Prognostic Features of Near-Infrared Spectroscopy Following Primary Radical Prostatectomy

Tijl Vermassen; Jonas Himpe; Renaat Coopman; Charles Van Praet; Nicolaas Lumen; Sylvie Rottey; Joris Delanghe

doi:10.3390/cancers13236034

Prognostic Features of Near-Infrared Spectroscopy Following Primary Radical Prostatectomy

Cancers (Basel). 2021 Nov 30;13(23):6034. doi: 10.3390/cancers13236034.

Authors

Tijl Vermassen¹, Jonas Himpe², Renaat Coopman³, Charles Van Praet⁴, Nicolaas Lumen⁴, Sylvie Rottey¹, Joris Delanghe²

Affiliations

¹ Department of Medical Oncology, University Hospital Ghent, 9000 Ghent, Belgium.
² Department of Clinical Chemistry, Microbiology and Immunology, Ghent University, 9000 Ghent, Belgium.
³ Department of Plastic, Reconstructive and Aesthetic Surgery, University Hospital Ghent, 9000 Ghent, Belgium.
⁴ Department of Urology, University Hospital Ghent, 9000 Ghent, Belgium.

Abstract

Background: Only a few biomarkers have been evaluated for their prognostic value with regard to biochemical recurrence (BCR) following primary radical prostatectomy. We explored the possibilities of using near-infrared (NIR) spectroscopy as a prognostic biomarker for BCR-free survival (BCR-FS).

Methods: Tissue specimens from 82 prostate cancer patients were obtained. Formalin-fixed paraffin-embedded slides (hematoxylin-eosin-stained) were analyzed using NIR spectroscopy. Prognostic features for BCR-FS were determined following normalization of the spectra.

Results: Several differences were found throughout the NIR spectrum for the patients with or without BCR, for both the first derivative and second derivative of the NIR spectrum. Following categorization and Cox regression analysis, spectral regions at 5236 cm^-1 (first derivative; median BCR-FS not reached versus 3.2 years; HR_high = 0.18 [0.08-0.39]; and p < 0.0001) and at 5956 cm^-1 (second derivative; median BCR-FS not reached versus 3.8 years; HR_low = 0.22 [0.10-0.48]; and p = 0.0002) showed prognostic properties for BCR-FS. The combination of both parameters further increased the prognostic value of NIR (p < 0.0001).

Conclusions: We demonstrated NIR spectral variations between patients with or without BCR, which have been shown to have prognostic value. This easy-to-use technique could possibly further improve post-primary radical prostatectomy monitoring and swift referral to adjuvant local therapies. Further elaboration is highly recommended to fully elucidate these variations and to gain a deeper insight into the changing chemical and physical compositions of the prostate tumor architecture.

Keywords: NIR; biochemical recurrence; local treatment; prognosis; prostate cancer; spectroscopy.