Proteomic Mass Spectrometry Imaging for Skin Cancer Diagnosis




Mass spectrometry imaging can be successfully used for skin cancer diagnosis, particularly for the diagnosis of challenging melanocytic lesions. This method analyzes proteins within benign and malignant melanocytic tumor cells and, based on their differences, which constitute a unique molecular signature of 5 to 20 proteins, can render a diagnosis of benign nevus versus malignant melanoma. Mass spectrometry imaging may assist in the differentiation between metastases and nevi as well as between proliferative nodules in nevi and melanoma arising in a nevus. In the difficult area of atypical Spitzoid neoplasms, mass spectrometry diagnosis can predict clinical outcome better than histopathology.


Key points







  • •

    Mass spectrometry imaging studies the proteome and identifies molecular signatures comprising a unique combination of 5 to 20 proteins that enable a specific diagnosis of cancer or disease.


  • •

    There are proteomic differences between the melanocytic cells of benign nevi and malignant melanomas, which can be identified by mass spectrometry imaging.


  • •

    This technology may assist in the classification of diagnostically challenging melanocytic lesions.


  • •

    Mass spectrometry imaging diagnosis of atypical Spitzoid neoplasms seems to predict the clinical outcome better than histopathology and is strongly associated with aggressive clinical behavior.


  • •

    Mass spectrometry analysis using a proteomic signature may improve the diagnosis and prediction of outcome/risk stratification for patients with atypical Spitzoid neoplasms and other types of challenging melanocytic lesions.






Introduction


Histopathologic examination is currently the gold standard for the diagnosis of melanocytic lesions. However, some problematic cases show overlapping features of both nevi and melanomas. One of the most difficult areas in dermatopathology is distinguishing Spitz nevi from Spitzoid melanomas using current histopathologic criteria. Diagnostically challenging cases that show features of both Spitz nevi and Spitzoid melanomas are termed atypical Spitzoid neoplasms. It has been shown in several studies that it can be difficult to predict the clinical behavior of atypical Spitzoid neoplasms from assessment of histopathologic features. The interobserver reproducibility among pathologists for the diagnosis of these tumors is generally poor.


Mass spectrometry imaging can be successfully used for skin cancer diagnosis, particularly for the diagnosis of challenging melanocytic lesions. Mass spectrometry imaging analysis is a bioanalytical method for identifying the nature and spatial distribution of biomolecules, including peptides and proteins, DNA segments, lipids, and metabolites, from tissue samples. It has been used to determine the molecular signatures of various types of cancer and other diseases. Molecular signatures usually present as a unique combination of 5 to 20 proteins and enable a specific diagnosis. Protein and peptide analysis is superior to gene expression analysis, because it represents the functional state of the disease or tumor rather than the potential risk of developing it. Furthermore, posttranslational modification is thus also considered.


Methodology


Formalin-fixed, paraffin-embedded 5-μm-thick consecutive tissue sections from each patient are used for mass spectrometry analysis. The first section is placed on a charged slide for mass spectrometry imaging ( Fig. 1 A). The following consecutive section is stained with hematoxylin and eosin (see Fig. 1 B). A targeted approach is used, in which only discrete areas within a tissue section are analyzed. A dermatopathologist selects and marks these areas, which are 100 μm in diameter (see Fig. 1 B). On average, 20 different areas containing only melanocytes without admixed dermis or epidermis are marked for each case (see Fig. 1 B). Many different samples can be analyzed in 1 experiment (see Fig. 1 C). Images of stained and unstained sections are merged and coordinates of annotations are determined (see Fig. 1 D). Trypsin and matrix are applied after the sections are deparaffinized (see Fig. 1 E) and mass spectra are obtained (see Fig. 1 F).




Fig. 1


( A ) An unstained 5 μm-thick tissue section placed on a charged slide for mass spectrometry (MS). ( B ) A consecutive serial section is stained for hematoxylin and eosin. The blue dots in the superficial portion of the lesion represent the areas of pure melanocytic component chosen by the dermatopathologist for mass spectrometric analysis. Additional melanocytic areas colored in red have also been selected to compare differences between the proteomic composition of melanocytes in the superficial and deep portions of the lesion. ( C ) Two adjacent charged slides for mass spectrometry analysis containing 8 separate cases to be analyzed. ( D ) The images of the hematoxylin and eosin–stained sections and unstained sections are merged and this image is used to guide data acquisition for mass spectrometry analysis. ( E ) Trypsin and matrix are applied to the tissue in preparation for mass spectrometry analysis. ( F ) Mass spectra are collected, which show differences in the proteins from the melanocytic components of the benign nevi ( green ) and malignant melanoma ( red ).


The mass spectral profile is acquired for each of the areas of interest ( Fig. 2 ). The spectra are then loaded into a software for advanced statistical analysis, visualization, and interpretation of mass spectrometry imaging data. The spectra are baseline corrected, normalized, and peaks are picked and aligned. The samples are sorted into training and validation sets. Hypothesis testing and discriminative m/z (mass to charge ratio) value analysis (receiver operator characteristic curve) are performed on the training set. A linear discriminant analysis classification algorithm is generated and validated on a validation set.




Fig. 2


Tryptic peptide at m/z (mass to charge ratio) 1138.8 shows statistically significant differential expression in malignant melanomas and benign nevi. Mass spectrometry images of the peptide showing low intensity in malignant melanomas ( A ) and higher intensity in benign nevi ( B ). ( C ) Average spectra from benign nevi ( green ) and malignant melanomas ( red ). The peptide has 2-fold higher intensity in benign nevi than in malignant melanoma. ( D ) Box and whiskers plot of distributions of the intensity of the peptide at m/z 1138.8 showing complete separation of the boxes (1 standard deviation from the mean). ( E ) Receiver operator characteristic curve for the peptide at m/z 1138.8. The area under the curve (AUC) is 0.105, indicating that this peptide can classify the two groups with 89.5% specificity.


Application of Mass Spectrometry Imaging in the Diagnosis of Challenging Melanocytic Lesions


Atypical Spitzoid neoplasms


In a recent mass spectrometry imaging study, we identified differences on the proteomic level between Spitz nevi and Spitzoid melanomas. Five peptides, comprising a specific proteomic signature, were differentially expressed by the melanocytic component of Spitz nevi and Spitzoid melanomas in formalin-fixed, paraffin-embedded tissue samples. In a subsequent study, we sought to determine whether mass spectrometry imaging could assist in the diagnosis and risk stratification of atypical Spitzoid neoplasms.


In 2009, we founded the International Spitzoid Neoplasm Study Group. The mission statement of the group is to (1) establish cooperation among dermatopathologists throughout the world in order to research Spitzoid lesions; (2) share information and case material so that each participant benefits using a large database; and (3) allow all contributors to be rewarded by participation in publications. Members of the International Spitzoid Neoplasm Study Group together with other collaborators provided more than 200 cases of atypical Spitzoid neoplasms for a retrospective study, which involved centers from 11 countries and 11 US institutions.


We performed mass spectrometry imaging in a large series of patients atypical Spitzoid neoplasms and available clinical follow-up. In each case we compared the diagnosis rendered by mass spectrometry imaging with the histopathologic diagnosis and also correlated the diagnoses with clinical outcomes. Patients were divided into 4 clinical groups representing best to worst clinical behavior. The association among mass spectrometry imaging findings, histopathologic diagnoses, and clinical groups was assessed.


When analyzing atypical Spitzoid neoplasms for which neither melanoma nor nevus was favored histopathologically, mass spectrometry imaging seemed to be more accurate in predicting the benign character of atypical Spitzoid neoplasms than histopathology and correlated better with their clinical behavior ( Fig. 3 ). Histopathology had a tendency to overdiagnose either atypical features or malignancy. A strong association was found between the diagnosis of Spitzoid melanoma by mass spectrometry imaging and an adverse clinical outcome when clinical group 1 (no recurrence or metastasis beyond a sentinel node) was compared with groups 2, 3, and 4 (recurrence of disease, metastases or death). In addition, the diagnosis of Spitzoid melanoma by mass spectrometry imaging was statistically strongly associated with adverse clinical behavior. Mass spectrometry imaging analysis using a proteomic signature may be able to provide more reliable diagnosis and clinically useful and statistically significant risk assessment of atypical Spitzoid neoplasms, beyond the information provided by histology and other ancillary techniques.




Fig. 3


A dome-shaped and asymmetric proliferation of large epithelioid melanocytes in sheets. This patient was a 15-year-old girl with a lesion on the neck. A histopathologic diagnosis of atypical Spitzoid neoplasm. Because Spitzoid melanoma could not be excluded, in addition to wide excision, the patient underwent sentinel lymph node dissection and had negative nodes. Mass spectrometry imaging rendered a diagnosis of Spitz nevus. The patient is alive and with no evidence of disease at 4 years of follow-up.


Metastasis versus benign nevus


The distinction between a metastasis and a nevus is sometimes difficult to make, but it is very important for determining the patient’s clinical staging and management. Here is an example of a case in which mass spectrometry imaging was very helpful. A 37-year-old pregnant woman was diagnosed with a 2.2-mm-thick melanoma on the right upper arm in her late pregnancy. Ulceration was absent; there were 25 mitoses/mm 2 , and intravascular invasion was present. The patient had a term delivery by a cesarean section, at which time a wide excision and lymph node dissection were undertaken. The newborn baby had several atypical melanocytic lesions on the back and flank ( Fig. 4 ). It was of paramount importance to determine whether these were congenital nevi or metastases from the mother’s melanoma. In the latter case, the mother’s clinical stage had to be upgraded to stage IV.




Fig. 4


Two atypical melanocytic lesions in a newborn baby from the back ( left ) and right flank ( right ).


Biopsies from 2 of the baby’s atypical melanocytic lesions were sent for histopathologic examination. In the mother’s melanoma, there was a nodular area composed of small melanocytes with high nuclear to cytoplasmic ratio. The melanocytes within the 2 biopsied lesions from the baby resembled the melanocytes from the nodular portion of the mother’s melanoma. Because of the histologic similarities, it was difficult to make a definitive diagnosis of either benign nevi or melanoma metastases from the mother’s malignant melanoma. The mother’s melanoma and the baby’s 2 melanocytic lesions, were subjected to mass spectrometry imaging analysis and compared ( Fig. 5 ). The baby’s lesions were classified by mass spectrometry as nevi and had a different proteomic profile than the mother’s melanoma ( Fig. 6 ).




Fig. 5


The melanoma of the mother ( A ) and one of the atypical melanocytic lesions from the baby ( B ) with marked areas to be studied by mass spectrometry imaging.

Tags:
Feb 11, 2018 | Posted by in Dermatology | Comments Off on Proteomic Mass Spectrometry Imaging for Skin Cancer Diagnosis

Full access? Get Clinical Tree
Get Clinical Tree app for offline access