Eur Arch Otorhinolaryngol (1992) 249 : 364-369

European Archives of

Oto-RhinoLaryngology © Springer-Verlag 1992

Yearly chest radiography in the early detection of lung cancer following laryngeal cancer A. M. Engelen 1, L. J. A. Stalpers 1, J. J. Manni 2, J. H. J. Ruijs 3, and W. A. J. van Daal 1 Departments of 1Radiotherapy, 20torhinolaryngology and 3Diagnostic Radiology, Nijmegen University Hospital St. Radboud, Postbox 9101, NL-6500 HB Nijmegen, The Netherlands Received May 26, 1992 / Accepted June 23, 1992

Summary. In a retrospective study of 556 patients (505 men, 51 women) with laryngeal cancer the incidence and prognosis of lung malignancies was studied in patients who were examined yearly by chest radiography. In 69 patients (12.4%) a lung malignancy was diagnosed, with 28 having a histologically confirmed second primary malignancy. All of these 69 patients were men. The incidence of radiologically detected lung malignancies, both second primary and metastatic cancer, is higher and m o r e prolonged following supraglottic carcinoma than following glottic carcinoma. In 47 patients (68%) without symptoms, the lung malignancy was detected by routine annual chest radiography. The survival rate in patients with lung cancer detected by the yearly radiography was significantly higher than in patients diagnosed after symptoms (median survival 10 and 4 months, respectively). However, taking into account the lead time between early radiologic diagnosis and the time a t u m o r would have been diagnosed following symptoms, the observed survival benefit of yearly radiography was much lower, or even nil. Key words: Laryngeal neoplasms - Lung cancer - Second primary neoplasms - Radiology - Early diagnosis

Introduction In Western countries the incidence of laryngeal cancer is 114 per 106 in men and 6 per 106 in w o m e n [27]. It is predominantly a disease of the sixth decade of life [30]. Symptoms m a y occur relatively early and especially involve persistent hoarseness. In such cases cure can be relatively well attained with radiotherapy and/or surgery [21]. As early as 1932, Warren and Gates [26] described an increased association between cancer in the head and neck and second primary tumors. Second primary tumors are predominantly seen in the lung, occurring in 2 - 1 2 % Correspondence to: L. J. A. Stalpers

of patients [11, 13]. This increased incidence of lung cancer following laryngeal cancer is probably due to cigarette smoking as a c o m m o n etiologic factor [28]. With improving results of the local treatment of laryngeal cancer, secondary tumors b e c o m e a major threat to the patients' well being. As such, early detection and early treatment of second primary malignancies m a y eventually improve the overall prognosis of these patients [6, 11]. This is the rationale for regular chest radiography in the follow-up of patients with laryngeal cancer. In the present study we examined: (1) the incidence of second primary malignancies in the lung and lung metastases in patients treated for laryngeal cancer and (2) the role of routine yearly chest radiography in the early detection of these lung malignancies.

Patients and methods From January 1979 to January 1986, 590 patients with squamous cell carcinoma of the larynx were admitted for treatment to the St. Radboud University Hospital, Nijmegen, The Netherlands. All patients admitted for primary treatment were eligible for this study. Thirty-four patients were excluded who came for treatment of known recurrent disease. In all, the medical cases of 556 patients were reviewed for this study. For estimation of the incidence of lung cancer, follow-up closed in July 1986. The median observation duration was 37 months. For estimation of the survival rate of patients with lung cancer, followup closed in July 1991, ensuring at least a 5-year follow-up period. The mean follow-up was 58 months. The male to female ratio of the 556 patients was 9.9 (505 men, 51 women). The patient distribution is shown in Fig. 1, with the median age at diagnosis of the laryngeal tumor 61 years (range 32-96 years). All tumors were staged according to the TNM classification by tumor site (Table 1). Treatment consisted of radiotherapy in T1 and T2 carcinomas (64-68 Gy in daily fractions of 2 Gy in 6-8 weeks). Surgical salvage was reserved for recurrent tumors. In the more advanced stages T3 and T4, primary surgery was performed with or without preoperative radiotherapy (25 Gy in five daily fractions of 5 Gy in the week preceding operation) [12]. The follow-up protocol for all patients is shown in Table 2. This included yearly chest radiography consisting of anteroposterior and lateral views. All X-rays were reviewed at least threefold before a "final" report was given, both by a resident radiologist and a staff radiologist and finally by a radiologist and the prescribing oncologist. If a chest X-ray was sugges-

365 Table 2. Protocol for the follow-up of patients with laryngeal cancer

Frequency

I. Clinical examination First year Second year Third year Fourth year > 4 years

150 '

1O0

Every 2 months Every 3 months Every 4 months Every 6 months Annual

II. Yearly chest radiography - Anteroposterior view - Lateral view III. Blood and/or urine analysis - if indicated

50

0 30-39

40-49

50-59

60-69

70-79

80-89

Age

Fig. 1. Distribution by age and sex of 556 patients with squamous cell carcinoma of the larynx, • male; N female

Table 1. Distribution of 556 patients treated for laryngeal cancer by tumor site and stage [1, 23] Site

Stage

All

I

II

III

IV

tire of malignancy, further radiologic examinations were performed. Depending on the feasibility of invasive diagnostic studies and on the lesion found, bronchoscopy and/or mediastinoscopy were performed. Surgery was performed in curable and operable lung cancer; radiotherapy and chemotherapy were reserved for palliative care. The incidence of a second primary malignancy or a pulmonary metastasis was calculated as a crude rate of all patients and as an actuarial incidence rate corrected for the duration of follow-up [5, 17]. The survival of patients with a pulmonary malignancy was estimated by an actuarial method; the survival curve of patients detected by the yearly chest radiography without symptoms and the survival curve of patients detected after symptoms were compared with the log-rank test [5],

Supraglottic Glottic Subglottic

35 173 0

42 62 1

58 58 1

106 23 4

241 309 6

Results

Total

208

105

110

133

556

I n 69 of 556 p a t i e n t s ( 1 2 . 4 % ) a m a l i g n a n c y of t h e l u n g was d i a g n o s e d f o l l o w i n g t r e a t m e n t for l a r y n g e a l c a n c e r .

50

40

8

30 i ......................

ca

,-' i

o

20

,--- ~.......... r .-----, ;-, '

>

"5 E 0

166

~

................

i

,,j _

113

6

~

J~

33

[

~.-:

~..;.._.,._.,.:.~. ...z-

lO

~'~':

0

0

-:

:

......

.. ~ ...-:''-'"-'":'"

I

I

I

I

I

I

I

1

2

3

4

5

6

7

Years

Fig. 2. Incidence of lung cancer in the follow-up of patients with laryngeal carcinoma (patient numbers at risk are indicated). Dotted lines: 95% confidence limits

366 50

40

supraglottic ¢, 0 ¢.. ¢1

30

0

77

_¢

51

r........

sl

glottic

110 >

,...:

20

..................

"5 E --I

$:

144 -J

o

..r-:

I

rJ

"186

10

a7

132 ...r -..

0

~5~

Fig. 3. Incidence of lung cancer following cancer of the glottic or supraglottic larynx (patient numbers at risk are indicated)

177

I

I

t

I

I

I

I

1

2

3

4

5

6

7

8

Years

i00

'L !

80

60

._> > Or)

40

....,

]

_

_

~...,

L

......... i "-i

]: ..... %

Detected by cheat X-rays (N=47) I

20

l__

0

L

All patients

(N=69)

i ...............................

Diagnosed folowing symptoms (N=22)

1

I

I

1

I

1

2

3

4

Ng. 4. Survival of patients with lung cancer following laryngeal cancer, by mode of lung cancer presentation

Years

Two patients with histologically confirmed metastases from breast cancer were not included in this number. Consequently, all 69 patients were men. In 28 patients (5.0%) the lung t u m o r was histologically confirmed as a second primary tumor; in 11 patients (2,0%) t u m o r was found histologically to be a metastasis. In 30 patients (5.4%) no histological confirmation could be made for the progressive lung mass and they were therefore con-

sidered to have metastatic lung cancer. Figure 2 shows the actuarial incidence rate for lung malignancy, which was 17% after 5 years and m o r e than 20% after 7 years. As shown in Fig. 3 a lung malignancy was seen m o r e frequently following supraglottic cancer (41/241, 17.0%) than following glottic cancer (28/309, 9.1%). This difference was highly significant (P = 0.001). The incidence of lung cancer after supraglottic cancer then increased to

367 i00

80

k7

'

[

1

i. 7 I I I

Surgery (N=17)

60 I I ....

"i

.>_

I

> :...,

if)

I

40

i. . . . .

7

i I [ i

20

. I

i..k~

i

No Treatment (N=38)

0

Radiotherapy (N=14)

Fig. 5. Survivalcurves of patients with lung cancer, by mode of treatment

~....

I

I

I

I

2

3

4

5

Years

32% after 7 years. In contrast, no further lung malignancies were seen in the 67 patients with glottic cancer after 5 years. Of the 69 patients with a lung malignancy, 47 had no symptoms and the malignancy was detected first by the routine yearly chest radiography. The tumors in the other 22 patients were diagnosed after they had produced symptoms between the yearly radiologic examinations. Figure 4 shows the survival curve of all 69 patients with lung malignancies and survival curves after stratification by mode of presentation. The survival of patients in whom tumor was detected without symptoms was significantly better than that of the patients with symptoms (P < 0.05), the median survival being 10 and 4 months, respectively. Figure 5 gives the survival curves after stratification by lung cancer treatment. Patients who had surgery did significantly better than patients who had radiotherapy or no treatment at all (P < 0.05), the median survivals being 26, 11 and 3 months respectively.

Discussion Second primary lung cancer was defined by the three criteria of Warren and Gates as modified by Gluckman et al. [9] in 1980: 1. The neoplasms must be clearly malignant as determined by histologic evaluation. 2. Each neoplasm must be geographically separate and distinct. The lesion should be separated by normalappearing mucosa. If a second neoplasm is contiguous to the initial primary tumor or is separated by mucosa with intraepithelial neoplastic change, the two should be con-

sidered as confluent growths rather than multicentric carcinomas. 3. The possibility that the second neoplasm represents a metastasis should be excluded. The observation that the invasive carcinoma arises from an overlying epithelium demonstrating a transition of carcinoma in situ to invasive carcinoma is helpful, and when the separate loci have significant differences in histology the diagnosis of separate primary cancers is appropriate. Patients with no histologically proven second primary in the lung were defined as patients with metastatic lung cancer. In this study patients with a mass that was seen radiologically to progress in the presence of clinical deterioration but without histologic confirmation of a malignancy were also defined as patients with metastatic lung cancer. In general, these patients were in too poor a condition to allow histologic confirmation or curative surgery. In contrast, patients with a radiographic lesion suggestive of a malignancy but not showing radiographic or clinical progression within a year were identified as having no pulmonary malignancy. As described in Table 3, the incidence of lung cancer following laryngeal cancer and lung cancer has been reported to range from 1.1% to 12.5%. This wide range may be explained both by different lengths of follow-up and - probably more important - by different criteria for diagnosis of a second primary malignancy [4, 15]. To minimize the latter confusion, several authors stress the need for histologic verification of a secondary lung tumor [4, 9, 15]. The requirement of histological confirmation of second primary malignancies may result in selection of "pure" second primaries but may result in underestimation of the real incidence. Moreover, overemphasis of the need for a histological distinction be-

368 Table 3. Review of the literature on second primary malignancies following laryngeal cancer

Author

Reference

Period

Patients (n)

Sherman et al. (1967) Goffinet et al. (1973) Brown (1978) Martin et al. (1979) Gluckman et al. (1980) Wagenfeld et al. (1980, 1981) Hordijk and de Jong (1983) Rodrfguez et al. (1984) Miyahara et al. (1985) De Vries et al. (1986) Present study (1992)

[20] [10] [3] [13] [9] [24, 25] [11] [19] [14] [6]

1955-65 1957-70 1958-74 1960-77 1974-78 1965-74 1958-79 1979-82 1958-81 1963-84 1979-86

187 535 1600 496 189 903 691 286 1389 748 556

Second primary malignancies (%) All sites Lung a 16.0 4.1 9.1 17.9 7.5 20.5 9.9 13.9 7.7

7.4 1.9 1.1 2.0 6.8 4.2 11.7 12.5 1.7 8.6 5.0

a Excluding metastatic lung cancer

tween metastatic and primary lung cancer can distract from the clinical purpose of yearly chest radiography, namely to detect a lung malignancy at a curable stage. Therefore, we evaluated the role of annual radiography in detecting both metastatic and primary lung cancer. Our present findings showed that the risk of developing lung cancer following supraglottic carcinoma is higher than following glottic cancer. This finding agrees with other studies by Wagenfeld et al. [24, 25]. Moreover, we found that while an increased lung cancer risk persists for many years following supraglottic carcinoma, the incidence curve slopes down after 2 years following glottic carcinoma and becomes asymptotic after less than 5 years. A similar pattern, although with lower incidence rates, was found if we confined the population to the 28 patients wiht histologically confirmed second primary lung cancer (not presented here). These findings support the hypothesis of Wagenfeld et al. that different mechanisms underlie tumor development in patients with supraglottic and glottic carcinoma, despite common etiologic factors. Second primary lung cancer following laryngeal cancer has been described as a feature predominantly seen in males [3, 20]. We found no lung malignancy associated with laryngeal cancer at all in 51 women, even though 7 cases (13%) were expected. This finding is highly unlikely attributed to chance alone (P = 0.01). The overall survival of patients with lung malignancy following laryngeal cancer is poor [20, 29] and similar to lung cancer survival in the general population [16, 18]. The higher survival rate of patients detected by yearly chest radiography compared with patients diagnosed after symptoms suggests a benefit of yearly chest radiography. However, median survivals differ by only 6 months. Considering the lead time between early diagnosis of lung cancer and the time a tumor would be diagnosed following symptoms, the observed survival benefit of yearly chest radiography is much less [7, 22]. If the survival of patients detected by routine chest radiography is corrected for only several months lead time, the survival curve would converge with the curve of patients diagnosed after symptoms. As a consequence,

the difference observed tends to disappear. However, since almost all patients diagnosed following symptoms die within 2 years while almost 20% of patients "without symptoms" show long-term survival, there is probably a small benefit from routine yearly chest radiography. Nonetheless, our finding that yearly chest radiography hardly contributes to improved survival of patients with lung cancer (and thus with laryngeal cancer) agrees with the disappointing results in general of population screening for lung cancer [2, 8]. Although patients who had surgery for lung cancer show a better survival than those without surgery, the benefit of surgery is probably spurious. Thus our current policy is to reserve surgery for "cure." Therefore, patients with a poor prognosis due to poor general health or to large or multiple lung tumors are not eligible for surgery. For those patients with a prolonged enhanced risk of developing lung cancer, one must intensify the search for both better early detection methods and the underlying causes and measures to prevent secondary cancers. References

1. American Joint Committee for Cancer Staging and End Results Reporting (1978) Clinical staging system for carcinoma of the larynx. American Joint Committee for Cancer Staging and End Results Reporting, Chicago 2. Bailar JC (ed) (t984) Screening for lung cancer - where are we now? Am Rev Respir Dis 130 : 541-542 3. Brown M (1978) Second primaries in cases of the larynx. J Laryngol Otol 92 : 991-996 4. Cahan WG, Montemayor PB (1962) Cancer of the larynx and lung in the same patient. J Thorac Cardiovasc Surg 44:309320 5. Cutler SJ, Ederer F (1958) Maximum utilization of the life table method in analyzing survival. J Crhonic Dis 8 : 699-712 6. De Vries N, Waal I van der, Snow GB (1986) Multiple primary tumours in laryngeal cancer. J Laryngol Otol 100 : 915-918 7. Feinleib M, Zelen M (1969) Some pitfalls in the evaluation of screening programs. Arch Environ Health 19 : 412-415 8. Fontana RS, Sanderson DR, Taylor WF, Woolner LB, Miller E, Muhm JR, Uhlenhopp MA (1984) Early lung cancer detection: results of the initial (prevalence) radiologic and cytologic

369

9. 10.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

screening in the Mayo Clinic study. Am Rev Respir Dis 130: 561-565 Gluckman JL, Crissman JD, Donegan JO (1980) Multicentric squamous cell carcinoma of the upper aerodigestive tract. Head Neck Surg 3 : 90-96 Goffinet DR, Eltringham JR, Glatstein E, Bagshaw MA (1973) Carcinoma of the larynx. Results of radiation therapy in 213 patients. Am J Roentgenol Radium Ther Nucl Med 117: 553-564 Hordijk GJ, Jong JMA de (1983) Synchronous and metachronous tumors in patients with head and neck cancer. J Laryngol Otol 97 : 619-621 Kazem I, Broek P van den, Huygen PLM (1982) Planned preoperative radiation therapy for advanced laryngeal carcinoma. Int J Radiat Oncol Biol Phys 8 : 1533-1537 Martin G, Glanz H, Kleinsasser O (1979) Multiple maligne Tumoren bei Patienten mit Larynxcarcinomen. Laryngol Rhinol Otol 58 : 756-763 Miyahara H, Yoshino K, Umatani K, Sato T (1985) Multiple primary tumours in laryngeal cancer. J Laryngol Otol 99 : 9991004 Moertel CG (1977) Multiple primary malignant neoplasms. Historical perspectives. Cancer 40 : 1786-1792 Nou E (1984) The natural five-year course of bronchial carcinoma. Cancer 53:2211-2216 Peto R, Pike MC, Armitage P (1977) Design and analysis of randomized clinical trials requiring prolonged observation of each patient, 2. Analysis and examples. Br J Cancer 35 : 1-39 Ries LG, Pollack ES, Young JL (1983) Cancer patients survival: surveillance, epidemiology and end results program, 1973-79. J Natl Cancer Inst 70 : 693-707 Rodrfguez E, Castella J, Andr6s L de, Cornudella R (1984) Lung cancer in patients with tracheostomy due to cancer of the larynx. Respiration 46 : 323-327 Sherman JO, Staley CJ, Shields TW (1967) Double primary tumors of larynx and lung. Arch Surg 94 : 550-558

21. Stalpers LJA, Verbeek ALM, Daal WAJ van (1987) Results of radiotherapy and surgery for glottic carcinoma. Cancer Treat Rev 14 : 131-141 22. Stalpers LJA, Vierzen PBJ van, Brouns JJA, Bruaset I, Manni JJ, Verbeek ALM, Ruys JHJ, Daal WAJ van (1989) The role of yearly chest radiography in the early detection of lung cancer following oral cancer. Int J Oral Maxillofac Surg 18 : 99-103 23. UICC-International Union Against Cancer (1978) UICCTNM classification of malignant tumours. UICC, Geneva, pp 33-37 24. Wagenfeld DJH, Harwood AR, Bryce DP, Nostrand AWP van, Boer G de (1980) Second primary respiratory tract malignancies in glottic carcinoma. Cancer 46 : 1883-1886 25. Wagenfeld DJH, Harwood AR, Bryce DP, Nostrand AWP van, Boer G de (1981) Second primary respiratory tract malignant neoplasms in supraglottic carcinoma. Arch Otolaryngol 107 : 135-137 26. Warren S, Gates O (1932) Multiple primary malignant tumors, a survey of the literature and statistical study. Am J Cancer 16 : 1358-1414 27. Waterhouse J, Muir C, Powell J, Shanmugaratnan K (eds) (1982) Cancer incidence in five continents, vol 4. WHO-IARC scientific publications no. 42. International Agency for Research on Cancer, Lyon 28. Wynder EL, Mushinsky MH, Spivak JC (1977) Tobacco and alcohol consumption in relation to the development of multiple primary cancers. Cancer 40:1872-1878 29. Yellin A, Hill LR, Benfield JR (1986) Bronchogenic carcinoma associated with upper aerodigestive cancers. J Thorac Cardiovasc Surg 91 : 674-683 30. Zagars G, Norante JD (1983) Head and neck tumors. In: Rubin P (ed) Clinical oncology for medical students and physicians - a multidisciplinary approach, 6th edn. American Cancer Society, Rochester