Вы находитесь на странице: 1из 6

Thoracic M e t a s t a s i s in M a l i g n a n t Melanoma*

A Radiographic S u r v e y of 6 5 Patients

W. Richard Webb, M.D., and Gordon Gamsu, M.D.

Sixty-five patients with malignant melanoma metastatic nosis than other patterns of pulmonary involvement.
to the thorax were evaluated retrospectively. Chest x-ray Enlargement of mediastinal lymph nodes, bronchial ob-
films showed abnormalities in 6 3 patients and provided the struction with atelectasis, pleural effusion, bone metas-
first evidence of dissemination in 4 2 of the 6 2 with wide- tasis with an extrapleural mass, and cardiomegaly were
spread metastasis. Frequent radiographic follow-up eval- also seen. Follow-up chest x-ray films are of limited value
uations of patients with primary melanoma is necessary in evaluating patients once they are found to have thoracic
to detect metastasis at an early stage. Pulmonary metas- metastasis. The rate of growth of metastatic lesions or the
tasis was seen radiographically in almost 9 0 percent of the regression of the metastasis does not correlate with sur-
patients. Snowstorm metastasis denotes a poorer prog- vival.

TV/T elanomas are uncommon and represent only 1 1974. There were 4 4 male and 2 1 female patients, all of
whom were white. At the time of diagnosis of their primary
percent of all malignant tumors. 1,2
tumor, the ages of patients ranged from 1 6 to 7 6 years, as
their relative rarity, melanomas account for a size-
follows: 16 to 1 9 years, four patients ( 6 p e r c e n t ) ; 2 0 to 2 9
able percentage of the lesions that metastasize to the years, seven patients ( 1 1 p e r c e n t ) ; 3 0 t o 3 9 years, 12 patients
thorax. Five to 15 percent of all cases of pulmonary ( 1 9 p e r c e n t ) ; 4 0 to 4 9 years, 1 9 patients ( 2 9 p e r c e n t ) ; 5 0
metastasis and 5 percent of cases of metastasis to to 5 9 years, 1 3 patients ( 2 0 p e r c e n t ) ; 6 0 to 6 9 years, six
patients ( 9 p e r c e n t ) ; and 7 0 to 7 9 years, four patients ( 6
lung and mediastinum are due to this tumor. " 3 6

Nearly all patients with disseminated melanoma
The primary tumors were most commonly of the chest,
have thoracic involvement, 7
and various patterns back, or abdomen, but sites were widely distributed, as
of metastasis have been identified radiographically. shown by the following listing: head and neck, 12 patients
Solitary, multiple, and miliary pulmonary nod- { 1 8 p e r c e n t ) ; upper extremity, 17 patients ( 2 6 p e r c e n t ) ;
trunk, 2 4 patients ( 3 7 p e r c e n t ) ; lower extremity, nine pa-
ules, 8 1 2
lymphangitic spread, 13
enlargement of hilar
tients ( 1 4 p e r c e n t ) ; and unknown, three patients ( 5 per-
and mediastinal lymph n o d e s , 8 1 0 , 1 4
bronchial ob-
c e n t ) . All were diagnosed histologically as malignant mela-
struction with atelectasis, " 14 16
pleural effusion, ' 7 17
noma. Nineteen tumors had been classified using Clark's
cardiac metastasis, 11,1819
and rib destruction 20,21
c r i t e r i a . - ' Twelve were of the nodular type, and seven
2 23

have all been seen. were superficial spreading. All but one of the 19 had invaded
levels 4 or 5 (subcutaneous f a t ) of the dermis. Thoracic
Our report is the first to describe the frequency
metastasis was proved by examinaton of a tissue specimen,
of radiographic patterns of thoracic metastasis in a primarily at autopsy, in 3 1 patients; for the other 3 4 , patho-
large series of patients with melanoma, and to cor- logic data were not available, and the diagnosis of thoracic
relate these patterns with the symptoms, clinical involvement was based on abnormal findings on the chest
course, and survival of the patients. In addition, the x-ray film and pathologic, radiographic, or clinical evidence
of extrathoracic metastasis.
role of chest radiographic studies in evaluating pa-
Fifty-six of the 6 5 patients received systemic chemother-
tients with this disease is assessed.
apy, immunotheraphy, or both. Eleven patients had local
radiation t o the lung and mediastinum, and seven under-
MATERIALS AND METHODS went partial or complete lobectomies.
Forty-nine patients died. Thirteen have been observed
The medical records and chest x-ray films of 6 5 consecu- from one month to almost three years (mean, 9 months)
tive patients having thoracic metastasis from melanoma were since detection of thoracic metastasis. Three patients did not
reviewed. All of the patients were seen at the University of return for follow-up examinations.
California, San Francisco, from June 1 9 6 8 through September

" F r o m the Department of Radiology, University of Califor- RESULTS

nia School of Medicine, San Francisco.
Supported in part by National Institutes of Health Grant Sixty-three of the 65 patients demonstrated ab-
GM 0 1 2 7 2 from the National Institute of General Medical
Science. normalities on chest x-ray films; as shown by the
Manuscript received March 8 ; revision accepted July 7. following tabulation listing numbers of patients
Reprint requests: Dr. Webb, Radiology, 380-M, University of
California School of Medicine, San Francisco 94143 with various findings on plain chest x-ray films:

176 WEBB, GAMSU CHEST, 7 1 : 2 , FEBRUARY, 1 9 7 7

Normal chest x-ray film 2 T a b l e 1 — C o m p a r i s o n of Results of Plain Films
Pulmonary metastasis 57 and Tomograms in 11 Patients
Solitary nodules 14
Multiple nodules 41 Plain Film Tomogram Examinations
Miliary (snowstorm) nodules 8
Normal Lung nodules 2
Lymphangitic spread 5
Lung nodule (?) Normal 1
Enlargement of lymph nodes 28
Lung nodules Additional nodules seen 5
Pleural effusion 10
Lung nodules No additional nodules 2
Atelectasis and bronchial obstruction 8
Atelectasis Bronchial narrowing 2
Lytic bone metastasis 6
Cardiomegaly 4
patients exhibiting enlargement of hilar or medias-
The two patients with no radiographic abnormali- tinal lymph nodes or bronchial obstruction with
ties were found at autopsy to have small pulmonary atelectasis. The average survival of symptomatic
nodules. Sixty-two patients exhibited both intra- patients was five months, not significantly less than
thoracic and extrathoracic metastases; in 42 the that of patients without symptoms.
chest x-ray films provided the first objective evi-
dence of dissemination beyond regional lymph Pulmonary Metastasis
nodes. In two instances, pulmonary metastasis was
identified radiographically before detection of the Fifty-seven patients (88 percent) had pulmonary
primary tumor. metastasis detected radiographically. Metastases
were visible on plain chest x-ray films in 55. In five
The period of time from diagnosis of the primary
patients with radiographically normal lungs, multi-
tumor to radiographic recognition of thoracic metas-
ple small pulmonary nodules were found at au-
tasis averaged 32 months, but the range was broad,
topsy. Twelve tomographic examinations were per-
and in two instances, pulmonary metastasis ap-
formed on 11 patients, and ten of these examina-
peared 12 and 15 years after the initial diagnosis of
tions were helpful in determining the extent of
melanoma. The length of this period did not cor-
thoracic involvement (Table 1 ) . Tomograms were
relate with the length of subsequent survival. Sur-
particularly useful in detecting small nodules in the
vival after radiographic recognition of thoracic
lung, and in two patients with normal radiographic
metastasis averaged 7 months and was not greater
findings, tomograms disclosed pulmonary nodules
in patients with primary tumors that were slow to
for the first time. Despite the frequent finding of
central necrosis and hemorrhage in pulmonary
Thirty-eight patients had more than one radio-
nodules at autopsy, no instance of cavitation was
graphic examination. Of these 38 patients, 29 dem-
documented radiographically or pathologically.
onstrated progression of metastatic disease, five
Solitary nodules were identified radiographically
showed no change during a follow-up period of one
in 14 patients (Fig 1 ) ; multiple discrete nodules
to two months, and four patients had a decrease in
the size and number of visible nodules. Of the 29
patients with progression of metastasis, 20 were
receiving chemotherapy or immunotherapy. The
four patients showing regression of nodules were
also being treated.
Among the patients with progressive metastatic
disease, the time period over which nodules dou-
bled in volume ranged from less than one week

to five months, with a mean of approximately two

months. The rate of growth of the metastatic le-
sions or the presence of regression had little effect
on survival of patients. Metastatic lesions that
showed regression, were unchanged in size, or
doubled in a period of more than two months were
associated with a slightly longer survival period
(seven months) than those more rapidly growing
tumors (six months), but differences were sta-
tistically insignificant.
Twenty-three patients had respiratory symptoms,
most commonly cough, chest pain, and shortness of F I G U R E 1. Well-defined solitary nodule in left lower lobe. No
breath. Symptoms were most frequently found in other nodules were found at autopsy.


F I G U R E 2. Multiple large nodules located predominantly at
bases of lungs. These were detected four years after removal
of primary tumor.

were seen in 41 patients (Fig 2 ) . Lengths of sur-

vival after detection of pulmonary metastasis
averaged nine and seven months, respectively, for
these two groups and were not significantly dif-
ferent from each other. Innumerable miliary (or
snowstorm) nodules were recognized in eight pa-
tients (Fig 3 ) . Survival of patients with this pat-
tern of metastasis averaged only five weeks from F I G U R E 4 . Snowstorm metastasis with linear interstitial den-
the time of detection, significantly less than that of sities and Kerley's B lines. Lymphatic obstruction and inter-
stitial pulmonary edema were found at autopsy.
patients with either solitary or discrete multinodu-
lar lesions. Five patients demonstrated ill-defined,
Lymph Node Metastasis
linear interstitial densities, suggesting lymphangitic
spread ( F i g 4 ) . Survival of this group averaged Radiographically identifiable enlargement of in-
only four weeks, but three of these patients had trathoracic lymph nodes was present in 28 of the
superimposed miliary pulmonary metastases. 65 patients. Nodal involvement was associated with

F I G U R E 3 . Innumerable, small pulmonary nodules (snowstorm

metastasis). This man had been observed for 1 8 months be- F I G U R E 5 . Asymmetric enlargement of right hilar and right
cause of slowly enlarging solitary nodule but died six weeks paratracheal lymph nodes, with tracheal displacement to
after appearance of this pattern. left.

178 WEBB, GAMSU CHEST, 7 1 : 2 , FEBRUARY, 1977

in ten patients; in two patients, it was bilateral.
Most effusions were small and did not affect treat-
ment or survival. Enlargement of mediastinal or
hilar lymph nodes was seen in eight of the patients,
and in all eight, it was present only on the side of
effusion. The association of pleural effusion with
lymph node metastasis has also been noted in other
types of malignant neoplasms. 25

Atelectasis and Bronchial Obstruction

Bronchial obstruction in melanoma may be due

to endobronchial metastasis or compression by en-
larged lymph nodes. Eight patients demonstrated
segmental or lobar atelectasis. Enlargement of
ipsilateral hilar lymph nodes was present in five of
the patients; two had visible bronchial narrowing
(Fig 6 ) .
At autopsy, two patients who demonstrated ra-
diographic findings of atelectasis were found to
have tracheal or bronchial metastasis. In one pa-
tient, a fungating intratracheal mass 1 cm in diam-
eter, was found just proximal to the carina.

Bone Metastasis

Lytic bone metastasis was visible on the chest

x-ray films of six patients. Rib metastasis was pres-
ent in five patients; in three of the five, more than

F I G U R E 6. Large subearinal and left hilar mass with marked

narrowing (arrows) of left main-stem bronchus and volume
loss in left lower lobe.

radiographic evidence of pulmonary parenchymal

metastasis in 25 instances (Fig 5 ) . The average
survival period of patients with radiographic intra-
thoracic adenopathy and associated pulmonary
metastasis was not different from that of patients
showing a similar pulmonary parenchymal pattern
without mediastinal or hilar involvement. The
average survival period of the three patients having
only lymph node metastasis was three months.
Thirteen patients with no radiographically visible
enlargement of intrathoracic lymph nodes were ex-
amined at autopsy. Lymph nodes were involved by
melanoma in seven of these patients.

Pleural Effusion
Pleural effusion was identified radiographically F I G U R E 7. Lytic metastasis of rib and large extrapleural mass.


one rib was involved, and in four patients, it was metastasis was most frequent, but solitary nodules,
associated with an extrapleural soft tissue mass snowstrom nodules, and lymphangitic spread were
(Fig 7 ) . In radiographic studies of melanoma also recognized as distinct patterns. Cavitation of
metastatic to bone, rib involvement occurred in nodules has been reported but did not occur

more than one-third of the cases. - Two patients

20 21
among our patients and was not observed in 30 pa-
demonstrated compression of thoracic vertebral tients with pulmonary melanoma studied by Dodd
bodies, and Boyle. The radiographic pattern of pulmonary

metastasis has not been stressed in the literature as

Cardiac Metastasis useful for determining prognosis in this disease;
however, our patients with snowstorm pulmonary
Rapid cardiac enlargement has been reported as involvement had a shorter survival period than
the most significant radiographic sign of metastasis those with other patterns of disease. The many pul-
to the heart in patients with melanoma, but it is un- monary nodules reflect the diffuse dissemination
common. In our series, four patients had cardio-
that is responsible for the patient's death.
megaly. Autopsy was performed on two of these
Intrathoracic lymph node metastasis is frequently
patients, and cardiac metastasis was found in both;
identified in both surgical and autopsy series of pa-
however, cardiomegaly could not be attributed
tients with melanoma. - Almost half of our patients
7 9

solely to melanoma because in these two patients,

demonstrated radiographic enlargement of lymph
there was associated atherosclerotic myocardial dis-
nodes; nevertheless, more than 50 percent of our
ease. Four patients with autopsy-proved cardiac
patients without radiographically enlarged lymph
metastasis had a normal-appearing heart radio-
nodes had hilar or mediastinal metastasis at au-
topsy. Therefore, chest radiographic studies must
be considered of limited value in detecting nodal
DISCUSSION metastasis in patients with melanoma.
Chest x-ray films are essential in the evaluation Despite its value in detecting and delineating
and follow-up of patients with malignant mela- metastasis, chest x-ray films are of limited useful-
noma. Intrathoracic metastasis is usually recog- ness in the follow-up study of patients with known
nized before metastasis elsewhere becomes clini- intrathoracic disease. It has been suggested that
cally apparent. In 42 of our 62 patients with wide- chest x-ray films be used to assess systemic therapy
spread metastatic disease, the dissemination was in patients with melanoma by allowing an easy de-
first diagnosed on the basis of chest x-ray films. termination of change in size of metastatic de-
posits. In our study, regression of pulmonary

Because aggressive local or systemic measures

nodules was uncommon and had no discernible ef-
may be of benefit in treating metastatic melanoma,
fect on survival. In addition, prognosis was not
chest radiographic studies should be performed at
altered by the length of time between diagnosis of
frequent intervals to ensure the early detection of
the primary tumor and detection of thoracic metas-
metastasis. Since the average survival of our pa-
tasis, the pattern of metastasis (other than snow-
tients after radiographic recognition of thoracic
storm), or the rate of growth of metastatic lesions.
metastasis was only seven months, examinations
should be spaced no more than three months apart.
Full pulmonary tomographic studies should be
performed on all patients at the time of initial stag- 1 Attie JN, Khafif RA: Melanotic Tumors: Biology, Pathol-
ing and whenever metastasis is suspected, particu- ogy and Clinical Features. Springfield, III, Charles C
larly if local irradiation or surgery is planned. Small Thomas, 1964, p 122
2 Nathanson L , Hall T C , Farber S: Biological aspects of
pulmonary nodules not visible on plain chest x-ray
human malignant melanoma. Cancer 2 0 : 6 5 0 - 6 5 5 , 1967
films may be detected with tomograms. Respiratory 3 Dodd GD, Boyle J J : Excavating pulmonary metastases.
symptoms may be important in calling attention to Am J Roentgenol 8 5 : 2 7 7 - 2 9 3 , 1961
thoracic metastasis in some instances, - but only
7 8 4 Johnson RM, Lindskog C E : One hundred cases of tumor
11 of our 65 patients had related symptoms at the metastatic to lung and mediastinum. JAMA 2 0 2 : 1 1 2 - 1 1 6 ,
time metastasis was detected.
5 Minor G R : A clinical and radiologic study of metastatic
Pulmonary metastasis is common in malignant pulmonary neoplasms, J Thorac ,Surg 2 0 : 3 4 - 4 2 , 1950
melanoma and was found in 82 percent of the 6 Steele J D : The solitary pulmonary nodule: Report of a
patients examined at autopsy in one study. Pul- 2 cooperative study of resected asymptomatic solitary pul-
monary nodules in males. J Thorac Cardiovasc Surg 4 6 :
monary metastasis was radiographically identified
2 1 - 3 9 , 1963
in almost 90 percent of our patients with intra-
7 Das Gupta T , Brasfield R: Metastatic melanoma: A clin-
thoracic dissemination. Multinodular pulmonary icopathological sUidy. Cancer 1 7 : 1 3 2 3 - 1 3 3 9 , 1964

180 WEBB, GAMSU CHEST, 7 1 : 2 , FEBRUARY, 1977

8 Cahan W G : Excision of melanoma metastases to lung: 18 Cohen GU, Peery T M , Evans J M : Neoplastic invasion
Problems in diagnosis and management. Ann Surg 178: of the heart and pericardium. Ann Intern Med 4 2 : 1 2 3 8 -
7 0 3 - 7 0 9 , 1973 1245, 1 9 5 5
9 Freundlich IM, Capp M P : Granulomatous disease of 19 Glancy D L , Roberts W C : T h e heart in malignant mela-
the lungs. Radiol Clin North Am 1 1 : 2 9 5 - 3 1 6 , 1973 noma: A study of 7 0 autopsy cases. Am J Cardiol 2 1 : 5 5 5 -
10 Hilbish T F : Roentgen manifestations of malignant mela- 571, 1968
noma. Am J Roentgenol 7 8 : 7 6 9 - 7 7 9 , 1957 2 0 Selby H M , Sherman RS, Pack G T : A roentgen study of
11 Prichard R W : Tumors of the heart: Review of the sub- bone metastases from melanoma. Radiology 6 7 : 2 2 4 -
ject and report of 150 cases. Arch Pathol 5 1 : 9 8 - 1 2 8 , 1 9 5 1 2 2 8 , 1956
12 Reed RJ, Kent E M : Solitary pulmonary melanomas: T w o 21 Sreiner GM, MacDonald J S : Metastases to bone from
case reports. J Thorac Cardiovasc Surg 4 8 : 2 2 6 - 2 3 1 , 1 9 6 4 malignant melanoma. Clin Radiol 2 3 : 5 2 - 5 7 , 1972
13 Rode I: Clinical and Radiobiological Properties of Mela- 22 Clark W H , From L, Bernardino EA, et al: The histo-
noblastoma. Budapest, Akademiai Kiado, 1968, p 1 1 8 genesis and biologic behavior of primary human malig-
14 Sutton F D Jr, Vestal R E , Creagh C E : Varied presenta- nant melanomas of the skin. Cancer Res 2 9 : 7 0 5 - 7 2 6 , 1969
tions of metastatic pulmonary melanoma. Chest 6 5 : 4 1 5 -
2 3 McGovern V J : The classification of melanoma and its
419, 1974
relationship with prognosis. Pathology 2 : 8 5 - 9 8 , 1 9 7 0
15 King DS, Castleman B : Bronchial involvement in m e -
2 4 Nathan M i l , Collins VP, Adams RA: Differentiation of
tastatic pulmonary malignancy. J Thorac Surg 1 2 : 3 0 5 -
benign and malignant pulmonary nodules by growth rate.
3 1 5 , 1943
Radiology 7 9 : 2 2 1 - 2 3 2 , 1962
16 Schoenbaum S, Viamonte M : Subepithelial endobronchi-
al metastases. Radiology 1 0 1 : 6 3 - 6 9 , 1971 2 5 Meyer P C : Metastatic carcinoma of the pleura. Thorax
17 Anderson C B , Philpott GW, Ferguson T B : The treat- 21:437-443, 1966
ment of malignant pleural effusions. Cancer 3 3 : 9 1 6 - 26 Holland J F , Frei E : Cancer Medicine. Philadelphia, L e a
922, 1974 and Febiger, 1973, p 1 8 3 5

Advances in Clinical I m m u n o l o g y and A s t h m a
Beth Israel Medical Center (New York) and the Medical Center. For further information, contact the
National Jewish Hospital and Research Center (Den- Postgraduate School Office, Beth Israel Medical Cen-
ver) will present the course, Advances in Clinical ter, Ten Nathan D. Perlman Place, New York, New
Immunology and Asthma April 27 at the Beth Israel York 10003.

Clinical Cytopathology for Pathologists

The 18th Postgraduate Institute for Pathologists in formation, write John K. Frost, M.D., 610 Pathology
Clinical Cytopathology will be held April 11-22 at Building, The Johns Hopkins Hospital, Baltimore,
The Johns Hopkins University School of Medicine Maryland 21205.
and The Johns Hopkins Hospital, Baltimore. For in-

Current Problems in Intensive C a r e — M e c h a n i c a l Ventilation

The Department of Anesthesia, in cooperation at the Stanford Court Hotel, San Francisco. For in-
with extended programs in medical education, Uni- formation, contact Extended Programs in Medical
versity of California School of Medicine, San Fran- Education, Room 569-U, University of California,
cisco, will sponsor the course, Current Problems in San Francisco 94143,
Intensive Care—Mechanical Ventilation, April 28-29

A m e r i c a n Heart Association/Virginia Affiliate

The American Heart Association/Virginia Affiliate tails, write the American Heart Association, Virginia
will present a scientific program at the 1776 Inn, Affiliate, PO Box 12365, Richmond, Virginia 23241.
Williamsburg, Virginia, April 22-23. For further de-