Clinical spectral characterisation of colonic mucosal lesions using autofluorescence and delta aminolevulinic acid sensitisation

Abstract

Background and aims: Laser induced fluorescence (LIF) from colonic mucosa was measured in vivo with and without delta aminolevulinic acid (ALA) in an attempt to differentiate between neoplasia and non-neoplasia in real time during colonoscopy.

Methods: Spectra from 32 adenomas, 68 normal sites, and 14 hyperplastic polyps in 41 patients were obtained with a point monitoring system. Twenty one of the patients had been given a low dose of ALA as a photosensitiser before the examination. Light of 337, 405, or 436 nm wavelength was used as excitation. Stepwise multivariate linear regression analysis was performed.

Results: With 337 nm excitation, 100% sensitivity and 96% specificity was obtained between normal mucosa and adenomas. Seventy seven per cent of the hyperplastic polyps were classified as non-neoplastic. When exciting with 405 and 436 nm, the possibility of distinguishing different types of tissue was considerably better in the ALA patients than in the non-ALA patients.

Conclusions: The in vivo point measurements imply that a good discrimination between normal tissue and adenomatous polyps can be obtained using the LIF technique. Excitation at 337 nm and at 405 nm or 436 nm using ALA gives good results. LIF also shows potential for distinguishing adenomatous from hyperplastic polyps. The number of detection wavelengths could be reduced if chosen properly.

Figure 1

Experimental set up for in vivo LIF recording.

Figure 2

Typical in vivo spectra from normal and adenomatous colon tissue. The spectra are normalised to a total intensity of 1. Error bars represent standard deviations. (A) 337 nm excitation, both ALA and non-ALA patients; (B) 405 nm excitation, non-ALA patients; (C) 405 nm excitation, ALA patients; (D) 436 nm excitation, non-ALA patients; (E) 436 nm excitation, ALA patients.

Figure 3

Fluorescence score derived using stepwise MVLR analysis. The circles with error bars represent mean scores with standard errors. (A) 337 nm excitation—evaluation at 409, 440, 498, 572, and 661 nm; (B) 405 nm excitation, non-ALA patients—evaluation at 498, 543, 560, 590, 602, and 661 nm; (C) 405 nm excitation, ALA patients—evaluation at 498, 543, 560, 572, 602, and 661 nm; (D) 436 nm excitation, non-ALA patients—evaluation at 498, 543, 560, 572, 590, and 602 nm; (E) 436 nm excitation, ALA patients—evaluation at 498 and 661 nm.

Figure 4

MVLR results using two evaluation wavelengths. The circles with error bars represent mean scores with standard errors. (A) 337 nm excitation, evaluation at 409 and 543 nm; (B) 405 nm excitation, ALA patients, evaluation at 498 and 661 nm.