- Email: [email protected]
Multiple primary malignant tumors of the head and neck
Multiple Primary Malignant Tumors of the Head and Neck ARNOLD M. COHN, M.D.,* AND SEAN B. PEPPARD, M.D.*
Multiple primary malignant tumors associated with an index cancer in the head and neck are gaining attention as more patients survive their initial primary disease, the longevity of the population at large increases, health care delivery becomes more readily available, and sophisticated diagnostic technology is developed. Forty-four patients with multiple primary cancer in whom the index neoplasm was in the head and neck are reported. This group comprises 17 per cent of the patients with head and neck cancer treated at the Detroit General Hospital in the last 10 years. Forty-seven per cent of the cancers occurred simultaneously, 10 per cent were synchronous (discovered within six months), and 43 per cent were metachronous (discovered later than six months). Eighty-six per cent of the patients are dead; 75 per cent died within one year after the diagnosis of the secondary primary cancer. Patients with simultaneously occurring malignant tumors appeared to have a somewhat poorer prognosis than those with metachronously occurring tumors. Although the highest incidence of multiple primary malignant tumors occurs in the first year, a consistent high incidence puts the patient with a solitary neoplasm at risk well beyond the three to five years of survival usually considered as "cure." Although no factor other than coincidence has been proved to be involved in the pattern of occurrence of multiple primary neoplasms of different tissues or organs, this study does corroborate other reported data that emphasize that patients with head and neck cancer are at greater risk of developing a second primary malignant tumor in the multicentric vicinity of the original lesion, and in the esophagus and lung than the general population. A program of management and post-treatment surveillance is presented. The potential for new primary malignant neoplasms to develop in the multicentric vicinity of the original lesion and in remote organs cannot be approached with detached curiosity.
Multiple primary cancers can no longer be considered an exception. The flurry of recent attention given their incidence attests to this point. This recognition is a measure, first, of the increased longevity of the population at large; second, mare patients are surviving the initial primary malignant tumor in particular; third, there is greater availability of advanced diagnostic technology; and fourth, there is increased awareness by physicians of the problem.
The incidence of multiple primary malignant tumors has ranged from 1 to 20 per cent. In one series recording a 9 per cent incidence of multicentric oral cavity cancer, 55 per cent of the separate lesions were recognized grossly and involved contiguous tissues of adjacent organs: the lips, pharynx, larynx, and esophagus. However, many multicentric lesions could be confirmed only when biopsy of suspicious appearing mucosa was carried out and tissue was
Accepted far publication July 15, 198o. *Professor, Department of Otolaryngology, WayneState University School of Mediclne. Clinical Staff, Hm'per Hospital, Children's Hospital of Michigan, Detroit Receiving Hospital, and Veterans Administration Hospital, Detroit, Michigan. ~rAssistunt Professor, Department of Otolaryngology, Wayne State University School of Medicine. Clinical Staff, Harper Hospital, Children's Hospital of Michigan, Detroit Recelving Hospital, and Veterans Administration Hospital, Detroit,Michigan. American Journal of Otolaryngology--Volume 1, Number 5, November1980
411
American
Journa[ of
OfoJaryngology
TABLE 1. Classification of Multiple Primary Neoplasms I, Multiple primary neoplasms of multicentric origin A. The same tissue or organ B. A common contiguous tissue shared by different organs C. The same tissue in bilaterally paired organs II, Multiple primary malignant neoplasms of different tissues or organs III, Multiple primary malignant neoplasms of mulficentric origin and lesions of different tissues or organs
examined microscopically during posttreatment surveillance? The known occurrence of multiple primary cancers in unrelated organ systems suggests the importance of local and systemic examination in the initial evaluation and of serial local and systemic examination in the follow-up management. A prospective classification of multiple primary malignant neoplasms becomes helpful in evaluating the problem. The classification developed by MoerteP appears particularly useful (Table 1). However, bias enters into most studies of multiple primary malignant tumors. ~,2 Reports of hospital series are of a selected population. Close surveillance makes possible a greater frequency of cancer detection and diagnosis and greater effectiveness of follow-up and reporting. Distortion and artifact increase ff a specialized referral center manages a disproportionately large population of patients with particular disorders, such as head and neck cancer; statistics within such series may not reflect the general population. This kind of artifact is enhanced if synchronous multiple primary malignant tumors are studied as compared to metachronous multiple neoplasms; an asymptomatic second neoplasm might never be found were it not for the evaluation stimulated by the symptomatic index neoplasm. Sample size also may not be sufficient to yield meaningful data. Finally the application of national incidence rates to one geographic area may not provide accurate data for meaningful analysis or sufficient data to explain variables.
CLINICAL MATERIAL
The records of all patients with histiologically proven head and neck carcinoma admitted to the Detroit General Hospital during the 10 years from 1970 to 1979 were analyzed, Patients admitted to the Otolaryngology Service routinely underwent panendoscopy if head and neck can412
cer was suspected. During this period there were 274 patients with this diagnosis. Clinical and pathological data were examined to satisfy the criteria of Warren and Gates 3 for the diagnosis of multiple primary cancer: (1) The neoplasms must be clearly malignant and confirmed histologically. (2) Each neoplasm must be geographically distinct. If the cancers occur in the same tissue or organ or in c o m m o n contiguous tissue shared at different organs and are of the same cell type, they should be separated by normal mucous membrane. Multiple tumors may be "colliding" if they are of different cell types. (3) The possibility that the second lesion is a metastasis from the other must be excluded; this determination is the most difficult to make. Fifty-one patients met these criteria. Most of these patients were diagnosed in advanced stages of their diseases; advanced disease is another population artifact that may preclude direct comparison to other head and neck cancer series. Treatment included surgery, irradiation, or chemotherapy according to the judgment of the managing team.
RESULTS
During the period from 1970 to 1979, 274 patients were diagnosed as having a malignant tumor of the head and neck region; of these, 51 (19 per cent) developed multiple cancers. Neoplasms previously diagnosed at other hospitals may not be recorded in discharge summaries. Seven patients who developed head and neck cancer after the occurrence of initial malignant tumors in remote areas of the body were excluded from further analysis. Patients with neoplasms of the nose and nasopharynx were also excluded. Of the remaining 267 patients with an initial index primary cancer in the head and neck, 44 (17 per cent) were found to have a second primary malignant tumor simultaneously or subsequent to the initial examination (Table 2). Simultaneously occurring neoplasms were those diagnosed at the same time; synchroTABLE 2. Multiple Primary Malignant Tumors, Detroit General Hospital, 1970-1979 Patients with index primary head and neck cancer 267 Patients developing multiple primary cancer 44 (17%) (33 males, 11 females) Simultaneous 47%
Synchronous
10%
Metachronous
43%
MULTIPLE PRIMARY MALIGNANT TUMORS OF THE HEAD AND NECK
25
Vo.lu me 1 Number 5
21 20
November 1980
r-
15 r~
I0
E Z
5~ ~
~
~
4 2
0
I-6
7-12
~
15-24 25-56
57-48
5
2 49-60
m 61-72
73-84
2 >85
/ntervat in Months Figure 1. The interval between initial diagnosis and subsequent cancer diagnoses in patients with index malignant tumors in the head and neck,
nous malignant tumors were those diagnosed within six months after the initial diagnosis; metachronous cancers were those diagnosed at intervals greater than six months after diagnosis of the initial lesion. Thirty-three patients were males, and 11 were females. The mean age at the time of initial diagnosis was 56 years, with a range from 35 to 80 years. Ninety-two malignant tumors were diagnosed i n these 44 patients; 40 patients had two primary cancers, a n d four patients had three primary malignant tumors. In 21 patients (47 per cent) the multiple malignancies were diagnosed simultaneously; one patient with three primary tumors developed simultaneously occurring cancers four years after the diagnosis of the first primary malignant tumor and is included in the data on metachronously and simultaneously occurring neoplasms. In five patients (10 per cent) the neoplasms were synchronous. In 19 patients (43 per cent) the neoplasms were metachronous; the mean interval in these patients between diagnosis of the first neoplasm and diagnosis of the subsequent cancer was 40 months with a range of six months to eight years (Fig. I). Lesions were designated anatomically as defined by the American Joint Committee on Can-
cer Staging and End Results in 1978. Initial primary lesions occurred in the oral cavity in eight patients (18 per cent), the oropharynx in 21 patients (48 per cent), the hypopharynx in four patients (9 per cent), and the larynx in 11 patients (25 per cent), (Table 3). The site of the second primary cancer was the oral cavity in four patients (8 per cent), the oropharynx in three patients (6 per cent), the hypopharynx in three (6 per cent), and the larynx in six patients (13 per cent), but more commonly the esophagus (12 patients, 25 per cent), the lung (10 patients, 21 per cent), and miscellaneous remote sites (10 other patients, 21 per cent) were involved (Table 4). hi the 12 patients in whom the esophagus was the second site, the lesion was discovered simultaneously in six, synchronously in two, and after more than six months in four patients. Of 10 patients with second primary tumors in the lung, diagnoses were made simultaneously in two, synchronously in one, and metachronously in seven. In five of 11 patients with initial laryngeal carcinoma the site of the second primary cancer was the lung. At the time of analysis six patients (14 per cent) were alive with a minimum of two years' TABLE 4. Regional Site of Second Primary Cancer NUMBER
TABLE 3. Regional Site of Index Primary Cancer NUMBER (N = 44) Oral cavity Oropharynx Hypopharynx Larynx
ARNOLD
M. C O H N
PERCENTAGE
8
21 4 11
AND
S E A N B. P E P P A R D
18 48 9 25
(N = 4S}
Oral cavity Oropharyn.x Hypopharynx Larynx Esophagus Lung Remote tissues
4 3 3 6 12 I0
10
PERCENTAGE
8 6 6 13 25 21 21
413
American Journal
TABLE 5. Multiple Primary Index Head and Neck Carcinoma, Cumulative Mortality Rate No.
of
PATIENTS
6 MONTHS
1 YEAR
YEARS
>2 YEARS
2
Otolaryngology All patients
44
64% 28/44
75% 33144
82% 36/44
86% 38t44
Simultaneous
21
76% 16/21
81% 17121
81% 17121
81% 17121
Syrmh.ronous
5
80%
100%
100%
100%
415
515
5/5
515
44~
61% 11118
77% 14118
89% 16118
Motachronous
18
8/18
post-treatment surveillance; 38 patients (86 per cent) were dead. Thirty-three patients (75 per cent) died within one year after diagnosis of the second cancer; three others had died by two years and two patients died after three years. A somewhat poorer prognosis exists for patients with simultaneously occurring neoplasms as compared to those with metachronously occurring neoplasms. Seventy-six per cent (16 of 21) of the patients with simultaneous neoplasms died within six months and 81 per cent were dead by one year; four patients (19 per cent) were alive after two years (Table 5). Forty-four per cent of patients with metachronous neoplasms were dead at six months and 61 per cent succumbed by one year; two patients (11 per cent) were alive after t w o years.
DISCUSSION
Despite advanced technology and improved diagnostic and therapeutic techniques, the etiology of cancer remains obscure. Although there is a greater possibility of development of a second malignant tumor in patients cured of an initial malignant tumor than that expected in the general population, still less is known regarding the reason for this phenomenon. Multiple primary malignant tumors fall into two general categories: those involving the same tissue in the same or contiguous organs and multiple tumors involving unrelated organ systems; this categorization suggests differing potential etiologic and pathogenic mechanisms. It does appear that a combination of genetic, immunologic, and environmental carcinogenic factors, such as smoking, alcohol, and viruses, have an effect on this risk. Moerte] 4 suggests that carcinogenesis results from specific intrinsic and extrinsic carcinogenic stimuli acting o n a susceptible tissue for a sufficient period of time to initiate irreversible
414
anaplastic changes. He further suggests t h a t carcinoma is not a unicellular disease but a disease of an entire tissue. This concept of "field can, cerization" is not limited to a localized area. '~In many cases, after an initial carcinoma has developed where the stimulus has been maximal, further evolution of the carcinogenic p r o c e s s in the surrounding tissue contributes to l o c a l extension or multiple lesions, w h i c h can appear simultaneously or at various times after a n initial lesion. Thus, local extension, multicentri. city, or both m a y be the result of either expansion and destrucUon from pre-existent cancer cells or further evolution and progressive change in a common tissue e n d o w e d w i t h common susceptibility exposed to a c o m m o n carcinogen. Although elimination of the carcinogen may lessen the risk of another malignant tumor, cellular change rendering the tissue vulnerable t o neoplastic induction m a y not be eliminated completely after "cure" of the first malignant neoplasm. What also remains unclear is the role o f immunologic incompetence in this equation. Studies have produced conflicting conclusions as to whether surviving one cancer m a y p r o d u c e relative cellular immunity to the d e v e l o p m e n t of another cancer in another tissue, or w h e t h e r the risk is actually increased for remote tissues owing to immunologic incompetence. 6"r A n e r g y to skin recall antigens, depression of in vitro lymphocyte stimulation b y phytohemagglutinene, and depression of absolute T cell l e v e l s are parameters usually cited. What is clear is that the potential for new primary malignant neoplasms to develop i n both the multicentric vicinity of the original lesion and in remote organs cannot be a p p r o a c h e d with detached curiosity. An incidence of 27 p e r cent for multiple primary malignant tumors i n patients with index tumors in the head and n e c k is not unusual. Several recent sets of data record
MULTIPLE PRIMARY MALIGNANT TUMORS OF THE HEAD AND NECK
incidences of 6.5 to 20 per cent. s-L~ The risk of developing a new carcinoma approximates the risk of developing regional or distant metastases from a solitary neoplasm. Several authors have also surmised that these figures may actually be lowY u Patients with solitary lesions in the head and neck are often lost to follow-up by the original physician; incidence figures are derived from patients surviving their index tumor and do not consider person-years of risk of developing additional lesions among those dying from an initial cancer. There is also frequent indecision regarding the interpretation of recurrence versus a new primary lesion. Our series, as well as other recently reported data, demonstrates that multiple primary malignant tumors develop most commonly within the first two years after treatment and, among this group, most frequently within the first year. In our series 47 per cent were discovered simultaneously and 60 per cent within one year of the index neoplasm diagnosis. These findings agree with those of Moertel, 'l Vrabec, u and others. 9"~ The implication of these data is that simultaneously occurring cancers must be discovered at the initial patient evaluation. This interference is reinforced if cumulative mortality is examined. Although our series may be biased with patients presenting with advanced disease, 75 per cent of our patients died within the first year after diagnosis of the second primary malignant tumor. Odette and his colleagues ~a reported that 50 per cent of their patients with multiple primary cancer died within one year after the diagnosis of the second malignant tumor. Furthermore, a somewhat poorer prognosis is seen in patients diagnosed simultaneously as compared to those diagnosed metachronously, an observation also made by Fee and his associates ~'~ in a report discussing carcinoma of the palate in particular. Although the highest incidence of multiple primary malignancy occurs in the first year and the incidence tapers in the second year and stabilizes in the third year after treatment, a consistently high incidence puts the patient with a solitary neoplasm at risk well beyond the three
to five years usually considered for "cure," This finding is supported by other studies. 4, 9, ,1, i~, is Indeed Kogelnik, Fletcher, and Iesse 17 studied 1163 patients with solitary head and neck cancer who developed neither recurrent cancer nor another cancer within five years after the initial treatment; they observed a new cancer of the head and neck in 11 per cent of the patients after this five year interval and in 15 per cent if the lung and esophagus were also included. Thus, an intensive search for additional primary malignant tumors must be made not only at the time of initial diagnosis of an index cancer, but throughout the post-treatment surveillance. The majority of our patients died of the second malignant tumor. This observation is most likely exceptional in that other incurrent diseases did not claim some of them. However, Brown ~6 reports that 71 per cent of his patients with cancer in the larynx and multiple primary malignant tumors also died of the second cancer. Early death is related in part to the extent of disease and also to the indirect effects of t h e patient's pretreatment general health and metabolic status and the indirect effects of treatment in maintaining a positive nutritional, metabolic, and immunologic status for tissue repair a n d defense. Increasing attention to these needs m a y alter the present dismal outlook. Although no factor other than coincidence has been proved to be involved in the pattern of occurrence of multiple primary tumors of different tissues or organs, there can be little doubt that multiple cancers of the same tissue occur more frequently than coincidence allows?' 4 Extending the concept of multieentricity to contiguous stratified squamous cell epithelium exposed to common carcinogenic stimuli, such as that of the oral cavity, pharynx, lung, and esophagus, further exaggerates this increased incidence. Although our data support the presence of an increased risk o f developing a second primary cancer, there does not appear to be any particular site association between the areas of the two primary tumors as has been suggested by others.~8 Our data parallel several sets of data in this regard. 9"~t,~9 Table 6 indicates t h e regional
TABLE 6. Regional Site of Second Primary WETCHERT AND SlffUMRICK~ Oral cavity, p h a r y n x Lung Esophagus Remote tissues
61 22 7 1
ARNOLD M. COHN AND SEAN B. PEPPARD
BROWNS O N ET AL, 19 40 10 5 45
Cancer
Volume 1 Number 5 November 1980
IPercentage)
WEAWER1~
VRABEC ~1
PRESENT SERIES
62 25 13 --
28 28 14 30
33 21 25 21
415
American Journal o~C Otolaryngology
416
sites of secondary cancer abstracted from data in several series. In surveying these results it is apparent that witl~ an index neoplasm in the head and neck, a second primary malignant tumor will arise most commonly anywhere in the same tissue and in the multicentric zone of contiguous tissue, but remote primary malignant tumors cannot be excluded in terms of site or interval. Several observations should be further emphasized. Our series, as those of others, emphasizes the frequency of synchronous and metachronous development of a second primary carcinoma in the esophagus among patients with head and neck cancer2' ~0,1~.e0.2~ In two series, 1 per cent of the patients with head and neck cancer developed carcinoma of the esophagus, which was estimated as nine times the risk to the general population)5, 21 The incidence was 5 per cent in our series. The fact that patients with head and neck cancer have dysphagia following surgery or irradiation may delay the clinical suspicion of a primary carcinoma of the esophagus. Indeed 70 to 80 per cent of the patients developing a second primary malignant tumor in the esophagus may be expected to be metachronous, with an average interval of occurrence of five years) 5. ~-~ Previous negative radiographic studies do not preclude development of a carcinoma. Indeed Vrabec u reports that in seven of 11 patients with carcinoma of the esophagus found at endoscopy pro-endoscopic upper gastrointestinal x-ray views were normal. Thus, although the greatest incidence of the second primary malignant tumor may be within the first two to three years, experience with patients surviving three years and longer supports the wisdom of periodic life-long re-examination. Three and five year "cure" rates should not make us feel secure. Awareness of an increased association between neoplasms of the head and neck and those of the esophagus should lead to detection of small neoplasms that are potentially curable. '~ Periodic surveillance with radiography and endoscopy should allow discovery of early neoplasms as well as specific dysplastic areas of preinvasive potential for special follow-up. However, if these lesions occur simultaneously, alternative modalities of treatment may need to be considered. It remains to be seen whether discovery of asymptomatic lesions in the esophagus will be associated with increased survival. A similar trend may be cited for carcinoma of the larynx and the development of a second primary malignant tumor in the lung. Lung cancer has been reported in 10 to 15 per cent of
all patients with laryngeal cancer, usually following the laryngeal lesion."" From the opposite perspective, the larynx has been reported to be the site of second primary cancer in 20 per cent of the patients with cancer of the lung, of which 30 per cent were synchronous. 2~.24 These data suggest a greater incidence than expected in a random selection of individual cancer patients of comparable age. Two recent studies, however, conclude that metachronous development of carcinoma cannot be attributed to previous irradiation utilized in the treatment of an initial malignant tumor of the m u c o u s membranes within the multicentric vicinity, t". ~7 Twenty-one per cent of our series of patients with multiple primary head and neck malignant tumors developed carcinoma of the lung; Vrabec 11 reported an incidence of 28 per cent, furthor noting that when the index neoplasm was in the larynx, the lung was the most c o m m o n site of the second lesion. The difficulty sometimes is to establish w h e t h e r one neoplasm is a metastasis from the other, or both are primary cancers. There are radiologic criteria for distinguishing a primary lung neoplasm from a pulmonary metastasis, but it is not always easy. The addition of polytomography and computerized tomography has also proven helpful, but they are not always conclusive. If the lung cancer is solitary, of the same cell type as the laryngeal cancer, but amenable to resection, the question of metastasis may have little practical importance with five year survival rates approximating the 30 per cent observed for primary lung lesions. 22 The important consideration is that in a patient in w h o m the laryngeal lesion is the initial lesion, the signs and symptoms of the lung cancer m a y be masked as a complication of the tracheostomy or treatment, i.e., persistent cough, hemoptysis, or pulmonm7 infiltrate assumed to be the result of tracheitis. A program of post-treatment chest radiography is also needed in our protocols. Furthermore, five of nine of Vrabec's patients had pro-endoscopic chest radiographs reported as normal." Thus bronchoscopy as well as esophagoscopy should be applied in the routine post-treatment followup.
It may be said that routine endoscopy in post-treatment surveillance is prohibitive because of cost-yield dividends; also the number of patients cured of lung cancer is small. However, the potential benefits from treatment of early lesions of the esophagus and lung are still under investigation and indeed show encouraging results. ~0Furthermore it is difficult for us to place a
MULTIPLE PRIMARY MALIGNANT TUMORS OF THE H E A D A N D NECK
monetary value among the criteria for diagnostic evaluation of potential lethal disease in high risk patients; this consideration appears to be more a social than a medical one. In the management of patients with simultaneous multiple primary malignant tumors, if both neoplasms are considered curable, the treatment is carried out as if each neoplasm were a solitary primary malignant tumor. If one neoplasm is in the head and neck and the other is remote, therapy is directed toward the more lethal of the neoplasms with priority of treatment established by a team approach. Consideration is given to the cellular diagnosis and the neoplasm's location, extent, and potential behavior in evaluating the prognosis for control and the choice of treatment. If the remote malignant neoplasm is clearly uncontrollable, palliative treatment is given for the head and neck tumor, Early recognition of a second primary cancer is as important as the discovery of metastases. New primary tumors should be discovered before they become symptomatic because of the factors cited. We now recommend that routine chest radiography be carried out at three month intervals, barium esophagograms carried out every six months, and routine endoscopy performed annually for the first three years after treatment in addition to the usual surveillance given cancer patients. After three years this schedule may be adjusted perhaps to an annual basis, but we must accept the observation that subsequent primary carcinoma continues to be a risk well beyond the traditional five year survival period. Any minor change in symptoms must be approached aggressively and not attriubted to the consequences of treatment.
References 1. Moertel, C, G,: Multiple primary neoplasms. Cancer, 40:1786-1792, 1977. 2. Berg, J. W., and Schottenfeld, D.: Multiple primary cancers at Memorial Hospital, 1949-1962. Cancer, 40:1801-1805, 1977. 3, Warren, S., and Gates, O.: Multiple malignant tumors: a survey of literature and statistical study. Am. f. Cancer, 51:1358-1414, 1932. 4. Moartel, C. G.: Incidence and significance of multiple primary malignant neoplasms. Ann. N.Y. Acad. Sci., 114'.886-895, 1964.
ARNOLD M. COHN AND SEAN B. PEPPARD
5. Slaughter, D. T.: Multicentric origin of intra-oral carcinoma. Surgery, 20:133-146, 1946. 6. Dellon, A. L., et el,: Multiple primary malignant neoplasms. Arch, Surg., 110:156-160, 1975. 7. Bartal, A., ot el.: Cellular immunity in patients with laryngeal cancer developing additional primary malignant tumors. Ann. Otol. Rhinol. Laryngol., 81:311315, 1979. 8. Gluckman, J. L.: Synchronous multiple primary lesions of the upper aerodigestive system. Arch. Otolaryngol., 105:597-598, 1979. 9. Weichert, K. A., and Shumrick, D.: Multiple malignancies in patients with primary carcinomas of the head and neck. Laryngoscope, 89:988-989, 1979. 10. Weaver, A., et al.: Triple endoscopy', a neglected essential in head and neck cancer. Surgery, 86:493-496, 1979. 11. Vrabeck, D. T.: Multiple primary malignancies of the upper aerodigestive tract. Ann. Otol. Rhinol. Lmsmgel., 88:846-854, 1979. 12. Hadju, J. I., and Hadju, E. O.: Multiple primary malignant tumors. J. Am. Gartatr. Soc., 16:16-26, 1968. 13. Odette, J., Szymanowski, T., and Nichols, R. D.'. Multiple head and neck malignancies. Trans. Am. Acad, Ophthalmol. Otolaryngol., 84:806-813, 1977. 14. Fee, W. E., et al.: Squamous cell carcinoma of the soft palate. Arch, Otolaryngol., 105:710-718, 1979. 15. Thompson, W. N, et al.: Synchronous and metachronous squamous cell carcinomas of the head and neck and esophagus, Gastrointest. Radiol., 3:123-127, 1978. 16. Brown, M.: Second primaries in cases of cancer of the larynx. J, LarynBol., 92:991-996, 1978. 17. Kogelnik, H. D., Fletcher, G. H., and Jesse, R. H,: Clinical course of patients with squamous cell carcinoma of the upper respiratory and digestive t~act with no evidence of disease five years after initial treatment. Radiology, 115:423-427, 1975. 18. Newall, G, R., and Kremantz, E. T.: Multiple malignant neoplasms in the Charity Hospital of Louisiana Tumor Registry. Cancer, 40:1812-1820, 1977. 19, Brownson, R. J., et el.: Simultaneous malignant tumors in patients with head and neck cancer. Arch. Otolaryngol., 97:347-349, 1973. 20. Cahan, W. B., ot al.: Separate primary carcinomas of the esophagus and head and neck region in the same patient, Cancer, 37:85-89, 1976, 21. Goldstein, H. M., and Zornoza, J.: Association of squamous cell carcinomas of the head and r~eck with cancer of the esophagus. Am. J. Roentgenol,, 131:791794, 1978. 22. Sherman, J. O., Staley, J., end Shields, T. W.: Double primary tumors of the laryux and lung. Arch. Surg., 94:550-558, 1967. 23. Campbell, L. D., Jr., and Watney, A. L.: Multiple primary malignant neoplasms. Arch. Surg., 99:401-405, 1969. 24. Cahan, W. B., and Moatemayor, P. B.: Cancer of the larynx and lung in the same patient: a report of sixty patients. Surgery, 44:209-230, 1962.
Volume 1 Number 5 November 1980
Department of Otolaryngology, 5E-UCB Way21e State University School of Medicine 540 E. Canfield Avenue Detroit, Michigan 48201 (Dr. Cohn]
417
Multiple primary malignant tumors associated with an index cancer in the head and neck are gaining attention as more patients survive their initial primary disease, the longevity of the population at large increases, health care delivery becomes more readily available, and sophisticated diagnostic technology is developed. Forty-four patients with multiple primary cancer in whom the index neoplasm was in the head and neck are reported. This group comprises 17 per cent of the patients with head and neck cancer treated at the Detroit General Hospital in the last 10 years. Forty-seven per cent of the cancers occurred simultaneously, 10 per cent were synchronous (discovered within six months), and 43 per cent were metachronous (discovered later than six months). Eighty-six per cent of the patients are dead; 75 per cent died within one year after the diagnosis of the secondary primary cancer. Patients with simultaneously occurring malignant tumors appeared to have a somewhat poorer prognosis than those with metachronously occurring tumors. Although the highest incidence of multiple primary malignant tumors occurs in the first year, a consistent high incidence puts the patient with a solitary neoplasm at risk well beyond the three to five years of survival usually considered as "cure." Although no factor other than coincidence has been proved to be involved in the pattern of occurrence of multiple primary neoplasms of different tissues or organs, this study does corroborate other reported data that emphasize that patients with head and neck cancer are at greater risk of developing a second primary malignant tumor in the multicentric vicinity of the original lesion, and in the esophagus and lung than the general population. A program of management and post-treatment surveillance is presented. The potential for new primary malignant neoplasms to develop in the multicentric vicinity of the original lesion and in remote organs cannot be approached with detached curiosity.
Multiple primary cancers can no longer be considered an exception. The flurry of recent attention given their incidence attests to this point. This recognition is a measure, first, of the increased longevity of the population at large; second, mare patients are surviving the initial primary malignant tumor in particular; third, there is greater availability of advanced diagnostic technology; and fourth, there is increased awareness by physicians of the problem.
The incidence of multiple primary malignant tumors has ranged from 1 to 20 per cent. In one series recording a 9 per cent incidence of multicentric oral cavity cancer, 55 per cent of the separate lesions were recognized grossly and involved contiguous tissues of adjacent organs: the lips, pharynx, larynx, and esophagus. However, many multicentric lesions could be confirmed only when biopsy of suspicious appearing mucosa was carried out and tissue was
Accepted far publication July 15, 198o. *Professor, Department of Otolaryngology, WayneState University School of Mediclne. Clinical Staff, Hm'per Hospital, Children's Hospital of Michigan, Detroit Receiving Hospital, and Veterans Administration Hospital, Detroit, Michigan. ~rAssistunt Professor, Department of Otolaryngology, Wayne State University School of Medicine. Clinical Staff, Harper Hospital, Children's Hospital of Michigan, Detroit Recelving Hospital, and Veterans Administration Hospital, Detroit,Michigan. American Journal of Otolaryngology--Volume 1, Number 5, November1980
411
American
Journa[ of
OfoJaryngology
TABLE 1. Classification of Multiple Primary Neoplasms I, Multiple primary neoplasms of multicentric origin A. The same tissue or organ B. A common contiguous tissue shared by different organs C. The same tissue in bilaterally paired organs II, Multiple primary malignant neoplasms of different tissues or organs III, Multiple primary malignant neoplasms of mulficentric origin and lesions of different tissues or organs
examined microscopically during posttreatment surveillance? The known occurrence of multiple primary cancers in unrelated organ systems suggests the importance of local and systemic examination in the initial evaluation and of serial local and systemic examination in the follow-up management. A prospective classification of multiple primary malignant neoplasms becomes helpful in evaluating the problem. The classification developed by MoerteP appears particularly useful (Table 1). However, bias enters into most studies of multiple primary malignant tumors. ~,2 Reports of hospital series are of a selected population. Close surveillance makes possible a greater frequency of cancer detection and diagnosis and greater effectiveness of follow-up and reporting. Distortion and artifact increase ff a specialized referral center manages a disproportionately large population of patients with particular disorders, such as head and neck cancer; statistics within such series may not reflect the general population. This kind of artifact is enhanced if synchronous multiple primary malignant tumors are studied as compared to metachronous multiple neoplasms; an asymptomatic second neoplasm might never be found were it not for the evaluation stimulated by the symptomatic index neoplasm. Sample size also may not be sufficient to yield meaningful data. Finally the application of national incidence rates to one geographic area may not provide accurate data for meaningful analysis or sufficient data to explain variables.
CLINICAL MATERIAL
The records of all patients with histiologically proven head and neck carcinoma admitted to the Detroit General Hospital during the 10 years from 1970 to 1979 were analyzed, Patients admitted to the Otolaryngology Service routinely underwent panendoscopy if head and neck can412
cer was suspected. During this period there were 274 patients with this diagnosis. Clinical and pathological data were examined to satisfy the criteria of Warren and Gates 3 for the diagnosis of multiple primary cancer: (1) The neoplasms must be clearly malignant and confirmed histologically. (2) Each neoplasm must be geographically distinct. If the cancers occur in the same tissue or organ or in c o m m o n contiguous tissue shared at different organs and are of the same cell type, they should be separated by normal mucous membrane. Multiple tumors may be "colliding" if they are of different cell types. (3) The possibility that the second lesion is a metastasis from the other must be excluded; this determination is the most difficult to make. Fifty-one patients met these criteria. Most of these patients were diagnosed in advanced stages of their diseases; advanced disease is another population artifact that may preclude direct comparison to other head and neck cancer series. Treatment included surgery, irradiation, or chemotherapy according to the judgment of the managing team.
RESULTS
During the period from 1970 to 1979, 274 patients were diagnosed as having a malignant tumor of the head and neck region; of these, 51 (19 per cent) developed multiple cancers. Neoplasms previously diagnosed at other hospitals may not be recorded in discharge summaries. Seven patients who developed head and neck cancer after the occurrence of initial malignant tumors in remote areas of the body were excluded from further analysis. Patients with neoplasms of the nose and nasopharynx were also excluded. Of the remaining 267 patients with an initial index primary cancer in the head and neck, 44 (17 per cent) were found to have a second primary malignant tumor simultaneously or subsequent to the initial examination (Table 2). Simultaneously occurring neoplasms were those diagnosed at the same time; synchroTABLE 2. Multiple Primary Malignant Tumors, Detroit General Hospital, 1970-1979 Patients with index primary head and neck cancer 267 Patients developing multiple primary cancer 44 (17%) (33 males, 11 females) Simultaneous 47%
Synchronous
10%
Metachronous
43%
MULTIPLE PRIMARY MALIGNANT TUMORS OF THE HEAD AND NECK
25
Vo.lu me 1 Number 5
21 20
November 1980
r-
15 r~
I0
E Z
5~ ~
~
~
4 2
0
I-6
7-12
~
15-24 25-56
57-48
5
2 49-60
m 61-72
73-84
2 >85
/ntervat in Months Figure 1. The interval between initial diagnosis and subsequent cancer diagnoses in patients with index malignant tumors in the head and neck,
nous malignant tumors were those diagnosed within six months after the initial diagnosis; metachronous cancers were those diagnosed at intervals greater than six months after diagnosis of the initial lesion. Thirty-three patients were males, and 11 were females. The mean age at the time of initial diagnosis was 56 years, with a range from 35 to 80 years. Ninety-two malignant tumors were diagnosed i n these 44 patients; 40 patients had two primary cancers, a n d four patients had three primary malignant tumors. In 21 patients (47 per cent) the multiple malignancies were diagnosed simultaneously; one patient with three primary tumors developed simultaneously occurring cancers four years after the diagnosis of the first primary malignant tumor and is included in the data on metachronously and simultaneously occurring neoplasms. In five patients (10 per cent) the neoplasms were synchronous. In 19 patients (43 per cent) the neoplasms were metachronous; the mean interval in these patients between diagnosis of the first neoplasm and diagnosis of the subsequent cancer was 40 months with a range of six months to eight years (Fig. I). Lesions were designated anatomically as defined by the American Joint Committee on Can-
cer Staging and End Results in 1978. Initial primary lesions occurred in the oral cavity in eight patients (18 per cent), the oropharynx in 21 patients (48 per cent), the hypopharynx in four patients (9 per cent), and the larynx in 11 patients (25 per cent), (Table 3). The site of the second primary cancer was the oral cavity in four patients (8 per cent), the oropharynx in three patients (6 per cent), the hypopharynx in three (6 per cent), and the larynx in six patients (13 per cent), but more commonly the esophagus (12 patients, 25 per cent), the lung (10 patients, 21 per cent), and miscellaneous remote sites (10 other patients, 21 per cent) were involved (Table 4). hi the 12 patients in whom the esophagus was the second site, the lesion was discovered simultaneously in six, synchronously in two, and after more than six months in four patients. Of 10 patients with second primary tumors in the lung, diagnoses were made simultaneously in two, synchronously in one, and metachronously in seven. In five of 11 patients with initial laryngeal carcinoma the site of the second primary cancer was the lung. At the time of analysis six patients (14 per cent) were alive with a minimum of two years' TABLE 4. Regional Site of Second Primary Cancer NUMBER
TABLE 3. Regional Site of Index Primary Cancer NUMBER (N = 44) Oral cavity Oropharynx Hypopharynx Larynx
ARNOLD
M. C O H N
PERCENTAGE
8
21 4 11
AND
S E A N B. P E P P A R D
18 48 9 25
(N = 4S}
Oral cavity Oropharyn.x Hypopharynx Larynx Esophagus Lung Remote tissues
4 3 3 6 12 I0
10
PERCENTAGE
8 6 6 13 25 21 21
413
American Journal
TABLE 5. Multiple Primary Index Head and Neck Carcinoma, Cumulative Mortality Rate No.
of
PATIENTS
6 MONTHS
1 YEAR
YEARS
>2 YEARS
2
Otolaryngology All patients
44
64% 28/44
75% 33144
82% 36/44
86% 38t44
Simultaneous
21
76% 16/21
81% 17121
81% 17121
81% 17121
Syrmh.ronous
5
80%
100%
100%
100%
415
515
5/5
515
44~
61% 11118
77% 14118
89% 16118
Motachronous
18
8/18
post-treatment surveillance; 38 patients (86 per cent) were dead. Thirty-three patients (75 per cent) died within one year after diagnosis of the second cancer; three others had died by two years and two patients died after three years. A somewhat poorer prognosis exists for patients with simultaneously occurring neoplasms as compared to those with metachronously occurring neoplasms. Seventy-six per cent (16 of 21) of the patients with simultaneous neoplasms died within six months and 81 per cent were dead by one year; four patients (19 per cent) were alive after two years (Table 5). Forty-four per cent of patients with metachronous neoplasms were dead at six months and 61 per cent succumbed by one year; two patients (11 per cent) were alive after t w o years.
DISCUSSION
Despite advanced technology and improved diagnostic and therapeutic techniques, the etiology of cancer remains obscure. Although there is a greater possibility of development of a second malignant tumor in patients cured of an initial malignant tumor than that expected in the general population, still less is known regarding the reason for this phenomenon. Multiple primary malignant tumors fall into two general categories: those involving the same tissue in the same or contiguous organs and multiple tumors involving unrelated organ systems; this categorization suggests differing potential etiologic and pathogenic mechanisms. It does appear that a combination of genetic, immunologic, and environmental carcinogenic factors, such as smoking, alcohol, and viruses, have an effect on this risk. Moerte] 4 suggests that carcinogenesis results from specific intrinsic and extrinsic carcinogenic stimuli acting o n a susceptible tissue for a sufficient period of time to initiate irreversible
414
anaplastic changes. He further suggests t h a t carcinoma is not a unicellular disease but a disease of an entire tissue. This concept of "field can, cerization" is not limited to a localized area. '~In many cases, after an initial carcinoma has developed where the stimulus has been maximal, further evolution of the carcinogenic p r o c e s s in the surrounding tissue contributes to l o c a l extension or multiple lesions, w h i c h can appear simultaneously or at various times after a n initial lesion. Thus, local extension, multicentri. city, or both m a y be the result of either expansion and destrucUon from pre-existent cancer cells or further evolution and progressive change in a common tissue e n d o w e d w i t h common susceptibility exposed to a c o m m o n carcinogen. Although elimination of the carcinogen may lessen the risk of another malignant tumor, cellular change rendering the tissue vulnerable t o neoplastic induction m a y not be eliminated completely after "cure" of the first malignant neoplasm. What also remains unclear is the role o f immunologic incompetence in this equation. Studies have produced conflicting conclusions as to whether surviving one cancer m a y p r o d u c e relative cellular immunity to the d e v e l o p m e n t of another cancer in another tissue, or w h e t h e r the risk is actually increased for remote tissues owing to immunologic incompetence. 6"r A n e r g y to skin recall antigens, depression of in vitro lymphocyte stimulation b y phytohemagglutinene, and depression of absolute T cell l e v e l s are parameters usually cited. What is clear is that the potential for new primary malignant neoplasms to develop i n both the multicentric vicinity of the original lesion and in remote organs cannot be a p p r o a c h e d with detached curiosity. An incidence of 27 p e r cent for multiple primary malignant tumors i n patients with index tumors in the head and n e c k is not unusual. Several recent sets of data record
MULTIPLE PRIMARY MALIGNANT TUMORS OF THE HEAD AND NECK
incidences of 6.5 to 20 per cent. s-L~ The risk of developing a new carcinoma approximates the risk of developing regional or distant metastases from a solitary neoplasm. Several authors have also surmised that these figures may actually be lowY u Patients with solitary lesions in the head and neck are often lost to follow-up by the original physician; incidence figures are derived from patients surviving their index tumor and do not consider person-years of risk of developing additional lesions among those dying from an initial cancer. There is also frequent indecision regarding the interpretation of recurrence versus a new primary lesion. Our series, as well as other recently reported data, demonstrates that multiple primary malignant tumors develop most commonly within the first two years after treatment and, among this group, most frequently within the first year. In our series 47 per cent were discovered simultaneously and 60 per cent within one year of the index neoplasm diagnosis. These findings agree with those of Moertel, 'l Vrabec, u and others. 9"~ The implication of these data is that simultaneously occurring cancers must be discovered at the initial patient evaluation. This interference is reinforced if cumulative mortality is examined. Although our series may be biased with patients presenting with advanced disease, 75 per cent of our patients died within the first year after diagnosis of the second primary malignant tumor. Odette and his colleagues ~a reported that 50 per cent of their patients with multiple primary cancer died within one year after the diagnosis of the second malignant tumor. Furthermore, a somewhat poorer prognosis is seen in patients diagnosed simultaneously as compared to those diagnosed metachronously, an observation also made by Fee and his associates ~'~ in a report discussing carcinoma of the palate in particular. Although the highest incidence of multiple primary malignancy occurs in the first year and the incidence tapers in the second year and stabilizes in the third year after treatment, a consistently high incidence puts the patient with a solitary neoplasm at risk well beyond the three
to five years usually considered for "cure," This finding is supported by other studies. 4, 9, ,1, i~, is Indeed Kogelnik, Fletcher, and Iesse 17 studied 1163 patients with solitary head and neck cancer who developed neither recurrent cancer nor another cancer within five years after the initial treatment; they observed a new cancer of the head and neck in 11 per cent of the patients after this five year interval and in 15 per cent if the lung and esophagus were also included. Thus, an intensive search for additional primary malignant tumors must be made not only at the time of initial diagnosis of an index cancer, but throughout the post-treatment surveillance. The majority of our patients died of the second malignant tumor. This observation is most likely exceptional in that other incurrent diseases did not claim some of them. However, Brown ~6 reports that 71 per cent of his patients with cancer in the larynx and multiple primary malignant tumors also died of the second cancer. Early death is related in part to the extent of disease and also to the indirect effects of t h e patient's pretreatment general health and metabolic status and the indirect effects of treatment in maintaining a positive nutritional, metabolic, and immunologic status for tissue repair a n d defense. Increasing attention to these needs m a y alter the present dismal outlook. Although no factor other than coincidence has been proved to be involved in the pattern of occurrence of multiple primary tumors of different tissues or organs, there can be little doubt that multiple cancers of the same tissue occur more frequently than coincidence allows?' 4 Extending the concept of multieentricity to contiguous stratified squamous cell epithelium exposed to common carcinogenic stimuli, such as that of the oral cavity, pharynx, lung, and esophagus, further exaggerates this increased incidence. Although our data support the presence of an increased risk o f developing a second primary cancer, there does not appear to be any particular site association between the areas of the two primary tumors as has been suggested by others.~8 Our data parallel several sets of data in this regard. 9"~t,~9 Table 6 indicates t h e regional
TABLE 6. Regional Site of Second Primary WETCHERT AND SlffUMRICK~ Oral cavity, p h a r y n x Lung Esophagus Remote tissues
61 22 7 1
ARNOLD M. COHN AND SEAN B. PEPPARD
BROWNS O N ET AL, 19 40 10 5 45
Cancer
Volume 1 Number 5 November 1980
IPercentage)
WEAWER1~
VRABEC ~1
PRESENT SERIES
62 25 13 --
28 28 14 30
33 21 25 21
415
American Journal o~C Otolaryngology
416
sites of secondary cancer abstracted from data in several series. In surveying these results it is apparent that witl~ an index neoplasm in the head and neck, a second primary malignant tumor will arise most commonly anywhere in the same tissue and in the multicentric zone of contiguous tissue, but remote primary malignant tumors cannot be excluded in terms of site or interval. Several observations should be further emphasized. Our series, as those of others, emphasizes the frequency of synchronous and metachronous development of a second primary carcinoma in the esophagus among patients with head and neck cancer2' ~0,1~.e0.2~ In two series, 1 per cent of the patients with head and neck cancer developed carcinoma of the esophagus, which was estimated as nine times the risk to the general population)5, 21 The incidence was 5 per cent in our series. The fact that patients with head and neck cancer have dysphagia following surgery or irradiation may delay the clinical suspicion of a primary carcinoma of the esophagus. Indeed 70 to 80 per cent of the patients developing a second primary malignant tumor in the esophagus may be expected to be metachronous, with an average interval of occurrence of five years) 5. ~-~ Previous negative radiographic studies do not preclude development of a carcinoma. Indeed Vrabec u reports that in seven of 11 patients with carcinoma of the esophagus found at endoscopy pro-endoscopic upper gastrointestinal x-ray views were normal. Thus, although the greatest incidence of the second primary malignant tumor may be within the first two to three years, experience with patients surviving three years and longer supports the wisdom of periodic life-long re-examination. Three and five year "cure" rates should not make us feel secure. Awareness of an increased association between neoplasms of the head and neck and those of the esophagus should lead to detection of small neoplasms that are potentially curable. '~ Periodic surveillance with radiography and endoscopy should allow discovery of early neoplasms as well as specific dysplastic areas of preinvasive potential for special follow-up. However, if these lesions occur simultaneously, alternative modalities of treatment may need to be considered. It remains to be seen whether discovery of asymptomatic lesions in the esophagus will be associated with increased survival. A similar trend may be cited for carcinoma of the larynx and the development of a second primary malignant tumor in the lung. Lung cancer has been reported in 10 to 15 per cent of
all patients with laryngeal cancer, usually following the laryngeal lesion."" From the opposite perspective, the larynx has been reported to be the site of second primary cancer in 20 per cent of the patients with cancer of the lung, of which 30 per cent were synchronous. 2~.24 These data suggest a greater incidence than expected in a random selection of individual cancer patients of comparable age. Two recent studies, however, conclude that metachronous development of carcinoma cannot be attributed to previous irradiation utilized in the treatment of an initial malignant tumor of the m u c o u s membranes within the multicentric vicinity, t". ~7 Twenty-one per cent of our series of patients with multiple primary head and neck malignant tumors developed carcinoma of the lung; Vrabec 11 reported an incidence of 28 per cent, furthor noting that when the index neoplasm was in the larynx, the lung was the most c o m m o n site of the second lesion. The difficulty sometimes is to establish w h e t h e r one neoplasm is a metastasis from the other, or both are primary cancers. There are radiologic criteria for distinguishing a primary lung neoplasm from a pulmonary metastasis, but it is not always easy. The addition of polytomography and computerized tomography has also proven helpful, but they are not always conclusive. If the lung cancer is solitary, of the same cell type as the laryngeal cancer, but amenable to resection, the question of metastasis may have little practical importance with five year survival rates approximating the 30 per cent observed for primary lung lesions. 22 The important consideration is that in a patient in w h o m the laryngeal lesion is the initial lesion, the signs and symptoms of the lung cancer m a y be masked as a complication of the tracheostomy or treatment, i.e., persistent cough, hemoptysis, or pulmonm7 infiltrate assumed to be the result of tracheitis. A program of post-treatment chest radiography is also needed in our protocols. Furthermore, five of nine of Vrabec's patients had pro-endoscopic chest radiographs reported as normal." Thus bronchoscopy as well as esophagoscopy should be applied in the routine post-treatment followup.
It may be said that routine endoscopy in post-treatment surveillance is prohibitive because of cost-yield dividends; also the number of patients cured of lung cancer is small. However, the potential benefits from treatment of early lesions of the esophagus and lung are still under investigation and indeed show encouraging results. ~0Furthermore it is difficult for us to place a
MULTIPLE PRIMARY MALIGNANT TUMORS OF THE H E A D A N D NECK
monetary value among the criteria for diagnostic evaluation of potential lethal disease in high risk patients; this consideration appears to be more a social than a medical one. In the management of patients with simultaneous multiple primary malignant tumors, if both neoplasms are considered curable, the treatment is carried out as if each neoplasm were a solitary primary malignant tumor. If one neoplasm is in the head and neck and the other is remote, therapy is directed toward the more lethal of the neoplasms with priority of treatment established by a team approach. Consideration is given to the cellular diagnosis and the neoplasm's location, extent, and potential behavior in evaluating the prognosis for control and the choice of treatment. If the remote malignant neoplasm is clearly uncontrollable, palliative treatment is given for the head and neck tumor, Early recognition of a second primary cancer is as important as the discovery of metastases. New primary tumors should be discovered before they become symptomatic because of the factors cited. We now recommend that routine chest radiography be carried out at three month intervals, barium esophagograms carried out every six months, and routine endoscopy performed annually for the first three years after treatment in addition to the usual surveillance given cancer patients. After three years this schedule may be adjusted perhaps to an annual basis, but we must accept the observation that subsequent primary carcinoma continues to be a risk well beyond the traditional five year survival period. Any minor change in symptoms must be approached aggressively and not attriubted to the consequences of treatment.
References 1. Moertel, C, G,: Multiple primary neoplasms. Cancer, 40:1786-1792, 1977. 2. Berg, J. W., and Schottenfeld, D.: Multiple primary cancers at Memorial Hospital, 1949-1962. Cancer, 40:1801-1805, 1977. 3, Warren, S., and Gates, O.: Multiple malignant tumors: a survey of literature and statistical study. Am. f. Cancer, 51:1358-1414, 1932. 4. Moartel, C. G.: Incidence and significance of multiple primary malignant neoplasms. Ann. N.Y. Acad. Sci., 114'.886-895, 1964.
ARNOLD M. COHN AND SEAN B. PEPPARD
5. Slaughter, D. T.: Multicentric origin of intra-oral carcinoma. Surgery, 20:133-146, 1946. 6. Dellon, A. L., et el,: Multiple primary malignant neoplasms. Arch, Surg., 110:156-160, 1975. 7. Bartal, A., ot el.: Cellular immunity in patients with laryngeal cancer developing additional primary malignant tumors. Ann. Otol. Rhinol. Laryngol., 81:311315, 1979. 8. Gluckman, J. L.: Synchronous multiple primary lesions of the upper aerodigestive system. Arch. Otolaryngol., 105:597-598, 1979. 9. Weichert, K. A., and Shumrick, D.: Multiple malignancies in patients with primary carcinomas of the head and neck. Laryngoscope, 89:988-989, 1979. 10. Weaver, A., et al.: Triple endoscopy', a neglected essential in head and neck cancer. Surgery, 86:493-496, 1979. 11. Vrabeck, D. T.: Multiple primary malignancies of the upper aerodigestive tract. Ann. Otol. Rhinol. Lmsmgel., 88:846-854, 1979. 12. Hadju, J. I., and Hadju, E. O.: Multiple primary malignant tumors. J. Am. Gartatr. Soc., 16:16-26, 1968. 13. Odette, J., Szymanowski, T., and Nichols, R. D.'. Multiple head and neck malignancies. Trans. Am. Acad, Ophthalmol. Otolaryngol., 84:806-813, 1977. 14. Fee, W. E., et al.: Squamous cell carcinoma of the soft palate. Arch, Otolaryngol., 105:710-718, 1979. 15. Thompson, W. N, et al.: Synchronous and metachronous squamous cell carcinomas of the head and neck and esophagus, Gastrointest. Radiol., 3:123-127, 1978. 16. Brown, M.: Second primaries in cases of cancer of the larynx. J, LarynBol., 92:991-996, 1978. 17. Kogelnik, H. D., Fletcher, G. H., and Jesse, R. H,: Clinical course of patients with squamous cell carcinoma of the upper respiratory and digestive t~act with no evidence of disease five years after initial treatment. Radiology, 115:423-427, 1975. 18. Newall, G, R., and Kremantz, E. T.: Multiple malignant neoplasms in the Charity Hospital of Louisiana Tumor Registry. Cancer, 40:1812-1820, 1977. 19, Brownson, R. J., et el.: Simultaneous malignant tumors in patients with head and neck cancer. Arch. Otolaryngol., 97:347-349, 1973. 20. Cahan, W. B., ot al.: Separate primary carcinomas of the esophagus and head and neck region in the same patient, Cancer, 37:85-89, 1976, 21. Goldstein, H. M., and Zornoza, J.: Association of squamous cell carcinomas of the head and r~eck with cancer of the esophagus. Am. J. Roentgenol,, 131:791794, 1978. 22. Sherman, J. O., Staley, J., end Shields, T. W.: Double primary tumors of the laryux and lung. Arch. Surg., 94:550-558, 1967. 23. Campbell, L. D., Jr., and Watney, A. L.: Multiple primary malignant neoplasms. Arch. Surg., 99:401-405, 1969. 24. Cahan, W. B., and Moatemayor, P. B.: Cancer of the larynx and lung in the same patient: a report of sixty patients. Surgery, 44:209-230, 1962.
Volume 1 Number 5 November 1980
Department of Otolaryngology, 5E-UCB Way21e State University School of Medicine 540 E. Canfield Avenue Detroit, Michigan 48201 (Dr. Cohn]
417