Auris Nasus Larynx 42 (2015) 311–317
Contents lists available at ScienceDirect
Auris Nasus Larynx journal homepage: www.elsevier.com/locate/anl
Harmonic focus in thyroidectomy for substernal goiter Christoffer Holst Hahn *, Waldemar Trolle, Christian Hjort Sørensen ENT Department, Rigshospitalet/Gentofte University Hospital, Copenhagen, Denmark
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 January 2014 Accepted 29 December 2014 Available online 6 March 2015
Objectives: No previous prospective study has evaluated harmonic scalpel in thyroidectomy for substernal goiter. The objective of this study was to evaluate the use of harmonic scalpel (FOCUS shear, Ethicon Endo-Surgery) in thyroidectomy for substernal goiter for blood loss, operative time, hospital stay and complications. Materials and methods: Prospective non-randomised study of 242 consecutive patients with substernal goiter out of 2258 patients (11%) who underwent thyroidectomy. A total of 121 patients had thyroidectomy performed with bipolar electrocoagulation and knot-tying techniques and 121 patients had harmonic scalpel thyroidectomy. Results: The use of harmonic scalpel was associated with significant reduction in intraoperative blood loss (50 vs. 100 mL, p = 0.001), postoperative haemorrhage (4% vs. 12%, p = 0.03) and length of hospital stay (2 vs. 3 days, p = 0.001). The mean operative time was significantly longer in the harmonic group. Conclusion: Harmonic scalpel is a safe tool for thyroidectomy for substernal goiter. Its utilisation is associated with reduced blood loss, lower incidence of postoperative haemorrhage and shorter hospital stay. ß 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Thyroidectomy Substernal goiter Ultrasonic devise Electrocautery Complications
1. Introduction Thyroidectomy is a frequently performed operation in ENT surgery. The thyroid gland has one of the richest blood supplies, with numerous blood vessels entering the parenchyma [1]. Thus, meticulous haemostasis is important in order to avoid damage to adjacent structures, e.g. the recurrent laryngeal nerves and the parathyroid glands, and to prevent postoperative haematomas [2]. Complication rates are reported between 5 and 15% for transient hypocalcaemia and recurrent laryngeal nerve injury [3–5], 1% for permanent paralysis of the recurrent laryngeal nerve and hypocalcaemia [6,7], and 0–5% for postoperative haemorrhage in large cohort studies [8,9]. Haemostasis can be performed by suture, clip ligation, electrocoagulation or ultrasound technology. A harmonic scalpel uses ultrasound technology to denature proteins in vessel walls and tissues. Harmonic scalpel is able to seal vessels up to 5 mm in diameter, whereas bipolar electrosurgical devices only seal vessels reliably up to 2 mm in diameter [10,11]. It can be used for ligation, dissection and cutting. In cervical goiters, the harmonic scalpel is
* Corresponding author at: Gjorslevvej 14, 2720 Vanløse, Denmark. Tel.: +45 22994416; fax: +45 35452071. E-mail address: [email protected] (C.H. Hahn). http://dx.doi.org/10.1016/j.anl.2014.12.007 0385-8146/ß 2015 Elsevier Ireland Ltd. All rights reserved.
safe and confers some advantages over conventional dissections methods, including decreased intraoperative blood loss and lower operative time [2]. The complication rate is higher in substernal than cervical thyroidectomy [6,12–14], as dissection of the mediastinal component sometime forces the surgeon to perform blind manoeuvres. Furthermore, intraoperative bleeding complicates the surgical dissection, stains and obscures important structures. In a large study of over 19,000 patients the risk of haematoma, hypocalcaemia and laryngeal recurrent nerve palsy was roughly double in substernal goiters compared to cervical goiters [6]. Until now, no study has evaluated harmonic scalpel in substernal goiter. Therefore, we sought to determine whether the use of harmonic scalpel is a safe alternative to conventional haemostasis method in surgery for substernal goiter. We compared operative blood loss, operative time and rate of complications. 2. Materials and methods In this prospective non-randomised trial, all patients scheduled for substernal thyroid surgery at Copenhagen University Hospital, Gentofte, Denmark were enrolled from January 2001 to April 2012. The study was a part of the THYKIR Database [8], approved by the Danish Data protection Agency (journal number 2000-41-0010). In 2008 the department started to use the
312
C.H. Hahn et al. / Auris Nasus Larynx 42 (2015) 311–317
harmonic scalpel. We included patients with benign and malign thyroid disease. A goiter was considered substernal if any part of the gland extended through the thoracic inlet. All patients with goiters were examined with ultrasound examination. If the caudal border of the thyroid grand was not visible on ultrasound examination, a CT scan without contrast was performed throughout the whole study period. The following parameters were recorded in the database: age, gender, earlier thyroid surgery, degree of substernal involvement, weight of surgical specimen, surgical procedure, histology of thyroid specimens, operative time, bleeding during surgery, postoperative haemorrhage, temporary and permanent recurrent laryngeal nerve injury, infection, hypoparathyroidism, surgeon, and length of hospital stay. Degree of substernal involvement was divided into three categories: (1) 1–4 cm, (2) 5–9 cm and (3) >10 cm below the thoracic inlet. Operative time was recorded as the time interval between skin incision and closure. Mobility of the vocal cords was checked by laryngoscopy before and after surgery. Patients with postoperative vocal cord palsy were examined 3, 6 and 12 months after the operation or until full recovery had been confirmed. Bleeding during surgery was measured as weight of sponge and blood in suction. Postoperative haemorrhage was defined as bleeding after wound closure that required reoperation in general anaesthesia. Four senior surgeons performed all operations. Three surgeons (Surgeons 2–4, Table 4) used the conventional method until 2007, and harmonic scalpel from 2008 until 2012. The last surgeon (Surgeon 1, Tables 4 and 5) used the conventional method in the whole period from 2001 until 2012. SPSS 20.0 was used for statistical analysis. Length of hospital stay and bleeding during surgery were not normally distributed (Shapiro–Wilk test of normality), and were analysed with Mann– Whitney’s test and Kruskal–Wallis test when appropriate. Age, weight of surgical specimen, operative time and parathyroid hormone level were all normally distributed, and were tested with the unpaired t test and ANOVA test when appropriate. The Chi-square test was used to test gender, surgical procedure, histology, degree of substernal goiter, postoperative haemorrhage and hypoparathyroidism. Fisher’s exact test was used to test infection and recurrent laryngeal nerve injury.
2.1. Surgical procedure Following a Kocher incision, the skin flaps were constructed. The strap muscles were divided in the midline and retracted laterally. Capsular dissection was conducted. In the harmonic group, the inferior, middle and superior vessels were divided with harmonic scalpel (FOCUS shear, Ethicon Endo-Surgery). In the conventional group, all vessels were ligated with suture ties, clips and electrocautery. In both groups, the recurrent laryngeal nerves and parathyroid glands were identified and protected, and the thoracic component of the thyroid was retracted manually to the cervical region. If sternotomy was necessary, a thoracic surgeon was called. 3. Results During the study period a total of 2258 patients underwent thyroidectomy. Of these, 242 patients (11%) were operated for substernal goiter. 3.1. Substernal goiter versus cervical goiter Patients operated for substernal goiter were older compared to patients operated for cervical goiter (59 vs. 48 years, p = 0.001) and more often male (29% vs. 18%, p = 0.001). Furthermore, the incidence of total thyroidectomy was higher (35% vs. 25%, p = 0.04), the mean size of the goiter was over three times larger (140 g vs. 30 g, p = 0.001), the mean operative time was significantly longer and bleeding during surgery and postoperative haemorrhage was significantly larger in substernal goiters (Table 1). 3.2. Harmonic versus conventional haemostasis in substernal goiter A total of 121 patients (50%) underwent harmonic thyroidectomy and 121 patients (50%) underwent conventional thyroidectomy (Table 2). No patients needed a sternotomy. Most patients (76%) had a goiter reaching 1–4 cm substernally. Twenty-two percent had a goiter 5–9 cm substernally, and only 3 patients (1%) had goiter extending more than 10 cm substernally. No significant
Table 1 Substernal vs. cervical goiter. Characteristics Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Histology Malign, n (%) Benign, n (%) Mean weight, g (SD) Mean operative time, min (SD) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR) Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%) IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
Cervical goiter n = 2016
Substernal goiter n = 242
363 (18) 1653 (82) 48 (14)
70 (29) 172 (71) 59 (14)
1518 (75) 498 (25)
158 (65) 84 (35)
298 1718 30 92 103 50 34 37 72 2
28 214 140 112 20 75 7 9 12 2
p-value 0.001
0.001 0.05
0.2 (15) (85) (105) (42) (5) (70) (1.7) (1.4) (2.8) (1)
n = 498 4.1 (3.4) 23 (4.6)
(12) (88) (121) (42) (8) (100) (3) (2.7) (3.6) (1)
n = 84 3.8 (2) 4 (4.9)
0.001 0.001 0.03 0.001 0.2 0.08 0.08 0.001
0.3 0.7
C.H. Hahn et al. / Auris Nasus Larynx 42 (2015) 311–317
313
Table 2 Substernal goiter. Harmonic vs conventional. Characteristics Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Histology Malign, n (%) Benign, n (%) Mean weight, g (SD) Degree of substernal goiter 1–4 cm, n (%) 5–9 cm, n (%) >10 cm, n (%) Mean operative time, min (SD) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR) Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%)
Harmonic n = 121
Conventional n = 121
Total n = 242
p-value 0.4
32 (26) 89 (74) 58 (14)
38 (31) 83 (69) 60 (15)
70 (29) 172 (71) 59 (14)
79 (65) 42 (35)
79 (65) 42 (35)
158 (65) 84 (35)
7 (6) 114 (94) 132 (105)
21 (17) 100 (83) 148 (135)
28 (12) 214 (88) 140 (121)
93 26 2 117 5 50 3 3 6 2
92 28 1 106 15 100 4 6 6 3
185 54 3 112 20 75 7 9 12 2
0.3 1
0.008
(77) (22) (2) (42) (4) (70) (2.5) (1.8) (3.6) (1)
n = 42 3.7 (1.7) 1 (2.6)
(76) (23) (1) (42) (12) (150) (3.3) (3.6) (3.6) (2)
n = 42 3.9 (2.3) 3 (7.1)
(76) (22) (1) (42) (8) (100) (3) (2.7) (3.6) (1)
n = 84 3.8 (2) 4 (4.9)
0.3 0.8
0.04 0.03 0.001 0.7 0.6 1 0.001
0.6 0.3
IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
differences were seen between the study groups with respect to age, gender, surgical procedure, weight or degree of substernal goiter (Table 2). A total of 28 patients (12%) had malignant diagnosis. In the conventional group significantly more had a malignant diagnosis compared to the harmonic group (p = 0.005). The mean operative time was significantly shorter in the conventional group (106 min vs. 117 min, p = 0.04). The use of harmonic scalpel was associated with a significant reduction in intraoperative blood loss (50 mL vs. 100 mL, p < 0.001) and postoperative haemorrhage (4% vs. 12%, p = 0.03). The median length of hospital stay was shorter in the harmonic group (3 days
vs. 2 days, p = 0.001). There was a non-significant trend towards less permanent recurrent laryngeal nerve palsy in the harmonic group (3 vs. 6 patients, p = 0.6) and fewer patients requiring permanent calcium consumption (1 vs. 3 patients, p = 0.3). No significant differences were observed between the two groups in relation to infection, postoperative parathyroid hormone level or transient recurrent laryngeal nerve injury (Table 2). Table 3 shows subgroup analysis for thyroidectomy of benign substernal goiter. In this subgroup the use of harmonic scalpel was associated with significant reduction in intraoperative blood loss (50 vs. 100 mL, p = 0.001), postoperative haemorrhage (4 vs. 14%,
Table 3 Substernal benign goiter. Harmonic vs. conventional haemostasis. Characteristics Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Mean weight, g (SD) Degree of substernal goiter 1–4 cm, n (%) 5–9 cm, n (%) >10 cm, n (%) Median operative time, min (IQR) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR) Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%) IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
Harmonic n = 114
Conventional n = 100
Total n = 214
p-value 0.5
30 (26) 84 (74) 58 (14)
31 (31) 69 (69) 59 (15)
74 (65) 40 (35) 134 (106)
66 (66) 34 (34) 153 (140)
140 (65) 74 (35) 144 (121)
87 25 2 116 5 50 3 3 6 2
76 23 1 104 14 100 2 4 4 3
163 48 3 110 19 75 5 7 10 2
(77) (22) (2) (39) (4.4) (70) (2.6) (1.9) (3.8) (1)
n = 40 4.0 (2.5) 1 (2.5)
(76) (23) (1) (37) (14) (150) (2) (3.0) (3.0) (2)
n = 34 3.8 (1.7) 2 (5.9)
61 (29) 153 (71) 59 (14)
(76) (22) (1) (38) (9) (100) (2.3) (2.4) (3.5) (1)
n = 74 3.8 (2) 3 (4)
0.7 0.9
0.3 0.8
0.03 0.02 0.001 1 0.7 0.8 0.02
0.6 0.6
314
C.H. Hahn et al. / Auris Nasus Larynx 42 (2015) 311–317
p = 0.02) and length of hospital stay (2 vs. 3 days, p = 0.02). The operative time was longer in the harmonic group (116 vs. 104 min, p = 0.03). 3.3. Difference by surgeons in substernal goiter Table 4 illustrates differences in operative characteristics between the surgeons. No significant differences were observed in relation to gender, age, surgical procedure, histology and weight. The mean operative time was significantly different between the four surgeons (p = 0.001). Post hoc multiple comparisons (Bonferroni) showed that Surgeon 1 was significantly faster than Surgeon 2 and Surgeon 4. No differences were observed between the four surgeons in relation to bleeding, infections, hospital stay or recurrent laryngeal nerve injury. Table 5 shows data from Table 4 disaggregated for each surgeon according to surgical method. For Surgeon 2 the use of harmonic scalpel was associated with a significant reduction in intraoperative blood loss (75 mL vs. 200 mL, p < 0.001), postoperative haemorrhage (4% vs. 23%, p = 0.006), wound infections (1% vs.10%, p = 0.04) and median length of hospital (3.5 days vs. 2 days, p < 0.001). No significant differences were observed between the harmonic group and the conventional group for Surgeons 3 and 4. 4. Discussion This is the first study comparing harmonic scalpel with the conventional haemostasis method in surgery for substernal goiter. Our results suggest that the use of harmonic scalpel results in significantly less intraoperative bleeding, postoperative haemorrhage and shorter hospital stay. Surprisingly, operative time was longer. There is no standard definition of substernal goiter. Some studies only include cases with 50% or more of the gland situated in the thorax, whereas other studies include patients with any part of the gland extending through the thoracic inlet. The reported incidence of substernal goiter is highly variable, ranging from 3 to 30% [3,6,15]. In our study the incidence of substernal goiter was 11%. The substernal goiter was larger, and the patients were older.
Mean operative time was longer, and bleeding during surgery was larger. This resulted in more postoperative bleeding and a tendency to more complications. This is in accordance with most studies in substernal goiters, where surgery is associated with increased complication rates [6,12–14]. A few studies conclude that the risk of complications is not increased [15]. Previous randomised prospective studies reported the safety and usefulness of harmonic focus in thyroidectomy [2,16–22]. A meta-analysis of randomised clinical trials comparing harmonic scalpel with conventional techniques showed a significant reduction in operative time equivalent to 23 min (nearly 25% of the operating room time). The reduction in operative time ranges from 15% to 30%, and is more apparent when the surgical procedure is long and complex [18,23]. The reduced surgical time is due to a decrease in intraoperative instrument exchange by 70% [18]. One prospective nonrandomized study on lobectomy showed no benefit in terms of operative time [23]. Our study demonstrated a significant increase in operative time from 106 min with the conventional technique to 117 min with harmonic focus, corresponding to an increase of 9%. Different length of operative time between the four surgeons may explain this finding. Surgeon 1, who used the conventional method in the whole study period, was significantly faster than Surgeon 2 and Surgeon 4, who used harmonic focus in the last part of the study period. Surgeons 2–4 had a mean operative time of 131 min with the conventional method and 117 min with the harmonic focus (p = 0.1). This corresponds to a 12% decrease in operative time. Unfortunately, due to our small study population, we may have failed to demonstrate a possible significant difference in operative time in our disaggregated analysis. However, the operative time of surgeon 2 was 15% shorter with the harmonic scalpel (p = 0.06). Bleeding in the neck is a potentially life-threatening complication. The thyroid is a highly vascular organ; thus, haemostasis is very important. In our study, intraoperative bleeding was significantly reduced from 100 mL in the conventional group to 50 mL in the harmonic group. This 50% reduction is consistent with all other randomised studies concerning harmonic scalpel, showing a relative reduction in intraoperative bleeding between 13 and 61% [24,25]. A meta-analysis demonstrated a 20 mL
Table 4 Difference by the surgeons. Characteristics
Surgeon 1 (n = 65)
Surgeon 2 (n = 99)
Surgeon 3 (n = 38)
Surgeon 4 (n = 40)
Total (n = 242)
P-value
20 (31) 45 (69) 61 (15)
31 (32) 68 (68) 62 (12)
12 (32) 26 (68) 56 (14)
12 (30) 28 (70) 60 (15)
70 (29) 172 (71) 59 (14)
1
39 (60) 26 (40)
52 (53) 47 (47)
28 (73) 10 (27)
25 (63) 15 (37)
158 (65) 84 (35)
Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Histology Malign, n (%) Benign, n (%) Mean weight, g (SD) Mean operative time, min (SD) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR)
10 55 142 90 6 80 1 6 0 2
Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%)
n = 26 3.7 (2.6) 2 (8)
0.3 0.3
0.7
IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
(15) (85) (103) (30) (10) (100) (2) (9) (2)
7 93 162 123 10 100 4 2 11 3
(7) (93) (137) (45) (10) (150) (4) (2) (11) (2)
n = 47 3.8 (1.9) 1 (2)
8 30 147 111 2 80 1 0 0 3
(21) (79) (123) (55) (5) (178) (3)
(2)
n = 10 4.3 (3) 1 (10)
3 37 141 123 2 50 1 1 1 2
(7%) (93%) (117) (37) (5) (79) (3) (2.5) (2.5) (1)
n = 15 3.9 (2) 0
28 (12) 214 (88) 150 (121) 112 (42) 20 (8) 75(100) 7 (3) 9 (2.7) 12 (3.6) 2 (1)
0.7 0.001 0.7 0.1 0.9 0.1 1 0.2
n = 84 3.8 (2) 4 (4.9)
0.6 0.5
Table 5 Harmonic vs conventional method for each surgeon. Characteristics
Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%)
Surgeon 2
Surgeon 3
Conventional
Harmonic
Conventional
Harmonic
n = 65
n=0
n = 31
n = 68
20 (31) 45 (69) 61 (15)
0 0 0
7 (23) 24 (77) 62 (14)
24 (35) 44 (65) 62 (12)
39 (60) 26 (40)
0 0
24 (77) 7 (23)
28 (41) 40 (59)
0 0 0 0 0 0 0 0 0 0
6 (19) 25 (81) 183 (180) 137 (47) 7 (23) 200 (125) 3 (10) 0 6 (19) 3.5 (3)
1 67 151 116 3 75 1 2 5 2
n = 26 0 0
n=0 4 (2) 0
n = 10 3.8 (1.9) 1
10 55 142 90 6 80 1 6 0 2
(15) (85) (103) (30) (10) (100) (2) (9) (2)
3.7 (2.6) 2 (8)
(1) (99) (110) (43) (4) (61) (1) (3) (7) (1)
p-value
0.8 0.2
0.3 0.06 0.006
Contents lists available at ScienceDirect
Auris Nasus Larynx journal homepage: www.elsevier.com/locate/anl
Harmonic focus in thyroidectomy for substernal goiter Christoffer Holst Hahn *, Waldemar Trolle, Christian Hjort Sørensen ENT Department, Rigshospitalet/Gentofte University Hospital, Copenhagen, Denmark
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 January 2014 Accepted 29 December 2014 Available online 6 March 2015
Objectives: No previous prospective study has evaluated harmonic scalpel in thyroidectomy for substernal goiter. The objective of this study was to evaluate the use of harmonic scalpel (FOCUS shear, Ethicon Endo-Surgery) in thyroidectomy for substernal goiter for blood loss, operative time, hospital stay and complications. Materials and methods: Prospective non-randomised study of 242 consecutive patients with substernal goiter out of 2258 patients (11%) who underwent thyroidectomy. A total of 121 patients had thyroidectomy performed with bipolar electrocoagulation and knot-tying techniques and 121 patients had harmonic scalpel thyroidectomy. Results: The use of harmonic scalpel was associated with significant reduction in intraoperative blood loss (50 vs. 100 mL, p = 0.001), postoperative haemorrhage (4% vs. 12%, p = 0.03) and length of hospital stay (2 vs. 3 days, p = 0.001). The mean operative time was significantly longer in the harmonic group. Conclusion: Harmonic scalpel is a safe tool for thyroidectomy for substernal goiter. Its utilisation is associated with reduced blood loss, lower incidence of postoperative haemorrhage and shorter hospital stay. ß 2015 Elsevier Ireland Ltd. All rights reserved.
Keywords: Thyroidectomy Substernal goiter Ultrasonic devise Electrocautery Complications
1. Introduction Thyroidectomy is a frequently performed operation in ENT surgery. The thyroid gland has one of the richest blood supplies, with numerous blood vessels entering the parenchyma [1]. Thus, meticulous haemostasis is important in order to avoid damage to adjacent structures, e.g. the recurrent laryngeal nerves and the parathyroid glands, and to prevent postoperative haematomas [2]. Complication rates are reported between 5 and 15% for transient hypocalcaemia and recurrent laryngeal nerve injury [3–5], 1% for permanent paralysis of the recurrent laryngeal nerve and hypocalcaemia [6,7], and 0–5% for postoperative haemorrhage in large cohort studies [8,9]. Haemostasis can be performed by suture, clip ligation, electrocoagulation or ultrasound technology. A harmonic scalpel uses ultrasound technology to denature proteins in vessel walls and tissues. Harmonic scalpel is able to seal vessels up to 5 mm in diameter, whereas bipolar electrosurgical devices only seal vessels reliably up to 2 mm in diameter [10,11]. It can be used for ligation, dissection and cutting. In cervical goiters, the harmonic scalpel is
* Corresponding author at: Gjorslevvej 14, 2720 Vanløse, Denmark. Tel.: +45 22994416; fax: +45 35452071. E-mail address: [email protected] (C.H. Hahn). http://dx.doi.org/10.1016/j.anl.2014.12.007 0385-8146/ß 2015 Elsevier Ireland Ltd. All rights reserved.
safe and confers some advantages over conventional dissections methods, including decreased intraoperative blood loss and lower operative time [2]. The complication rate is higher in substernal than cervical thyroidectomy [6,12–14], as dissection of the mediastinal component sometime forces the surgeon to perform blind manoeuvres. Furthermore, intraoperative bleeding complicates the surgical dissection, stains and obscures important structures. In a large study of over 19,000 patients the risk of haematoma, hypocalcaemia and laryngeal recurrent nerve palsy was roughly double in substernal goiters compared to cervical goiters [6]. Until now, no study has evaluated harmonic scalpel in substernal goiter. Therefore, we sought to determine whether the use of harmonic scalpel is a safe alternative to conventional haemostasis method in surgery for substernal goiter. We compared operative blood loss, operative time and rate of complications. 2. Materials and methods In this prospective non-randomised trial, all patients scheduled for substernal thyroid surgery at Copenhagen University Hospital, Gentofte, Denmark were enrolled from January 2001 to April 2012. The study was a part of the THYKIR Database [8], approved by the Danish Data protection Agency (journal number 2000-41-0010). In 2008 the department started to use the
312
C.H. Hahn et al. / Auris Nasus Larynx 42 (2015) 311–317
harmonic scalpel. We included patients with benign and malign thyroid disease. A goiter was considered substernal if any part of the gland extended through the thoracic inlet. All patients with goiters were examined with ultrasound examination. If the caudal border of the thyroid grand was not visible on ultrasound examination, a CT scan without contrast was performed throughout the whole study period. The following parameters were recorded in the database: age, gender, earlier thyroid surgery, degree of substernal involvement, weight of surgical specimen, surgical procedure, histology of thyroid specimens, operative time, bleeding during surgery, postoperative haemorrhage, temporary and permanent recurrent laryngeal nerve injury, infection, hypoparathyroidism, surgeon, and length of hospital stay. Degree of substernal involvement was divided into three categories: (1) 1–4 cm, (2) 5–9 cm and (3) >10 cm below the thoracic inlet. Operative time was recorded as the time interval between skin incision and closure. Mobility of the vocal cords was checked by laryngoscopy before and after surgery. Patients with postoperative vocal cord palsy were examined 3, 6 and 12 months after the operation or until full recovery had been confirmed. Bleeding during surgery was measured as weight of sponge and blood in suction. Postoperative haemorrhage was defined as bleeding after wound closure that required reoperation in general anaesthesia. Four senior surgeons performed all operations. Three surgeons (Surgeons 2–4, Table 4) used the conventional method until 2007, and harmonic scalpel from 2008 until 2012. The last surgeon (Surgeon 1, Tables 4 and 5) used the conventional method in the whole period from 2001 until 2012. SPSS 20.0 was used for statistical analysis. Length of hospital stay and bleeding during surgery were not normally distributed (Shapiro–Wilk test of normality), and were analysed with Mann– Whitney’s test and Kruskal–Wallis test when appropriate. Age, weight of surgical specimen, operative time and parathyroid hormone level were all normally distributed, and were tested with the unpaired t test and ANOVA test when appropriate. The Chi-square test was used to test gender, surgical procedure, histology, degree of substernal goiter, postoperative haemorrhage and hypoparathyroidism. Fisher’s exact test was used to test infection and recurrent laryngeal nerve injury.
2.1. Surgical procedure Following a Kocher incision, the skin flaps were constructed. The strap muscles were divided in the midline and retracted laterally. Capsular dissection was conducted. In the harmonic group, the inferior, middle and superior vessels were divided with harmonic scalpel (FOCUS shear, Ethicon Endo-Surgery). In the conventional group, all vessels were ligated with suture ties, clips and electrocautery. In both groups, the recurrent laryngeal nerves and parathyroid glands were identified and protected, and the thoracic component of the thyroid was retracted manually to the cervical region. If sternotomy was necessary, a thoracic surgeon was called. 3. Results During the study period a total of 2258 patients underwent thyroidectomy. Of these, 242 patients (11%) were operated for substernal goiter. 3.1. Substernal goiter versus cervical goiter Patients operated for substernal goiter were older compared to patients operated for cervical goiter (59 vs. 48 years, p = 0.001) and more often male (29% vs. 18%, p = 0.001). Furthermore, the incidence of total thyroidectomy was higher (35% vs. 25%, p = 0.04), the mean size of the goiter was over three times larger (140 g vs. 30 g, p = 0.001), the mean operative time was significantly longer and bleeding during surgery and postoperative haemorrhage was significantly larger in substernal goiters (Table 1). 3.2. Harmonic versus conventional haemostasis in substernal goiter A total of 121 patients (50%) underwent harmonic thyroidectomy and 121 patients (50%) underwent conventional thyroidectomy (Table 2). No patients needed a sternotomy. Most patients (76%) had a goiter reaching 1–4 cm substernally. Twenty-two percent had a goiter 5–9 cm substernally, and only 3 patients (1%) had goiter extending more than 10 cm substernally. No significant
Table 1 Substernal vs. cervical goiter. Characteristics Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Histology Malign, n (%) Benign, n (%) Mean weight, g (SD) Mean operative time, min (SD) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR) Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%) IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
Cervical goiter n = 2016
Substernal goiter n = 242
363 (18) 1653 (82) 48 (14)
70 (29) 172 (71) 59 (14)
1518 (75) 498 (25)
158 (65) 84 (35)
298 1718 30 92 103 50 34 37 72 2
28 214 140 112 20 75 7 9 12 2
p-value 0.001
0.001 0.05
0.2 (15) (85) (105) (42) (5) (70) (1.7) (1.4) (2.8) (1)
n = 498 4.1 (3.4) 23 (4.6)
(12) (88) (121) (42) (8) (100) (3) (2.7) (3.6) (1)
n = 84 3.8 (2) 4 (4.9)
0.001 0.001 0.03 0.001 0.2 0.08 0.08 0.001
0.3 0.7
C.H. Hahn et al. / Auris Nasus Larynx 42 (2015) 311–317
313
Table 2 Substernal goiter. Harmonic vs conventional. Characteristics Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Histology Malign, n (%) Benign, n (%) Mean weight, g (SD) Degree of substernal goiter 1–4 cm, n (%) 5–9 cm, n (%) >10 cm, n (%) Mean operative time, min (SD) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR) Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%)
Harmonic n = 121
Conventional n = 121
Total n = 242
p-value 0.4
32 (26) 89 (74) 58 (14)
38 (31) 83 (69) 60 (15)
70 (29) 172 (71) 59 (14)
79 (65) 42 (35)
79 (65) 42 (35)
158 (65) 84 (35)
7 (6) 114 (94) 132 (105)
21 (17) 100 (83) 148 (135)
28 (12) 214 (88) 140 (121)
93 26 2 117 5 50 3 3 6 2
92 28 1 106 15 100 4 6 6 3
185 54 3 112 20 75 7 9 12 2
0.3 1
0.008
(77) (22) (2) (42) (4) (70) (2.5) (1.8) (3.6) (1)
n = 42 3.7 (1.7) 1 (2.6)
(76) (23) (1) (42) (12) (150) (3.3) (3.6) (3.6) (2)
n = 42 3.9 (2.3) 3 (7.1)
(76) (22) (1) (42) (8) (100) (3) (2.7) (3.6) (1)
n = 84 3.8 (2) 4 (4.9)
0.3 0.8
0.04 0.03 0.001 0.7 0.6 1 0.001
0.6 0.3
IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
differences were seen between the study groups with respect to age, gender, surgical procedure, weight or degree of substernal goiter (Table 2). A total of 28 patients (12%) had malignant diagnosis. In the conventional group significantly more had a malignant diagnosis compared to the harmonic group (p = 0.005). The mean operative time was significantly shorter in the conventional group (106 min vs. 117 min, p = 0.04). The use of harmonic scalpel was associated with a significant reduction in intraoperative blood loss (50 mL vs. 100 mL, p < 0.001) and postoperative haemorrhage (4% vs. 12%, p = 0.03). The median length of hospital stay was shorter in the harmonic group (3 days
vs. 2 days, p = 0.001). There was a non-significant trend towards less permanent recurrent laryngeal nerve palsy in the harmonic group (3 vs. 6 patients, p = 0.6) and fewer patients requiring permanent calcium consumption (1 vs. 3 patients, p = 0.3). No significant differences were observed between the two groups in relation to infection, postoperative parathyroid hormone level or transient recurrent laryngeal nerve injury (Table 2). Table 3 shows subgroup analysis for thyroidectomy of benign substernal goiter. In this subgroup the use of harmonic scalpel was associated with significant reduction in intraoperative blood loss (50 vs. 100 mL, p = 0.001), postoperative haemorrhage (4 vs. 14%,
Table 3 Substernal benign goiter. Harmonic vs. conventional haemostasis. Characteristics Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Mean weight, g (SD) Degree of substernal goiter 1–4 cm, n (%) 5–9 cm, n (%) >10 cm, n (%) Median operative time, min (IQR) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR) Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%) IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
Harmonic n = 114
Conventional n = 100
Total n = 214
p-value 0.5
30 (26) 84 (74) 58 (14)
31 (31) 69 (69) 59 (15)
74 (65) 40 (35) 134 (106)
66 (66) 34 (34) 153 (140)
140 (65) 74 (35) 144 (121)
87 25 2 116 5 50 3 3 6 2
76 23 1 104 14 100 2 4 4 3
163 48 3 110 19 75 5 7 10 2
(77) (22) (2) (39) (4.4) (70) (2.6) (1.9) (3.8) (1)
n = 40 4.0 (2.5) 1 (2.5)
(76) (23) (1) (37) (14) (150) (2) (3.0) (3.0) (2)
n = 34 3.8 (1.7) 2 (5.9)
61 (29) 153 (71) 59 (14)
(76) (22) (1) (38) (9) (100) (2.3) (2.4) (3.5) (1)
n = 74 3.8 (2) 3 (4)
0.7 0.9
0.3 0.8
0.03 0.02 0.001 1 0.7 0.8 0.02
0.6 0.6
314
C.H. Hahn et al. / Auris Nasus Larynx 42 (2015) 311–317
p = 0.02) and length of hospital stay (2 vs. 3 days, p = 0.02). The operative time was longer in the harmonic group (116 vs. 104 min, p = 0.03). 3.3. Difference by surgeons in substernal goiter Table 4 illustrates differences in operative characteristics between the surgeons. No significant differences were observed in relation to gender, age, surgical procedure, histology and weight. The mean operative time was significantly different between the four surgeons (p = 0.001). Post hoc multiple comparisons (Bonferroni) showed that Surgeon 1 was significantly faster than Surgeon 2 and Surgeon 4. No differences were observed between the four surgeons in relation to bleeding, infections, hospital stay or recurrent laryngeal nerve injury. Table 5 shows data from Table 4 disaggregated for each surgeon according to surgical method. For Surgeon 2 the use of harmonic scalpel was associated with a significant reduction in intraoperative blood loss (75 mL vs. 200 mL, p < 0.001), postoperative haemorrhage (4% vs. 23%, p = 0.006), wound infections (1% vs.10%, p = 0.04) and median length of hospital (3.5 days vs. 2 days, p < 0.001). No significant differences were observed between the harmonic group and the conventional group for Surgeons 3 and 4. 4. Discussion This is the first study comparing harmonic scalpel with the conventional haemostasis method in surgery for substernal goiter. Our results suggest that the use of harmonic scalpel results in significantly less intraoperative bleeding, postoperative haemorrhage and shorter hospital stay. Surprisingly, operative time was longer. There is no standard definition of substernal goiter. Some studies only include cases with 50% or more of the gland situated in the thorax, whereas other studies include patients with any part of the gland extending through the thoracic inlet. The reported incidence of substernal goiter is highly variable, ranging from 3 to 30% [3,6,15]. In our study the incidence of substernal goiter was 11%. The substernal goiter was larger, and the patients were older.
Mean operative time was longer, and bleeding during surgery was larger. This resulted in more postoperative bleeding and a tendency to more complications. This is in accordance with most studies in substernal goiters, where surgery is associated with increased complication rates [6,12–14]. A few studies conclude that the risk of complications is not increased [15]. Previous randomised prospective studies reported the safety and usefulness of harmonic focus in thyroidectomy [2,16–22]. A meta-analysis of randomised clinical trials comparing harmonic scalpel with conventional techniques showed a significant reduction in operative time equivalent to 23 min (nearly 25% of the operating room time). The reduction in operative time ranges from 15% to 30%, and is more apparent when the surgical procedure is long and complex [18,23]. The reduced surgical time is due to a decrease in intraoperative instrument exchange by 70% [18]. One prospective nonrandomized study on lobectomy showed no benefit in terms of operative time [23]. Our study demonstrated a significant increase in operative time from 106 min with the conventional technique to 117 min with harmonic focus, corresponding to an increase of 9%. Different length of operative time between the four surgeons may explain this finding. Surgeon 1, who used the conventional method in the whole study period, was significantly faster than Surgeon 2 and Surgeon 4, who used harmonic focus in the last part of the study period. Surgeons 2–4 had a mean operative time of 131 min with the conventional method and 117 min with the harmonic focus (p = 0.1). This corresponds to a 12% decrease in operative time. Unfortunately, due to our small study population, we may have failed to demonstrate a possible significant difference in operative time in our disaggregated analysis. However, the operative time of surgeon 2 was 15% shorter with the harmonic scalpel (p = 0.06). Bleeding in the neck is a potentially life-threatening complication. The thyroid is a highly vascular organ; thus, haemostasis is very important. In our study, intraoperative bleeding was significantly reduced from 100 mL in the conventional group to 50 mL in the harmonic group. This 50% reduction is consistent with all other randomised studies concerning harmonic scalpel, showing a relative reduction in intraoperative bleeding between 13 and 61% [24,25]. A meta-analysis demonstrated a 20 mL
Table 4 Difference by the surgeons. Characteristics
Surgeon 1 (n = 65)
Surgeon 2 (n = 99)
Surgeon 3 (n = 38)
Surgeon 4 (n = 40)
Total (n = 242)
P-value
20 (31) 45 (69) 61 (15)
31 (32) 68 (68) 62 (12)
12 (32) 26 (68) 56 (14)
12 (30) 28 (70) 60 (15)
70 (29) 172 (71) 59 (14)
1
39 (60) 26 (40)
52 (53) 47 (47)
28 (73) 10 (27)
25 (63) 15 (37)
158 (65) 84 (35)
Gender Male, n (%) Female, n (%) Mean age, years (SD) Surgical procedure Lobectomy, n (%) Total thyreoidectomy, n (%) Histology Malign, n (%) Benign, n (%) Mean weight, g (SD) Mean operative time, min (SD) Post-operative haemorrhage, n (%) Median bleeding during surgery, ml (IQR) Infection, n (%) Permanent recurrent laryngeal nerve injury, n (% NAR) Temporary recurrent laryngeal nerve injury, n (% NAR) Median hospital stay, days (IQR)
10 55 142 90 6 80 1 6 0 2
Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%)
n = 26 3.7 (2.6) 2 (8)
0.3 0.3
0.7
IQR, Interquartile range; SD, Standard deviation; NAR, nerve at risk.
(15) (85) (103) (30) (10) (100) (2) (9) (2)
7 93 162 123 10 100 4 2 11 3
(7) (93) (137) (45) (10) (150) (4) (2) (11) (2)
n = 47 3.8 (1.9) 1 (2)
8 30 147 111 2 80 1 0 0 3
(21) (79) (123) (55) (5) (178) (3)
(2)
n = 10 4.3 (3) 1 (10)
3 37 141 123 2 50 1 1 1 2
(7%) (93%) (117) (37) (5) (79) (3) (2.5) (2.5) (1)
n = 15 3.9 (2) 0
28 (12) 214 (88) 150 (121) 112 (42) 20 (8) 75(100) 7 (3) 9 (2.7) 12 (3.6) 2 (1)
0.7 0.001 0.7 0.1 0.9 0.1 1 0.2
n = 84 3.8 (2) 4 (4.9)
0.6 0.5
Table 5 Harmonic vs conventional method for each surgeon. Characteristics
Total thyroidectomy Mean parathyroidea hormone, pmol/l (SD) Hypoparathyroidism requiring Calcium and D-vit, n (%)
Surgeon 2
Surgeon 3
Conventional
Harmonic
Conventional
Harmonic
n = 65
n=0
n = 31
n = 68
20 (31) 45 (69) 61 (15)
0 0 0
7 (23) 24 (77) 62 (14)
24 (35) 44 (65) 62 (12)
39 (60) 26 (40)
0 0
24 (77) 7 (23)
28 (41) 40 (59)
0 0 0 0 0 0 0 0 0 0
6 (19) 25 (81) 183 (180) 137 (47) 7 (23) 200 (125) 3 (10) 0 6 (19) 3.5 (3)
1 67 151 116 3 75 1 2 5 2
n = 26 0 0
n=0 4 (2) 0
n = 10 3.8 (1.9) 1
10 55 142 90 6 80 1 6 0 2
(15) (85) (103) (30) (10) (100) (2) (9) (2)
3.7 (2.6) 2 (8)
(1) (99) (110) (43) (4) (61) (1) (3) (7) (1)
p-value
0.8 0.2
0.3 0.06 0.006
