Autofluorescence for parathyroid glands detection: a multicentric analysis

Detection of parathyroid autofluorescence using near-infrared imaging: a multicenter analysis of concordance between different surgeons
Bora Kahramangil, MD1, Fernando Dip, MD2,3, Fares Benmiloud, MD4, Jorge Falco, MD2,  Martin de La Fuente, MD2, Silvina Verna, MD2, Raul Rosenthal, MD3, and Eren Berber, MD1

1Department of Endocrine Surgery, Cleveland Clinic, Cleveland, OH; 2Instituto Argentino de Diagnostico y tratamiento
(IADT), Buenos Aires, Argentina; 3Department of Surgery, Cleveland Clinic Florida, Weston, FL; 4Department of  Endocrine Surgery, Hôpital Européen, Marseille, France
Parathyroid glands (PGs) exhibit autofluorescence  (AF) when excited by near-infrared laser. This multicenter study aims to analyze how this imaging could facilitate the detection of PGs during thyroidectomy and parathyroidectomy procedures.

This was a retrospective Institutional Review  Board-approved analysis of prospectively collected data at three centers. Near-infrared fluorescence imaging (NIFI) was used to detect AF from PGs during thyroidectomy and parathyroidectomy procedures. Logistic regression analysis was performed to assess the utility of NIFI to identify PGs and concordance at these centers.

Overall, 210 patients underwent total thyroidectomy  (n = 95), thyroid lobectomy (n = 41), and parathyroidectomy (n = 74) (n = 70 per center). Using NIFI, AF was detected from 98% of visually identified PGs. Upon initial exploration, 46% of PGs were not visible to the naked eye due to coverage by soft tissue, but AF from these glands could be detected by NIFI without any further dissection. Overall, a median of one PG per patient was detected by NIFI in this fashion before being identified visually (p = nonsignificant between centers). On logistic regression, smaller PGs were more likely to be missed visually, but localized by AF on NIFI (odds ratio with increasing size, 0.91; p = 0.02).

In our experience, NIFI facilitated PG  identification by detecting their AF, before conventional recognition by the surgeon, in 37–67% of the time. Despite the variability in this rate across centers, there was a concordance in detecting AF from 97 to 99% of the PGs using NIFI. We suggest the incorporation of AF on NIFI alongside conventional visual cues to aid identification of PGs 
during neck operations.