BACKGROUND: Early diagnosis of nasopharyngeal carcinoma (NPC) remains a challenge. Serum protein profiling is a promising approach for the classification of cancer versus noncancer samples. The objective of the current study was to assess the feasibility of mass spectrometry-based protein profiling and a classification tree algorithm for discriminating between patients with NPC and noncancer controls. METHODS: Serum samples from patients with NPC and noncancer controls were analyzed by using surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS). The study was divided into a preliminary training set and a blind test set: A preliminary training set and a classification tree of spectra derived from 55 patients with NPC and a group of 60 noncancer controls were used to develop a proteomic model that discriminated cancer from noncancer effectively. Then, the validity of the classification tree was challenged with a blind test set, which included another 25 patients with NPC and 28 noncancer controls. RESULTS: Four protein peaks at 4097 daltons (Da), 4180 Da, 5912 Da, and 8295 Da were chosen automatically as a biomarker pattern in the training set that discriminated cancer from noncancer with sensitivity of 94.5% and specificity of 96.7%. When the SELDI marker pattern was tested with the blinded test set, it yielded a sensitivity of 92%, a specificity of 92.9%, and an accuracy rate of 92.5%. The accuracy of 2 protein peaks (4581 Da and 7802 Da) was 80% for predicting stage I and II NPC and 86% for predicting stage III and IV NPC. CONCLUSIONS: The high sensitivity and specificity obtained by the serum protein profiling approach demonstrated that SELDI-TOF-MS combined with a tree analysis model both can facilitate discriminating between NPC and noncancer controls and can provide an innovative clinical diagnostic platform to improve the detection of NPC.
BACKGROUND: Early diagnosis of nasopharyngeal carcinoma (NPC) remains a challenge. Serum protein profiling is a promising approach for the classification of cancer versus noncancer samples. The objective of the current study was to assess the feasibility of mass spectrometry-based protein profiling and a classification tree algorithm for discriminating between patients with NPC and noncancer controls. METHODS: Serum samples from patients with NPC and noncancer controls were analyzed by using surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS). The study was divided into a preliminary training set and a blind test set: A preliminary training set and a classification tree of spectra derived from 55 patients with NPC and a group of 60 noncancer controls were used to develop a proteomic model that discriminated cancer from noncancer effectively. Then, the validity of the classification tree was challenged with a blind test set, which included another 25 patients with NPC and 28 noncancer controls. RESULTS: Four protein peaks at 4097 daltons (Da), 4180 Da, 5912 Da, and 8295 Da were chosen automatically as a biomarker pattern in the training set that discriminated cancer from noncancer with sensitivity of 94.5% and specificity of 96.7%. When the SELDI marker pattern was tested with the blinded test set, it yielded a sensitivity of 92%, a specificity of 92.9%, and an accuracy rate of 92.5%. The accuracy of 2 protein peaks (4581 Da and 7802 Da) was 80% for predicting stage I and II NPC and 86% for predicting stage III and IV NPC. CONCLUSIONS: The high sensitivity and specificity obtained by the serum protein profiling approach demonstrated that SELDI-TOF-MS combined with a tree analysis model both can facilitate discriminating between NPC and noncancer controls and can provide an innovative clinical diagnostic platform to improve the detection of NPC.
Authors: Lixia Diao; Charlotte H Clarke; Kevin R Coombes; Stanley R Hamilton; Jack Roth; Li Mao; Bogdan Czerniak; Keith A Baggerly; Jeffrey S Morris; Eric T Fung; Robert C Bast Journal: Cancer Inform Date: 2011-03-14