In 1945, Harry Rosenwasser reported a case of a 36-year-old male presenting to Mount Sinai Hospital with a growth of the left ear and a 10-year history of hearing loss (Fig. 1.2). “He had a facial paralysis and a large, red mass in the ear.” Standard roentgen exams showed “good pneumatization, but clouding of the periantral region and decalcification of the septums in the perisinal region.” On April 18, 1942, Rosenwasser performed an exploratory radical mastoidectomy and encountered “excessive bleeding” of the bone; the facial nerve was embedded in “vascular granulation tissue .” When the canal wall was taken down, the middle ear was filled with “a large purple mass, which appeared continuous with the mastoid tip.” He felt no bone on the floor of the middle ear and believed that the tumor could be involved with the jugular dome. Only the middle ear portion of the tumor was excised, leaving the hypotympanic disease behind [2].

Based on his operative findings, Rosenwasser felt that this must be some sort of aggressive malignant tumor. So while waiting for final pathology, the patient received radiation therapy. The pathology report returned showing “large tumor cells in groups, separated by dense fibrous tissue septums, which contain dilated veins. The groups of tumor cells are bordered by capillaries and form small alveolar structures. The general histologic picture, however, is that of a carotid body tumor.” Referring to Guild’s report in 1941, Rosenwasser states “The possibility of the tumor herein reported may have developed from the glomus jugularis, first named by Guild, is tentatively proposed” [2]. Dr. Rosenwasser sent printed reports of this paper to Stacey R. Guild. The two men then began a correspondence, and Guild agreed that the tumor indeed may have grown from the glomus jugularis [3].

Over the next several years, multiple case reports of glomus jugularis tumors were reported in the literature in an attempt to further elucidate their character and behavior. In 1949, Lattes and Waltner reported eight cases in New York in addition to the ten cases found in the literature. They propose a correction to the naming of the tumor. “The correct Latin diction should be “glomus jugulare,” not “glomus jugularis,” because the Latin noun “glomus” is neuter.” Within this report, we get the first clinical characteristics of glomus jugulare tumors; the tumors are slow growing, present with pulsatile tinnitus, and bleed easily. Two of the reported tumors metastasized, one to a lymph node and one to the liver. In these early reports, tumors were generally treated with a combination of radical mastoidectomy and radiation therapy (Fig. 1.3). Pathologically, they describe the tumors similar to Guild as “highly vascular, and the walls of most of the blood vessels are limited to an endothelial lining. In the vascular network, are ribbons and round nests of cells that are usually referred to as epithelioid” (Fig. 1.4) [4]. Until this point, the anatomic origin of these tumors remained unclear, and there was still speculation that glomus jugulare tumors may represent metastasis from carotid body tumors. Lattes and Waltner report that serial sections show glomus bodies in the dome of the jugular bulb, reported by Guild, but also “along the course of the tympanic nerve, in its bony canal, and in the bone of the promontorium, very near the mucosal lining of the middle ear.” They postulate that some of these tumors originate in the middle ear and argue for recognizing the entity of “paraganglia tympanicum ” (Fig. 1.5) [4].

In 1951, Weille and Lane published their experience from the Massachusetts Eye and Ear Infirmary detailing the operative difficulties of glomus tumors, including the risk of bleeding, facial paralysis, and the difficulty of identifying anatomic landmarks during the operation. They assumed that the tumor originated from the dome of the jugular bulb, “from which the most violent bleeding is expected to occur.” In order to account for this, they report leaving a “small bit of neoplasm in the floor of the hypotympanum until the tumor is widely exposed.” Then they proceed with rapid dissection and removal of all gross tumor except disease in the floor of the hypotympanum. Weille and Lane felt that compressing the lateral sinus and tying off the internal jugular vein would not control the blood flow from the inferior petrosal veins, or the arterial supply from the ascending pharyngeal artery; “therefore, we feel that the bulb must be spared” [5].

In their report, Weille and Lane relay a conversation held between the surgeons and radiation oncologists when discussing the merits of adjuvant radiation following subtotal tumor removal: “The specialist from the X-ray department there made the following answer ‘The histology of glomus tumor arising in the ear is essentially that of glomus tumors elsewhere in the body. They are notoriously radioresistant, and there is no reason to presume that those in this location should be more responsive. The literature is quite wanting in any enthusiastic statement regarding the value of X-ray treatments, although a number have been recorded. Treatment is not without hazard to the normal tissues in which the tumor lies, and I would not recommend this procedure.’”

In 1952, Winship performed a comprehensive review of the literature up to that point, including 65 cases of histologically proven glomus jugulare tumors. Clinically, most patients presented with a long history of hearing loss and a mass in the external auditory canal. In advanced cases, further symptoms of tinnitus, dizziness, pain, and paralysis of the cranial nerves occur. The average age of the patients in his review was 44 years, with a range from 17 to 80 years, and 85% were women. By this time, the pathology was classically described as “cuboidal cells with eosinophilic, finely granular cytoplasm and prominent vesicular nuclei. The nests of cells are separated by thin, vascular strands of fibrous tissue. Mitosis is rare even in the obviously malignant cast” (Fig. 1.6) [6]. In this publication, Winship reports on early treatment and outcomes. Radiation following biopsy was used to treat only six cases. Of these six, one patient died soon after treatment, but others were still living and one had even been followed for 20 years. The remaining 59 patents were treated with surgery, and most were followed for at least 4 years with an overall mortality rate of 21.5%. Based on these findings, Winship concluded that glomus jugulare tumors might be controlled for long periods of time with radiation and that early recognition was needed in order to facilitate prompt treatment and reduce patient mortality [6].

Brown’s seminal report published in 1953 provides us with the first description of the “pulsation sign ” with tumor blanching to aid in diagnosis, early descriptions of X-ray findings, as well as the first description of angiography in a glomus jugulare tumor. The “pulsation sign” is meant to diagnose a glomus tumor on otoscopy: “The patient is placed in the sitting position. By fitting the external auditory canal with an aural speculum large enough to seal off the canal, and while looking through the magnifying otoscope, if the air pressure is increased in the external auditory canal, through the otoscope, by means of compressing the attached rubber bulb, the tumor will be seen to pulsate, sometimes almost violently. As the air pressure is increased by more compression of the rubber bulb, the tumor will blanch, and the pulsation will decrease. Then, as the air pressure is reduced by less compression of the rubber bulb, the pulsation of the tumor will return, until the normal room pressure has been reached, at which time the pulsation will again subside, or return to the pretest spontaneous pulsation” [7].

In Brown’s original description of early imaging characteristics of glomus tumors, he admits that an X-ray may not be helpful in the diagnosis, as “in the earliest stages of the disease, no changes are seen in the films.” However, “as the disease progresses, the mastoid cellular structure becomes Roentgenographically cloudy…If there is any characteristic X-ray evidence of the glomus jugulare tumor, it is the picture of the mastoid and petrous cells being ‘rubbed into coalescence.’ There is not so much demineralization as is seen in acute surgical infection; there is not sclerosis that appears in long-standing chronic infection of the mastoid bone; there is not the complete obliteration of the bone as is seen in cancer of the temporal bone.”

Brown pioneered the arteriogram by injecting X-ray opaque solution into the carotid system. Unfortunately, in his experiments, diagnostic imaging studies did not demonstrate any radiologic abnormalities. In an effort to reduce operative blood loss, Brown injected sclerosing solution into the tumor in two cases; however, no changes were noted in the tumors after the injection. He also tried to ligate the external carotid artery, distal to the superior thyroid, but proximal to the ascending pharyngeal artery in one patient; however “bleeding seemed to be as severe as it was in each case where the artery was not ligated” [7]. In his surgically treated cases, Brown describes preparing the radical mastoid cavity first, then teasing the tumor from the periphery of the middle ear with an elevator, and then finally lifting the tumor out of the middle ear. Following rapid tumor extirpation, Brown controlled bleeding with Gelfoam packing . Each of his five cases required 500 cc of transfusion. Seven to 10 days after surgery, adjuvant radiation was routinely given in doses of 2500–4800 rads [7].

In 1953, Guild reports again on the anatomy of the normal jugular body in 88 ears. In this report, Guild confirms that glomus bodies are associated with the tympanic branch of the glossopharyngeal nerve and with the auricular branch of the vagus nerve and are found in the adventitia over the jugular bulb and the mucosa of the cochlear promontory (Fig. 1.7). He further confirms that the blood supply is from the ascending pharyngeal artery from the tympanic branch (Fig. 1.8). He notes that the tympanic branch “accompanies the nerve of Jacobson and in the jugular fossa often gives off a branch that accompanies the nerve of Arnold. Given these anastomoses it may not be the sole source of blood” [8].

In 1955, Shambaugh, from the University of Chicago, presented the new endaural “hypotympanotomy ” surgical approach to glomus tumors that would preserve hearing. Up until this point, radical mastoidectomy was the treatment of choice. The first two patients had significant bleeding when they encountered the tumor at the jugular bulb, so they then began ligating the internal jugular vein and tying off the external carotid artery. They then exposed the tumor via an endaural hypotympanotomy. “The hypotympanum and jugular bulb were then exposed by resecting the entire tympanic bone forming the anterior, inferior, and posterior osseous meatal wall until the ascending portion of the carotid artery in its bony canal, the facial nerve emerging from the stylomastoid foramen and the jugular bulb with the attached tumor protruding into the tympanic cavity lay fully exposed. This was done without any attempt to remove or mobilize the tumor. Only when the tumor lay fully exposed was it rapidly dissected away from the dome of the jugular bulb, using a small periosteal elevator, with constant suction for the profuse bleeding. After this the point of origin at the jugular bulb is cauterized using a topical sclerosing solution” (Fig. 1.9). Both patients have no evidence of recurrence 2 and 5 years from surgery.

In 1955, Williams reported a series of patients who received external beam radiation therapy from the Mayo Clinic [9]. The patients all had extensive jugular tumors presenting with cranial nerve deficits and received between 1080 and 4000 roentgens over 1–2 weeks. Five cases were treated with radiation alone and eight received adjuvant radiation following subtotal resection. Among those patients receiving primary radiation therapy, they reported “little evidence of regression of the tumor”; however, there was growth of the lesion in only one case. This patient initially received 1230 roentgens and was given an additional 1500 roentgens establishing tumor arrest. In the eight cases receiving radiation therapy following incomplete resection, the authors reported: “In all cases when operation alone was used, the advance of the tumor seemed to continue unabated until irradiation was used…We believe that it is fair to conclude that radiation therapy has a definite therapeutic effect on chemodectomas of the glomus jugulare. The results seem to indicate that if surgical procedures are used, they should always be combined with radiation therapy….It seems probable in the future that cases of chemodectomas may be best treated with radiation therapy alone” [9].

In 1962, Alford published a literature review of all 316 published cases in the literature, including 11 from their own experience at Baylor University in Houston. They note the “slow, incipient growth of the tumor.” In their report, the average interval of symptoms prior to diagnosis was 6 years. Again, a female preponderance is reported with 66% of cases occurring in women. Presenting symptoms were similar to those previously reported, with the most common being hearing loss and tinnitus (Fig. 1.10) [10].

Of the 316 cases, 183 included results of various imaging studies. Seventy-eight percent had abnormal findings including 36% with clouding, 37% had erosion of the petrous bone, 33% with erosion of the mastoid, 10% with erosion of the jugular foramen, and two patients had erosion of the foramen lacerum or foramen magnum. Angiography was performed in nine cases and showed the tumor in five of those cases. Seven patients also had concomitant carotid body tumors. Metastatic lesions were reported in six patients: two to the liver, two to the neck, and two to the ribs. Based on his review, Alford devised a classification scheme in an attempt to correlate the results of treatment with location and size of the tumor. He identified five stages based on symptoms, physical exam findings, and imaging results (Fig. 1.11). Based on this staging schema, Alford was able to show that there was a significant difference in outcome based on stage (Fig. 1.12) and proposed the following treatment algorithm: “For stage 0 tumors, surgical removal by tympanotomy or hypotympanotomy technique of Shambaugh can usually result in complete removal. If surgical exploration reveals a true glomus jugulare tumor, radiotherapy followed by surgery is indicated. For stage 1 and 2 lesions radical mastoidectomy followed by radiotherapy is recommended. Stage 3 lesions appear to be controlled by therapeutic doses of radiation, and we believe this to be the preferred treatment as it would be unusual for all the tumor to be removed by present surgical techniques. For stage 4 lesions, it appears that palliation by radiotherapy and neurosurgical decompression when applicable is all that is indicated” [10].