Systematic esophageal endoscopy screening in patients previously treated for head and neck squamous-cell carcinoma

Annals ofOncology 12 643-646,2001. © 2001 Kluwer Academic Publishers Printed in the Netherlands.

Original article Systematic esophageal endoscopy screening in patients previously treated for head and neck squamous-cell carcinoma T. Petit,1 C. Georges,1 G.-M. Jung,1 C. Borel,1 G. Bronner,2 H. Flesch,2 G. Massard,3 M.Velten,1 P. Haegele1 & S. Schraub1

Summary Background- An attempt was made to improve metachronous oesophageal cancer prognosis through bi-annual systematic esophageal endoscopy screening in patients treated for head and neck cancer. Patients and methods: Bi-annual esophageal endoscopy, without a staining procedure, was performed in 1560 patients from 1987 to 1997. The distribution of previous head and neck cancer was oral cavity (20%), oropharynx (30%), hypopharynx (34%), and larynx (16%). All patients had initial panendoscopic inspection before HNSCC treatment. Esophageal tumors were considered to be second synchronous primaries when discovered within the first six months of initial tumor diagnosis. Results: Fifty metachronous esophageal asymptomatic cancers (42 T| and 7 in situ carcinomas) were diagnosed by endoscopy. The median time between the HNC and the esophageal carcinoma was 43 months (7-137 months). Metachronous esophageal carcinoma was discovered in 2.6% of patients with oral cavity tumor, 5.7% of patients with oropharynx

Introduction The phenomenon of multiple upper aerodigestive tract cancers was described as early as 100 years ago by Billroth and Von Winiwarter [1]. The field cancerization hypothesis suggests that the entire upper aerodigestive tract undergoes mutations by exogenous carcinogenic factors such as tobacco and alcohol, and is prone to multifocal cancers [2]. Second primary tumors in head and neck patients occur primarily in the upper aerodigestive tract, constituting a major threat to the survival of these patients. Moreover, as the survival rate of patients with upper aerodigestive tract cancers has increased, the detection of a second cancer has also increased. The constant annual risk of treated head and neck cancer patients developing a second primary tumor ranges from 1.5% to 5% [3-5], On average, one fourth of all patients treated for head and neck carcinoma will suffer from metachronous cancer in the same field [6-9]. In a retrospective cohort analysis of the Epidemiology

tumor, 2.3% of patients with hypopharynx tumor, and 1.7% of patients with larynx tumor. Causes of death were: 41.1% related to esophageal tumor with tumor progression, metastatic evolution, or treatment toxicity; 28.9%o related to non malignant causes; 26.6%o related to a cancer that was not of esophageal origin. Conclusions: Over a 10-year period, systematic bi-annual esophageal endoscopy uncovered metachronous esophageal tumors in 3.2% of 1560 patients originally treated for head and neck carcinoma, developing in a median time of 47 months. Patients with initial oropharyngeal tumors had a significantly higher risk of metachronous esophageal SCC, compared to the other tumor sites (P < 0.02 with Fisher exact test). Given the elevated death rate not related to the esophageal cancer and the median survival of 16 months, any potential benefit from this time-consuming procedure is debatable.

Key words: esophageal endoscopy, head and neck cancer, metachronous cancer

Program of the US National Cancer Institute, the relative risk of developing a metachronous esophageal squamouscell carcinoma following head and neck cancer was 9.5 in men and 38.8 in women [10]. As early diagnosis of esophagus cancer has a major impact on prognosis, we therefore tried to improve metachronous esophageal cancer prognosis through bi-annual systematic esophageal endoscopy screening in patients treated for head and neck cancer.

Patients and methods From 1987 to 1997, 1560 patienb were treated for primary head and neck squamous-cell carcinoma (HNSCC) in our institution. The median age of these patients was 60 years (range 29-85 years). The distribution of HNSCC localization was . 20% for oral cavity, 30% for oropharynx. 34% for hypopharynx, 16% for larynx. Distribution by AJC stage was as follows 8% of stage I, 21% of stage II, 36% of stage III, 35% of stage IV. All patients had initial panendoscopic inspection of the pharynx, larynx, esophagus and tracheobronchial tree before HNSCC treatment One hundred and seventy four patients (10.6%)

Downloaded from https://academic.oup.com/annonc/article-abstract/12/5/643/168930 by guest on 30 January 2019

Centre de Lulte Contrele Cancer Paul Strauss; ~ Department of Head and Neck Surgery. Climque Samte Barbe. Department of Thoracic Surgery. Hopitaux Universttaires. Strasbourg. France

644 Table 1 Time lapse from first HNSCC to esophageal SCC

Table 2 Esophageal carcinoma treatment

Time lapse

Treatment

Number of patients (%)

Surgery Radiotherapy Chemotherapy Chemo-radiotherapy Symptomatic treatment

5(10.9) 4(8 7) 2(4.3) 31 (67.4) 4(8.7)

6-12 months 1-2 years 2-3 years 3-4 years 4—5 years 5-10 years > 10 years

Number of esophageal SCC

6 5 5 18 2

Table 3. Causes of death.

Esophageal SCC progression Metastasis Related to treatment HNSCC NSCLC Hepatocarcinoma Pneumopathy Other non malignant Unknown

Number of patients (%)

Median survival (months)

11 (24.4) 4(9.0) 3(6.7) 6(13.3) 5(11.1) 1 (2.2) 6(13.3) 7(15.6) 2(4.4)

14 6 4 12 30 52 15 10 8

HNSCC treatment, it was impossible to use echo-endoscopy to find the TNM value and we therefore evaluated the tumor size based on the fraction of esophageal Results circumference covered by the tumor. Eight per cent of the tumors covered the whole circumference, 2% of the Fifty-four metachronous tumors tumors covered more than half of it, 26% of the tumors covered between a half and a quarter of it, and 64% Out of these 1560 patients, systematic endoscopy revealed 50 esophageal carcinomas, 2 gastric adenocarcinomas, 1 covered less than a quarter of the circumference. Locogastric lymphoma, and 1 gastric metastasis of squa- regional status of lymph nodes was determined by CT mous-cell carcinoma. The compliance to this systematic scan in 37 patients, and 16% had nodal involvement. procedure was excellent, without any major complica- Four patients had metastases when esophageal tumor was diagnosed. tions. The esophageal carcinoma treatments were as indiFifty esophageal carcinomas cated in Table 2. Among the 7 in situ tumors, 2 disappeared with an anti-acid medication, 3 became invasive According to the initial HNSCC anatomic site, 2.6% of with a delay of transformation of 8, 13 and 28 months, patients with oral cavity tumor, 5.7% of patients with and 2 patients died of non-malignant causes while having oropharynx tumor, 2.3% of patients with hypopharynx in situ esophageal tumors. Five patients had surgical tumor, and 1.7% of patients with larynx tumor, had a treatment and one patient is still alive with a seven-year metachronous esophageal carcinoma. The median age follow-up. Thirty-one patients were treated with an of patients with metachronous esophageal carcinomas association of chemotherapy and radiation, and their was 59 years (range 37-84 years). The median time lapse median survival was twelve months (range 1-78 months) from the first HNSCC was 47 months (range 7-168 Two of these patients show complete remission with a months) (Table 1). In addition to these 50 esophageal follow-up of, respectively, 52 and 78 months. tumors, another 22 tumors were diagnosed during the The causes of death are reported in Table 3. Causes follow-up of these 50 patients (10 HNSCC, 9 non-small- related to the esophageal tumor, with esophageal procell lung cancer (NSCLC), 1 bladder cancer, 1 anal cancer, gression, metastasis evolution, or treatment toxicity, were 1 hepatocarcinoma). responsible for 41.1% of the deaths, all metastasis being All these esophageal tumors were squamous-cell can- arbitrarily related to esophageal tumor evolution. Non cers (SCC), and 86% were invasive (43 of 50). The tumor malignant causes were responsible for 28.9% of deaths. localization was as follows: upper third 35%, middle Cancers that were not of esophageal origin, were rethird 52%, lower third 13%. Five tumors were bifocal. sponsible for 26.6% of deaths. The histologic grade distribution was 33% well differThe median survival of the 50 patients after the initial entiated, 25% moderately differentiated, and 42% poorly HNSCC diagnosis was 70.5 months. The median survival differentiated tumors. Due to after-effects from previous after the esophageal tumor diagnosis was 16 months

Downloaded from https://academic.oup.com/annonc/article-abstract/12/5/643/168930 by guest on 30 January 2019

had multiple synchronous HNSCC. Tumors were considered to be second synchronous primaries when the subsequent tumor was located at a site distant from the margin of excision of the first primary tumor, separated by normal mucosa and discovered within the first 6 months of initial tumor diagnosis. During this 10 year follow-up period, the numbers of patients with quadruple, triple, and double metachronous HNSCC sites were respectively 10, 35, and 123. A bi-annual systematic esophageal endoscopy screening, beginning six months after treatment of the primary tumors, was proposed to all these patients. All esophageal endoscopies were performed with flexible scopes under local anaesthesia by the same gastro-enterologist. No staining procedure was used during these endoscopies. During this 10 year-follow-up period, 1560 patients underwent a total number of 9682 esophageal endoscopies.

645 I 00 0J» 0JO 070

|

0,60

\

1 E \

040 OJO 0.20

OIO

0

12

24

36

41

60

72

84

%

IDS

I

Months since (Ligiwsis

Figure 1 Survival after diagnosis of esophageal cancer

(range 1-120 months) (Figure 1). Three patients with initial invasive tumor are presently alive and in complete remission, with one patient having been treated through surgery and two patients treated through the combination of chemotherapy-radiotherapy. Their follow-ups are, respectively, 52, 78, and 84 months.

Discussion In the present study, over a 10-year period, systematic bi-annual esophageal endoscopy uncovered metachronous esophageal tumors in 3.2% of 1560 patients originally treated for head and neck carcinoma, developing in a median time of 47 months. Patients with initial oropharyngeal tumors had a significantly higher risk of metachronous esophageal SCC, compared to the other tumor sites {P < 0.02 with Fisher exact test). The median survival of the 50 patients was 70.5 months after the primary HNSCC diagnosis, and 16 months after the esophageal SCC diagnosis. We should note here that optimal treatment for these patients often required adaptation, due to the previous HNSCC treatment. For example, 47% of the upper third esophageal were located in the supraclavicular radiation zone of the initial HNSCC. Large prospective panendoscopy series with routine inspections of the pharynx, larynx, esophagus, and tracheobronchial tree in patients undergoing initial evaluation of head and neck cancer found a 1.8%-8% incidence of synchronous esophagus cancer [11, 12]. There have been few studies which have focused on metachronous esophagus cancer. There were no cases of metachronous esophageal tumors in the retrospective serie of the Regina Elena Cancer Institute, with a median follow-up of 3.2 years [4]. Kinzie JJ et al. reported that 0.93% (3 of 321) of their HNSCC patients developed a metachronous esophageal cancer over a two-year period [13]. This incidence was 1% (2 of 200) over a four-year period for Shaha et al. [14]. Among 1294 patients with primary HNSCC cancer diagnosed at the Memorial

Downloaded from https://academic.oup.com/annonc/article-abstract/12/5/643/168930 by guest on 30 January 2019

1

000

Sloan-Kettering Cancer Center between 1970 and 1979, 21 patients (1.6%) developed a metachronous esophageal tumor [15]. Over a period of time similar to that of our study, this investigation uncovered fewer metachronous esophageal tumors (1.6% compared to 3.2%). We assume that this observation might be secondary to different epidemiologic risk factors, with more alcoholic abuse in the French study, and more tobacco abuse in the American study. In the Memorial Sloan-Kettering Cancer Center study, the median time lapse between the appearance of the primary tumor and the metachronous lung or esophageal tumor was 43 months, and the median survival after diagnosis of the metachronous lung and esophageal tumor was seven months. Abrasive esophageal cytology was evaluated for HNSCC metachronous cancer screening [16]. Twenty premalignant or early malignant lesions of the esophagus were revealed in three hundred twenty patients during a mean follow-up of four years. The risk for surviving patients of developing a metachronous tumor does not seem to decrease with time, according to data from the Radiation Therapy Oncology Group and the Connecticut Tumor [6, 17]. This risk is thought to be independent of the stage of the primary tumor, and the prognostic value of the index tumor anatomic site remains controversial [9, 17]. For Licciardello et al., both synchronous and metachronous second cancers occur at higher rates when the index primary tumor is located in the oral cavity [9]. On the contrary, Cooper et al. showed minor differences in frequency for different primary sites [6]. In the present study, metachronous esophageal SCC occurred at significantly higher rate when the initial tumor was located in the oropharynx. It was suggested that patients with the earliest stage of disease have the greatest risk of developing metachronous cancers since patients with more advanced stages of disease have a shorter survival [17]. Of the 50 patients in the present study, 32% of the initial HNSCC localizations were in stage I, 24% in stage II, 36% in stage III, and 8% in stage IV. There are few studies on the effect of initial HNSCC treatment on second primary tumor risk. Cases of radiation-induced carcinoma of the hypopharynx were reported in the long-term follow-up of patients treated with irradiation for larynx carcinoma [18], and predisposition toward head and neck cancer was reported in patients exposed to radiation upon treatment for benign disease during childhood [19]. In two different series, no difference in the respective incidence of metachronous tumors was observed in groups of patients treated with radiotherapy or surgery [20, 21]. On the contrary, radiation was demonstrated to reduce the prevalence of second cancer in the same field by Kogelnik et al., presumably by sterilizing dysplasia and microscopic neoplasia [22]. Some studies on second primary carcinomas showed a significantly lower survival rate when metachronous cancers arose within prior irradiated tissue [23, 24]. This might reflect that treatment given to the second tumor was less effective because of previous radiotherapy. All

646

Acknowledgements We thank Diana Pursglove and Gabrielle Taglang for excellent editorial assistance. References 1. Billroth T, Von Winiwarter A. A General Surgical Pathology and Therapeutics. New York. Appleton 1883. 2. Slaughter DP, Southwick HW, Smejkal W. Field cancenzation in oral stratified squamous epithelium: Clinical implications of multicentnc origin. Cancer 1953, 6: 963-8. 3 Lippman SM, Hong WK. Second malignant tumors in head and neck squamous cell carcinomas: The overshadowing threat for patients with early stage disease Int J Radiat Oncol Biol Phys 1989; 17: 691^1. 4. Cianfriglia F, Di Gregorio DA, Manieri A. Multiple primary tumours in patients with oral squamous-cell carcinoma Oral Oncol 1999.35: 157-63. 5 Leon X, Quer M. Diez S et al. Second neoplasm in patients with head and neck cancer. Head Neck 1999: 21. 204-10. 6. Cooper JS. Pajak TF, Rubin P et al. Second malignancies in patients who have head and neck cancer: Incidence, effect on survival and implications based on the RTOG experience. Int J Radiat Oncol Biol Phys 1989: 17. 449-56. 7. Larson JT, Adams GL, Fattah HA. Survival statistics for multiple primaries in head and neck cancer. Otolaryngol Head Neck Surg 1990, 103- 14-24. 8. McGarry W, MacKenzie K, Periasamy P et al. Multiple primary malignant tumours in patients with head and neck cancer' The implications for follow-up. Clin Otolaryngol 1992. 17: 558-62.

9. Licciardello JT, Spitz MR, Hong WK Multiple primary cancer in patients with cancer of the head and neck: Second cancer of the head, esophagus, and lung. Int J Radiat Oncol Biol Phys 1989: 17: 467-76. 10. Ahsan H, Neugut Al, Gammon MD Association of adenocarcinoma and squamous cell carcinoma of the esophagus with tobacco-related and other malignancies. Cancer Epidemiol Biomarkers Prev 1997; 6: 779-82. 11. Leipzig B, Zellmer JE, Klug D. The role of endoscopy in evaluating patients with head and neck cancer. Arch Otolaryngol 1985; 111: 589-94. 12. McGuirt WF. Panendoscopy as a screening examination for simultaneous primary tumors in head and neck- A prospective sequential study and review of the literature. Laryngoscope 1982; 92: 569-76. 13. Kinzie JJ, Evans RB, Ragan D. Double and multiple primary cancers in an adult head and neck radiation therapy clinic Int J Radiat Oncol 1984; 10. 2037-9. 14. Shaha AR, Hoover EL, Mitrani M et al. Synchromcity, multicentricity, and melachronicity of head and neck cancer Head Neck Surg 1988, 10.225-8 15. Robinson E, Zauber A, Fuks Z, Strong E. Clinical characteristics of patients with epidermoid carcinoma of the upper aerodigestive tract who develop second malignant tumors. Cancer Detect Prev 1992; 16- 297-303. 16 Pellanda A, Grosjean P, Leoni S et al. Abrasive esophageal cytology for the oncological follow-up of patients with head and neck cancer. Laryngoscope 1999; 109. 1703-8. 17. Winn DM, Blot WJ. Second cancer following cancers of the buccal cavity and pharynx in Connecticut, 1935-1982. Natl Cancer Inst Monogr 1985; 68' 25-48. 18. Lawson W, Som M. Second primary cancer after irradiation of laryngeal cancer. Ann Otol R Laryngol 1975; 84. 771-5. 19 Sirota DK., Eden AR, Biller HF. Multiple head and neck neoplasia following radiation for benign disease during childhood. J Surg Oncol 1988; 38: 101-3 20. Parker RG, Enstrom JE. Secondary cancer following primaries of head and neck. Inl J Rad Oncol Biol 1988; 14: 561—4 21 Shons AR, McQuarroe DG Multiple primary epidermoid carcinomas of the upper aerodigestive tract Arch Surg 1985. 12. 1007-9. 22 Kogelnik HD, Fletcher GH, Jesse RH. Clinical course of patients with squamous cell carcinoma of the upper respiratory and digestive tracts with no evidence of disease five years after initial treatment. Radiology 1975; 115: 423-7. 23 Robinson E, Neugut Al. Second primary tumors in patients with head and neck squamous-cell carcinoma. Cancer 1995; 76: 1684. 24. Dolan R, Vaughan C, Fuleihan N. Metachronous cancer Prognostic factors including prior irradiation. Otolaryngol Head Neck Surg 1998; 119. 619-23. 25. Day G L Blot WJ, Shore RE et al. Second cancers following oral and pharyngeal cancers' Role of tobacco and alcohol. J Natl Cancer Inst 1994; 86: 131-7 26. Tachimori Y, Watanabe H, Kato H et al. Treatment for synchronous and metachronous carcinomas of the head and neck and esophagus. J Surg Oncol 1990: 45' 43-5. Received 24 July 2000; accepted 10 November 2000. Correspondence lo T Petit. MD. PhD Department of Medical Oncology Centre de Lutte Contre le Cancer Paul Strauss 3 rue de la Porte de I'Hopital 67085 Strasbourg Cedex France E-mail, tpetitfastrasbourg.fnclcc.fr

Downloaded from https://academic.oup.com/annonc/article-abstract/12/5/643/168930 by guest on 30 January 2019

the same, the greatest risk factor for second primary head and neck tumors is certainly continuous abuse of alcohol and tobacco after initial head and neck tumor treatment [25]. Tachimori et al. have reported a study of 41 patients with initial HNSCC and metachronous esophageal tumors [26]. These metachronous esophageal tumors were not discovered by systematic esophageal endoscopy screening, and the average time interval between HNSCC and esophageal tumor diagnosis was 42 months, ranging from 9 months to 9 years. The two-year survival rate was 27.5% and five-year survival rate was 10% after treatment of the esophageal tumor. In our study, these survival rates were respectively 37% and 14.5%, thanks to consistent systematic endoscopy imposed on all patients. In more than half of the patients, the cause of death was not related to the esophageal tumor or its treatment (28.9% related to non malignant causes and 26.6% related to cancers that were not of esophageal origin). Given the short survival of these patients and the elevated non neoplastic co-morbidity, any potential benefit from this time-consuming procedure of systematic esophageal endoscopy is debatable. Finally, this early detection program might be limited to patients with oropharynx tumors of whom nearly 6% will develop a metachronous esophageal tumor according to our study.