Does fluoroscopy prevent inadvertent parathyroidectomy in thyroid surgery?

Abstract

Introduction

Near-infrared imaging may facilitate intraoperative identification of parathyroid glands by causing autofluorescence but its clinical value has not been established. Inadvertent parathyroidectomy occurs in 5–22% of thyroidectomies and is associated with temporary and permanent hypoparathyroidism. The aim of this study was to determine whether near-infrared imaging prevents inadvertent parathyroidectomy and early hypocalcaemia as a surrogate for permanent hypoparathyroidism.

Materials and methods

Near-infrared imaging was used in a prospective cohort of consecutive thyroidectomies. Thyroidectomies performed prior to the introduction of near-infrared imaging formed a control group. The thyroid bed and specimen were scanned with near-infrared imaging. Areas of autofluorescence on the specimen were examined and any parathyroid tissue found was autotransplanted. Inadvertent parathyroidectomy was therefore recorded as established intraoperatively by near-infrared imaging (allowing autotransplantation) or on subsequent histology (missed). Serum calcium and parathyroid hormone were measured on day one and at two weeks and six months postoperatively.

Results

A total of 269 patients were included: 106 near-infrared imaging and 163 controls. Inadvertent parathyroidectomy was detected by near-infrared imaging in two (and autotransplantation performed) and histologically (i.e. missed by near-infrared imaging in 13, 12.3% vs 17, 10.4% controls). Neither result was statistically significant (P = 0.08, 0.89). There was no significant difference in serum calcium or parathyroid hormone between near-infrared imaging and control groups at one day, two weeks or thereafter.

Discussion

Near-infrared imaging may detect inadvertent parathyroidectomy and may allow autotransplantation. It did not, however, reduce the incidence of missed inadvertent parathyroidectomy and no difference was seen in early hypocalcaemia or late hypoparathyroidism. Current near-infrared imaging technology does not appear to confer a clinical benefit sufficient to justify its use.

Introduction

Permanent hypoparathyroidism is defined as the loss of parathyroid function resulting in the dependence on calcium and vitamin-D analogues six months or more postoperatively. It is associated with a reduction in the quality of life of affected patients and, in one study, an increase in mortality.^1,2 Estimates of the incidence of permanent hypoparathyroidism range from 1% to 25%.^3–5 Loss of parathyroid function may be caused by their inadvertent removal, damage to their arterial supply and/or venous drainage, direct injury from manipulation or thermal energy, or a combination of these mechanisms.

Thyroidectomy operative technique has evolved to minimise the risk to parathyroid glands, using careful dissection, early identification and avoidance of excessive manipulation or thermal energy in close proximity to the parathyroid vascular pedicle with the aim of leaving preserved, well-vascularised parathyroid glands in place. As might be expected, outcomes including hypoparathyroidism correlate with surgical experience.^6,7 However, even with meticulous technique, some parathyroid glands will not be viable if left in place and others may be removed inadvertently. Standard practice at the end of thyroidectomy therefore includes visual review of the thyroid bed to inspect for the viability of the identified parathyroids and the inspection of the thyroid specimen for inadvertently removed parathyroid glands. Parathyroid glands which are not viable or have been inadvertently removed should be autotransplanted at the time of surgery. Autotransplantation is a simple process by which parathyroid fragments are suspended in a balanced salt solution and injected into the sternocleidomastoid muscle and, if performed correctly, contributes to preventing hypoparathyroidism.^3,8 ‘Missed’ inadvertent parathyroidectomy is discovered subsequently by the pathologist reviewing the thyroid specimen and occurs in between 5% to 22% of thyroidectomies with most towards the higher end of the spectrum.^9–12 At this point, salvage of the parathyroid gland is not possible.

The use of intraoperative near-infrared imaging (NIRI) has been shown to assist in the identification of parathyroid glands during neck surgery.^13,14 Parathyroid tissue emits light in the near infrared spectrum (700–900 nm) in the absence of any contrast administration, a property known as ‘autofluorescence’. Parathyroid tissue reliably shows autofluorescence both in vivo and after excision, with a signal intensity greater than that of thyroid autofluorescence and to maintain its emission for at least one hour after resection.¹⁵ Although feasibility studies have shown that NIRI (with and without fluorescent contrast) permits parathyroid identification and preservation,^15–18 long-term clinical outcomes have not been reported.

The aim of this study was to determine whether the use of intraoperative NIRI for parathyroid autofluorescence could reduce the incidence of post-thyroidectomy hypoparathyroidism.

Materials and methods

The primary outcome was the rate of inadvertent parathyroidectomy and secondary outcomes were early postoperative hypocalcaemia/hypoparathyroidaemia following total thyroidectomy. A sample size calculation, based on a background rate of inadvertent parathyroidectomy of 10% and estimated reduction to one fifth, with a power of 0.8, suggested that a minimum of 78 patients would be required in the NIRI group to reach a significant result at 0.05. With ethical review board approval, NIRI was used intraoperatively in consecutive patients undergoing thyroidectomy (hemi- and total). A control group was formed from thyroidectomies performed prior to the introduction of NIRI. The study was funded by a grant from the Butterfly Thyroid Cancer Trust (#0316) and an additional independent financial donor.

Thyroidectomy was performed by one of two consultant surgeons (FP or NT) or by a specialist trainee (one of three specialty trainees at year 4+) under their direct supervision in a department undertaking more than 600 neck endocrine procedures (predominantly thyroidectomy and parathyroidectomy) per annum. A standardised approach was used in all cases with general anaesthesia, access via a collar incision, capsular dissection, early identification of parathyroid glands in situ and inspection of the specimen for inadvertently removed parathyroid glands or parathyroid tissue. Any excised parathyroid tissue was reimplanted using a standard technique of suspension in 1 ml balanced salt solution and injection into the sternocleidomastoid muscle using wide-bore Braun Micropin needle.⁸ Following total and completion thyroidectomy, prophylactic oral calcium at 1 g twice daily was commenced and serum calcium and parathyroid hormone (PTH) measured on the morning after surgery, at outpatient review two weeks later. Prophylactic oral calcium was continued for one week in patients with hypocalcaemia alone, 1-alfa-calcidol was added in those with biochemical evidence of hypoparathyroidism (PTH < 1 pmol/l) and continued until two-week clinic review. Confirmation of normalisation of the biochemistry at two-week review prompted tapering of supplements. All patients were reviewed again at three to six months and those with persistently abnormal biochemistry or still requiring calcium/vitamin D supplementation were followed thereafter.

NIRI was performed with the Fluobeam® 800 (Fluoptics, Grenoble, France) device. This proprietary system was selected because it is licensed for use in thyroid surgery, does not require the administration of fluorescent contrast, thus avoiding any risk of contrast allergy/anaphylaxis and requires no contact between the camera and tissue. This system consists of a standard laptop with Fluoptics© software installed, a near-infrared generator/processor and a hand-held camera head comprising a combined near-infrared emitter (which produces light of 750 nm) and detector (collecting light of 850–900 nm).

The flow of patients through the study is shown in Figure 1: thyroidectomy was performed entirely as standard until the specimen was removed and inspected. At this point, NIRI was used to scan the specimen. The camera head was held in a sterile sheath by the operating surgeon at around 20 cm from the tissue to be visualised, all ambient lights were turned off and images were processed and displayed on the laptop screen. Any areas of autofluorescence were reinspected by the surgeon to confirm or refute the presence of parathyroid tissue (fig 2 shows normal parathyroids in the thyroid bed under fluorescent light). Autotransplantation of inadvertently removed parathyroid tissue was performed and the thyroid bed was then scanned to confirm the location of remaining parathyroids. Thyroid specimens were sent as standard for histological analysis and the presence of any parathyroid tissue reported as routine.

Figure 1.

Patient flow through the study.

Figure 2.

The thyroid bed after right lobectomy with parathyroid glands circled(left); viewed under NIRI, the two parathyroid glands showing autofluorescence (right).

Descriptive statistics were used for demographics of the study participants and the Fisher’s exact test, calculated in SPSS Statistics for Windows, Version 24.0, for comparison of the inadvertent parathyroidectomy and postoperative serum calcium/PTH between NIRI and control groups.

Results

A total of 269 patients were included between January 2016 and October 2017: 106 NIRI and 163 controls. Demographics of the participants are as shown in Table 1: there were 221 women and 48 men split evenly between the two groups, median age was 48 years and there was no significant difference in the number of procedures performed for malignancy in the two groups. Total thyroidectomy was performed in 140 patients and hemithyroidectomy in 129. The Fluobeam equipment was found to be quick and easy to use, resulting in minimal increase (two to five minutes) in operating time.

Table 1.

Demographics of study participants.

Demographic	NIRI group	Control group	P-valuea
Patients (n)	106	163	–
Sex: female : male (%)	88 : 18 (83% : 17%)	133 : 30 (81.5 : 18.5%)	0.84
Mean age at operation (years ± SD)	48.0 ± 14.2	47.9 ± 15.8	0.95
Total vs hemithyroidectomy (%)	54 : 52 (51% : 49%)	86 : 77 (53% : 47%)	0.69
Nodal dissection, n (%)	3 (2.8%)	11 (6.7%)	0.16
Indication: suspicious for malignancy n (%)	57 (54%)	79 (48.5%)	0.39
Histological diagnosis of malignancy n (%)	38 (36%)	47 (29 %)	0.23
Mean specimen weight (g)	74.1	60.7	0.72

^a All not significant at P < 0.05.

NIRI, near-infrared imaging; SD, standard deviation.

Figure 3 shows the incidence of inadvertent parathyroidectomy, which was detected by NIRI in two cases (and autotransplantation performed) but missed at operation and detected histologically in 17 (10.2%) of controls and 13 (12.3%) NIRI. Neither result was statistically significant (P = 0.08, 0.89). The lower limit of normal for albumin-corrected serum calcium was defined as 2 mmol/l and PTH 1 pmol/l. Patients requiring calcium and vitamin D supplementation beyond two weeks were included in the definition of hypoparathyroidism, even if the calcium and PTH were within the normal range. Hypocalcaemia on day 1 post-thyroidectomy occurred in 10.5% (n = 9) controls and 9.3% (n = 5) in the NIRI group (P = 0.53) and hypoparathyroidism in 11.6% (n = 10) and 11.1% (n = 6) (P = 0.38). There were no cases of hypoparathyroidism persisting at beyond six months post-thyroidectomy in either group. Table 2 shows the number of patients with hypocalcaemia and/or low PTH at day 1, two weeks and six months or beyond, none of which differed significantly between the NIRI and control groups.

Figure 3.

Primary outcome measure incidence of inadvertent parathyroidectomy (IP).

Table 2.

Secondary outcome measures in total thyroidectomy.

Secondary outcomes	NIRI group (n = 54)	Control group (n = 86)	P-valuea
Hypocalcaemia:
Day 1 (Corr Ca < 2.00 mmol/l)	5 (9.3%)	9 (10.5%)	0.53
Week 2 (Corr Ca < 2.00 mmol/l)	0	2 (2.3%)	0.57
Hypoparathyroidaemia (< 1 pmol/l):
Day 1	6 (11.1%)	10 (11.6%)	0.38
Week 2	0	5 (5.8%)	0.08
> 6 months or still requiring supplementation	0	0	NA

^a All not significant at P < 0.05.

Discussion

The use of near infrared imaging in neck endocrine surgery has grown in popularity since it was first reported in 2011.¹⁴ The appeal is clear: in parathyroidectomy any aid in the identification of an elusive parathyroid gland to prevent persistent hyperparathyroidism is welcome; and in thyroidectomy the failure to recognise and preserve parathyroids in place exposes to the risk of temporary and permanent hypoparathyroidism. However, the intraoperative identification of parathyroids is normally dependent solely upon the ability of the surgeon with outcomes dependent upon a surgeon’s operative volume.^7,19,20 Permanent hypoparathyroidism is the most common major complication of total thyroidectomy with rates in the published literature varying between 1% and 25%.^3–5 The rate of late hypocalcaemia is used a surrogate marker for permanent hypoparathyroidism by the British Association of Endocrine and Thyroid Surgeons and a rate of 6.5% reported following total thyroidectomy in their fifth national audit.²¹ Although, for the majority of patients, treatment with oral calcium and vitamin D analogues prevents hypocalcaemia, a minority have brittle disease requiring regular blood tests and hospital appointments. Consequently, health-related quality of life is reduced and, in one study, there was an association with increased mortality.^1,22 The rate of inadvertent parathyroidectomy (10–12%) and absence of permanent postoperative hypoparathyroidism in this study is equivalent to that of other high-volume units.²³

‘Autofluorescence’ is defined as the emission of light in the near infrared spectrum by a biological agent and occurs due to excitation by light of certain wavelengths.²⁴ This property has been exploited in several clinical specialties including dermatology, ophthalmology and surgery for breast cancer.^25–27 NIRI of parathyroid tissue requires excitation at around 785 nm and results in a signal in the 800–950 nm wavelength, 1.2–18 times brighter than that from thyroid tissue and persisting for several hours after excision of the tissue.^13,14 The addition of an intravenous contrast agent, such as indocyanine green, which emits autofluorescence, has also been described to map vasculature in colorectal anastomoses,²⁸ and in parathyroidectomy,^16,17,29,30 in an attempt to confirm their viability.

Several studies have been published using NIRI in thyroidectomy with promising results such as an autofluorescence signal from 100% of parathyroid glands in one study of 45 mixed thyroidectomy and parathyroidectomy,¹³ 98% in one multicentre study of 210 patients,³¹ and 90% in another study of 41 patients.³² However, to date, no study has shown any definite evidence of an improvement in clinical outcomes. Although the technique of NIRI is low risk, with no change in surgical approach required, no contact between the imaging camera and tissue in all but one system,^13,14,33 and no requirement for intravenous contrast with its small risk of anaphylaxis, the cost is not insignificant. In a publicly funded healthcare system, the appraisal of new technical adjuncts such as this, in terms of their clinical outcomes, is essential. By way of an example, the system used in this study costs approximately £35,000 for use of the equipment plus consumables for one year.

The type of NIRI camera and method of use has differed between studies and may impact upon the outcomes. The NIRI system of the Mahadevan–Jansen group, which first developed NIRI for clinical use in the neck, requires contact between the imaging probe and tissues and reports a numerical fluorescence output.^18,33 The three proprietary systems in most widespread clinical use (the Fluobeam 800 used in this study, the Pinpoint® by Novadaq, Ontario, Canada, and the OPAL1® by Karl Storz) are contact free and provide images but no quantification of the autofluorescence. The greatest sensitivity for parathyroid autofluorescence has been reported by the Mahadevan–Jansen group, with 100% of parathyroids showing autofluorescence,³³ whereas the proprietary systems show rates of 90–98%.^31,32 One study has compared the detection rate of parathyroids by NIRI with and without indocyanine green contrast and has shown similar rates at 95% and 98%, respectively.³⁴

NIRI has been used prior to definite identification of the parathyroids by the surgeon and reported as facilitating detection before traditional inspection in 52% with autofluorescence alone,³⁴ or a mean of one gland per patient,³¹ compared with 6% with ICG contrast.³⁴ Correlation of autofluorescence with outcomes at total thyroidectomy have, thus far, only reported early markers. One study has attempted to correlate early postoperative PTH with the level of indocyanine green fluorescence and has shown a correlation between these two.³⁰ However, only 27 patients were included and only 84% of glands fluoresced. Another study of 70 patients visualised with indocyanine green showed a correlation between day 1 hypocalcaemia, decline in PTH and brightness of the indocyanine green signal.³⁵ Correlation has also been shown between indocyanine green-proven vascularity of at least one parathyroid gland and absence of day one hypoparathyroid hormone.¹⁶ However, no study, either with or without indocyanine green has reported on longer-term outcomes and no comparative studies have been reported.

In conclusion, in this study, two parathyroid glands were identified after inadvertent removal, permitting autotransplantation. However, there was a false negative rate of 12.3% and no significant difference in the rate of inadvertent parathyroidectomy or in early or late postoperative outcomes. NIRI does show some promise with regard to parathyroid identification, and it is possible that fine tuning of the technical aspects of NIRI and adaptation of surgical technique to optimise its impact may improve its performance. However, the lack of definite benefit in clinical outcomes precludes it from being a cost-effective adjunct in its current form.