Carotid Body Tumors, Inheritance, and a High Incidence of Associated Cervical

Paragangliomas Phillip Gardner, MD, Michael Dalsing, MD, Edward Weisberger, MD, Alan Sawchuk, MD,

Richard Miyamoto, MD, /ndianapo/is, lndiana

BACKGROUND: Current experience with carotid body tumors suggesting a high prevalence of as- sociated cervical paragangliomas prompted this review.

PATIENTS AND METHODS: An 8-year retrospective study of patients with carotid body tumors was undertaken, detailing presentation, diagnosis, and treatment.

RESULTS: Eleven patients harboring 17 carotid body tumors were discovered. All patients had a neck mass. Seven patients (64%) had bilateral carotid body tumors. Six (55%) reported a posi- tive family history-4 were first-generation rela- tives, 5 had bilateral tumors, and 3 had other head and neck paragangliomas. Angiography documented 4 associated vagal and 2 glomus jugulare paragangliomas in addition to the ca- rotid body tumors. Precise surgical care limited blood loss to an average of 590 cc. The carotid artery was repaired during 5 resections (29%). Cranial nerve injury occurred in 3 cases, all fol- lowing vagal body or glomus jugulare resection. Every patient is currently alive, stroke free, and functioning without major disability.

CONCLUSIONS: Patients with carotid body tumors have a propensity for multiple head and neck paragangliomas. Angiography is diagnostic. The need for associated paraganglioma resection dramatically increases the risk of cranial nerve injury. Am J Surg. 1996;172:196-199.

0 ur current experience with carotid body tumors was reviewed in response to a perceived high incidence of associated cervical paragangliomas and to clarify

aspects of familial inheritance in our cohort of patients.

PATIENTS AND METHODS This retrospective review ( 1986 to 1995) was conducted

at two hospitals associated with the Indiana University

From the Departments of Surgery (MD, AS) and Otolaryngology (PG, EW, RM), Indiana University Medical Center, Indianapolis, Indiana.

Requests for reprints should be addressed to Michael C. Dals- ing, MD, Indiana University Medical Center, Wishard Memorial Hospital,1001 West 10th Street, OPE 303, Indianapolis, Indiana 46202.

Presented at the 24th Annual Meeting of The Society for Clinical Vascular Surgery, Ranch0 Mirage, California, March 20-24, 1996.

Medical Center. Patients with cervical paragangliomas not involving the carotid body were excluded from this study. The medical records for review were generated by computer search using carotid body tumor diagnostic codes (ICD9 237.3, 194.5). The majority of patients were actually being observed by one of the investigators.

Routine patient demographics in addition to the initial mode of presentation were tabulated. Any familial occur- rence was noted, and a detailed pedigree was actively sought. Methods of diagnosis used to confirm the clinical impression or to clarify the anatomic nature of the tumor were reviewed with the hope of defining the accuracy of each method. Treatment considerations involved preoper- ative tumor emholization, surgical technique, and overall blood loss. Surgical results included completeness of resec- tion, the incidence of stroke and cranial nerve injury, and any other operative complications. Follow-up highlighted residual disability, disease-free status, and stroke-free status.

RESULTS Eleven patients, 8 women and 3 men, were found. Their

mean age was 40 years (range 19 to 60). Six patients re- ported a family history positive for relatives having a similar condition; 4 of these patients were from the family A ped- igree (Figure 1)) 1 came from family B (Figure 2)) and 1 came from family C (Figure 3 ) .

All patients presented with a neck mass. In 1 case, a con- tralateral vagal body tumor was palpable and the carotid body tumor was found incidentally following further studies. Two carotid body tumors were discovered by physical ex- amination of patients with known contralateral neck masses. Only 3 patients complained of headaches or local intermittent cervical neck pain. One patient had had a pre- vious neck exploration and radiation therapy for a neck mass (etiology undetermined), resulting in ipsilateral vocal cord paralysis. Another patient had previously undergone an uncomplicated carotid body tumor resection on the op- posite side. Five patients had mild to moderate hypertension that was evaluated by vanillymandelic acid (VMA) and metanephrine urine sampling. These studies failed to dem- onstrate a functioning tumor.

All patients were evaluated h\; angiography (aortic arch and four-vessel study) hut only after an unsuccessful per- cutaneous biopsy failed to make the diagnosis in one case and after surgical exposure defined the vascular nature of the tumor in another. Angiography confirmed the diagnosis of a carotid body tumor in all cases as demonstrated by a tumor “blush” located in the carotid bifurcation (Figure 4). Furthermore, 13 previously unsuspected paragangliomas

196 0 1996 by Excerpta Medica, Inc. 0002-961 O/96/$1 5.00 All rights reserved. PII SOOO2-9610(96)00151-l

Figure 1. The pedigree of family A with familial paragangliomas represents an autosomal dominant inheritance pattern with vari- able penetrance.

Figure 2. The pedigree from family B suggests an autosomal re- cessive inheritance pattern.

0 Male n l Bilatera Carotid Body Tumor

0 Female q 0 Llnilah’al Carotid Body Tumor

Figure 3. The pedigree from family C represents an autosomal dominant inheritance pattern with incomplete penetrance.

and 1 internal carotid artery pseudoaneurysm were discov- ered. The unsuspected paragangliomas included 7 carotid body tumors, 4 vagal body tumors, and 2 glomus jugulare tumors. Therefore, there were 17 carotid body tumors in this review in addition to 4 vagal body tumors and 2 glomus jugulare tumors. There were no procedural complications associated with angiography.

I -/ CAROTID BODY TUMORS/GARDNER ET AL !

Nine patients were studied by contrast-enhanced com- puted tomography (CT). In 2 cases, this study failed to recognize the carotid body tumor. C1ne tumor was small (4 mm). The other- was a 4 X 4 cm tumor; however, wash-out of the contrast injection had occurred prior to filming. This patient was found to have a vascular tumor at operation, the procedure was aborted, and ,ulgiography clarified the diagnosis. In all other instances, the tumors demonstrated by angiography were also seen by this modality. Therefore, CT specificity ior carotid body tumors was lOO?h, but its sensitivity was 78%. Carotid artery duplex imaging was con- ducted in only 2 cases and demonstrated an interesting im- age of the tutnor’s vascularity in each case. This was gen- erally performed to evaluate a questionable mash in familial cases or an associated bruit.

Seven of 11 patients (64(X,) had bilateral carotid body tumors, and 5 of 6 patients with a positive family history were so afflicted (83%). Fifty percent of patients with a positive family history had associated cervical vagal body or glomus jugulare paraganglioma, with only 20% of sponta- neous cases affected.

Six carotid body tumors underwent prcopcratlve emboli- zation through vessels originating from the external carotid artery. One attrtllpt was aborted due to difficult anatomy. Temporary halloon occlusion of the internal carotid artery with the patient fully heparinized and with clectroenceph- alograph (EEG) monitoring to assehs the collateral circu- lation was performed prior to five tumor resections. No neu- rologic symptom or EEG change was noted during these test occlusions, nor were any sequelae from the emholization procedures observed.

Surgical resection routinely involved a platsymal flap for exposure, loupe magnification during the dissection, the use of bipolar cautery particularly when working near the ca- rotid artery, and postoperative sucti<)n drainage. Fourteen isolated carotid body tumor resections were performed, and every resection was complete. Three external carotid arter- ies were ligated, and five internal carotid arteries required primary suture repair (one was an end-to-end anastamosis.) Heparin was utilized in all repairs, and a carotid artery shunt was used in the case requiring an anastomosis. The other four repairs required less than 10 mmutes of carotid artery occlusion. The average blood loss W~XS 590 cc (range 25 to 1,800 cc). The size of the resected tumors ranged from 0.6 x 0.4 cm to 10 x 5 cm. It should be noted that 2 patients harboring a solitary carotid body tumor also had a contra- lateral paraganglioma, which of course did not affect the results of the isolated carotid body tumor resections.

Three patients had bilateral carotid body tumors with one side also having an associated cervical paraganglioma. Sur- gical parameters for these combined resections were consid- ered separately from those requiring isolated carotid body tumor resection. Preoperative emholization was not used prior to any of these cases. There was complete resection of all tumors, and cranial nerve injury occurred in all 3 patients. The average blood loss for these procedures was 1,100 cc, and no internal carotid required repair.

There were no deaths or cerebrovascular events observed in this series. At a mean follow-up of 3 years, there has heen no clinical recurrences. Cranial nerve injury occurred in 3 cases ( 18%); each of these injuries w,ls associated with ei- ther glomus jugulare or vagal body tumor resection. Cranial

THE AMERICAN JOURNAL OF SURGERY@ VOLUME 172 AUGUST 1996 197

Figure 4. This angiogram of the carotid artery system displays a “blush” consistent with a carotid body tumor. The tumor is betwee :n the internal and external carotid arteries with feeder ves.5 jels originating from the latter.

lot :ated

nerve X was involved in all 3 cases, whereas cranial nerve XII was also injured in 1 case, cranial nerve XI in 1 case, and cranial nerve Vll in 1 case. A Horner’s syndrome was noted postoperatively in 1 patient, and 1 patient was reex- plored for a hematoma on postoperative day 1. Symptoms of basorecepter failure were noted postoperatively in 1 pa- tient. Cranial nerve X injury resulted in the need for ad- ditional surgical procedures: Each of these patients received either a gastrostomy or jejunostomy feeding tube for dys- phasia, and also underwent vocal cord medialization pro- cedures. All patients are currently functioning without ma- jor disability, eating on their own, and able to speak without difficulty.

COMMENTS A middle-aged patient with an otherwise asymptomatic

palpable neck mass is the typical patient.’ A complere history and physical exammation is invaluable. Two tu- mors in this series were found on physical examination alone. Possibly more importantly, the identification of a family history can lead to the screening of other patients at risk. The potential for catecholamine secretion> in hy- pertensive patients was considered and evaluated in 5 of our patients by urinary VMA and metanephrine analysis. Although no functioning tumor was discovered in our cohort of patients and is only rarely noted with carotid body tumors in general ( 1% to 3%) ,3 simultaneous pheo- chromocytomas have been reported, especially in the fa- milial variety,4 and therefore we felt justified in investi- gating this possibility.

Angiography is the gold standard for diagnosis. Splaying of the carotid bifurcation and the characteristic vascular “blush” are virtually pathognomonic of a carotid body tu- mor.5 In our series, an additional 13 unsuspected paragan- gliomas of the 23 total observed and 1 internal carotid ar- tery pseudoaneurysm were discovered. Overall, 7 (64%) of our patients had bilateral carotid body tumors, highlighting the need for bilateral carotid imaging. The frequent reyuire- ment for bilateral operations led us to evaluate the collateral circulation and to strive to localize precise tumor blood sup-

ply for a more controlled operation. These needs translate clinically into a study consisting of aortic arch imaging plus four-vessel angiography with intracranial views, an ap- proach we recommend. In addition, the fiact that the carotid body tumor often obtains its blood supply from the external carotid artery allowed preoperative embolization of six tu- mors.6 Collateral circulation was further evaluated prior to five tumor resections by internal carotid artery test occlu- sion, a useful adjuvant for large tumors when the possible need for ligation is a concern. Preoperative needle biopsy is to be actively condemned, as highhghted by the carotid pseudoaneurysm noted following biopsy in 1 of our patients.

Contrast-enhanced CT scanning can define the compart- mental extent of the tumor.’ However, in our experience, 2 tumors in 9 patients who harbored 15 tumors were mis- diagnosed, making this study less than optimal for the pri- mary diagnosis of a carotid body tumor. Carotid body duplex imaging provides a very impressive picture of the vascular nature of the mass but is unlikely to image associated para- gangliomas high in the neck. We do not have sufficient data to know if this test can be used for screening.

Our current experience is skewed with a high proportion of familial cases. Generally, familial cases account for only about 10% of all cases in the literature.’ Transmission by autosomal dominance with variable penetrance5 could ex- plain the inheritance in our family A and family C patients. The inheritance pattern of f&lily B may be more readily explained hy an autosomal recessive trait.

The high degree of multiple paragangliomas, either bilat- eral carotid body tumors or associated paragangliomas, in addition to a carotid body tumor appears to be unique to this series. The incidence of cervical paragangliomas from origins other than the carotid body in association with ca- rotid body tumors has not been stressed previously in the literature. However, based upon the known neural crest or- igin and wide distribution of such cells in the neck,’ it is not a very surprising finding. Patients with a positive family history demonstrated an 83% incidence of bilateral carotid body tumor, and those without a family history harbored a 40% incidence. Previous studies describe a 30% incidence

198 THE AMERICAN JOURNAL OF SURGERY” VOLUME 172 AUGUST 1996

of bilateral carotid body tumors in familial cases and a 5% to 10% incidence in the sporadic type.4

A more recent study involving 30 patients demonstrated an 87% incidence of multiple paragangliomas in familial cases and 43% incidence in the sporadic group, an experi- ence much more compatible to our own.’ The presence of a carotid body tumor, either single or bilateral, appears to have little impact on surgical success. However, associated paraganglioma resection greatly increased the incidence of cranial nerve injury ( 100%) and the need for further cor- rective surgical intervention. Knowledge of the impact these associated tumors have on the risk of surgery is nec- essary for optimal patient risk/benefit stratification. Whether the current cohort of patients represents a unique group with carotid body tumors cannot be completely an- swered but certainly suggests that bilaterality and associated cervical paraganglioma may be more common than cur- rently believed, and it does affect the surgical care of these patients.

Surgical correction in this series utilized many of the tech- niques found helpful by previous investigators. Overall, pre- operative tumor embolization has decreased the average blood loss, but the use of digital subtraction angiography makes accurate preoperative tumor size estimates invalid, and therefore direct comparisons by size to other series is difficult. Proximal and distal vascular control is attempted in all cases. The use of loupe magnification for visualization of the small feeding vessels is considered helpful. The use of bipolar electrocautery for precise dissection around the carotid artery has been very useful to help decrease blood loss.’ Furthermore, beginning the dissection on the poste- rior and lateral side of the carotid vessels often defines a clear dissection plane in cases where no other free plane appears to be present. Vascular repair was required by meth- ods common for carotid artery surgery in 29% of cases, and no cerebrovascular events were noted. Ligation of the ex- ternal carotid artery was performed if it helped in tumor removal. There have been no deaths. Using these routine techniques, the postoperative complications of stroke re- ported in the literature have decreased from 23% (prior to 1966) to approximately 5% (1966 to 1986)‘@ and are more of a problem with large tumors. Cranial nerve damage still remains a problem, being noted in 0% to 40% of cases.9m12 Cranial nerve dysfunction (18%) was significant in our ex- perience but only when an associated vagal body tumor or glomus jugulare tumor was resected. No cranial nerve def- icits were noted following resection of an isolated carotid body tumor. The cranial nerve most commonly involved was cranial nerve X, and the associated morbidity usually led to further surgical procedures such as feeding tube place- ment and vocal cord medialization.

In conclusion, carotid body tumors provide a challenge for the cliniciar,. An asymptomatic neck mass should make one consider tE e diagnosis. Angiography (arch aortogram and four-vessel) is needed for diagnosis and for optimal sur- gical planning. Contrast-enhanced CT scanning is not a substitute for angiography but may help to localize the tu- mor in relationship to surrounding structures. Preoperative tumor embolizaiion and precise surgical technique can help to minimize patient risk and the need for blood transfusions. Resection of moderate-sized isolated carotid body tumors should be associated with minimal risk of stroke, death, or cranial nerve injury. The presence of associated paragan- gliomas, howev,?r, changes the patient risk for cranial nerve injury dramatically. Familial cases are at specifically high risk for both multiple paraganglioma and bilateral carotid body tumors, and have been relatively common in this se- ries. Therefore, special effort to look for these occurrences should be part of standard care. Certainly, relatives of pa- tients with these tumors should be screened for paragan- gliomas to improve the chance of finding them when they are small and more easily resected.

REFERENCES 1. Frey CF, Karol1 RP. Management of chemodectomas. Am] Surg. 1966,1:536 2. Pryse-Davies J, Dawson IMP, Westbury G. Some morphologic, histochemical, and chemical observations on chemodectomas and the normal carotid hody, including a study of the chromaffin reac- tion and possible ganglion cell elements. Cancer. 1964; 17: 185-202. 3. ReMine WH, Weiland LH, ReMine SC;. Carotid body tumors: chemodectomas. Curr Prob Cancer. 1978;2:1-26. 4. Grufferman S, Gillman MW, Pasternak R, et al. Familial carotid body tumors: case report and epidemiologic review. Cancer. 1980;46:2116-2122. 5. Lack EE, Cubilla AL, Woodruff JM, et al. Paragangliomas of the head and neck region. Cancer. 1977;39:397-409. 6. Lees CD, Levme HL, Beven EC, et al. Tumors of the carotid body: experience with 41 operative cases. AmJ Surg. 1981;142:362- 365. 7. Rangwala AF, Sylvia LC, Becker SM. Soft tissue metastasis of a chemodectoma. A case report and review of the literature. Cancer. 1978;42:2865-2869. 8. Netterville JL, Reilly KM, Robertson D, et al. Carotid body tu- mors: a review of 30 patients with 46 tumors. Laryngoscope. 1995;105:115-126. 9. LaMuraglia GM, Fabian RL, Brewster DC, et al. The current surgical management of carotid body paragangliomas. J Vusc Surg. 1992;15:1038-1045. 10. Shamblin WR, ReMine WH, Sheps SC, et al. Carotid body tumor (chemodectoma). Clinicopathologlc analysis of 90 cases. Am J Surg. 1971;122:732-739. 11. Schuaber MK. Diagnosis and management of catecholamine secreting glomus tumors. Lqngosco@. 1984;94:1008-1015. 12. Farr HW. Carotid body tumors: a thirty year experience at Me- morial Hospital. Am J Surg. 1967;114:614-619.

THE AMERICAN JOURNAL OF SURGERY” VOLUME 172 AUGUST 1996 199