-grade serous ovarian cancer is a chemotherapy-resistant disease in which patients have limited systemic treatment options. Our results suggest that the finding of a BRAF mutation predicts for a favorable outcome in surgically treated patients. Testing for BRAF mutation in newly diagnosed patients with serous borderline histology may serve as a powerful prognostic tool for identifying those patients who are unlikely to progress to more aggressive histology and advanced disease.

Our finding that BRAF mutations are rarely present in patients who require systemic therapy suggests that highly selective RAF inhibitors may have limited utility in this disease. Further studies of MEK and ERK inhibitors are warranted given the promising early clinical results with the selective MEK inhibitor AZD6244. A detailed exploration of the genetic basis of BRAF/KRAS WT serous borderline and low-grade serous ovarian cancer also is warranted, because such efforts may result in the identification of novel therapeutic targets in the cohort of patients with the greatest risk of ovarian cancer-specific mortality.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

This research was funded in part by a Stand Up to Cancer Dream Team Translational Research Grant, a Program of the Entertainment Industry Foundation (SU2C-AACR-DT0209; Dr. Solit and Dr. Aghajanian) and by a Cycle for Survival Grant (Dr. Grisham, Dr. Aghajanian, and Dr. Iyer).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES
  • 1
    Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature. 2011; 474: 609615.
  • 2
    Ho CL, Kurman RJ, Dehari R, Wang TL, Shih IeM. Mutations of BRAF and KRAS precede the development of ovarian serous borderline tumors. Cancer Res. 2004; 64: 69156918.
  • 3
    Jones S, Wang TL, Kurman RJ, et al. Low-grade serous carcinomas of the ovary contain very few point mutations. J Pathol. 2012; 226: 413420.
  • 4
    Mayr D, Hirschmann A, Lohrs U, Diebold J. KRAS and BRAF mutations in ovarian tumors: a comprehensive study of invasive carcinomas, borderline tumors and extraovarian implants. Gynecol Oncol. 2006; 103: 883887.
  • 5
    Sieben NL, Macropoulos P, Roemen GM, et al. In ovarian neoplasms, BRAF, but not KRAS, mutations are restricted to low-grade serous tumours. J Pathol. 2004; 202: 336340.
  • 6
    Singer G, Oldt R 3rd, Cohen Y, et al. Mutations in BRAF and KRAS characterize the development of low-grade ovarian serous carcinoma. J Natl Cancer Inst. 2003; 95: 484486.
  • 7
    Wong KK, Tsang YT, Deavers MT, et al. BRAF mutation is rare in advanced-stage low-grade ovarian serous carcinomas. Am J Pathol. 2010; 177: 16111617.
  • 8
    Bodurka DC, Deavers MT, Tian C, et al. Reclassification of serous ovarian carcinoma by a 2-tier system: a Gynecologic Oncology Group study. Cancer. 2012; 118: 30873094.
  • 9
    Singer G, Stohr R, Cope L, et al. Patterns of p53 mutations separate ovarian serous borderline tumors and low- and high-grade carcinomas and provide support for a new model of ovarian carcinogenesis: a mutational analysis with immunohistochemical correlation. Am J Surg Pathol. 2005; 29: 218224.
  • 10
    Vang R, Shih IeM, Kurman RJ. Ovarian low-grade and high-grade serous carcinoma: pathogenesis, clinicopathologic and molecular biologic features, and diagnostic problems. Adv Anat Pathol. 2009; 16: 267282.
  • 11
    Vang R, Shih IeM, Salani R, Sugar E, Ayhan A, Kurman RJ. Subdividing ovarian and peritoneal serous carcinoma into moderately differentiated and poorly differentiated does not have biologic validity based on molecular genetic and in vitro drug resistance data. Am J Surg Pathol. 2008; 32: 16671674.
  • 12
    Russell SE, McCluggage WG. A multistep model for ovarian tumorigenesis: the value of mutation analysis in the KRAS and BRAF genes. J Pathol. 2004; 203: 617619.
  • 13
    Shih IeM, Kurman RJ. Ovarian tumorigenesis: a proposed model based on morphological and molecular genetic analysis. Am J Pathol. 2004; 164: 15111518.
  • 14
    Smith Sehdev AE, Sehdev PS, Kurman RJ. Noninvasive and invasive micropapillary (low-grade) serous carcinoma of the ovary: a clinicopathologic analysis of 135 cases. Am J Surg Pathol. 2003; 27: 725736.
  • 15
    Bell DA, Longacre TA, Prat J, et al. Serous borderline (low malignant potential, atypical proliferative) ovarian tumors: workshop perspectives. Hum Pathol. 2004; 35: 934948.
  • 16
    Silva EG, Gershenson DM, Malpica A, Deavers M. The recurrence and the overall survival rates of ovarian serous borderline neoplasms with noninvasive implants is time dependent. Am J Surg Pathol. 2006; 30: 13671371.
  • 17
    Crispens MA, Bodurka D, Deavers M, Lu K, Silva EG, Gershenson DM. Response and survival in patients with progressive or recurrent serous ovarian tumors of low malignant potential. Obstet Gynecol. 2002; 99: 310.
  • 18
    Gershenson DM, SunCC, Bodurka D, et al. Recurrent low-grade serous ovarian carcinoma is relatively chemoresistant. Gynecol Oncol. 2009; 114: 4852.
  • 19
    Schmeler KM, Sun CC, Bodurka DC, et al. Neoadjuvant chemotherapy for low-grade serous carcinoma of the ovary or peritoneum. Gynecol Oncol. 2008; 108: 510514.
  • 20
    Shvartsman HS, Sun CC, Bodurka DC, et al. Comparison of the clinical behavior of newly diagnosed stages II–IV low-grade serous carcinoma of the ovary with that of serous ovarian tumors of low malignant potential that recur as low-grade serous carcinoma. Gynecol Oncol. 2007; 105: 625629.
  • 21
    Janakiraman M, Vakiani E, Zeng Z, et al. Genomic and biological characterization of exon 4 KRAS mutations in human cancer. Cancer Res. 2010; 70: 59015911.
  • 22
    Colanta A, Lin O, Tafe L, et al. BRAF mutation analysis of fine-needle aspiration biopsies of papillary thyroid carcinoma: impact on diagnosis and prognosis. Acta Cytol. 2011; 55: 563569.
  • 23
    Faris JE, Ryan DP. Trees, forests, and other implications of a BRAF mutant gene signature in patients with BRAF wild-type disease. J Clin Oncol. 2012; 30: 12551257.
  • 24
    Kalady MF, Dejulius KL, Sanchez JA, et al. BRAF mutations in colorectal cancer are associated with distinct clinical characteristics and worse prognosis. Dis Colon Rectum. 2012; 55: 128133.
  • 25
    Popovici V, Budinska E, Teipar S, et al. Identification of a poor-prognosis BRAF-mutant-like population of patients with colon cancer. J Clin Oncol. 2012; 30: 12881295.
  • 26
    Roth AD, Teipar S, Delorenzi M, et al. Prognostic role of KRAS and BRAF in stage II and III resected colon cancer: results of the translational study on the PETACC-3, EORTC 40993, SAKK 60-00 trial. J Clin Oncol. 2010; 28: 466474.
  • 27
    Wang W, Zhao W, Wang H, et al. Poorer prognosis and higher prevalence of BRAF (V600E) mutation in synchronous bilateral papillary thyroid carcinoma. Ann Surg Oncol. 2012; 19: 3136.
  • 28
    Joseph EW, Pratilas CA, Poulikakos PI, et al. The RAF inhibitor PLX4032 inhibits ERK signaling and tumor cell proliferation in a V600E BRAF-selective manner. Proc Natl Acad Sci U S A. 2010; 107: 14903-14908.
  • 29
    Farley J, Brady W, Birrer M, et al. Gynecologic Oncology Group. A phase II trial of AZD6244 in women with recurrent low-grade serous carcinoma of the ovary or peritoneum. Paper presented at: 17th International Meeting of the European Society of Gynaecological Oncology; September 11–14, 2011; Milan, Italy.