- Email: [email protected]
Neurovascular complications following carotid body paraganglioma resection
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Original article
Neurovascular complications following carotid body paraganglioma resection E. Lamblin a,b,∗ , I. Atallah a,b,c , E. Reyt a,b , S. Schmerber a,b , J.-L. Magne b,d , C.A. Righini a,b,c a Unité d’oto-rhino-laryngologie et chirurgie cervico-faciale, centre hospitalo-universitaire de Grenoble, 1, avenue du Maquis-du-Grésivaudan, 38043 Grenoble cedex 9, France b Université Joseph-Fourier, Grenoble I, BP 53, 38041 Grenoble cedex 9, France c Inserm U823, institut Albert-Bonniot, BP 170, 38042 Grenoble cedex 9, France d Unité de chirurgie endocrinienne, thoracique et vasculaire, centre hospitalo-universitaire de Grenoble, 1, avenue du Maquis-du-Grésivaudan, 38043 Grenoble cedex 9, France
a r t i c l e
i n f o
Keywords: Carotid body paraganglioma Surgery Neurovascular complications Shamblin classification
a b s t r a c t Objective: The present consecutive case series reports our experience in the management of carotid body paraganglioma and aims to assess whether the Shamblin classification or tumor size are predictive of early and late postoperative neurovascular complications. Material and methods: A retrospective study included 54 carotid body tumor resections in 49 patients, between 1980 and 2011. Data comprised early (< 1 month) and late (18 months) postoperative neurovascular complications. Results: Early postoperative complications occurred in 31 cases, including 30 cases of cranial nerve deficit (56%). Cranial nerve deficit occurred in 83% of Shamblin III carotid body paragangliomas and was associated with significantly larger mean tumor size (4 ± 1.4 cm versus 2.9 ± 1.3 cm; P < 0.01). Shamblin III tumor and tumor size > 3.2 cm emerged as predictive factors for early postoperative peripheral neurological complications. Eight patients (17%) showed no cranial nerve deficit recovery, even after 18 months’ follow-up; no predictive factors could be identified for this. Conclusion: Surgical resection remains the only curative treatment in carotid body paraganglioma, with low vascular morbidity. However, early postoperative nerve deficit remains frequent (56%), although mostly temporary, with 17% definitive sequelae at 18 months. Tumor size and Shamblin classification are predictive of early neurovascular complications. © 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction Paraganglioma is rare and usually benign, but certain genetic mutations (SDHB gene) may be associated with malignancy. It develops from diffuse neuroendocrine system cells between the skull base and pelvic floor. Cervical paraganglioma accounts for 0.6% of head and neck tumors [1]. Carotid body paraganglioma (CBP) is the most frequent parasympathetic form, with incidence of 1/30,000 to 1/100,000 in the general population. Surgical resection is the sole curative treatment in CBP. Growth is slow, but with risk of local compression and a slight risk of associated malignancy [2,3]. Progress in genetic screening, imaging
∗ Corresponding author. Unité d’oto-rhino-laryngologie et chirurgie cervicofaciale, centre hospitalo-universitaire de Grenoble, 1, avenue du Maquis-duGrésivaudan, 38043, Grenoble cedex 9, France. E-mail address: [email protected] (E. Lamblin).
(such as 3D time-of-flight MRI angiography) [4] and surgery have significantly reduced perioperative mortality (< 1%) and vascular morbidity (stroke: < 1%). Postoperative peripheral neurological morbidity, mainly involving the cranial nerves, however, is high, with a 14–49% rate of early deficit [2,3] and 6–23% persistent deficit, reported only in some recent studies with median follow-up of 6 months to 1 year [5–7]. Progress in alternative treatment strategies, such as adjuvant or exclusive, conventional, stereotactic or intensity-modulated radiation therapy, however, allows new attitudes in certain cases [8,9]. The Shamblin classification (Fig. 1), based on carotid vessel encasement, correlates with risk of postoperative vascular lesion [10,11]. Previous studies, however, disagreed on the relation between Shamblin group and onset of postoperative peripheral neurological complications [5–7,11,12]. The present study reports our experience in CBP, and assessed the risk of onset of early and late postoperative neurovascular complications according to Shamblin group and tumor size.
http://dx.doi.org/10.1016/j.anorl.2016.05.006 1879-7296/© 2016 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
2
2. Patients and methods A consecutive case series included all patients operated on for CBP in our university hospital center between January 1980 and December 2011. Demographic, clinical and paraclinical and surgical data, early (< 1 month) and late (> 18 months) postoperative neurovascular complications and disease progression were recorded. Tumors were graded on the Shamblin classification based on preoperative CT or MRI. Tumor size was defined by the long axis on pathologic examination. Resection was performed by a double ENT and vascular surgery team. Sufficiently large cervicotomy allowed good control of neurovascular structures. Carotid axis and the proximal and distal parts of the carotid arteries were first determined, then the vagus, spinal and hypoglossal nerves were located. Craniocaudal subadventitial dissection was performed up to the carotid sinus, which is the area at greatest risk of bifurcation tearing due to the presence of the carotid body artery [13]. An associated vascular procedure (simple suture, patch closure, carotid anastomosis resection or commonto-internal carotid bypass) was performed under effective heparin therapy and hemodynamic control by the vascular surgery team if arterial continuity was interrupted; this was either pre-planned, in case of large tumor or tumor adhering to the internal carotid artery (ICA), requiring carotid resection, or else in case of lesion during dissection. Lymph tissue adjacent to the CBP was systematically dissected after the year 2000. Patient characteristics were reported as median and interquartile range (IQR: 25th and 75th percentiles) for continuous variables and as percentages for categoric variables. Univariate analysis assessed whether Shamblin group or tumor size correlated with postoperative neurovascular complications risk. Categoric variables were analyzed on Chi2 or Fisher exact test, and continuous variables on Kruskal-Wallis or Mann-Whitney-Wilcoxon test. ROC curves were plotted to determine whether a threshold tumor size
was predictive of postoperative peripheral neurologic complications. The significance threshold was set at P < 0.05. All analyses were conducted on R software version 3.0.1. 3. Results 3.1. Description of population Baseline demographic and clinical and paraclinical characteristics are shown in Table 1. Fifty-four CBP resections were performed in 49 patients. In 36 cases, there was lateral cervical swelling for a median 38 months before medical consultation (IQR, 1–48 months). Median follow-up was 17 months (IQR, 10–67 months). Forty-three patients (48 CBPs) had a minimum 18 months’ follow-up. In 98% of cases, diagnosis was founded on arteriography or angio-CT; ultrasonography or fine-needle aspiration were the primary examination in 12 patients (10 before 1992, most referred from other centers), who then underwent exploratory cervicotomy, without biopsy, and without complications, before CBP resection following confirmation of diagnosis on complete imaging assessment. Most patients with multiple forms had bilateral CBP (9 out of 12); other associated locations comprised: 3 jugulotympanic, 1 vagal, 1 mediastinal and 1 pancreatic paragangliomas and 1 pheochromocytoma (adrenal). Six patients with bilateral CBP underwent bilateral resection, with the smaller tumor generally treated in the first step. In some cases, other treatments were proposed for the contralateral CBP: radiation therapy (due to ipsilateral vagus nerve lesion during the first procedure), or medical surveillance. Genetic assessment was systematically offered to patients and family, with screening for germline mutations in succinate dehydrogenase subunits D, B and C (SDHD, SDHB, SDHC). More recently, screening included other mutations: in succinate dehydrogenase
Fig. 1. The Shamblin classification of carotid body tumor enables the difficulty of resection to be predicted. Group I CBP is very localized and easy to resect. Group II adheres to or partially encases the arteries. Group III completely encases at least one artery. ECA: external carotid artery; ICA: internal carotid artery; X: vagal nerve; XII: hypoglossal nerve.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model
ARTICLE IN PRESS
ANORL-565; No. of Pages 6
E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
3
Table 1 Baseline demographic, clinical and paraclinical data. Variable Demographics CBP resection Age, yearsa Female gender Right CBP Multiple CBP Bilateral CBP Malignant CBP Follow-up, monthsa Mode of discovery Duration of symptoms before diagnosis, monthsa Neck mass Lesion staging Neurovascular work-up Flush Genetic assessment Associated symptoms Hypertension Malaise Tinnitus Cough and dysphonia Related to preoperative vagus nerve palsy Complementary examinations Genetic studies Germline mutation SDHD gene SDHB gene Catecholamine urine or blood assay Significant elevation Significant elevation related to a functional CBP Arteriography or CT-angiography CT MRI Scintigraphy
Number of patient (n = 49) 54 49 (38.2–57.7) 27 (55) 32 (59) 12 (24) 9 (83) 2 (4) 17 (10–67) 38 (1–48) 36 (67) 7 (13) 7 (13) 2 (4) 1 (2) 18 (37) 10 (20) 2 (4) 12 (25) 3 (25) 14 (29) 10 (71) 7 (70) 3 (30) 38 (78) 7 (18) 4 (11) 48 (98) 32 (65) 16 (33) 16 (33)
Values in brackets are percentages unless indicated otherwise; CBP: carotid body paraganglioma; SDHD-B: succinate dehydrogenase subunit D-B; CT: computed tomography; MRI: magnetic resonance imaging. a Values are medians (interquartile range).
subunit A and Von Hippel Lindau genes (SDHA, VHL). Ten patients showed mutations, mainly on the SDHD gene; 8 (7 with SDHD mutation) had multiple CBP. Four patients showed significant catecholamine elevation, related to secreting CBP. All specimens underwent histologic examination, systematically confirming diagnosis. Two patients had incidentally discovered metastatic adenopathies on pathologic examination of adjacent lymph nodes removed along with the tumor. Resection was incomplete in 4 patients with group-III CBP, due either to extension around skull-base vessels (2 patients with > 6 cm tumor) or (in 2 patients) to positive margins on pathologic examination. Tumor size was significantly greater in higher Shamblin grades (P < 0.01). 3.2. Surgery All patients underwent surgical resection. Preoperative embolization was performed in 2 tumors exceeding 4 cm, treated before 1992. Median operative time was 135 minutes (IQR, 120–180 minutes), and was significantly longer in case of higher Shamblin grade and greater tumor size (Fig. 2). There were 20 cases of intraoperative complications (37%). Seven patients (13%) suffered nerve sacrifice: 3 vagus, 3 superior laryngeal and 1 hypoglossal. Vascular surgery was performed in 19 cases (35%): 18 external carotid ligations, and 13 internal carotid procedures (9 bypasses, 2 venous patches, 1 ligation and 1 direct suture). Shamblin group and tumor size correlated with risk of
Fig. 2. Correlation of (a) Shamblin group and (b) tumor size with operative time, a: box-plot of surgery time according to Shamblin group; b: surgery time according to tumor size. Median operative time correlated with Shamblin group (Kruskal-Wallis test: *, P < 0.05) and tumor size (Pearson correlation test: r = 0.50; P < 0.001).
intraoperative complications, including nerve sacrifice and vascular surgery (P < 0.05, Tables 2 and 3).
3.3. Early (< 1 month) postoperative complications Postoperative neurovascular complications comprised: 3 cases of stroke (1 cerebral, after ICA ligation, and 2 transient ischemic attacks), and 30 of peripheral neurologic deficit (56%). Peripheral neurologic involvement concerned several nerves in 18 cases, with total deficit in 11. The nerves most often involved were the vagus (37% of cases of deficit) and hypoglossal (32%), followed by cervical sympathetic (13%), glossopharyngeal (10%) and superior laryngeal (6%). Non-neurological early postoperative complications were local (1 abscess, and 2 hematomas requiring surgical revision), vascular (1 bypass thrombosis, requiring revascularization) or general (1 case of pulmonary embolism, 1 of gastrointestinal hemorrhage, 1 of pneumonia and 2 of non-controlled hypertension). There were no early postoperative deaths. Incidence of early postoperative complications overall was significantly greater with higher Shamblin grade or greater tumor size, except for central
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model
ARTICLE IN PRESS
ANORL-565; No. of Pages 6
E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
4
Table 2 Neurovascular complications according to Shamblin group. Variable
Shamblin I (n = 17)
Shamblin II (n = 25)
Shamblin III (n = 12)
P value
Intraoperative complications Cranial nerve adherence Cranial nerve resection Vascular procedure Internal carotid repair Early postoperative complications Peripheral neurologic deficit Central neurologic deficit Non-neurologic complication Late postoperative complications (> 18 months) Permanent cranial nerve deficit
1 (6) 3 (18) 1 (6) 0 (0) 0 (0) 5 (29) 5 (29) 0 (0) 0 (0) n = 17 1 (6)
8 (32) 7 (28) 1 (4) 9 (36) 4 (13) 16 (64) 15 (60) 0 (0) 2 (8) n = 22 4 (18)
11 (92) 6 (50) 5 (42) 10 (83) 9 (75) 10 (83) 10 (83) 3 (25) 5 (42) n=9 3 (33)
< 0.001a 0.201a 0.008a < 0.001a < 0.001a 0.003b 0.013b 0.008a 0.003a 0.179a
Values in brackets are percentages unless indicated otherwise. a Fisher test. b 2 test.
neurologic complications, which correlated only with Shamblin group (Tables 2 and 3). Incidence of peripheral neurologic deficit was compared according to tumor size, with a view to determining a threshold size predictive of early peripheral neurologic deficit. Sensitivity (63%; 95% CI [57–70]) and specificity (75%; 95% CI [69–81]) were optimal for a threshold of ≥ 3.2 cm (area under the curve = 0.74), with a positive predictive value of 76% (95% CI [70%; 82%]).
3.4. Late (> 18 months) postoperative complications Follow-up exceeded 18 months in 43 patients (48 CBPs), with 5 patients initially showing deficit being lost to follow-up. Speech therapy was systematically prescribed in case of cranial nerve involvement. Peripheral neurologic deficit recovered completely in 17 cases (68% of patients with early deficit). Median time to recovery was 6 months (IQR, 2–12 months). Finally, 8 cases (17% of those followed up for ≥ 18 months) showed permanent peripheral neurologic deficit, with persistent symptoms: 5 with swallowing disorder and 3 with dysphonia. Univariate analysis found no predictive factor for permanent peripheral neurologic deficit (Tables 2 and 3).
3.5. Disease progression One patient died 10 months postoperatively, after stroke following ICA ligation: he had had group-III CBP of 7 cm; as there was no intraoperative skull-base arterial reflux, carotid bypass had not been performed. One patient showed local progression of residual tumor 1 year postoperatively. Resection had been incomplete, without complementary treatment. Resection was repeated, with no signs of recurrence at 6 years’ subsequent follow-up. Two patients, treated before 1992, showed local recurrence associated with remote tumor. The first showed local recurrence associated with mediastinal tumor 1 year after resection of groupIII bilateral CBP with incidentally discovered metastatic cervical adenopathies on pathologic examination of lymph nodes adjacent to the tumor. The cervical recurrence was managed by radiation therapy (due to contralateral vagus deficit), and the mediastinal paraganglioma was resected, without recurrence at 10 months’ follow-up. The second patient also showed local recurrence, but associated with multiple cervical locations, 27 years after primary surgery; there had been positive margins and metastatic cervical adenopathies, but no complementary treatment in view of advanced age and the multiplicity of the cervical locations.
Table 3 Neurovascular complications according to tumor size. Variable (n = 56)
No complications, n Median tumor size, cm (IQR)
Complications, n Median tumor size, cm (IQR)
P valuea
Intraoperative complications
34 (63) 3 (2–4) 38 (70) 3 (2–4) 47 (87) 3.3 (2.9, 3.7) 35 (65) 3 (2–4) 41 (76) 3 (2–4) 24 (44) 2.5 (2–3.2) 24 (44) 2.75 (2–3.25) 51 (98) 3 (2.2–4) 47 (87) 3 (2–4)
20 (37) 4 (3.1–6) 16 (30) 3.3 (3–5) 7 (23) 4.8 (3, 6.6) 19 (35) 4 (3.3–5.5) 13 (24) 4 (3.2–5) 30 (56) 4 (3–5) 30 (56) 4 (3–5) 3 (2) 5 (4–6) 7 (23) 6 (3.5–6)
< 0.001
40 (83) 3 (2–4)
8 (17) 3.25 (3–4.25)
Cranial nerve adherence Cranial nerve resection Vascular procedure Internal carotid repair Early postoperative complications (< 1 month) Peripheral neurologic deficit Central neurologic deficit Non-neurologic complication Postoperative complications (> 18 months) (n = 48) Permanent cranial nerve deficit
0.153 0.04 0.001 0.03 0.006 0.007 0.13 0.001
0.3
Values in brackets are percentages unless indicated otherwise. a Mann-Whitney-Wilcoxon test.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
One patient developed bone and liver metastases 4 years after resection of bilateral dopamine-secreting CBP. Clinical course rapidly deteriorated despite radiation therapy, and the patient died some weeks later.
4. Discussion Clinical assessment of any lateral neck mass should be meticulous, to avoid missing a rare differential diagnosis such as CBP. In more than two thirds of patients with CBP, diagnosis is based on discovery of a pulsatile lateral neck mass, usually with restricted vertical mobility [14] and often silently present for several years. Doppler ultrasound lacks sensitivity for diagnostic work-up of CBP. MRI (especially 3D time-of-flight MR angiography [4]) allows definitive diagnosis when angio-CT data are insufficient. However, the gold standard to detect small CBPs is digital subtraction angiography [4], because it reveals the specific vascular supply of the paraganglioma by the ascending pharyngeal artery. Fine-needle aspiration or exploratory cervicotomy are no longer indicated after well-conducted clinical and paraclinical assessment. The contribution of preoperative embolization is controversial [5,9,15–17]. We do not usually perform preoperative embolization (only 2 patients before 1992, for > 4 cm tumor, for which embolization was only partial), although it has recently been reported not to be associated with increased risk of stroke [17]. Most authors report little benefit in postoperative course, with no evidence of a decrease in morbidity in general or in cranial nerve deficit in particular [15]. It may, however, according to some authors, facilitate the resection of large, very adherent tumors, especially when located near the skull base, as it reduces clamping time and blood loss [5,16]. According to recent studies, 30% of CBPs are hereditary, mainly involving SDHD gene mutation [5,18]. Genetic assessment should be systematic, due to the high prevalence (6–25%) of mutations in patients presenting with apparently sporadic forms [8]. Several authors [19,20] reported that multifocal CBT is frequently associated with SDHD mutation (71% in the present study). A recent meta-analysis showed that the risk of malignant CBP was higher in case of SDHB gene mutation [20]. Hence systematic genetic assessment enables CBP to be detected at an earlier stage, and facilitates management of multifocal and malignant forms. In bilateral CBP, the smaller lesion is generally resected first and, if there are no postoperative complications, the second can be treated subsequently. In multiple forms, functional tumors (pheochromocytoma) are treated first. Two patients had cervical lymph node metastases and 1 had remote metastases. As there are no specific histopathological features of malignancy, diagnosis may be confirmed only when there is metastasis to non-neuroendocrine tissue [20,21]. Cervical metastasis was diagnosed incidentally in both cases of adjacent lymph nodes removed along with the tumor. We therefore now resect peritumoral lymph nodes to identify possible occult metastasis in all patients with apparently benign CBP [8,21]. Boedeker [8] recommended selective neck dissection for levels II and III, but dissection extension remains to be defined. Our results were consistent with those of previous studies, with malignancy around 6%, with less remote metastasis than in other head and neck paraganglioma sites [21,22]. The authors of two large studies of malignant head and neck paraganglioma reported significantly better 5-year overall survival (76.8–82.4%) in patients with regional than distant metastasis (11.8–41.4%). In case of metastasis, there is no evidence that postoperative radiotherapy improves the 5-year overall survival. In case of incomplete resection, however, adjuvant radiation therapy may slow tumor growth and improve survival [21,22]. Three patients showed local progression of residual tumor or recurrence, 1 and 27 years after surgery. Elshaikh et al. [23] likewise
5
reported long median time to recurrence of head and neck paraganglioma (36 months; range, 15–350 months), and recommended long follow-up to detect late recurrence. Their results confirmed that surgery provided control of local recurrence, but few studies assessed the role of radiation therapy in control of CBP recurrence [23]. Complete surgical excision is the aim of any operative procedure and is the only curative treatment. Nevertheless, radiation therapy may be considered for patients in whom surgery is contraindicated by advanced age, comorbidity or aggressive or unresectable tumor [8,24]. The goal of radiotherapy is to control the disease or inhibit growth rather than to eliminate the tumor. A recent review of CBP management [24] reported no significant difference in tumor control or mortality between surgery and radiotherapy, but a higher risk of major complications and cranial nerve palsy following surgery. Radiation side effects should also be taken into account in deciding treatment, with a 1% risk of radiation-induced malignancy at 10 years, and more difficult surgery with higher risk of neurovascular complication. However, some adverse effects have been reduced by new techniques such as intensity-modulated radiation therapy, making surgery no longer the obvious gold standard in CBP. Tumor doubling time is very slow, and conservative treatment, such as a clinical and radiological wait-and-see attitude, may be envisaged for elderly patients with small tumor size. The most important criterion in the Shamblin classification [10] is carotid wall encasement, which is related to tumor size [14]. However, the relationship has been reported to be non-linear in some studies [12]. Shamblin group was significantly associated with tumor size in the present study. The Shamblin classification was initially based on intraoperative and pathological findings [10]; the present study took account of preoperative radiological data, but Shamblin groups were nevertheless distributed as usually reported: one third in group I, almost a half in group II, and a quarter in group III [5,6]. Other radiologic criteria may also reflect artery encasement: ICA encasement angle, or radiolucent line between tumor and ICA [12]. We were not able to analyze these features due to lack of data in preoperative radiographs. Tumor size was assessed according to the longest axis in pathology reports; the assessment of tumor volume on imaging could provide a useful objective preoperative criterion [5,18]. Twenty-four percent of cases required ICA surgery, in agreement with the literature, reporting rates of 6–52% [9,25]. The present data included all types of vascular procedure, even direct suture and venous patch, which may account for the discrepancy with recent reports, which did not always detail the type of vascular procedure [12]. Vascular surgery risk has decreased over the past few decades, especially for ICA ligation [2]. Consequently, the risk of postoperative stroke is now very low: only one patient in the present study, in 1991. Seventeen percent of patients suffered stroke in an older large series published in 1988 [2]. Furthermore, vascular repair has become safer with improved surgical techniques: early control of proximal and distal carotid vessels, subadventitial tumor dissection, effective heparin therapy, precise hemodynamic control and reduced clamping time have all reduced postoperative risk. In the present study, higher Shamblin grade and larger tumor size correlated with higher risk of intraoperative vascular lesion; thus both allowed prediction of risk [7]. Peripheral neurological morbidity has remained high and unchanged for 50 years [2] despite improved surgical techniques. Early peripheral neurological deficit occurred in 56% of the present series, and was shown to correlate with both Shamblin group and tumor size, both of which could thus be used to predict the risk of early neurologic deficit [5]. Some authors suggested a 4 cm cutoff to predict increased risk of neurological morbidity [1,13] the present data, however, show that ≥ 3.2 cm was the best threshold on ROC analysis. Since the larger the tumor the higher the risk
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
6
of neurologic deficit, systematic genetic assessment should enable earlier diagnosis of small CBP [9]. In 68% of cases, peripheral neurologic deficit was transient, with permanent deficit at 18 months in 17%. Early appropriate speech therapy is essential. No cases of deficit at 18 months subsequently recovered. This finding is difficult to compare with the literature, as authors do not systematically report recovery time or else limit follow-up to 12 months [5]. We were unable to find any predictive factors for permanent neurologic deficit. Power et al. [5] reported that permanent cranial nerve deficit at 1 year correlated with Shamblin group. Further studies should take 18 months as the cut-off value for permanent neurologic deficit. 5. Conclusion Surgical resection remains the only curative treatment in CBP. Adjacent lymph node dissection is recommended. Early peripheral neurologic deficit is frequent (56%); only 17% fail to recover within 18 months. Tumor size and Shamblin group correlate with risk of early neurovascular complications. Genetic assessment should be systematic to enable early detection of CBP and improve outcome in multifocal and malignant forms. Disclosure of interest The authors declare that they have no competing interest. Acknowledgments The authors warmly thank Dr B. Boussat of the Clinical Investigation Center of the Grenoble University Hospital for the statistical analysis, and Mr P.-E. Colle for re-editing the English-language version. References [1] Maves MD. Vascular tumors of the head and neck. Head and neckotolaryngology. In: Bailey BJ, Jonhnson JT, Kohut RI, Pillsbury HC, editors. Tardy medical management. Philadelphia: JB Lippincott; 1993. p. 1397–409. [2] Hallett JW, Nora JD, Hollier LH, Cherry KJ, Pairolero PC. Trends in neurovascular complications of surgical management for carotid body and cervical paragangliomas: a fifty-year experience with 153 tumors. J Vasc Surg 1988;7(2):284–91. [3] Sajid MS, Hamilton G, Baker DM. A multicenter review of carotid body tumour management. Eur J Vasc Endovasc Surg 2007;34(2):127–30.
[4] Berg R. Imaging and management of head and neck paragangliomas. Eur Radiol 2005;15(7):1310–8. [5] Power AH, Bower TC, Kasperbauer J, et al. Impact of preoperative embolization on outcomes of carotid body tumor resections. J Vasc Surg 2012;56(4):979–89. [6] Makeieff M, Raingeard I, Alric P, Bonafe A, Guerrier B, Marty-Ane C. Surgical management of carotid body tumors. Ann Surg Oncol 2008;15(8):2180–6. [7] Lim JY, Kim J, Kim SH, et al. Surgical treatment of carotid body paragangliomas: outcomes and complications according to the Shamblin classification. Clin Exp Otorhinolaryngol 2010;3(2):91–5. [8] Boedeker CC. Paragangliomas and paraganglioma syndromes. GMS Curr Top Otorhinolaryngol Head Neck Surg 2011;10:3. [9] Fruhmann J, Geigl JB, Konstantiniuk P, Cohnert TU. Paraganglioma of the carotid body: treatment strategy and SDH-gene mutations. Eur J Vasc Endovasc Surg 2013;45(5):431–6. [10] Shamblin WR, ReMine WH, Sheps SG, Harrison Jr EG. Carotid body tumor (chemodectoma). Clinicopathologic analysis of ninety cases. Am J Surg 1971;122(6):732–9. [11] Luna-Ortiz K, Rascon-Ortiz M, Villavicencio-Valencia V, Herrera-Gomez A. Does Shamblin’s classification predict postoperative morbidity in carotid body tumors? A proposal to modify Shamblin’s classification. Eur Arch Otorhinolaryngol 2005;263(2):171–5. [12] Arya S, Rao V, Juvekar S, Dcruz AK. Carotid body tumors: objective criteria to predict the Shamblin group on MR Imaging. Am J Neuroradiol 2008;29(7):1349–54. [13] Makeieff M, Thariat J, Reyt E, Righini C-A. Treatment of cervical paragangliomas. Eur Ann Otorhinolaryngol Head Neck Dis 2012;129(6):308–14. [14] Davidge-Pitts K, Pantanowitz D. Carotid body tumor. Surg Annu 1984;16:203–27. [15] Litle VR, Reilly LM, Ramos TK. Preoperative embolization of carotid body tumors: when is it appropriate? Ann Vasc Surg 1996;10(5):464–8. [16] Zhang T, Jiang W, Li Y, Li B, Yamakawa T. Perioperative approach in the surgical management of carotid body tumors. Ann Vasc Surg 2012;26(6):775–82. [17] Vogel TR, Mousa AY, Dombrovskiy VY, Haser PB, Graham AM. Carotid body tumor surgery: management and outcomes in the nation. Vasc Endovascular Surg 2009;43(5):457–61. [18] Kruger AJ, Walker PJ, Foster WJ, Jenkins JS, Boyne NS, Jenkins J. Important observations made managing carotid body tumors during a 25-year experience. J Vasc Surg 2010;52(6):1518–23. [19] Neumann HP, Pawlu C, Peczkowska M, et al. Distinct clinical features of paraganglioma syndromes associated with SDHB and SDHD gene mutations. JAMA 2004;292(8):943–51. [20] Van Hulsteijn LT, Dekkers OM, Hes FJ, Smit JWA, Corssmit EPM. Risk of malignant paraganglioma in SDHB-mutation and SDHD-mutation carriers: a systematic review and meta-analysis. J Med Genet 2012;49(12):768–76. [21] Lee JH, Barich F, Karnell LH, et al. National cancer data base report on malignant paragangliomas of the head and neck. Cancer 2002;94(3):730–7. [22] Sethi RV, Sethi RKV, Herr MW, Deschler DG. Malignant head and neck paragangliomas: treatment efficacy and prognostic indicators. Am J Otolaryngol 2013;34(5):431–8. [23] Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, et al. Recurrent head-and-neck chemodectomas: a comparison of surgical and radiotherapeutic results. Int J Radiat Oncol 2002;52(4):953–6. [24] Suárez C, Rodrigo JP, Mendenhall WM, et al. Carotid body paragangliomas: a systematic study on management with surgery and radiotherapy. Eur Arch Otorhinolaryngol 2014;271(1):23–34. [25] Luna-Ortiz K, Rascon-Ortiz M, Villavicencio-Valencia V, Granados-Garcia M, Herrera-Gomez A. Carotid body tumors: review of a 20-year experience. Oral Oncol 2005;41(1):569–661.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
ARTICLE IN PRESS European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
Available online at
ScienceDirect www.sciencedirect.com
Original article
Neurovascular complications following carotid body paraganglioma resection E. Lamblin a,b,∗ , I. Atallah a,b,c , E. Reyt a,b , S. Schmerber a,b , J.-L. Magne b,d , C.A. Righini a,b,c a Unité d’oto-rhino-laryngologie et chirurgie cervico-faciale, centre hospitalo-universitaire de Grenoble, 1, avenue du Maquis-du-Grésivaudan, 38043 Grenoble cedex 9, France b Université Joseph-Fourier, Grenoble I, BP 53, 38041 Grenoble cedex 9, France c Inserm U823, institut Albert-Bonniot, BP 170, 38042 Grenoble cedex 9, France d Unité de chirurgie endocrinienne, thoracique et vasculaire, centre hospitalo-universitaire de Grenoble, 1, avenue du Maquis-du-Grésivaudan, 38043 Grenoble cedex 9, France
a r t i c l e
i n f o
Keywords: Carotid body paraganglioma Surgery Neurovascular complications Shamblin classification
a b s t r a c t Objective: The present consecutive case series reports our experience in the management of carotid body paraganglioma and aims to assess whether the Shamblin classification or tumor size are predictive of early and late postoperative neurovascular complications. Material and methods: A retrospective study included 54 carotid body tumor resections in 49 patients, between 1980 and 2011. Data comprised early (< 1 month) and late (18 months) postoperative neurovascular complications. Results: Early postoperative complications occurred in 31 cases, including 30 cases of cranial nerve deficit (56%). Cranial nerve deficit occurred in 83% of Shamblin III carotid body paragangliomas and was associated with significantly larger mean tumor size (4 ± 1.4 cm versus 2.9 ± 1.3 cm; P < 0.01). Shamblin III tumor and tumor size > 3.2 cm emerged as predictive factors for early postoperative peripheral neurological complications. Eight patients (17%) showed no cranial nerve deficit recovery, even after 18 months’ follow-up; no predictive factors could be identified for this. Conclusion: Surgical resection remains the only curative treatment in carotid body paraganglioma, with low vascular morbidity. However, early postoperative nerve deficit remains frequent (56%), although mostly temporary, with 17% definitive sequelae at 18 months. Tumor size and Shamblin classification are predictive of early neurovascular complications. © 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction Paraganglioma is rare and usually benign, but certain genetic mutations (SDHB gene) may be associated with malignancy. It develops from diffuse neuroendocrine system cells between the skull base and pelvic floor. Cervical paraganglioma accounts for 0.6% of head and neck tumors [1]. Carotid body paraganglioma (CBP) is the most frequent parasympathetic form, with incidence of 1/30,000 to 1/100,000 in the general population. Surgical resection is the sole curative treatment in CBP. Growth is slow, but with risk of local compression and a slight risk of associated malignancy [2,3]. Progress in genetic screening, imaging
∗ Corresponding author. Unité d’oto-rhino-laryngologie et chirurgie cervicofaciale, centre hospitalo-universitaire de Grenoble, 1, avenue du Maquis-duGrésivaudan, 38043, Grenoble cedex 9, France. E-mail address: [email protected] (E. Lamblin).
(such as 3D time-of-flight MRI angiography) [4] and surgery have significantly reduced perioperative mortality (< 1%) and vascular morbidity (stroke: < 1%). Postoperative peripheral neurological morbidity, mainly involving the cranial nerves, however, is high, with a 14–49% rate of early deficit [2,3] and 6–23% persistent deficit, reported only in some recent studies with median follow-up of 6 months to 1 year [5–7]. Progress in alternative treatment strategies, such as adjuvant or exclusive, conventional, stereotactic or intensity-modulated radiation therapy, however, allows new attitudes in certain cases [8,9]. The Shamblin classification (Fig. 1), based on carotid vessel encasement, correlates with risk of postoperative vascular lesion [10,11]. Previous studies, however, disagreed on the relation between Shamblin group and onset of postoperative peripheral neurological complications [5–7,11,12]. The present study reports our experience in CBP, and assessed the risk of onset of early and late postoperative neurovascular complications according to Shamblin group and tumor size.
http://dx.doi.org/10.1016/j.anorl.2016.05.006 1879-7296/© 2016 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
2
2. Patients and methods A consecutive case series included all patients operated on for CBP in our university hospital center between January 1980 and December 2011. Demographic, clinical and paraclinical and surgical data, early (< 1 month) and late (> 18 months) postoperative neurovascular complications and disease progression were recorded. Tumors were graded on the Shamblin classification based on preoperative CT or MRI. Tumor size was defined by the long axis on pathologic examination. Resection was performed by a double ENT and vascular surgery team. Sufficiently large cervicotomy allowed good control of neurovascular structures. Carotid axis and the proximal and distal parts of the carotid arteries were first determined, then the vagus, spinal and hypoglossal nerves were located. Craniocaudal subadventitial dissection was performed up to the carotid sinus, which is the area at greatest risk of bifurcation tearing due to the presence of the carotid body artery [13]. An associated vascular procedure (simple suture, patch closure, carotid anastomosis resection or commonto-internal carotid bypass) was performed under effective heparin therapy and hemodynamic control by the vascular surgery team if arterial continuity was interrupted; this was either pre-planned, in case of large tumor or tumor adhering to the internal carotid artery (ICA), requiring carotid resection, or else in case of lesion during dissection. Lymph tissue adjacent to the CBP was systematically dissected after the year 2000. Patient characteristics were reported as median and interquartile range (IQR: 25th and 75th percentiles) for continuous variables and as percentages for categoric variables. Univariate analysis assessed whether Shamblin group or tumor size correlated with postoperative neurovascular complications risk. Categoric variables were analyzed on Chi2 or Fisher exact test, and continuous variables on Kruskal-Wallis or Mann-Whitney-Wilcoxon test. ROC curves were plotted to determine whether a threshold tumor size
was predictive of postoperative peripheral neurologic complications. The significance threshold was set at P < 0.05. All analyses were conducted on R software version 3.0.1. 3. Results 3.1. Description of population Baseline demographic and clinical and paraclinical characteristics are shown in Table 1. Fifty-four CBP resections were performed in 49 patients. In 36 cases, there was lateral cervical swelling for a median 38 months before medical consultation (IQR, 1–48 months). Median follow-up was 17 months (IQR, 10–67 months). Forty-three patients (48 CBPs) had a minimum 18 months’ follow-up. In 98% of cases, diagnosis was founded on arteriography or angio-CT; ultrasonography or fine-needle aspiration were the primary examination in 12 patients (10 before 1992, most referred from other centers), who then underwent exploratory cervicotomy, without biopsy, and without complications, before CBP resection following confirmation of diagnosis on complete imaging assessment. Most patients with multiple forms had bilateral CBP (9 out of 12); other associated locations comprised: 3 jugulotympanic, 1 vagal, 1 mediastinal and 1 pancreatic paragangliomas and 1 pheochromocytoma (adrenal). Six patients with bilateral CBP underwent bilateral resection, with the smaller tumor generally treated in the first step. In some cases, other treatments were proposed for the contralateral CBP: radiation therapy (due to ipsilateral vagus nerve lesion during the first procedure), or medical surveillance. Genetic assessment was systematically offered to patients and family, with screening for germline mutations in succinate dehydrogenase subunits D, B and C (SDHD, SDHB, SDHC). More recently, screening included other mutations: in succinate dehydrogenase
Fig. 1. The Shamblin classification of carotid body tumor enables the difficulty of resection to be predicted. Group I CBP is very localized and easy to resect. Group II adheres to or partially encases the arteries. Group III completely encases at least one artery. ECA: external carotid artery; ICA: internal carotid artery; X: vagal nerve; XII: hypoglossal nerve.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model
ARTICLE IN PRESS
ANORL-565; No. of Pages 6
E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
3
Table 1 Baseline demographic, clinical and paraclinical data. Variable Demographics CBP resection Age, yearsa Female gender Right CBP Multiple CBP Bilateral CBP Malignant CBP Follow-up, monthsa Mode of discovery Duration of symptoms before diagnosis, monthsa Neck mass Lesion staging Neurovascular work-up Flush Genetic assessment Associated symptoms Hypertension Malaise Tinnitus Cough and dysphonia Related to preoperative vagus nerve palsy Complementary examinations Genetic studies Germline mutation SDHD gene SDHB gene Catecholamine urine or blood assay Significant elevation Significant elevation related to a functional CBP Arteriography or CT-angiography CT MRI Scintigraphy
Number of patient (n = 49) 54 49 (38.2–57.7) 27 (55) 32 (59) 12 (24) 9 (83) 2 (4) 17 (10–67) 38 (1–48) 36 (67) 7 (13) 7 (13) 2 (4) 1 (2) 18 (37) 10 (20) 2 (4) 12 (25) 3 (25) 14 (29) 10 (71) 7 (70) 3 (30) 38 (78) 7 (18) 4 (11) 48 (98) 32 (65) 16 (33) 16 (33)
Values in brackets are percentages unless indicated otherwise; CBP: carotid body paraganglioma; SDHD-B: succinate dehydrogenase subunit D-B; CT: computed tomography; MRI: magnetic resonance imaging. a Values are medians (interquartile range).
subunit A and Von Hippel Lindau genes (SDHA, VHL). Ten patients showed mutations, mainly on the SDHD gene; 8 (7 with SDHD mutation) had multiple CBP. Four patients showed significant catecholamine elevation, related to secreting CBP. All specimens underwent histologic examination, systematically confirming diagnosis. Two patients had incidentally discovered metastatic adenopathies on pathologic examination of adjacent lymph nodes removed along with the tumor. Resection was incomplete in 4 patients with group-III CBP, due either to extension around skull-base vessels (2 patients with > 6 cm tumor) or (in 2 patients) to positive margins on pathologic examination. Tumor size was significantly greater in higher Shamblin grades (P < 0.01). 3.2. Surgery All patients underwent surgical resection. Preoperative embolization was performed in 2 tumors exceeding 4 cm, treated before 1992. Median operative time was 135 minutes (IQR, 120–180 minutes), and was significantly longer in case of higher Shamblin grade and greater tumor size (Fig. 2). There were 20 cases of intraoperative complications (37%). Seven patients (13%) suffered nerve sacrifice: 3 vagus, 3 superior laryngeal and 1 hypoglossal. Vascular surgery was performed in 19 cases (35%): 18 external carotid ligations, and 13 internal carotid procedures (9 bypasses, 2 venous patches, 1 ligation and 1 direct suture). Shamblin group and tumor size correlated with risk of
Fig. 2. Correlation of (a) Shamblin group and (b) tumor size with operative time, a: box-plot of surgery time according to Shamblin group; b: surgery time according to tumor size. Median operative time correlated with Shamblin group (Kruskal-Wallis test: *, P < 0.05) and tumor size (Pearson correlation test: r = 0.50; P < 0.001).
intraoperative complications, including nerve sacrifice and vascular surgery (P < 0.05, Tables 2 and 3).
3.3. Early (< 1 month) postoperative complications Postoperative neurovascular complications comprised: 3 cases of stroke (1 cerebral, after ICA ligation, and 2 transient ischemic attacks), and 30 of peripheral neurologic deficit (56%). Peripheral neurologic involvement concerned several nerves in 18 cases, with total deficit in 11. The nerves most often involved were the vagus (37% of cases of deficit) and hypoglossal (32%), followed by cervical sympathetic (13%), glossopharyngeal (10%) and superior laryngeal (6%). Non-neurological early postoperative complications were local (1 abscess, and 2 hematomas requiring surgical revision), vascular (1 bypass thrombosis, requiring revascularization) or general (1 case of pulmonary embolism, 1 of gastrointestinal hemorrhage, 1 of pneumonia and 2 of non-controlled hypertension). There were no early postoperative deaths. Incidence of early postoperative complications overall was significantly greater with higher Shamblin grade or greater tumor size, except for central
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model
ARTICLE IN PRESS
ANORL-565; No. of Pages 6
E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
4
Table 2 Neurovascular complications according to Shamblin group. Variable
Shamblin I (n = 17)
Shamblin II (n = 25)
Shamblin III (n = 12)
P value
Intraoperative complications Cranial nerve adherence Cranial nerve resection Vascular procedure Internal carotid repair Early postoperative complications Peripheral neurologic deficit Central neurologic deficit Non-neurologic complication Late postoperative complications (> 18 months) Permanent cranial nerve deficit
1 (6) 3 (18) 1 (6) 0 (0) 0 (0) 5 (29) 5 (29) 0 (0) 0 (0) n = 17 1 (6)
8 (32) 7 (28) 1 (4) 9 (36) 4 (13) 16 (64) 15 (60) 0 (0) 2 (8) n = 22 4 (18)
11 (92) 6 (50) 5 (42) 10 (83) 9 (75) 10 (83) 10 (83) 3 (25) 5 (42) n=9 3 (33)
< 0.001a 0.201a 0.008a < 0.001a < 0.001a 0.003b 0.013b 0.008a 0.003a 0.179a
Values in brackets are percentages unless indicated otherwise. a Fisher test. b 2 test.
neurologic complications, which correlated only with Shamblin group (Tables 2 and 3). Incidence of peripheral neurologic deficit was compared according to tumor size, with a view to determining a threshold size predictive of early peripheral neurologic deficit. Sensitivity (63%; 95% CI [57–70]) and specificity (75%; 95% CI [69–81]) were optimal for a threshold of ≥ 3.2 cm (area under the curve = 0.74), with a positive predictive value of 76% (95% CI [70%; 82%]).
3.4. Late (> 18 months) postoperative complications Follow-up exceeded 18 months in 43 patients (48 CBPs), with 5 patients initially showing deficit being lost to follow-up. Speech therapy was systematically prescribed in case of cranial nerve involvement. Peripheral neurologic deficit recovered completely in 17 cases (68% of patients with early deficit). Median time to recovery was 6 months (IQR, 2–12 months). Finally, 8 cases (17% of those followed up for ≥ 18 months) showed permanent peripheral neurologic deficit, with persistent symptoms: 5 with swallowing disorder and 3 with dysphonia. Univariate analysis found no predictive factor for permanent peripheral neurologic deficit (Tables 2 and 3).
3.5. Disease progression One patient died 10 months postoperatively, after stroke following ICA ligation: he had had group-III CBP of 7 cm; as there was no intraoperative skull-base arterial reflux, carotid bypass had not been performed. One patient showed local progression of residual tumor 1 year postoperatively. Resection had been incomplete, without complementary treatment. Resection was repeated, with no signs of recurrence at 6 years’ subsequent follow-up. Two patients, treated before 1992, showed local recurrence associated with remote tumor. The first showed local recurrence associated with mediastinal tumor 1 year after resection of groupIII bilateral CBP with incidentally discovered metastatic cervical adenopathies on pathologic examination of lymph nodes adjacent to the tumor. The cervical recurrence was managed by radiation therapy (due to contralateral vagus deficit), and the mediastinal paraganglioma was resected, without recurrence at 10 months’ follow-up. The second patient also showed local recurrence, but associated with multiple cervical locations, 27 years after primary surgery; there had been positive margins and metastatic cervical adenopathies, but no complementary treatment in view of advanced age and the multiplicity of the cervical locations.
Table 3 Neurovascular complications according to tumor size. Variable (n = 56)
No complications, n Median tumor size, cm (IQR)
Complications, n Median tumor size, cm (IQR)
P valuea
Intraoperative complications
34 (63) 3 (2–4) 38 (70) 3 (2–4) 47 (87) 3.3 (2.9, 3.7) 35 (65) 3 (2–4) 41 (76) 3 (2–4) 24 (44) 2.5 (2–3.2) 24 (44) 2.75 (2–3.25) 51 (98) 3 (2.2–4) 47 (87) 3 (2–4)
20 (37) 4 (3.1–6) 16 (30) 3.3 (3–5) 7 (23) 4.8 (3, 6.6) 19 (35) 4 (3.3–5.5) 13 (24) 4 (3.2–5) 30 (56) 4 (3–5) 30 (56) 4 (3–5) 3 (2) 5 (4–6) 7 (23) 6 (3.5–6)
< 0.001
40 (83) 3 (2–4)
8 (17) 3.25 (3–4.25)
Cranial nerve adherence Cranial nerve resection Vascular procedure Internal carotid repair Early postoperative complications (< 1 month) Peripheral neurologic deficit Central neurologic deficit Non-neurologic complication Postoperative complications (> 18 months) (n = 48) Permanent cranial nerve deficit
0.153 0.04 0.001 0.03 0.006 0.007 0.13 0.001
0.3
Values in brackets are percentages unless indicated otherwise. a Mann-Whitney-Wilcoxon test.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
One patient developed bone and liver metastases 4 years after resection of bilateral dopamine-secreting CBP. Clinical course rapidly deteriorated despite radiation therapy, and the patient died some weeks later.
4. Discussion Clinical assessment of any lateral neck mass should be meticulous, to avoid missing a rare differential diagnosis such as CBP. In more than two thirds of patients with CBP, diagnosis is based on discovery of a pulsatile lateral neck mass, usually with restricted vertical mobility [14] and often silently present for several years. Doppler ultrasound lacks sensitivity for diagnostic work-up of CBP. MRI (especially 3D time-of-flight MR angiography [4]) allows definitive diagnosis when angio-CT data are insufficient. However, the gold standard to detect small CBPs is digital subtraction angiography [4], because it reveals the specific vascular supply of the paraganglioma by the ascending pharyngeal artery. Fine-needle aspiration or exploratory cervicotomy are no longer indicated after well-conducted clinical and paraclinical assessment. The contribution of preoperative embolization is controversial [5,9,15–17]. We do not usually perform preoperative embolization (only 2 patients before 1992, for > 4 cm tumor, for which embolization was only partial), although it has recently been reported not to be associated with increased risk of stroke [17]. Most authors report little benefit in postoperative course, with no evidence of a decrease in morbidity in general or in cranial nerve deficit in particular [15]. It may, however, according to some authors, facilitate the resection of large, very adherent tumors, especially when located near the skull base, as it reduces clamping time and blood loss [5,16]. According to recent studies, 30% of CBPs are hereditary, mainly involving SDHD gene mutation [5,18]. Genetic assessment should be systematic, due to the high prevalence (6–25%) of mutations in patients presenting with apparently sporadic forms [8]. Several authors [19,20] reported that multifocal CBT is frequently associated with SDHD mutation (71% in the present study). A recent meta-analysis showed that the risk of malignant CBP was higher in case of SDHB gene mutation [20]. Hence systematic genetic assessment enables CBP to be detected at an earlier stage, and facilitates management of multifocal and malignant forms. In bilateral CBP, the smaller lesion is generally resected first and, if there are no postoperative complications, the second can be treated subsequently. In multiple forms, functional tumors (pheochromocytoma) are treated first. Two patients had cervical lymph node metastases and 1 had remote metastases. As there are no specific histopathological features of malignancy, diagnosis may be confirmed only when there is metastasis to non-neuroendocrine tissue [20,21]. Cervical metastasis was diagnosed incidentally in both cases of adjacent lymph nodes removed along with the tumor. We therefore now resect peritumoral lymph nodes to identify possible occult metastasis in all patients with apparently benign CBP [8,21]. Boedeker [8] recommended selective neck dissection for levels II and III, but dissection extension remains to be defined. Our results were consistent with those of previous studies, with malignancy around 6%, with less remote metastasis than in other head and neck paraganglioma sites [21,22]. The authors of two large studies of malignant head and neck paraganglioma reported significantly better 5-year overall survival (76.8–82.4%) in patients with regional than distant metastasis (11.8–41.4%). In case of metastasis, there is no evidence that postoperative radiotherapy improves the 5-year overall survival. In case of incomplete resection, however, adjuvant radiation therapy may slow tumor growth and improve survival [21,22]. Three patients showed local progression of residual tumor or recurrence, 1 and 27 years after surgery. Elshaikh et al. [23] likewise
5
reported long median time to recurrence of head and neck paraganglioma (36 months; range, 15–350 months), and recommended long follow-up to detect late recurrence. Their results confirmed that surgery provided control of local recurrence, but few studies assessed the role of radiation therapy in control of CBP recurrence [23]. Complete surgical excision is the aim of any operative procedure and is the only curative treatment. Nevertheless, radiation therapy may be considered for patients in whom surgery is contraindicated by advanced age, comorbidity or aggressive or unresectable tumor [8,24]. The goal of radiotherapy is to control the disease or inhibit growth rather than to eliminate the tumor. A recent review of CBP management [24] reported no significant difference in tumor control or mortality between surgery and radiotherapy, but a higher risk of major complications and cranial nerve palsy following surgery. Radiation side effects should also be taken into account in deciding treatment, with a 1% risk of radiation-induced malignancy at 10 years, and more difficult surgery with higher risk of neurovascular complication. However, some adverse effects have been reduced by new techniques such as intensity-modulated radiation therapy, making surgery no longer the obvious gold standard in CBP. Tumor doubling time is very slow, and conservative treatment, such as a clinical and radiological wait-and-see attitude, may be envisaged for elderly patients with small tumor size. The most important criterion in the Shamblin classification [10] is carotid wall encasement, which is related to tumor size [14]. However, the relationship has been reported to be non-linear in some studies [12]. Shamblin group was significantly associated with tumor size in the present study. The Shamblin classification was initially based on intraoperative and pathological findings [10]; the present study took account of preoperative radiological data, but Shamblin groups were nevertheless distributed as usually reported: one third in group I, almost a half in group II, and a quarter in group III [5,6]. Other radiologic criteria may also reflect artery encasement: ICA encasement angle, or radiolucent line between tumor and ICA [12]. We were not able to analyze these features due to lack of data in preoperative radiographs. Tumor size was assessed according to the longest axis in pathology reports; the assessment of tumor volume on imaging could provide a useful objective preoperative criterion [5,18]. Twenty-four percent of cases required ICA surgery, in agreement with the literature, reporting rates of 6–52% [9,25]. The present data included all types of vascular procedure, even direct suture and venous patch, which may account for the discrepancy with recent reports, which did not always detail the type of vascular procedure [12]. Vascular surgery risk has decreased over the past few decades, especially for ICA ligation [2]. Consequently, the risk of postoperative stroke is now very low: only one patient in the present study, in 1991. Seventeen percent of patients suffered stroke in an older large series published in 1988 [2]. Furthermore, vascular repair has become safer with improved surgical techniques: early control of proximal and distal carotid vessels, subadventitial tumor dissection, effective heparin therapy, precise hemodynamic control and reduced clamping time have all reduced postoperative risk. In the present study, higher Shamblin grade and larger tumor size correlated with higher risk of intraoperative vascular lesion; thus both allowed prediction of risk [7]. Peripheral neurological morbidity has remained high and unchanged for 50 years [2] despite improved surgical techniques. Early peripheral neurological deficit occurred in 56% of the present series, and was shown to correlate with both Shamblin group and tumor size, both of which could thus be used to predict the risk of early neurologic deficit [5]. Some authors suggested a 4 cm cutoff to predict increased risk of neurological morbidity [1,13] the present data, however, show that ≥ 3.2 cm was the best threshold on ROC analysis. Since the larger the tumor the higher the risk
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006
G Model ANORL-565; No. of Pages 6
ARTICLE IN PRESS E. Lamblin et al. / European Annals of Otorhinolaryngology, Head and Neck diseases xxx (2016) xxx–xxx
6
of neurologic deficit, systematic genetic assessment should enable earlier diagnosis of small CBP [9]. In 68% of cases, peripheral neurologic deficit was transient, with permanent deficit at 18 months in 17%. Early appropriate speech therapy is essential. No cases of deficit at 18 months subsequently recovered. This finding is difficult to compare with the literature, as authors do not systematically report recovery time or else limit follow-up to 12 months [5]. We were unable to find any predictive factors for permanent neurologic deficit. Power et al. [5] reported that permanent cranial nerve deficit at 1 year correlated with Shamblin group. Further studies should take 18 months as the cut-off value for permanent neurologic deficit. 5. Conclusion Surgical resection remains the only curative treatment in CBP. Adjacent lymph node dissection is recommended. Early peripheral neurologic deficit is frequent (56%); only 17% fail to recover within 18 months. Tumor size and Shamblin group correlate with risk of early neurovascular complications. Genetic assessment should be systematic to enable early detection of CBP and improve outcome in multifocal and malignant forms. Disclosure of interest The authors declare that they have no competing interest. Acknowledgments The authors warmly thank Dr B. Boussat of the Clinical Investigation Center of the Grenoble University Hospital for the statistical analysis, and Mr P.-E. Colle for re-editing the English-language version. References [1] Maves MD. Vascular tumors of the head and neck. Head and neckotolaryngology. In: Bailey BJ, Jonhnson JT, Kohut RI, Pillsbury HC, editors. Tardy medical management. Philadelphia: JB Lippincott; 1993. p. 1397–409. [2] Hallett JW, Nora JD, Hollier LH, Cherry KJ, Pairolero PC. Trends in neurovascular complications of surgical management for carotid body and cervical paragangliomas: a fifty-year experience with 153 tumors. J Vasc Surg 1988;7(2):284–91. [3] Sajid MS, Hamilton G, Baker DM. A multicenter review of carotid body tumour management. Eur J Vasc Endovasc Surg 2007;34(2):127–30.
[4] Berg R. Imaging and management of head and neck paragangliomas. Eur Radiol 2005;15(7):1310–8. [5] Power AH, Bower TC, Kasperbauer J, et al. Impact of preoperative embolization on outcomes of carotid body tumor resections. J Vasc Surg 2012;56(4):979–89. [6] Makeieff M, Raingeard I, Alric P, Bonafe A, Guerrier B, Marty-Ane C. Surgical management of carotid body tumors. Ann Surg Oncol 2008;15(8):2180–6. [7] Lim JY, Kim J, Kim SH, et al. Surgical treatment of carotid body paragangliomas: outcomes and complications according to the Shamblin classification. Clin Exp Otorhinolaryngol 2010;3(2):91–5. [8] Boedeker CC. Paragangliomas and paraganglioma syndromes. GMS Curr Top Otorhinolaryngol Head Neck Surg 2011;10:3. [9] Fruhmann J, Geigl JB, Konstantiniuk P, Cohnert TU. Paraganglioma of the carotid body: treatment strategy and SDH-gene mutations. Eur J Vasc Endovasc Surg 2013;45(5):431–6. [10] Shamblin WR, ReMine WH, Sheps SG, Harrison Jr EG. Carotid body tumor (chemodectoma). Clinicopathologic analysis of ninety cases. Am J Surg 1971;122(6):732–9. [11] Luna-Ortiz K, Rascon-Ortiz M, Villavicencio-Valencia V, Herrera-Gomez A. Does Shamblin’s classification predict postoperative morbidity in carotid body tumors? A proposal to modify Shamblin’s classification. Eur Arch Otorhinolaryngol 2005;263(2):171–5. [12] Arya S, Rao V, Juvekar S, Dcruz AK. Carotid body tumors: objective criteria to predict the Shamblin group on MR Imaging. Am J Neuroradiol 2008;29(7):1349–54. [13] Makeieff M, Thariat J, Reyt E, Righini C-A. Treatment of cervical paragangliomas. Eur Ann Otorhinolaryngol Head Neck Dis 2012;129(6):308–14. [14] Davidge-Pitts K, Pantanowitz D. Carotid body tumor. Surg Annu 1984;16:203–27. [15] Litle VR, Reilly LM, Ramos TK. Preoperative embolization of carotid body tumors: when is it appropriate? Ann Vasc Surg 1996;10(5):464–8. [16] Zhang T, Jiang W, Li Y, Li B, Yamakawa T. Perioperative approach in the surgical management of carotid body tumors. Ann Vasc Surg 2012;26(6):775–82. [17] Vogel TR, Mousa AY, Dombrovskiy VY, Haser PB, Graham AM. Carotid body tumor surgery: management and outcomes in the nation. Vasc Endovascular Surg 2009;43(5):457–61. [18] Kruger AJ, Walker PJ, Foster WJ, Jenkins JS, Boyne NS, Jenkins J. Important observations made managing carotid body tumors during a 25-year experience. J Vasc Surg 2010;52(6):1518–23. [19] Neumann HP, Pawlu C, Peczkowska M, et al. Distinct clinical features of paraganglioma syndromes associated with SDHB and SDHD gene mutations. JAMA 2004;292(8):943–51. [20] Van Hulsteijn LT, Dekkers OM, Hes FJ, Smit JWA, Corssmit EPM. Risk of malignant paraganglioma in SDHB-mutation and SDHD-mutation carriers: a systematic review and meta-analysis. J Med Genet 2012;49(12):768–76. [21] Lee JH, Barich F, Karnell LH, et al. National cancer data base report on malignant paragangliomas of the head and neck. Cancer 2002;94(3):730–7. [22] Sethi RV, Sethi RKV, Herr MW, Deschler DG. Malignant head and neck paragangliomas: treatment efficacy and prognostic indicators. Am J Otolaryngol 2013;34(5):431–8. [23] Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, et al. Recurrent head-and-neck chemodectomas: a comparison of surgical and radiotherapeutic results. Int J Radiat Oncol 2002;52(4):953–6. [24] Suárez C, Rodrigo JP, Mendenhall WM, et al. Carotid body paragangliomas: a systematic study on management with surgery and radiotherapy. Eur Arch Otorhinolaryngol 2014;271(1):23–34. [25] Luna-Ortiz K, Rascon-Ortiz M, Villavicencio-Valencia V, Granados-Garcia M, Herrera-Gomez A. Carotid body tumors: review of a 20-year experience. Oral Oncol 2005;41(1):569–661.
Please cite this article in press as: Lamblin E, et al. Neurovascular complications following carotid body paraganglioma resection. European Annals of Otorhinolaryngology, Head and Neck diseases (2016), http://dx.doi.org/10.1016/j.anorl.2016.05.006