Early-onset postirradiation sarcoma of the head and neck: Report of three cases. (Original Article).
Postirradiation sarcomas of the head and neck are rare. When they do occur, most appear at least 10 years following radiation therapy. We report three cases of early-onset (1, 2, and 7 yr) postirradiation sarcoma. Physicians who care for previously irradiated patients should consider the possibility of a postirradiation sarcoma whenever they see a suspicious lesion, regardless of the amount of time that has passed since radiation therapy was administered. The original pathology should be reexamined to ensure that the original tumor was diagnosed correctly. Electron microscopy can be useful in differentiating sarcomatous-appearing epithelial lesions from true soft-tissue sarcomas, and thus can be helpful in guiding therapy.
Radiation therapy is a proven and highly effective treatment for head and neck malignancies. A rare complication of this treatment is the development of radiation-induced neoplasms, particularly sarcomas. Postirradiation sarcoma is typically a late event, occurring many years after the original treatment. Most reviews of postirradiation sarcoma report that the median interval from the completion of radiation therapy to the diagnosis of a secondary sarcoma is approximately 12 years. (1,2)
Criteria for identifying postirradiation sarcoma were first proposed by Cahan et al (4) and later modified by Arlen et al. (4) They include documentation of the presence of a sarcoma within the radiation field, histologic confirmation of the sarcoma, documentation that the affected tissue was normal prior to irradiation, and a latency period of several years' duration.
Although most postirradiation sarcomas arise more than a decade after radiotherapy, the nature of the radiation-induced transformation of cells does not rule out shorter latencies. Several reports in the literature describe cases that occurred within 3 years of radiation therapy. (2,5)
In this article, we report three unusual cases of postirradiation sarcoma of the head and neck. These cases are unusual in that their latency periods were short: 1, 2, and 7 years. We also demonstrate the utility of ultrastructural analysis in the histologic confirmation of tumors that are difficult to characterize.
Patient 1. A 61-year-old woman came to us with a complaint of hoarseness. Ten months earlier, she had undergone high-dose radiation therapy for a well-differentiated T2NOMO squamous cell carcinoma of the right true vocal fold.
We performed flexible endoscopic laryngoscopy, which demonstrated that a white, fleshy mass occupied the entire anterior half of the right true vocal fold and extended to the right false vocal fold and the anterior commissure. The mobility of the right vocal fold was impaired. Direct laryngoscopy revealed that the mass on the right vocal fold was fibrous and firm.
On initial pathologic examination, the biopsy specimen contained scar and granulation tissue but was negative for a neoplasm. A deeper biopsy was then taken, and analysis of it revealed that the mass was a high-grade leiomyo-sarcoma. Immunohistochemical staining for muscle-specific actin was positive, and stains for cytokeratin, CD68, and S-100 were negative, confirming the diagnosis (figure 1).
Because of the short l-year interval between the patient's radiation treatment and the development of the leiomyo-sarcoma at the same site, we initially questioned the first diagnosis. Therefore, we obtained the pathology of the initial tumor from the outside institution. Our own analysis confirmed that the original tumor was indeed a well differentiated squamous cell carcinoma (figure 1). The patient underwent a total laryngectomy for high-grade leiomyosarcoma.
Five months later, the patient developed a right-sided neck mass adjacent to her tracheostoma. She was treated with modified radical neck dissection, which demonstrated multiple foci of high-grade leiomyosarcoma. Eight months following surgical excision, the patient died of distant metastatic disease.
Patient 2. Hoarseness was also the complaint of a 58-year-old man who came to us for an evaluation. He had received 6,600 cGy of radiation for a T1N0M0 spindle-cell squamous carcinoma of the right vocal fold 2 years earlier. He reported that he had been persistently hoarse for approximately 1 year. During that time, he had undergone two laryngeal biopsies, both of which detected no neoplasm.
Indirect laryngoscopy at our clinic detected a 1.5-cm ulcerative mass that was centered on the anterior right vocal fold and extended into the ventricle. The mobility of the right vocal fold was diminished. Direct laryngoscopy confirmed these findings, and an analysis of a biopsy specimen identified the mass as a malignant fibrous histiocytoma.
Again, given the unusually short interval for the development of a postirradiation sarcoma, we considered the possibility that the original neoplasm had been misdiagnosed. Therefore, we obtained the pathology slides and paraffin-embedded tissue blocks from the patient's original biopsy. The slides were reviewed and new slides were cut from the tissue block for analysis under electron microscopy. Sections of the second neoplasm were also prepared for electron microscopy.
The tumors were compared by using standard histologic techniques and ultrastructural analyses. Although standard histology supported the original diagnosis, a conclusive differentiation between the two tumors could not be made (figure 2). However, electron microscopy definitively confirmed that the first neoplasm was epithelial in origin and that the second tumor was a sarcoma that had arisen in the radiation field rather than a recurrence of the original tumor (figure 2). The patient underwent total laryngectomy and was disease-free at 2.5 years of follow-up.
Patient 3. A 75-year-old woman came to us complaining of right-sided oral pan. Six years and 6 months earlier, she had been treated with 7,200 cGy of radiation following the excision of an extensive recurrence of a squamous cell carcinoma of the right facial skin. Our examination found that she had a large ulcerative lesion of the right mandibular alveolus. Analysis of a biopsy specimen showed that the lesion was an angiosarcoma.
Again, we reviewed the original pathology and confirmed that the first tumor had been a squamous cell carcinoma (figure 3). Our review of the second neoplasm confirmed that it was an angiosarcoma that had arisen in the radiation field. Immunohistochemical staining for CD34, a vascular endothelial marker, was positive (figure 3).
The patient underwent wide surgical excision, including composite resection. However, she died 1 year later of metastatic angiosarcoma.
Although it is impossible to distinguish a radiationinduced tumor from a second primary tumor, statistical excess of soft-tissue sarcomas following radiation therapy strongly suggests that there is a causal relationship between radiotherapy and new tumors. Fortunately, the cumulative incidence of postirradiation sarcomas is low--in the range of approximately 0.03 to 0.3%. (1) The most common radiation-induced soft-tissue sarcomas are malignant fibrous histiocytomas, which account for approximately 70% of cases. (6) Others that have been reported are osteosarcomas, fibrosarcomas, chondrosarcomas, and angiosarcomas. The 5-year survival rate for patients with postirradiation sarcomas is poor-less than 30%.2 When feasible, wide surgical extirpation remains the mainstay of care for these patients. Chemotherapy and further radiotherapy are reserved for unresectable cases.
Most authors have reported long latency periods prior to the development of postirradiation sarcomas, although cases that occurred within 3 years have been noted. (2 5) Mutagenesis resulting from radiation therapy could theoretically explain the development of a second neoplasm within a short interval.
Patient 1. Our first patient had undergone narrow-field radiation therapy for a T2N0M0 squamous cell carcinoma of the right true vocal fold. Our own pathologic analysis confirmed that this tumor was a well-differentiated squamous cell carcinoma. The patient had done well for approximately 1 year following radiotherapy, but then she developed recurrent hoarseness. Initially, the mass on her right vocal fold was thought to represent recurrent disease. However, pathologic examination revealed that it was a high-grade leiomyosarcoma. Immunohistochemical staining for muscle-specific actin was confirmatory.
Primary tumors of different histologic origins that arise at the same site are exceedingly uncommon. With the exception of the short period of latency, this tumor meets the modified Cahan criteria for a postirradiation sarcoma--that is, it arose within the radiation field, it was confirmed to be a sarcoma, and it was not evident prior to radiation therapy. We believe that this case represents an example of an early-onset postirradiation sarcoma.
Patient 2. The case of our second patient, who developed a laryngeal postirradiation sarcoma, is noteworthy not only because of the short latency period (2 yr), but also because of the ultrastructural characteristics of the two neoplasms. When we diagnosed the malignant fibrous histiocytoma 2 years following radiation therapy for a spindle-cell carcinoma, we questioned the original diagnosis. Therefore, blocks from both the original and the second neoplasm were pulled and recut for light and electron microscopy.
Light microscopy failed to satisfactorily differentiate between the two tumors, although the original diagnosis was favored (note the similar appearance of the two tumors on light microscopy in figure 2). Immunohistochemical staining for mono C, keratin CK, desmin, and S-100 on the second neoplasm were negative, suggesting that the tumor was not of epithelial origin. To further characterize the similarities and differences between the two tumors, electron microscopy was performed. Sections were prepared from the most sarcomatous-appearing area of the first tumor and from the most squamousappearing area of the second mass. The finding of epithelial elements, mature desmosomes, and tonofilaments in the original neoplasm secured the diagnosis of spindle-cell squamous carcinoma. The second tumor had no epithelial elements at all, which is consistent with a mesenchymal origin. Although it is possible that the sarcoma represents a metachronous primary tumor, we feel that aside from the short latency period, it too meets the criteria of Cahan and is indeed a radiation-induced neoplasm.
Patient 3. Our third case--an alveolar ridge postirradiation angiosarcoma--is histologically more typical. It also occurred within a relatively short latency period, but the original slides were reviewed to ensure that the first tumor was diagnosed correctly.
These three cases illustrate the importance of an accurate histologic diagnosis in the management of both primary and recurrent neoplasms. This is particularly important in early-appearing second primaries whose histology is different from that of the original tumor. Electron microscopy can be useful in cases where standard histologic and immunohistochemical techniques fail to secure a definitive diagnosis.
(1.) Mark RJ, Bailet JW, Poen J, et al. Postirradiation sarcoma of the head and neck. Cancer 1993;72:887-93.
(2.) Wiklund TA, Blomqvist CP, Raty J, et al. Postirradiation sarcoma. Analysis of a nationwide cancer registry material. Cancer 1991;68:524-31.
(3.) Cahan WG, Woodard HQ, Higinbotham NL, et al. Sarcoma arising in irradiated bone: Report of 11 cases. Cancer 1948;1:3-29.
(4.) Arlen M, Higinbotham NL, Huvos AG, et al. Radiation-induced sarcoma of bone. Cancer 1971;28:1087-99.
(5.) Kim JH, Chu FC, Woodard HQ, et al. Radiation-induced soft-tissue and bone sarcoma. Radiology 1978;129:501-8.
(6.) Enzinger FM, Weiss SW. Soft Tissue Tumors. 3rd ed. St. Louis: Mosby, 1995:3-4.
From the Department of Otolaryngology, University of Michigan, Ann Arbor (Dr. Johns and Dr. Teknos), the Department of Otolaryngology, Henry Ford Health System, Detroit (Dr. Concus), and the Department of Pathology, Veterans Health Administration, University of Michigan, Ann Arbor (Dr. Beals).
Reprint requests: Theodoros N. Teknos. MD, Assistant Professor, Department of Otolaryngology, University of Michigan, 1500 E. Medical Center Dr., TC 1904, Ann Arbor, MI48109. Phone: (734) 936-3172; fax: (734) 936-9625; e-mail: firstname.lastname@example.org
Originally presented at the annual meeting of the American Academy of Otolaryngology-Head and Neck Surgery; New Orleans; Sept. 25, 1999.