Current state of intraoperative use of near infrared fluorescence for parathyroid identification and preservation

Abstract

Background:

Finding and preserving normal parathyroid glands or localizing and removing diseased parathyroid glands are crucial steps to successful thyroid and parathyroid operations. Using near-infrared fluorescence detection to identify parathyroid glands during thyroid and parathyroid operations has lately gained widespread recognition, with 2 Food and Drug Administration-cleared devices currently in the market. We aim to update the endocrine surgery community on how near-infrared fluorescence detection can be most optimally used for rapid intraoperative parathyroid gland identification or preservation.

Methods:

A literature review was performed using the key terms: “parathyroid,” “near infrared,” and “fluorescence” in relevant search engines. Based on the reviewed literature and expert surgeons’ opinions, recommendations were formulated for applying near-infrared fluorescence detection to identify or preserve parathyroid glands during cervical endocrine surgery.

Results:

The scope of near-infrared fluorescence detection can be broadly categorized into (1) using near-infrared auto-fluorescence to identify or locate both healthy and diseased parathyroid glands, and (2) using contrast-enhanced near-infrared fluorescence to evaluate parathyroid gland perfusion. The benefits and pitfalls for both near-infrared-based approaches are described herein.

Conclusion:

Near-infrared fluorescence detection appears helpful for identification and likely preservation of parathyroid glands. We hope these recommendations will be valuable to the practicing endocrine surgeon as they consider incorporating these intraoperative adjuncts in their surgical practice.

Introduction

Finding and preserving healthy parathyroid glands (PGs) during neck operations can be challenging even for experienced surgeons. As a result, inadvertent damage to healthy PGs or its vasculature could lead to transient or permanent hypocalcemia. In comparison, inability to detect and remove hyperfunctioning PGs results in failed parathyroidectomy. Both are complications that are associated with patient morbidity, which may require additional evaluation, treatment, and expenses.¹ Currently PGs are usually identified visually by experienced surgeons according to their usual location and characteristic appearance. Frozen section and needle aspiration for parathyroid hormone (PTH) measurement can also be used intraoperatively to confirm parathyroid tissue, but they can be invasive, labor-intensive, costlier, and time-consuming. Neither approach can help the surgeon find the PG or preserve their perfusion.

Researchers have used light-based (optical) techniques intraoperatively to identify PGs because they are noninvasive and provide real-time results. These techniques include optical coherence tomography, dynamic optical contrast imaging and Raman spectroscopy.² Unfortunately, they are technically complex and not yet practical for use in most operating rooms. In comparison, fluorescence-based techniques use simpler and less expensive instrumentation while retaining high sensitivity.

Fluorescence typically occurs when a substance is illuminated or “excited” by light of higher energy (shorter wavelengths), which then subsequently emits light of lower energy (longer wavelengths) within nanoseconds. These “fluorophores” can be intrinsic (naturally occurring in the target tissue) or extrinsic (externally synthesized and injected into the patient). Presence of intrinsic fluorophores in some tissues can cause native fluorescence or “autofluorescence,” which makes autofluorescence detection in tissues a “label-free” technique. Extrinsic fluorophores, such as contrast agents or targeted-labels, can also be added to generate “contrast-enhanced” fluorescence. Both auto-fluorescence and contrast-enhanced fluorescence detection require: (1) an intense light source (usually lasers) of the required wavelength to excite the target fluorophore, (2) excitation filters that allow only specific wavelengths to excite the fluorophore, (3) emission filters that only permit wavelengths of fluorophore emissions through, (4) a detector (cameras, spectrographs, or photodiodes) that collects the filtered fluorophore emissions, and (5) software that gather signal or data from the detector and relay them as images, spectra or other information to the end-user. It must be noted that fluorescence is optimally detected in the dark, because most fluorescence systems cannot fully filter out interference from ambient room light, which has a very broad range of wavelengths (400–800 nm).

Near infrared fluorescence in the parathyroid glands

Light sources with shorter wavelength are attenuated more by the tissue than those with longer wavelengths. Fluorescence caused by short wavelengths such as ultraviolet-visible light can only be detected within a few microns from the surface of the tissue. In contrast, near infrared (NIR) light sources that have longer excitation wavelengths (>750 nm) allows tissue penetration as deep as 5 mm, thus making NIR light sources optimal for in vivo fluorescence imaging.

Fluorescence with NIR light in PGs can be detected as (1) autofluorescence from the gland itself or (2) contrast-enhanced fluorescence using injectables. Autofluorescence in PGs with NIR illumination was serendipitously discovered by researchers at Vanderbilt University.³ Parathyroid glands yielded much stronger NIR autofluorescence (NIRAF) relative to surrounding tissues including thyroid, fat, muscle, and trachea. NIRAF is able to detect both normal and diseased PGs whereas other preoperative (ultrasound, computed tomography, sestamibi scans) or intraoperative (gamma probe based-scintigraphy) modalities could only detect some abnormal PGs.

The NIRAF observed in parathyroid tissue has an emission peak at a wavelength of approximately 820 nm.³ Although NIRAF has also been described in other tissues, no specific endogenous fluorophore has been known to exhibit autofluorescence beyond 700 nm. Although the responsible fluorophore remains unknown, NIR autofluorescence of PGs depends on the cellularity of the gland, expression of calcium-sensing receptors, or presence of pseudo-colloid.⁴ The NIRAF in parathyroid tissue is robust and can persist ex vivo for as long as 150 hours to 2 years, and under extreme temperatures, proteinase activity, and formalin fixation.^2,5,6 This striking resilience of the responsible fluorophore also causes the NIRAF to remain unaffected by gland vascularity, making it unsuitable for determining PG perfusion or viability. To assess perfusion of the PGs, it is necessary to use contrast-enhanced fluorescence with dyes or parathyroid targeted-labels. Among the different contrast agents, indocyanine green (ICG) is a Food and Drug Administration (FDA)-approved NIR excitable dye that is used to evaluate parathyroid perfusion. With an emission peak at 830 to 835 nm, ICG provides a higher signal-to-background ratio than other contrast agents. Thus, ICG produces a brighter image (ie, greater quantum yield) and enables deeper tissue imaging than autofluorescence.

Current technologies for NIR parathyroid fluorescence detection

In 2018 the FDA granted clearance for marketing to two NIR detection class II devices: PTeye (probe-based) and Fluobeam-800 (image-based) for real time PG identification during surgery (https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-two-devices-detect-parathyroid-tissue-real-time-during-surgery). These devices are considered adjunct tools that use NIRAF to detect tissues or structures and are not intended to provide a diagnosis. In addition, there are other imaging-based systems that have been approved solely for contrast-enhanced NIR fluorescence detection.

1. PTeye (marketed by AiBiomed, Santa Barbara, CA): This system comprises of a disposable fiber-optic probe connected to a console housing a NIR light source and an interactive display (Fig 1, A). PTeye provides real-time visual and auditory feedback when the probe touches a PG (similar to using a probe for a nerve-monitoring device). Quantitative parameters, such as “detection level” and “detection ratio,” are generated at each measurement, with the threshold for PG identification set at a detection ratio >1.2. The surgeon initially sets the baseline background fluorescence by taking 5 measurements with the probe on different areas of the thyroid gland. After setting the baseline, the thyroid gland is surveyed again to find any area with a detection ratio higher than 1.2 that was missed during the initial survey. If areas of higher ratios are found on the thyroid gland, the baseline is accordingly readjusted upward. After the baseline is finally set, the probe is ready for PG detection. During the operation, when a suspected PG is found, it is touched with the probe (Fig 1, B). The surgeon then activates the NIR light source by stepping on the foot pedal. The device detects the NIRAF signal and displays the detection level and ratio. If the ratio is higher than the preset threshold for parathyroid tissue of 1.2, the machine provides an auditory beep (Fig 1, C and D). Ambient operating room lights, including ceiling lights, surgical lights, and headlights, do not interfere with the function of PTeye.

2. Fluobeam-800 and Fluobeam-LX (marketed by Fluoptics, France): The device consists of a handheld camera system that also encloses a NIR light source to illuminate the tissue of interest within the surgical field. The resulting auto-fluorescence signals are processed to generate grayscale images on a display monitor (Fig 2, A). The Fluobeam camera, covered by a sterile cover, is held by the surgeon at a defined distance from the surgical field (Fig 2, B) to generate the images with no tissue contact. The PGs are identified as the brighter areas on images generated, and thus providing a map of their locations (Fig 2, C and D).^4,7 The Fluobeam-800 requires that all ambient OR lights be off during NIRAF imaging. A newer version Fluobeam-LX can now be used with the room ceiling lights on, but brighter surgical lights or headlights still need to be off or turned away from the field.

3. Other imaging devices: There are other commercially available laparoscopic or handheld NIR camera systems that were developed for use with ICG. These devices are FDA-approved for imaging tissue perfusion, tumor and anatomy, and sentinel lymph nodes. These include: PDE Neo II (Hamamatsu, Japan), Stryker Spy-PHI (Stryker, USA), Karl Storz Opal-1 (Karl Storz, Germany), and Quest Spectrum (Quest, Netherlands). Although these devices have not been specifically cleared by the FDA to detect NIRAF in PGs, surgeons have been able to use some of these systems without ICG to find PGs by relying on NIRAF. As with Fluobeam-800, these devices also require all ambient room lights to be off during imaging.

Fig. 1.

FDA-cleared devices for parathyroid NIRAF detectioneprobe-based systems. (A) PTeye is a probe-based device that consists of (1) a console that contains a NIR light source and a detector, (2) a display to inform the surgeon if a tissue is a parathyroid or not, (3) a foot-pedal to activate the NIR light source, and (4) a sterile detachable fiber probe. (B) Surgeon places fiber probe of PTeye in contact with tissue for NIRAF detection. (C and D) Display interface when the tissue is a parathyroid (left), and when it is not (right). Figure adapted from unpublished images obtained with PTeye.

Fig. 2.

FDA-cleared devices for parathyroid NIRAF detectioneimage-based systems. (A) Fluobeam LX is an image-based system that has (1) a console to adjust NIR illumination of surgical field, (2) a handheld NIR camera, and (3) a display panel to spatially visualize the parathyroid glands. (B) Surgeon holds the NIR camera over the surgical field. (C) White light image of the surgical field where the superior pole of the right thyroid gland has been dissected and retracted with Babcock clamp. The right superior parathyroid gland is embedded in fatty tissue here. (D) The corresponding NIR image displays right superior parathyroid with high NIRAF intensity (white circle). Note that NIRAF is also seen from the superior pole of thyroid gland, Vicryl suture and surgical ink. Figure adapted from unpublished images obtained with Fluobeam LX. Figure 2, A was obtained from www.fluoptics.com.

Extent and feasibility of NIRAF parathyroid detection during neck endocrine procedures

NIRAF detection may help the surgeon with (1) real-time label-free confirmation of PGs, (2) finding or mapping of superficially located PGs, (3) differentiating normal from abnormal PGs, and (4) prevention of transient postoperative hypocalcemia during total thyroidectomy. Table I summarizes NIRAF detection for PG identification by probe- and image-based systems with their salient features.

Table I.

Probe-based versus image-based approaches for NIR fluorescence detection

	Probe-based	Image-based
NIR commercial devices	FDA-cleared for NIRAF detection: PTeye	FDA-cleared for NIRAF detection: Fluobeam-800 and LX, FDA-approved for contrast-enhanced NIR fluorescence detection: Fluobeam-800 and LX, PDE-Neo II, Firefly, Karl Storz Opal-1, Quest Spectrum, PinPoint, SPY, ENV, SPY-PHI, 1588 AIM, FLARE, IMAGE1 SPIES, EleVision, Olympus
Steps for use	Set the device up as instructed. Place the sterile probe in contact with tissue. Press foot-pedal to activate NIR light source for tissue NIRAF detection. First establish baseline for each patient by obtaining NIRAF measurements on 5 random spots on the thyroid. If thyroid is absent, acquire baseline on neck muscle. Survey the thyroid/muscle with probe to ensure accuracy of baseline NIRAF. Readjust the baseline as necessary such that detection ratio on thyroid/muscle does not exceed 1.2. After setting the baseline, adequately expose the tissue of interest before NIRAF measurement. When probe touches PG tissue, the device should typically display a detection ratio >1.2 and generate high frequency auditory beep.	Set the device up as instructed. Wrap the handheld NIR camera with a sterile drape before NIRAF imaging. Switch off all operating room lights: room ceiling lights, surgical lights and surgeon headlight before NIRAF imaging. For Fluobeam LX, room ceiling lights can remain on, while the remaining light sources should be switched off/turned away. Survey the surgical field for tissue of interest using the handheld NIR camera. Ensure holding the camera at a fixed distance from the surgical field during NIRAF imaging. PG tissue would typically appear as ovoid structures with higher NIRAF intensity compared with adjacent tissues.
Advantages	Provides real-time quantitative information with “detection level” and “detection ratios.” Compact device with audio-visual feedback. Can be used with very small surgical incisions. Probe can be easily placed on PGs in ectopic location/deeper planes for NIRAF detection. Fluorescence from other sources in surgical field does not interfere with NIRAF detection. Probe contact with tissue ensures that NIRAF is measured only from site of interest. Functional with all ambient OR lights—room ceiling lights, surgical lights, and surgeon’s head light.	PGs can be spatially localized/mapped due to NIR view of the surgical field. Tissue can be visualized in context with adjacent structures to guide the surgery and preserve PG vasculature. Approach does not require tissue contact. NIRAF imaging can be combined with ICG angiography for identifying and preserving PGs.
Disadvantages	No NIR view of the surgical field to spatially visualize NIRAF and localize PGs. Contact-based approach, requires use of a sterile disposable probe. Error in baseline NIRAF acquisition could generate inaccurate detection ratios; can cause false positives/negatives. Utility for PG angiography with contrast agents such as ICG are yet to be demonstrated. Cannot detect PGs that are deep seated-buried/hidden/missing.	Lack of real-time quantitative information. Surgical incision needs to be wider for optimal NIRAF imaging. NIRAF signal may not be detected well from ectopic or deeper plane PGs. Fluorescence from Vicryl sutures, surgical ink, kittners, or another PG in field may interfere during NIRAF imaging. Cannot detect PGs that are deep seated buried/hidden/missing. Ambient OR lights-room ceiling lights, surgical lights, or surgeon’s head light need to be turned off.

NIRAF, near infrared autofluorescence.

Probe-based systems

PTeye is the only FDA-cleared probe-based NIR fluorescence detector for PG identification. One of the largest studies using a laboratory-built probe-based system (the precursor to the commercially available PTeye) showed that 97% of PGs have higher NIRAF than the surrounding tissues.⁸ Although high calcium levels, low 25-hydroxyvitamin D levels, PG disease state and high body mass index affected NIRAF intensity, the probe-based system accurately identified PGs. PTeye was subsequently tested against the laboratory-built system by a single surgeon and showed higher accuracy at identifying PGs (97.7% vs 90.8%).⁹ As probe-based NIRAF detection can identify PGs in real time, it can aid in increasing the surgeon’s confidence in their identification. Although there has been no study to specifically address the efficacy of PTeye to “map” the locations of PGs, it is possible that PTeye may be used for “scanning” through the surgical field to locate parathyroid tissue-analogous to how one may use a nerve monitoring probe for mapping a nerve before nerve visualization. Although surgeons typically rely on experience to distinguish normal from abnormal PGs during parathyroidectomies, a device that can distinguish diseased glands from healthy ones may potentially be useful. Diseased PGs appear to have lower and more heterogenous NIRAF intensities than normal ones,¹⁰ but further studies are needed to determine if a probe-based NIRAF system can consistently differentiate normal from abnormal PGs. Randomized clinical trials using PTeye for thyroidectomy and parathyroidectomy are currently ongoing (https://clinicaltrials.gov/ct2/show/NCT04281875; https://clinicaltrials.gov/ct2/show/NCT04299425).

Image-based systems

Fluobeam-800, and its newer version Fluobeam-LX, are the only FDA-cleared image-based NIR fluorescence detectors for identifying PGs. Initial studies with this approach confirmed high PG detection rates ranging from 94% to 100%.^4,7 Subsequent studies showed that mapping the position of PGs before and during thyroid resection led to an earlier identification of PGs compared with white light (ie, plain visual inspection under normal OR lights). As a result, the “test” group that used NIRAF detection had less inadvertent resection of normal PGs after total thyroidectomy compared with historical controls,¹¹ which was then confirmed by subsequent single-center and multi-center randomized studies.^12–14 Although certain studies have reported on lower rate of transient hypocalcemia with use of NIRAF imaging,^11–13 other studies have observed no statistical significant difference in the postoperative hypocalcemia rates between the NIRAF versus control groups.^14,15 The contrast in these findings indicate the need for further trials to validate if NIRAF detection can truly minimize postoperative hypocalcemia rates. In addition, it is likely that preserving PG vasculature may have a more dominant effect on the overall postoperative PG function, as compared with merely detecting or visualizing all PGs. As discussed earlier, image-based systems have demonstrated that NIRAF in diseased PGs (particularly adenomas) tend to be more heterogeneous compared with normal glands.^10,16 It is possible that these optical traits may be used to distinguish diseased from normal PGs (Fig 3). However, more robust studies are required to determine whether this approach can help reliably find abnormal PGs and decrease the rate of persistent or recurrent hyperparathyroidism.

Fig. 3.

Heterogeneity of NIRAF observed in abnormal PGs. (A) NIRAF images obtained with image-based system for excised PGs. NIRAF in the excised parathyroid adenoma is more heterogenous compared with that of the adjacent normal PG. (B, C, and D) Note the heterogeneity of NIRAF in the resected parathyroid adenomas. These adenomatous glands tend to have a healthier appearing “cap” that have higher NIRAF compared with the rest of the diseased regions. (E and F) NIRAF intensity measured with probe-based system on the adenomatous portion of the excised PG. (G and H) NIRAF intensity measured on the healthier appearing “cap” of the diseased gland. NIRAF detection levels and detection ratios obtained with probe-based system were observed to be higher in the healthier appearing “cap” compared with the remaining portion of the gland, analogous to the heterogeneity of NIRAF in parathyroid adenomas observed with image- based system. Figure adapted from Fluobeam LX images obtained by Demarchi et al¹⁶ and unpublished images with PTeye.

Benefits and pitfalls of NIRAF detection for parathyroid gland identification

Because PGs can be accurately identified by NIRAF,¹⁷ the real-time and label-free nature of NIRAF detection can serve as a noninvasive “optical biopsy.” The technique may improve a surgeon’s confidence in identifying PGs earlier than just with plain visualization and thus reduce the need for frozen-section or PTH analysis.¹⁸ With an average turnaround time of 20 to 30 minute per frozen-section or tissue aspirate PTH analysis, the surgeon needs to wait to take the next surgical decision: (1) steps to preserve the healthy PG and its vasculature or (2) steps to identify and excise PG(s), if diseased. NIRAF detection can thus help surgeons make these decisions more quickly and cut down the operating time accordingly. In addition, improved identification of PGs and their vasculature would lessen the risk of injury and lower the rate of postoperative hypocalcemia/hypoparathyroidism and incidental parathyroidectomy.^12,17 Even when PGs are accidentally excised with thyroid glands, NIRAF detection can be useful to localize these excised PGs that could be autotransplanted quickly to minimize postoperative hypoparathyroidism. Additional large scale, multicentric trials using this technique may provide insight as to whether the ability to confirm PGs in real time can help shorten the learning curve of less-experienced surgeons. If this is found to be true with further studies, it is likely that less-experienced or trained surgeons, who may have a higher rate of hypoparathyroidism, may benefit more by using NIRAF to detect PGs. From the literature, sensitivity of NIRAF to identify PGs typically ranged from 80% to 100%.¹⁷ Although false negatives can occur, especially in secondary hyperparathyroidism, false positives have been reported with brown fat, colloidal thyroid nodules, and metastatic lymph nodes.¹⁹ These potential false negatives and false positive can be mitigated, as summarized in Table II. Although reports on potential false positives due to metastatic lymph nodes have only been anecdotal till date, autotransplanting a metastatic lymph node presumed as a PG can have catastrophic consequences. Thus, it is advisable to perform frozen section analyses during thyroidectomies of malignant thyroid diseases, to distinguish between a metastatic lymph node that needs to be removed and a devascularized parathyroid gland that requires an autotransplantation.^9,19 A key limitation of NIRAF is its inability to detect PGs buried under soft tissue.^17,19 Although subcapsular or superficial PGs on thyroid specimens can be detected with these modalities, deep-seated intrathyroidal PGs need to be dissected and exposed before device testing. Because NIR can only penetrate a few millimeters, it is therefore not possible to “find” PGs that are deeply buried or located using NIRAF detection. Another pitfall is that NIRAF detection cannot distinguish between well-perfused or viable and devascularized or nonviable PGs. Perfusion assessment is only possible when contrast agents such as ICG are used.

Table II.

Recommendations for use of NIR fluorescence detection to identify and preserve PGs, with mitigation strategies in case of false positives/negatives with these modalities

	NIRAF detection		Contrast-enhanced NIR fluorescence detection
	Probe-based	Image-based
Protocol recommendations and device result interpretation	Obtain initial baseline NIRAF on thyroid. After baseline acquisition, place the probe on random thyroid sites and ensure that detection ratios always stay <1.2. If ratios exceed 1.2 in certain thyroid regions, repeat baseline acquisition using those regions If thyroid is absent, baseline should be performed on muscle. In such cases, if detection ratios range from 1.2—2.0 for suspect tissues, the surgeon should interpret device output with caution and additionally rely on her/his surgical experience Prior to obtaining measurements, carefully expose the potential PG If detection ratios appear to be low or <1.2 on a probable diseased PG, place probe on the healthier or other portion of PG and repeat measurement; variable NIRAF signal has been reported in diseased PGs10	Hold the camera at a fixed distance (15–25 cm) from the surgical incision and look around the thyroid Minimize presence of Vicryl sutures, kittners, surgical ink, or surgical drape/sheets in the surgical field whenever possible during imaging If more than one potential. PG is seen in field of view, use imaging to identify one PG at a time. A brighter PG may obscure a dimmer PG, if both are in the same field of view Heterogeneity and low NIRAF in a PG might be suggestive of a diseased PG,10 mostly adenomas	Most authors used 5–7.5 mg of ICG solution. However, more sensitive equipment could allow for lower ICG doses (1–5 mg) ICG fluorescence can be typically detected within 20–30 seconds after intravenous injection The most frequently used scoring system is the 3-grade ICG score (ICG-0, ICG-1, and ICG-2)24 ICG scoring can be done at the end of the thyroidectomy or sequentially after each lobectomy
Mitigation strategies in the event of false positives or negatives	Perform frozen section analysis to differentiate between metastatic lymph nodes and devascularized PGs during thyroidectomies for malignant thyroid disease9 Interpret device findings with intraoperative PTH levels and surgical judgement during parathyroid surgeries. Intraoperative PTH assay criterion adopted by surgeon not being met would suggest either (a) presence of more diseased PGs or (b) excised tissue may be a false positive as detected by the device9		Adequate ICG wash-out duration (>10 min) should pass before trying to localize another PG or to evaluate PG perfusion
Notes of caution	NIRAF in secondary hyperparathyroidism cases is diminished and can result sometimes in false negatives9 Data should be interpreted cautiously if dyes (eg, methylene blue) for sentinel lymph node surveillance are injected during thyroid cancer surgeries. NIRAF detection should be performed before dye injection, for ICG angiography or lymph node surveillance Ex vivo measurements should be ideally performed on nondyed substrates. Blue/green dyes in surgical ink and surgical drapes/sheets may interfere with NIRAF detection Devices are not validated in pediatric patients and thus data must be interpreted with caution Probable false positives: Colloidal thyroid nodules, brown fat or fibro-adipose tissues, occasional metastatic lymph nodes May not localize/identify “missing” or “hidden” PGs that are deeper than a few millimeters from surface, eg intrathyroidal PGs		Not to be used in patients with potential allergies to ICG, possibly patients with history of allergies to iodide, shellfish

NIRAF, near infrared autofluorescence.

Extent and feasibility of contrast-enhanced NIR fluorescence parathyroid detection for neck endocrine procedures

Indocyanine green (ICG) is a fluorescent dye that was developed in the late 1950s. ICG has an excellent safety profile, with severe adverse reactions occurring only in <1 out of 200,000 patients. ICG typically binds to albumin and stays in the intravascular compartment until its uptake by hepatocytes and excreted into the bile. Because ICG has a half-life of 3 to 4 minutes, it allows for repeated injections (maximum dose of 5 mg/kg).

ICG can be used to provide real-time intraoperative angiography-like information without radiation exposure. It can be used both at the beginning of dissection to identify well-vascularized structures and at the end of dissection to confirm the presence or absence of their perfusion. Because PGs are highly vascularized organs, some surgeons have used ICG angiography to detect or identify them, in particular diseased PGs, both in primary hyperparathyroidism and in renal failure-induced secondary hyperparathyroidism patients.²⁰ When a PG is surrounded or embedded in neck fat or thymus, the contrast between the well vascularized PG and poorly vascularized fat or thymus is clear and can help localize the gland.

During neck operations, merely preserving the PGs is not enough to avoid hypoparathyroidism. It is equally essential to conserve the vascular supply of these glands to ensure optimal parathyroid function. PGs are usually perfused by an end-artery that comes from tiny branches of typically the inferior and less commonly the superior thyroid artery running along the surface of the thyroid gland. These vessels can be damaged during thyroidectomy, thus devascularizing the PGs.²¹ Because plain visual inspection of the PGs after thyroidectomy does not reliably predict their perfusion and function, ICG angiography can be a useful adjunct to evaluate the perfusion and function of the PGs after thyroid resection.²⁰ Although the term “angiography” is typically used in context of imaging blood vessels supplying an organ, it must be noted that “ICG angiography” described from this point forward refers mainly to the modality being used to visualize and evaluate PG perfusion.

Utility of contrast-enhanced NIR fluorescence detection for assessing perfusion of parathyroid glands

ICG angiography can be performed safely over multiple injection boluses during thyroidectomy. Initial studies indicated that when PGs were found to be well-perfused with ICG angiography at the end of thyroid resection, postoperative PTH levels were normal.^22,23 Subsequently a randomized study showed that patients with one or more well-perfused PGs with ICG did not develop hypoparathyroidism on the day after thyroidectomy, obviating the need for further postoperative measurements of calcium or PTH.²⁴ Most studies showed ICG uptake by the PGs after thyroid resection correlated with postoperative levels of PTH.^16,20 However, such correlations were more difficult to interpret when fewer than 4 PGs were identified. Perfusion of the PGs with ICG angiography (Fig 4) is usually graded visually by the surgeon (grade = 0 not vascularized, grade 1 = moderately well vascularized, grade 2 = well vascularized). Few studies have tried to quantify ICG perfusion with imaging software, but the quantitative analyses were performed postoperatively and not in real-time.²⁰

Fig. 4.

Contrast-enhanced NIR fluorescence detection. Images demonstrating NIR fluorescence enhanced after ICG administration to assess parathyroid gland perfusion. Figure represents white light image (top row), NIR fluorescence image enhanced with ICG (middle row), and a merged white light-NIR image for: (a) devascularized, (b) moderately vascularized, and (c) well-vascularized parathyroid gland. Note the absent uptake of ICG by the devascularized parathyroid gland. Figure adapted from Fortuny et al.²⁴

For patients undergoing total thyroidectomy, ICG angiography can also be used after lobectomy on one side before contralateral dissection. Depending on the patient and the thyroid pathology, the extensiveness of the dissection of the contralateral side can be adjusted according to the vascularization of PGs of the ipsilateral side.²⁴ The approach can be analogous to the “staged thyroidectomy” strategy used to prevent bilateral recurrent nerve injury when there is loss of the intraoperative nerve monitoring signal on the first side.²⁵ Similarly knowing the perfusion or function of the PGs on the first side could inform surgeons about the potential risk for hypoparathyroidism when dissecting the opposite side. This may be especially strategic while operating on patients with prior gastric bypass who have a higher risk for morbidity, even from transient hypoparathyroidism.

Benefits and pitfalls of contrast-enhanced NIR fluorescence detection for assessment of parathyroid perfusion

Using ICG contrast-enhanced NIR fluorescence PG detection can lower the rate of permanent hypoparathyroidism after thyroidectomy. Currently it has been demonstrated that patients with at least one well perfused PG do not develop permanent hypoparathyroidism.²⁴ Although variable transient postoperative hypocalcemia rates have been reported,²⁶ this is likely due to the subjective nature of the visual grading associated with ICG scoring. Nonetheless, it must again be emphasized that patients with at least some perfusion of the PGs did not develop permanent hypoparathyroidism. Furthermore, during parathyroidectomy, ICG angiography may help detect diseased PGs faster than with the naked eye and can be used in subtotal parathyroidectomies to assess whether the planned remnant is adequately perfused, while allowing for multiple attempts at remnant selection.²⁰

Another advantage of ICG angiography is its ability to analyze the perfusion and the function of individual PGs. Prior to ICG angiography, it was not technically possible to assess the function or viability of individual glands as intraoperative PTH assays only measure the collective function of all PGs. However, this can also be a disadvantage for ICG angiography, as all 4 glands would need to be evaluated with angiography to reliably predict overall parathyroid function. ICG angiography can be safely performed multiple times but the best images are obtained after the first ICG injection. With administration of successive ICG doses, some residual fluorescence from the previous ICG injection(s) persists in soft tissues of the neck. This mainly happens because the ICG contrast or gradient between noninjection phase (pre-ICG) and injection phase (post-ICG) reduces with each successive injection, depending on the time duration between each ICG dose administered. This in turn could decrease the fluorescence contrast in subsequent NIR images. In addition, because of the strong remnant ICG fluorescence, it is not possible to detect the NIRAF of PGs even after a single ICG injection. If NIRAF-based parathyroid detection or localization is planned, it needs to be performed before ICG angiography.

In conclusion, despite the existing constraints observed with NIR fluorescence detection, this technology could be further enhanced for easier integration into the current surgical workflow. Currently, image-based NIRAF is limited by intensity of the signal and image resolution. As a result, real-time quantitative assessment for image-based NIRAF detection can be challenging. Furthermore, as contrast-based NIR fluorescence detection using ICG can overwhelm autofluorescence detection, it is not possible to use NIRAF to find missing PGs once ICG is used. There are potential alternatives to evaluate tissue perfusion without using ICG. For example, laser speckle contrast imaging can quantify parathyroid perfusion with 91.5% accuracy (Fig 5, A–D).²⁷ Laser speckle contrast imaging can effectively be combined with NIR fluorescence detection to provide a label-free device that can both localize PGs and determine their perfusion status.²⁸ With contrast-enhanced NIR fluorescence detection, ICG angiography can also be used to identify parathyroid blood vessels before thyroid resection and assist the surgeon in preserving PG vasculature (Fig 5, E and F). This valuable technique of mapping and tracking the vessels that perfuse and drain the PGsehas been coined as “parathyroid vessel cartography” (communication with Dr Benmiloud at the Second Symposium of Parathyroid Fluorescence 2020, Lausanne).

Fig. 5.

Future directions to track PG perfusion and preserve its vasculature. (A, B, C, D) Laser speckle contrast imaging is a label-free modality that could be used to assess PG perfusion in real-time. White light (left) and laser speckle image (right) for a well-vascularized PG (encircled in A and B), compared with a devascularized PG (encircled in C and D). Although the PG vascularity is not apparent with white light (surgeon’s view), laser speckle contrast is markedly lower for a well-vascularized PG than a devascularized one. (E and F) Parathyroid vessel mapping or “cartography” is another valuable technique that could be used to visualize the PG vasculature beforehand to preserve it during thyroid dissection. White light (left) shows left thyroid lobe retracted medially to expose left inferior PG (encircled). After ICG injection, the parathyroid artery can be clearly visualized (dotted lines and arrows) in the merged white light-near infrared image (right). Figure adapted from Mannoh et al²⁷ and unpublished images obtained with Novadaq.

Currently surgeons see the fluorescence images on a display monitor that require them to stare away from the surgical field. Binocular surgical goggles can provide augmented reality and are more ergonomic for the surgeon than display monitors.²⁹ Similarly, OTIS (Overlay Tissue Imaging System) can detect and back- project NIR fluorescence signal as visibly bright images directly onto the surgical field,³⁰ allowing intuitive visual identification of PGs by augmented reality.

Our recommendations summarized in Tables II to IV show how NIR fluorescence detection can be integrated into the current workflow in thyroid and parathyroid surgeries. Additional investigations are required to examine if NIR fluorescence detection devices can assist younger surgeons in becoming better trained to identify PGs more quickly. Furthermore, these modalities could be potentially used as education tools for teaching fellows or residents to identify or locate and preserve PGs more efficiently during head and neck surgeries.

Table IV.

Possible indications for use of contrast-enhanced NIR fluorescence detection technologies to preserve PGs in thyroid and parathyroid operations

Procedures	Indications	Current alternatives
Thyroid lobectomies; hemithyroidectomies	To determine PG perfusion and therefore PG function after lobectomy, anticipating the risk of hypoparathyroidism in case a completion thyroidectomy is necessary	No alternative
Total thyroidectomies	To determine PG perfusion and therefore PG function after thyroidectomy, therefore predicting the risk (or absence thereof) of postoperative hypoparathyroidism To determine PG perfusion and therefore PG function after the first side lobectomy, allowing to adapt the extension of the resection of the second side (staged thyroidectomy) in case the first side PGs could not be preserved To search for PG vessels before resection, allowing their preservation during thyroidectomy To identify devascularized PGs to decide whether to leave them in situ or to auto-transplant them	IOPTH assays
Total thyroidectomies with central neck dissection or lateral neck dissection	To determine the perfusion and therefore the function of the PGs after thyroidectomy, therefore predicting the risk (or absence thereof) of postoperative hypoparathyroidism To determine PG perfusion and therefore PG function after the first side lobectomy, allowing to adapt the extension of the resection of the second side (staged thyroidectomy), in case the first side PGs could not be preserved To determine PG perfusion and therefore PG function after the thyroidectomy, assessing the additional risk of hypoparathyroidism if a prophylactic central neck dissection is performed To search for PG vessels before resection, allowing their preservation during dissection To identify devascularized PGs in order to decide whether to leave them in situ or to auto-transplant them	IOPTH assays
Completion/reoperative thyroidectomies	To determine PG perfusion and therefore PG function after thyroidectomy, therefore predicting the risk (or absence thereof) of postoperative hypoparathyroidism To search for PG vessels before resection, allowing their preservation during thyroidectomy To identify devascularized PGs in order to decide whether to leave them in situ or to auto-transplant them	IOPTH assays
Focused parathyroidectomies	To enhance the identification of abnormal PGs, including ectopic adenomas (particularly in the thymus)	IOPTH assays, frozen section analysis, tissue aspirate PTH analysis
Bilateral exploration parathyroidectomies	To enhance the identification of abnormal PGs including ectopic adenomas (particularly in the thymus) To identify and avoid injuring the normal PGs. To determine PG perfusion and function of the remaining PGs To determine PG perfusion and function of the parathyroid remnant in subtotal parathyroidectomy To prepare more than one parathyroid remnant and to leave the best one, avoiding postoperative hypoparathyroidism in subtotal parathyroidectomy	IOPTH assays, frozen section analysis, tissue aspirate PTH analysis
Completion/reoperative parathyroidectomies	To enhance the identification of abnormal PGs including adenomas (particularly in the thymus) To determine PG perfusion and function of the remaining PGs To determine PG perfusion and function of the parathyroid remnant in subtotal parathyroidectomy To prepare more than one parathyroid remnant and to leave the best one, avoiding postoperative hypoparathyroidism in subtotal parathyroidectomy	IOPTH assays, frozen section analysis, tissue aspirate PTH analysis
Concurrent thyroid-parathyroid procedures	To enhance the identification of abnormal PGs including adenomas (particularly in the thymus) To determine PG perfusion and function of the PGs after thyroidectomy, therefore predicting the risk (or absence thereof) of postoperative hypoparathyroidism To search for PG vessels before resection, allowing to preserve them during dissection To identify devascularized PGs to decide whether to leave them in situ or to auto-transplant them	IOPTH assays, frozen section analysis, tissue aspirate PTH analysis

IOPTH, intraoperative parathyroid hormone; NIRAF, near infrared autofluorescence.

Although NIR fluorescence detection is a safe technique that appears to be efficacious, more investigations are needed to assess its cost-effectiveness. For high-volume centers, investing in a NIRAF detection technology may potentially reduce frozen sections, operating time, and postoperative complications. Although low-volume surgeons could benefit the most from these technologies, the cost-efficacy of these modalities need to be further substantiated for eventual use in low-volume centers. In conclusion, NIR fluorescence detection has the potential to help identify and preserve PGs, improve surgical efficiency, and minimize hypoparathyroidism.

Table III.

Possible indications for use of NIRAF detection technologies to identify PGs in thyroid and parathyroid operations

Procedures	Indications	Current alternatives
Thyroid lobectomy; hemithyroidectomy	To identify and preserve PGs, by distinguishing from fat, thyroid nodules, or lymph nodes in cases when a completion thyroidectomy may be necessary	Frozen section analysis
Total thyroidectomy	Graves and Hashimoto thyroiditis: To identify and preserve PGs when dissecting through fibrous, highly vascular tissues and reactive lymph nodes. Toxic and nontoxic multinodular goiter: To distinguish between thyroid nodules and PGs in the setting of distorted anatomy Malignant thyroid disease: To identify and preserve as many PGs in case of central neck recurrence that may require reoperation later To identify devascularized PGs in situ or incidentally resected in the specimen before auto-transplant	Frozen section analysis
Total thyroidectomies with central neck dissection or lateral neck dissection	To identify and preserve as many PGs during extensive thyroid and lymph node dissection To distinguish PGs from lymph nodes Caution: Metastatic lymph nodes may provide occasional false positive with NIRAF detection technologies (see Mitigation Strategies in Table II)	Frozen section analysis
Completion/reoperative thyroidectomy	To identify and preserve PGs by distinguishing from fat, thyroid nodules, or lymph nodes due to fibrosis and distorted anatomy from previous neck operation(s) To identify devascularized PGs in situ or incidentally resected in the specimen before auto-transplant To monitor PG(s) that was/were previously auto-transplanted	Frozen section analysis
Focused parathyroidectomy	To identify abnormal or ectopic PGs To identify and confirm the abnormal PG when preoperative localization studies are positive To identify the other ipsilateral PG in unilateral explorations To identify possible PGs by distinguishing from fat, thyroid nodules, lymph nodes and during conversion to bilateral exploration (when IOPTH assays fails to decrease)	Frozen section analysis, tissue aspirate PTH analysis
Bilateral exploration parathyroidectomy	To identify abnormal or ectopic PGs To identify the diseased PG when preoperative localization studies are negative To identify possible PGs by distinguishing from fat, thyroid nodules, lymph nodes during bilateral exploration To identify other PGs when IOPTH assays fails to decrease To identify and preserve normal/remnant PG(s)	Frozen section analysis, tissue aspirate PTH analysis
Completion/reoperative parathyroidectomy	To identify and distinguish PGs from adjacent fat, thyroid nodules or lymph nodes due to distorted anatomy from previous surgeries To identify and preserve normal/remnant PG(s) To monitor PG(s) that was/were previously auto-transplanted	Frozen section analysis, tissue aspirate PTH analysis
Concurrent thyroid-parathyroid procedures	To identify and distinguish between PGs from adjacent fat, thyroid nodules, and lymph nodes To identify devascularized PGs in situ or incidentally resected in the specimen before auto-transplant	Frozen section analysis, tissue aspirate PTH analysis

IOPTH, intraoperative parathyroid hormone; NIRAF, near infrared autofluorescence.