MALIGNANT TUMORS OF THE TONSIL*

VoL. 114, No. I MALIGNANT Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For persona...
Author: Jasper Powers
2 downloads 2 Views 932KB Size
VoL.

114,

No.

I

MALIGNANT

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

ANALYSIS By CARLOS

OF A.

PEREZ, JOSEPH

TUMORS

FAILURES

AND

FACTORS

M.D., LAUREN V. H. OGURA, M.D., ST.

OF

UMEROUS reports have been pub lished on the treatment of carcinoma of the tonsil, either by a radical surgical resection2’7”4”7’27 or by radiation therapy.6’8”5’ 19,21-24,26,30 In the more recent past, a combined approach using preoperative irradiation and operation has been practiced at some institutions.’3”9’24 Five year cure rates have varied fiom 15 to 40 per cent. Although several publications have analyzed the curability of these lesions depending on clinical stage, only a few attempts have been made to systematically correlate the factors which may exist in the host or the tumor that will influence the results of treatment. Profound differences in pathologic classification are encountered in the literature22 and most of the reports deal with epithelial tumors, !ymphomas frequently being excluded. The present study attempts to correlate the various parameters with survival. The mechanisms of failure for treatment modalities were analyzed. The medical records, as well as the pathologic slides and reports of i. patients with various types of carcinoma and 25 patients with malignant lymphoma of the tonsil were reviewed in retrospect. The material comprises the experience in the treatment of carcinoma of the tonsil at the Mallinckrodt Institute of Radiology and Barnes Hospital Medical Center, Washington University School of Medicine, St. Louis, Missouri, from January, 1950 to December, 1967, all patients being available for a minimum 3 year follow-up and 132 for years. Included are a group of patients who received preoperative radiation therapy at our institution, Presented

at the

Fifty-third

Annual

Meeting

of the

M.D.,f WILLIAM B. E. POWERS, M.D.

MILL,

M.D.,

hut were operated upon at the Veterans Administration Hospital and St. Louis City Hospital by members of the Washington University Staff. Eight additional patients referred from other institutions for the treatment of recurrent tumors are not included, but 3 patients receiving incomplete irradiation and 4 patients in the preoperative group found to be inoperable are included. The results in the treatment of a significant portion of these patients have been published recently.’9 METHODS

As described tients

(2)

American

Radium

in

TREATMENT

that by

publication, different

the methods:

pa-

Radiation therapy alone. Irradiation was given to 6i patients with 5,ooo tads tumor dose (TD) in 6 weeks (orthovoltage-prior to 1959) or 6,ooo rads TD with cobalt 6o (after 1959). Usually the tonsillar fossa and adjacent tissues, as well as the entire neck were irradiated. Only a few cases with T, lesions were treated with an ipsilateral port and the lower necks were not irradiated in 4 patients. Preoperative irradiation. Forty-eight patients were treated by preoperative cobalt 6o irradiation to the primary tumor and the ipsilateral neck lymph nodes, usually to a dose of 2,000 or 3,000 rads, delivered at a rate of i,ooo rads per week. Patients with more advanced lesions on occasion received 4,000 to 5,000 rads preoperatively; 3-6 weeks after completion of the irradiation, a radical tonsillectomy, combined with a jawSociety,

Radiation Therapy, Mallinckrodt Institute of Radiology Washington University School of Medicine, St. Louis, Missouri. Address: Lester E. Cox Medical Center, Springfield, Missouri.

43

OF

treated

were (i)

From the Divisionsof Otolaryngology, § Present

PROGNOSIS

MISSOURI

N

*

TONSIL*

AFFECTING

ACKERMAN, and WILLIAM

LOUIS,

THE

Mexico, and

Surgical

D.F.,

Mexico, Pathologyj

March and

15-18,

the

1971.

Department

o

Perez,

44

Ackerman,

Mill,

neck

dissection was performed. Four initially admitted to this group were found to be inoperable at the completion of their therapy because of poor regression of massive oropharyngeai cancer or large fixed neck lymph nodes. Surgery. A radical tonsillectomy with resection of the contiguous tissues in the soft palate, lateral pharyngeal wall and base of tongue was performed in 19 patients, usually in combination with an ipsilateral neck dissection. Ri don implantation. A small group of i6 patients was treated by the Division of Plastic Suigery with radon implantation to the tonsillar fossa (10-20 one millicurie seeds) in the majority of the cases combined with an ipsilateral radical neck dis-

Ogura

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

(.)

Powers

T2-Lesions

patients

(,)

and

JANUARY,

3-5

tension

to

T,-Lesions limited

with

cm., adjacent

larger extension

1972

minimal

ex-

structures

than 5 cm., to adjacent

with struc-

tures

T4-Massive

tumors

nasopharynx tongue,

geal

the to the base of the or lateral phaiyn-

extending

down epiglottis

from

wall

Metastatic

Lymph

Nodes

in the Neck

N0-No clinical lymph node Ni-Single, less than 3 cm. mobile lymph node N2-Single, mobile lymph than 3 cm. or multiple non-fixed lymph nodes N3-Large fixed, unilateral or bilateral lymphadenopathy ANALYSIS

OF

involvement in

diameter,

node larger ipsilateral lymph

node

RESULTS

section.

The

The were

patients treated

lymphoma

with malignant

to

doses

ranging

from

2,000

in the orthovoltage era and 3,500 to 4,003 rads after the installation of a cobalt 6o unit in 1959. In the earlier years, no attempt was made to treat the entire neck, but this has been routinely done in the past ia years. No prophylactic radiation of the mediastinum has been done. In patients with abdominal or other to 3,500

rads

manifestation therapy

of

has

been

tumor

spread,

combined

radiation

with

absolute 3 year cure rate with the methods of treatment was about 50 per cent and at 5 years ranged from 31 to4o per cent (Fig. i,zl and B). Six patients lost to follow-up are considered dead with tumor. All patients have been followed up until December, 1970 or until death. various

chemo-

therapy.

AGE

No striking difference was noted in the 5 year absolute cure rates when correlated with the age of the patients at the time of treatment. The table with these results was analyzed but not published for the sake of brevity.

CLASSIFICATION

Using the description the lymph nodes on the

of the tumor and hospital and radiation therapy records at the time of the initial examination, the patients with epi theli al tumors were retrospectively staged according to the following classification Primary

:15 Tumor

T,-Lesions localized less than 3 cm.

to in

the diameter

tonsil,

but

SEX

of the tonsil is predominant about 20 per cent of the cases in women. Although some authors have observed a better prognosis in this latter group,17’25’27 in our experience, the survival rates were comparable in both sex groups. In the preoperative irradiation group, there is a larger proportion of women surviving without tumor at 3 years, but on closer analysis 4 of them had early Carcinoma

in men, presenting

only

VOL.

114,

lesions

Malignant

No.

(T,,

T2),

which

carry

a better

Tumors

of the

prog-

45

Radiation 22/48

nosis. CLINICAL

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Tonsil

STAGE

Primary Tumor. Some differences in the various methods of treatment were noted in the cure of the tumor in relation to clinical staging. The results are not statistically significant because of the small number of patients treated in each group. However, all of the patients with T, lesions treated by irradiation alone, about 50 per cent of the ones with T,, and 35 per cent of those with T, lesions were cured. Except for the T, lesions, the results are comparable with the other methods of treatment (Fig. 2). All methods of treatment were equally ineffective in the treatment of massive T4 lesions, and only rarely was a patient cured. Metastatic Lymph Nodes. As has been reported in the past, patients with no clinical evidence of metastatic neck lymph nodes at the time of treatment had a very good prognosis. The 3 year cure rate was 75-80 per cent for the patients treated with irradiation alone or combined with a radical surgical procedure and o-6o per cent for those treated by surgical resection or radon implantation (Table I). The 3 year survival rate of patients with ipsilateral lymph nodes was about 40 per cent and comparable for all types of treatment. Patients with bilateral lymphadenopathy did very poorly, only I surviving after

MALES

(47%)

FEMALES

5/13 (3B%j

P3,

P-e-0

Pre-Op

Surgery

16/38

8/16

7/14

(42%)

(50%)

(50%)

7/10 (70%)

1/3

1/2

1

1

S’ep

4/12 330’

4/5 (5/0)

Totil

P:

// 23

,03/ /2

3fl /2S (30

Radon

L p:a

3/

4/12 (330)

MALES

FEMALES

42

30

20

12

B0 I. Carcinoma of the tonsil. Graphs showing (A) 3 and (B) 5 year absolute cure rates. Essentially the same cure rates were noted in men and women.

FIG.

irradiation. PATHOLOGIC HISTOLOGIC

FINDINGS CLASSIFICATION

All the pathologic slides reviewed and the patients (Table ii). The pathology of cancer concerned with squamous lymphomas. The diagnosis cell carcinoma following made without difficulty, instances of so-called thelioma.”

and reports were were reclassified of the tonsil is cell carcinoma or of squamous biopsy is usually except in the few “lymphoepi-

The microscopic pattern of squamous cell carcinoma can vary within a wide range (Fig. 3, ii and B; and C and D). Squamous cell carcinoma, both in the primary lesion and within the lymph node, can be sterilized by radiation therapy. At times, this results in large masses of keratin surrounded by giant cells (Fig. 4), or there can be replacement of the lymph node by fibrous tissue (Fig. 5). It is difficult to decide in

46

Perez, Ackerman, RADIATION

No

5/8

1-_.L 3/6

2/3

Ni

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Total

N2

-

1/3

13PRE-OPERATIVE

fl1TT2

-Tal

Ni

1/5*

1/74

&

Rx

-2JI16

N2

O/2

1/2

3/11

Il/lfl

N3

I

I03I_--1 0/3

13/40 0/2

Mi

Mi

t3.6

8/14 5/5

I

(100%)

lotal

3/7 Total

(4

(32

(27%)

3%)

SURGERY

12

RADON

13

No

1 /2

fli-14iotai

2/2

1972

6/10

13/19)

1-L43/11

I/I

N3

JANUARY,

12

ALONE

13

No

and Powers

Mill, Ogura

Ni

LffI

&jotl 4/6

1/4

0/2 1/6

N2

N2

1/2

0/2 0/1

N3

N3

0/1

6/16

Mi 0/6 2/4 4,9 Io

FIG.

2.

2/7

tol

of the tonsil. Five

Carcinoma

year

t Postoperative death in each group; # no autopsy; 0’ patient inoperable-incomplete (6,ooo rads). Eight additional patients treated

None

14/19

(%) Ipsilateral

*

Six

plete

t

patients

I/TO with

.

distant

Metastasis-

Surgery

8/io

6/32

(8o%)

(50%)

(45%) #{176}/4t metastasis;

Four One

NED=

patients inoperable; postoperative death no evidence of disease.

3 postoperative in each group.

Radon

4/6 (66%)

3/7

4/10

(43%)

(40%)

-

-

3 patients

therapy. deaths.

stage

with

various

treatment

modalities.

postoperative deaths; * patient died suddenly before therapy (3,ooo rads); &I patient inoperable-complete less than 5 years ago: 3 are alive, tumor free.

2

TONSIL

15/33

12/32

(37%) Bilateral

Preoperative Irradiationt

by clinical

had

incom-

surgery, therapy

instances whether the tumor is viable (Fig. 6). Difficulties occur in the diagnosis of malignant lymphoma because often there is an accompanying inflammatory reaction which masks the picture of lymphoma. In the cases of lymphoma, we have used the classification widely accepted as outlined by Rappaport.2#{176} Table ii shows that the 5 year cure was slightly better for those patients with keratinizing squamous cell carcinoma (,j per cent) than for non-keratinizing (30 per cent). The patients with well-differentiated lesions had a somewhat better prognosis (50 per cent), but no significant difference was noted when the tumor was moderately or poorly differentiated (about 35 per cent cure rates). many

OF THE

Clinically Palpable Lymph Node 3 Yr. Survival-NED

Radiation Therapy*

rates

I

TABLE CARCINOMA

cure

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

VoL.

Malignant

No.

114,

Tumors

of the

Tonsil

47

Although “lymphoepithelioma” is common in the nasopharynx, where it carries a good prognosis, it is extremely unusual in the tonsillar area. Two out of 3 patients with this diagnosis survived years, which again may indicate a favorable host response to what is now known to be a nonkeratinizing squamous cell carcinoma. Most of the malignant lymphomas of the tonsil were classified as either lymphocytic or histiocytic

types

(former

lymphosarcoma

and The

reticulum

cell sarcoma,

prognosis

was

ent

for

2

iii).

Only

histologic variants (Table of Hodgkin’s disease was

OF

PATHOLOGIC

these i case

not

respectively).

significantly

differ-

observed. STATUS

SPECIMEN

The

3 and year survival rates were corwith the presence of “viable” tumor cells in the primary tumor or in the neck lymph nodes (Table iv). related

Radiation who had

(i)

.-

been



..

‘I

Therapy.Fourteen

patients

with

treated

radical

FIG. 3. (C) Rare pattern be called transitional (D) Undifferentiated (35#{176}x).

..

-

,.

.

#{149}--

#{149}..-

4

surviving.

1*1

4-,

that could (iso x). carcinoma

irradiation were operated upon: in 3 instances because of residual ulceration in the tonsil; and in the remaining ii because of persistent induration in the neck, which was suspected by the otolaryngologist to be residual tumor. Of the 7 patients without viable tumor cells, 6 survhed for 3 years with no evidence of tumor and #{231} for 5 years. Two patients out of 3 with non-viable residual carcinoma in the neck lymph nodes survived. However, the presence of intact tumor cells in the tonsil or the lymph nodes carried a very poor prognosis, only i out of patients

A’ -.

of squamous cell cell carcinoma squamous cell

‘A

.?Sl1 FIG. 3. (A) Photomicrograph of heavily keratinized squamous cell carcinoma (300 x). (B) Nonkeratinizing squamous cell carcinoma with a pattern of so-called “lymphoepithelioma” (350 x).

(2)

Preoperative tients with

Irradiation. All 8 panegative specimens are surviving 3 and 5 years after treatment. Of those with no tumor in the tonsil, but positive lymph nodes, 2 out of 6 survived 3 years and none

48

Ackerman,

Perez,

Mill,

Ogura

and

Powers

JANUARY,

1972

tonsil

in the survival of the patients by this modality. About per cent of the patients with positive treated

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

primary

tumor

lymph per

15

(,)

and

positive

nodes

survived 5 years.

cent

Surgery.

Radon

(4)

neck and

Implantation.

that the absence

it is noted

Again,

3 years

30

of

and negative neck lymph nodes in the specimen carries an excellent prognosis, most of the patients surviving (675 per cent). If there was evidence of residual tumor in the neck lymph nodes, 40 per cent of the patients survived 3 years and 20 per cent 5 recurrent

primary

tumor

years. NUMBER

OF

INVOLVED

OLOGIC

As

11G.

.

cinoma apy

(A) Undifferentiated squamous with very little keratinization, (350 x). (B) Following therapy,

large

masses

(150

x).

of

keratin

surrounded

cell

before there by

giant

car-

therare cells

out of , years. Of the patients with a positive primary tumor, but negative lymph nodes, out of 8 survived 3 years and 3 out of 7, 5 years. This suggests that the presence of positive lymph nodes is more significant than residual tumor in the

LYMPH

SI’ECIMEN

NODES AND

IN

THE

PATH-

PROGNOSIS

with no evidence of tumor in the neck lymph nodes at the time of radical neck dissections had a better outlook: 23 out of 32 patients survived 3 years (72 per cent) and i8 out of 29 survived years (62 per cent). Patients with a single positive lymph node also had a good prognosis, io out of 17 surviving 3 years and 5 years (59 per cent?). In contrast, only 8 out of 24 patients with 2 to 4 positive lymph nodes survived for 3 years (33 per cent) and only 2 out of 21 for 5 years ( per cent). Patients with more than 5 positive lymph nodes had an equally unfavorable prognosis, 4 out of 14 (28 per cent) being alive at 3 years and none out of patients

13

patients

expected,

5 years

at

(Tables

v

and

vi).

I-:

H’

--

-

6. After radiation therapy, keratinization is noted in an undifferentiated squamous cell carcinoma. There are masses of tumor cells which

11G.

11G.

5. Cervical

fibrous

tissue.

serially

sectioning

lymph No

node tumor

the

lymph

partially was node

replaced

found

(

after x).

by sub-

are

probably

ment

can

not

be made

viable,

(iso X).

but

no

definite

state-

VOL.

114,

Malignant

No.

Tumors

of the

CARCINOMA

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Radiation Therapy Alone

.

Squamous

Cell

Classification

Carcinoma

Keratinizing Well differentiated

3/3

49

II

TABLE

Histologic

Tonsil

OF THE TONSIL and

5 Year

Preoperative . Irradiation

Absolute

Cure

Surgery

.

Total

Radon

‘/3

2/4

Rates

c/

6/Il (54%)

Intermediate

6/15

6/23

4/8

3/5

6/14

i/

3/5

2/5

19/51 (37%)

Poorlydifferentiated undifferentiated

and

10/28

(%)

Non-Keratinizing Well differentiated

I/I i/6

Intermediate

5/’,

6/i6

9/3!

15/32

2/2

35/90

(%)

I/I

-

-

0/2

o/i

3/IT (30%)

Poorly differentiated undifferentiated

and

4/13

1/3

1/5

6/21 (27%)

4/8

5/19

0/2

1/4

!0/33

o/z

oh

Unspecified “Lymphoepithelioma” Undifferentiated

2/2

Transitional

i/i

o/z

Total

23/57

6/39

13/40

It

is worthwhile to point out that the of the patients with 2-4 positive lymph nodes surviving are in the group treated by combined preoperative irradiation and surgery. As was indicated in a previous publication,’9 this treatment modality has been characterized by a prolongation of tumor-free life span, which is confirmed by the significant number of patients surviving 3 years who died of their tumor by the fifth year. OF

0/4 i/i

majority

CAUSE

2/3

o/z

0/3

(30%)

DEATH

It is apparent that most deaths are due to uncontrolled tumor (about 30-40 per cent of all patients; Table vii). The probability of failure at various anatomic areas is analyzed in detail below. Since a significant proportion of the patients are in an advanced age, it is not uncommon to find intercurrent disease as a cause of death, particularly after several

years

of

6/,6

48/132

life. As was reported a greater proportion of complinoted in those patients treated

tumor-free

previously, cations was

III

TABLE LYMPHOMA

Histologic

Type

Lymphocytic Well differentiated Intermediate Poorly differentiated Hystiocytic Well differentiated Intermediate Poorly differentiated Hodgkin’s Total

disease

OF THE

TONSIL

5 Year Cure Rate

2/4

1/3 2/7

5/14

(%)

4/Jo

(40%)

I/I 1/5 2/4

0/i 9/25

(36%)

Perez,

50

Mill,

Ackerman,

Ogura

CARCINOMA

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

OF

THE

Primary*_Neg.

Lymph

Nodes

Neg. Primary-Pos. Lymph Pos. Primary-Neg. Lymph Pos.Primary-Pos.LymphNodes Neck Dissection Not Done

Nodes Nodes

Total No

*

t

clinical

evidence

of tumor

with

by preoperative irradiation which is probably related trauma previously

The is

7 patients

in tonsil;

Non-viable tumor in 2 patients. Three patients inoperable, treated

of

the range with

viously

of

patients

5 Yr.

6/7*

5/7*

8/8

6/6

2/3t

2/3t

2/6

o/5

0/2

0/2

/8

3/7

4/6

1/2

0/2

7/21

3/18

4/10

18/47

16/43

I/

i/4

27/61

23/57

23/48

radical other

many

some

years

ago,

patients the records

were were

No. of Positive Lymph Nodes

Radiation Therapy

neck

Positive Neck Specimen Correlated Preoperative . Irradiationt

6/7

0

Yr.

3/6

3/4

3/4

2/To

4/9

2/9

i/

1/3

i/

i/

9/19

6/19

8/i6

6/16

residual

neck

lymph

nodes.

operation.

ANALYSIS

OF

In our previous report,19 recurrence in approximately of the patients developed

treated incom-

of

*

before

suspected

3 Yr.

pre-

and

TABLE

V

OF

THE

CARCINOMA

Number

for

Yr.

and no known cause of death could be ascertained in 9 patients. Practically all the patients with a malignant lymphoma who died, had generalized spread of their disease (Table viii).

tumors.’#{176}’1’

Because

died

13/40

3 Yr.

Radon

plete;

cancers

head

dissection

patient

surgery,’9

reported

having

neck

to the added procedure in

that

Surgery

.

3 Yr.

alone;

a radical surgical irradiated tissues. incidence of second primary

within

Preoperative . Irradiation

5 Yr.

had

Lymph With

FAILURES

it was 20-30

in

noted per

the

TONSIL

Radon

3/6

3/4

Total 23/32 (72%)

3/6

3/5

2/4

10/17

(59%) 2-4

0/4

3/5

5/12

0/3

8/24

(33%) or More

i/i

3/TO

0/I

0/2

4/14

(28%) *

Radical

t One

neck

patient Non-viable

dissection had neck tumor.

for

residual

dissection

mass done

after

previously

radiation

therapy.

for carcinoma

of pyriform

sinus.

that cent

tonsillar

Nodes in Initial Pathologic 3 Year Cure Rates

Surgery

.

11/15

2/2

1972

at Initial Treatment Cure Rates

3 Yr.

irradiation

and

JANUARY,

TONSIL

of Pathologic Specimen 3 and 5 Year Absolute Radiation Therapy Alone

Neg.

Powers

IV

TABLE

Status

and

VOL.

No.

114,

Malignant

I

Tumors

of the

TABLE CARCINOMA

Tonsil

51

VI

OF THE TONSIL

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Number of Positive Neck Lymph Nodes in Initial Pathologic Specimen Correlated with Year Absolute Cure Rates No. of Positive Lymph Nodes

Radiation Therapy *

Preoperative . . Irradiationt

s/’

0

Surgery

Radon

2/6

3/4

8/12

Total

18/29 (62%)

I

3/6

2/2

3/5

2/4

10/17

(59%) 2-4

0/4

1/9

1/5

0/3

2/21

(9%) or More *

t

Neck

o/i

dissection

One patient Non-viable

for residual had neck tumor.

0/9 mass

dissection

after

radiation

done

o/i

for carcinoma

of pyriform

area, with or without evidence of neck lymph node involvement or distant metastasis. Marginal failures were more common in the patients treated with preoperative irradiation (i per cent). Approximately

Cause

Tumor

VII

Disease

Second Primary Esophagus Lung Nasopharynx Colon

following

One patient dead Jaw-neck dissection

Pts.= patients.

TONSIL

(Unlimited

Time)

Preoperative

Therapy Pts.)

.

Surgery (19 Pts.)

.

(16 Pts.) 6

17

3

2

0

2 *

2t

I

I

I

i

i

(3rd Yr.) (6th Yr.) I (64 Yr.)

2

i

I

2

2

(i and

2

I

Yr.) I

I

surgical

removal

after cryotherapy for persistent

of tonsillar for tonsillar neck lymph

recurrence,

10

years

recurrence. node and necrotic

3

2

after

treatment.

ulcer

in pharynx

I month

P

6

21

3

Lymh

Radon

Tumor

Unknown #{149} Death

of Death

10

Complications Postoperative pneumonia Carotid rupture Postoperative cerebrovascular accident Cardiac arrest

t

OF THE

31

Intercurrent

sinus.

i per cent of the patients showed lymph node metastasis in the neck without either local or marginal failures. In the present analysis, we attempted to correlate the stage of the disease and the

TABLE CARCINOMA

(6i

0/13

therapy.

previously

Radiation

0/2

after

treatment.

Pelez,

52

LYMPHOMA

Sites

Mill,

VIII

TABLE

MALIGNANT

Ackerman,

of

OF

THE

Ogura

irradiation

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

and No.

Total No. Failing -i6

of Patients-25

Tonsil

irradiation

few ease.

44

II

#{231}

20 20

6

Other

24

8 7 3 3

marginal

28

group

12

extension

12

4

I

4

Bone Marrow

4

16

Bone Metastasis

3

12

Unknown

2

counted

8

as i in every

location.

of failure according to the methods of treatment. This was done by tabulating the number of patients failing and each failure site was counted as i, so that a patient with local recurrence and positive neck lymph nodes would be represented in both groups. This way, the probability of a patient failing at a given site was determined (Fig. 7; and 8). With T, tumors, no failures were seen in patients treated by irradiation. Recurrences were noted in 3 out of 8 treated with mechanism different

preoperative

irradiation 40

per

cent

surgery alone or Most of the failures were either in the

ipsilateral With

and of

those

combined in the tonsillar

surgery

and

with latter fossa

in

with radon.

treated

2 groups or in the

neck lymph nodes. T2 lesions, the rate of failure

was

local

common

reported failures were

due

the 29

per

patients)

the

(about

majority treated by per

(23

(21

per

prognosis,

is in the incidence

of

cent).

only

a

disrange of 6 of distant their

cent.

previously,’9 some in the irradiation to

cent)

preoperative the marginal

control

failure

or higher, being

various in the

tumor

a poor

having

The

32

1

failure

more

have

patients

As

Viscera Spleen Liver Lung Brain Stomach Adrenal

about

were

per cent metastasis

recurrent

1972

group. a large pro-

It is noted that in the patients alone were local to those in the group in which

T4 lesions

Mediastinal

Each

opposed

failures

Lymph Nodes Retroperitoneal Neck

*

showed

cent). failures

irradiation

4

I

were local the preoperative with T3 lesions,

in

patients

6o per of the as

(

group

portion

Per Cent of Patients Failing at Each Site

No. of Failures

cent in the of the failures

per

Most

marginal

In

JANUARY,

35

approximately groups treated.

TONSIL

Failure*

Total

Poweis

and

of

nasophaiyngeal

the alone

tumor

which was not adequately covered by the radiation fields. The probability of failing in the ipsilateral neck lymph nodes is the same for the irradiation alone and the preoperative irradiation groups (i and i6 per cent, respectively) , but contralateral lymph node involvement was slightly more frequent in the preoperative irradiation group (12.5 per cent), where no treatment had been given. A large, but comparable proportion of patients died with distant metastasis in these 2 groups (19 per cent in the irradiation alone group and 21 per cent in the preoperative

The nant

irradiation

group).

the patients with maliglymphoma failed with known or strongly suspected retroperitoneal lymph nodes (. per cent) or with involvement of the liver, the spleen or both (Table yIn). Invariably, all the patients failing had generalized manifestations of the disease. However, the incidence of recurrence in the tonsil or in the neck lymph nodes was relatively low. Mediastinal spread was known in only 5 patients (20 per cent). Similar observations have been reported by Terz and Farr.28 In the past 5 years lymphangiograms as part

majority

of

have of

the

been

initial

obtained

routinely

work-up

in

an

ef-

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Vo,..

‘i,

Malignant

No.

Tumors

Radiation Pre-Op

23

Surgery

the Tonsil

TOTAL

NO.

PATI ENTS _______________ NuM3ER

15 (65%)

-

1

3

3

(13%)

1

2

2

FAI

-

6

5

4

(21%)

(16%)

(17%)

4

ENTS

2 MARGINAL

-_

2

(12%)

4

(16%)

IPSILATERAL

3 (12%) (21%)

NODES

‘LATERAL

NECK

(8%

2

(23%)

2

NODES

_____________

(20%)

inoperable committed 1 month after

2

_____________

NECK

2 patients .1 patient suicide

13 (76%)

8

(21%)

-

17

LING

LICAL

(24%)

2 -.-__

PATI

(60%)

4

53

______________Radiation

25

Radon

-

of

.

DISTANT

/TASTASIS OPPOSITE

operatio

FIG. 7. Carcinoma of the tonsil. Sites of failures by clinical stage of tumor with various treatment modalities (multiple failures counted as one in each location). Notice the excellent local control with radiation therapy alone in T, and T, lesions. In contrast, a high incidence of local failure is noted in T, and T4 lesions.

fort to better determine the extent of the disease. Because of the small number of patients, no conclusions can be drawn from these findings yet. DISCUSSION

Few reports are available in the literature analyzing the factors influencing the prognosis in carcinoma of the tonsil. However, fragmentary analysis of some of these is

encountered in the multiple papers with treatment results. In the present series, the age patients at the time of diagnosis and ment did not significantly influence over-all survival rates. Similar results been published by other authors.25 experience, the sex of the patients significantly affect the cure rate. et al.2’ reported a 5 year survival

dealing of the treatthe have In our did not Tapley rate

of

Perez,

54

Ackerman,

Mill,

Ogura oftime. of the

JANUARY,

1972

In our experience about 6o per cent recurrences appeared within i year

and

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Powers

and

8o

per

cent

between

2-3

years

after

treatment. Terz and Farr2T reported on 536 patients with carcinoma of the tonsillar area. After excluding i 14 for a variety ofreasons, the year survival for 303 patients treated with orthovoltage

with per

irradiation

was

tumor

control

a primary cent. In

treated

the

group

by surgical

rate

was

per

26

55 per

cent. In Stage cent with irradiation

per

*2

patients

in

to follow-up have died

lost

each

presumed with tumor

to

per

cent,

rate

of

resection, cent,

19.5

of

i86

42.9

patients

the

#{231} year

cure

with a local control i the cure rate was and

58

per

of 5i

cent

with surgery, Stage ii, 24 per cent, and 36 per cent and in Stage iii, 20 per cent with irradiation and 30 per cent with surgery. Calamel and Hoffmeister7 reported 53.1 per cent 5 year cure with surgical resection (29 patients) in contrast to 13.4 per cent with irradiation alone (5 patients). The

FIG. 8. Carcinoma of the tonsil. Total incidence of failures. Sites of failure in entire group of patients with carcinoma, according to treatment modalities. Notice the higher local failure rate with radiation therapy alone, as opposed to a higher number of marginal failures in the patients treated by preoperative irradiation and operation.

although excess higher

77 per cent in a group of 17 women treated, of whom 13 had been followed for 5 years. This was contrasted with the 21 per cent cure rate in a group of 64 men treated, of whom 52 had been followed for years. Terz and Farr27 found a better survival rate for women (37.8 per cent) than men

cent in the T1 group and 28 per cent in the T4 group. The absolute 5 year survival rate was 45.6 per cent for 57 patients with T,-T2 lesions, and 6 per cent for the 5 patients with T3-T4 tumors. Fletcher and Lindberg,’#{176} in 89 patients

technique

of

it

irradiation

is

and

with

tients, having

Lampe9

been

contrast

to

reported

per

72

cent

brought

38.7

carcinoma

ofthe

with

radical

irradiation

reports

out

cent)

in

a large

group

of

patients

at Memorial has been

on

comparable

or

superior

results

with radical irradiation have appeared in the literature.6’ 8-10,15,19.2123,25,26,30 Several reports indicating the adequacy of suigical resection for these lesions2’4”4 have follow-ups of i year, on the basis that most recurrences appear within this period

performed

on

T1 and under

pa-

102

T2 lesions control,

in

per

with

per

described,

mentioned

Hospital in New York. made of the effectiveness of various treatment modalities. Although some authors have advocated surgical resection as the treatment of choice for carcinoma of the tonsil,24’7’14’17’27 numerous

(24

treated Claim

not

that “doses in of 5,000 rads TD and frequently than 6,ooo rads TD were given.”

Fayos

is

tonsillar

fossa

alone

in 4 for local

or neck

treated

(operation lymph

node

surviving 2 years (48 per cent). In the patients treated with megavoltage 29 out of 66 survived 3 years recurrences)

(43.9

reported

43

per cent absolute of 52 for 5 years

cure rate). Scanlon irradiation,

et al.22 either

(36.5

cure per

rate), cent

and absolute

19

treated 131 patients by with continuous or split

course. The year survival rate was 72 per cent for Stage i, 6 per cent for Stage II, 54 per cent for Stage III and 22 per cent

VOL.

114,

for Stage viva! rate

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

23 patients

tion per

Malignant

No.

Tumors

with an over-all year surper cent. A small group of treated with split course irradiai ,#{231} surviving at years (6

Iv,

of

52

showed cent). Most of the tumors of the tonsil have been classified as epidermoid carcinoma with different degrees of differentiation, or as undifferentiated tumors.21’22 A small number of transitional cell carcinomas or lymphoepitheliomas have been reported. Scanlon et al. observed no significant correlation between the pathologic nature of the primary tumor and the cure rate. Their survival rates, in the range of o per cent, are somewhat better than those in our series (39 per cent 5 year cure rate for keratinizing carcinoma and 30 per cent for non-keratinizing). In some ofthe reports in which the site of failure has been carefully identified, a significant proportion oflocal or marginal failures has been noted.’9’2’ Rider2’ found the highest incidence of recurrence in the base of the tongue (,j per cent of232 cases) compared with 13 per cent failures in the tonsil or pillars, 14 per cent in the neck lymph nodes and 14 per cent with distant metastasis. The year survival rate in patients with tumor confined to the tonsillar area was 31 per cent as opposed to 19 per cent in those with added involvement of the base of the tongue. His belief that failure to control tumor spread into the tongue is the most common cause of failure in treatment of cancer of the tonsillar area has not been substantiated in our series, in which only about 10 per cent failed in this particular anatomic site. In contrast, in 20 per cent of 89 patients with tonsillar fossa tumors, Fletcher and Lindberg’#{176} reported i8 primary local or marginal failures, of which 8 were in the tonsillar fossa, 4 in the base of the tongue and the remainder in the palate, posterior pillar, floor of the mouth and adjacent tissues. In 17 patients there was recurrence in the neck lymph nodes (i per cent); 8 patients were known to have died with distant metastasis.

of the

Tonsil

55

In general, the over-all incidence of metastatic neck lymph nodes is in the range of 60-70 per cent, 15 per cent of the patients presenting with bilateral lymphadenopathy.19 The presence of cervical lymph nodes at the time of treatment invariably has been found to affect the prognosis unfavorably, except for the series reported by Tapley et al.25 In our experience, the surviva! rate was closely correlated with the incidence of lymph node metastasis. Patients without lymph node metastasis had survival rates in the range of 70-80 per cent. With ipsilateral positive lymph nodes, the 3 year survival rate was about 40 per cent and only I patient with bilateral lymphadenopathy was cured. Similar results were reported by Scanlon et al.22 with survival rates in the range of 40-50 per cent in patients with ipsilateral positive neck lymph nodes and 10 per cent with contralateral or bilateral lymphadenopathy. The level of lymph node metastasis did not significantly influence the prognosis as long as they remained ipsilateral. Although some surgical series raise the question of the effectiveness of irradiation in controlling metastatic tumor in the lymph nodes of the neck,2 a growing body of evidence shows that this can be satisfactorily accomplished.’0”2’2’ In a group of 89 patients with carcinoma of the tonsillar fossa reported by Fletcher and Lindberg,’#{176} ‘7 failed in the neck ( of them combined with primary recurrence), and Rider2’ reported about i per cent neck lymph node recurrences. These failure rates are comparable to those in the present series. In our patients the site of failures has varied somewhat, depending on the various methods of treatment used. In the irradiation alone group, failures have been more common in the local area (19 patients, 3! per cent), whereas marginal failures were less frequent (6 patients, 10 per cent). The reverse is true for the patients treated with preoperative irradiation and operation, 6 out 0148 (i 2.5 per cent) failing locally and JO in the marginal tissues (2! per cent). Local failures were quite common in the

Perez, Ackerman, patients implantation

The

incidence node

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

lymph ing

by surgery alone or radon per cent). of ipsilateral metastatic

treated

recurrence

per

14.8

(

cent

radiation

with

preoperative patients). Slightly per

(25

cent)

treated by implantation

patients

( 19

per

noted

in

the

patients 01

commonly

by

seen

irradiation

halfofthem

alone

were

examination, be treated with larger lymph

is present the

of

in patients at the initial

in the

lower

neck,

neck

of

with

for

the the

5,000

nodes

equally

satisfactory

with high combination

In

are

dose

in

results

radiation of moderate

followed a jaw-neck large

present

in the

by

a

can

radiation.29

the

radical

large neck,

or with of irradia-

a

tonsillectomy

dissection.

lesions

tieated

The

theoretically of a resistant

addition eliminate tumor.

irradiation,

remains high, the presence of sensitivity to of

a

resection

this possible However, the

The

Judicious

However,

the

residual

primary

site

particularly

therapy,

was In

found

our

was negative the cure rate

after to

series,

(tonsil was in the

or

in

radical

aggravate when and range

the lymph of 8o

but in

decreased when carcinoma the specimen. Although not significant (P.o), the prog-

was more unfavorable when the nodes contained viable tumor.

neck

It is important to emphasize that after radical irradiation, residual neck lymph nodes should be carefully evaluated for unequivocal evidence of tumor growth when a surgical procedure is contemplated. A

significant

shows by

of the regresprerequisite is are to be ex-

tonsil.

results.

nodes,

cent, found

nosis lymph

with

be obtained

therapy doses

this incidence of local failure and this is probably due to hypoxic cells with decreased would foci1s

nodes

after elective irradiation given (maximum) dose. T3 tumors, particularly if

With lymph

tion with

lymph

necks rads

predetermined

irradiation.

the

in

prognosis.27

specimen nodes),

rads

of metastatic

lower

the

radiation

statistically

5,coo

of

lymph

incidence

of

value

resection of a persistent primary or a radical neck dissection in the of a residual mass may contribute either

per was

doses

marginal The

a

preoperative

carcinoma

surgical lesion presence

TD should be carried out. This subject has been extensively reviewed by Million et al.18 and later by Berger et al.,’ who observed a very low to

of

higher incidence of complications, some of them fatal, may significantly offset the theoretic benefit of this combined approach. From the analysis of these patients, it is concluded that until proven otherwise, radiation therapy remains the best treatment

tumor,

nodes

the

combina-

volume of tissue to be initiation of radiation emphasized.’9 The sur-

remove

to improve

elective

lymph

sion

pected

only the upper necks should in small T, lesions. However, tumors, or if a positive neck

node

irradiation

node metastasis lymph nodes

incidence

of tissue, regardless the tumor. This if better results

paramount

present

the time of the initial treatment. Although the number of patients is not adequate for statistical analysis, it appears that radiation therapy alone (6,5oo-7,ooo rads TD) offers the greatest probability ofcure in patients with early lesions. In our experience, all T1 and 6o per cent of the T2 lesions were successfully treated by irradiation. Because of the extremely low mciof lymph palpable

should

of

of this

appreciably.

of “tatooing” the resected prior to therapy has been

amount in

the

and pro-

regression

purpose

increasing

1972

of irradiation, in the surgical

of partial the

therapy,

geon

radon

Distant

at

dence without

defeat

failures

more

but

because

tumor, tion

treated

cent),

cedure

out

resection alone neck dissection.

were

be-

JANUARY,

of moderate doses times a compromise

at

alone

were

radical and

metastases

the

is comparable,

use

of6i patients) with and i6 per cent irradiation (8 out of 48 higher neck failure rates

therapy

and Powers

Mill, Ogura

proportion

fibrosis

and

no viable tumor. lary gland with confused Although

with

of keratin

Occasionally, extensive a neck

a surgical

cure some patients practice has been insttution because

the

specimens

formation,

but

a submaxilfibrosis may be

lymph resection

node. alone

with early tumors, discontinued at of complications

can

this our and

Voi.

Malignant

No.

114,

Tumors

of the

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

better alternative methods of treatment. Likewise, radon implantation has not been performed for at least 8 years.

Tonsil REFERENCES

1.

L. V., and DEL REGATO, J. A. Cancer: Diagnosis, Treatment and Prognosis. Fourth edition. C. V. Mosby Company, St.

ACKERMAN,

Louis, SUMMARY

The nosis

various in 169

factors patients

of the tonsil have cal stage of the presence

affecting with

are

the

lymph

most

prog.. tumors

been reviewed. primary tumor

of metastatic

neck

the

malignant

The cliniand the

nodes

important

in

the

prognostic

Age of the patient) sex and histologic classification of the tumor did not affect the prognosis. Patients with small primary tumors or no evidence ofmetastatic neck lymph nodes exhibited cure rates in the range of 70-80 per cent, in contrast to about 40 per cent factors.

for patients clinically

with palpable

more advanced lymph nodes.

of viable tumor in the tonsil lymph nodes of the surgical cated

a good

with

more

the

neck

prognosis, tion

positive

2

The

of

given

methods

most

of

the

However, lymph the

of failures of

patients nodes

carried dose

neck mdi-

treatment

radiation

of

Although

comparable

over-all

FAYOS,

cure

G. H., and mous cell carcinomas palatine arch. AM. THERAPY & NUCLEAR

FLETCHER,

II.

FLETCHER,

12.

HANKS,

13.

HENSCHKE,

reviewed, being

J. V., and LAMPE, I. Radiation of carcinoma of tonsillar region. ROENTGENOL., RAD. THERAPY & MED., 1971, III, 85-94.

10.

vanlocal

therapy AM. NUCLEAR

J.

LINDBERG, R. D. Squaof tonsillar area and J. ROENTGENOL., RAD. MED., 1966, 96, 574-

587.

rates

various methods of therapy alone, to doses of 6,oo to 7,000 rads, remains the best treatment for carcinoma of the tonsil. A higher incidence of complications was observed in patients treated by preoperative irradiation and operation. The importance of judicious post-treatment follow-up of these patients and of careful evaluation before performing radical surgery after high dose radiation therapy is stressed.

Carlos A. Perez, M.D. Mallinckrodt Institute of Radiology 510 South Kingshighway St. Louis, Missouri 63110

9.

1970.

G. W., and HEMMENWAY, W. G. Carciof tonsil: surgical treatment. LaryngoI 960, 70, 246-257. 3. R. R., and WEINER, S. Clinical manageof tonsillar carcinoma. Surg., Gynec. & 1965, 121, 1035-1038. 4. J. F. Cancer of retromolar area. I1.M.I1. Arch. Otolaryng., 1959, 69, 19-30. 5. BERGER, D. S., FLETCHER, G. H., LINDBERG, R. D., and JESSE, R. H., JR. Elective irradiation of neck lymphatics for squamous cell carcinomas of nasopharynx and oropharynx. AM. J. ROENTGENOL., RAD. THERAPY & NUCLEAR MED., 1971, III, 66-72. 6. BERVEN, E. G. Malignant tumors of tonsil. zicta radio!., 1931, Suppl. i I. 7. CALAMEL, P. M., and HOFFMEISTER, F. S. Carcinoma of tonsil: comparison of surgical and radiation therapy. Am. 7. Surg., 1967, “1 582-586. 8. DALY, J. F., and FRIEDMAN, M. Carcinoma of tonsil. Laryngoscope, 1960, 70, 595-615.

radia-

those in the preoperative irradiation surgery groups being marginal.

were observed with treatment, radiation

ALLEN, noma scope, BAKER, ment Obst., BARBOSA,

a poor

for the are

failures

2.

in

preoperatively.

mechanisms

Ous

or in the specimen

specimen

regardless

therapy

and and

prognosis.

than dissection

lesions or Absence

57

G. H., and MACCOMB, tion Therapy in Management the Oral Cavity and Oropharynx. Thomas, Publisher, Springfield,

W. S. Radiaof Cancers of Charles C Ill., 1962, pp.

249-262.

G. E., BAGSHAW, M. A., and KAPLAN, H. S. Management of cervical lymph node metastasis by megavoltage radiotherapy. AM. J. ROENTGENOL., RAD. THERAPY & NuCLEAR MED., 1969, 105, 74-8 2. E. L., HILARI5, B. S., H. R., and STRONG, E. W. Value of preoperative x-ray therapy as adjunct to radical neck dissection. Radiology, 1966, 86, 450-453. NICKSON,

U. K., FRAZELL,

J.

J.,

TOLLEFSEN,

04.

KLOPP, C. T., and SCHURTER, ment of cancer of soft Cancer, 1956, 9, 1239-1243.

15.

MACCOMB, W. S., and FLETCHER, G. H. Cancer of the Head and Neck. Williams & Wilkins Company, Baltimore, 1967, pp. 179-212. MARTIN, H. Case for prophylactic neck dissection. Cancer, 1951, , 92-97.

16.

M. Surgical palate and

treattonsil.

Perez,

58

Mill,

H. E., and SUGARBAKER, E. L. Cancer of tonsil. Am. 7. Surg., 1941,52, 158-196. i8. MILLION, R. R., FLETCHER, G. H., and J .SSE, R. H., JR. Evaluation of elective irradiation of neck for squamous cell carcinoma of nasopharynx, tonsillar fossa and base of tongue. Radiology, 1963, 8o, 973-988. 19. PEREZ, C. A., MILL, W. B., OGURA, J. H., and POWERS, W. E. Carcinoma of tonsil: sequential comparison of four treatment modalities. Radiology, 1970, 94, 649-659. 20. RAPPAPORT, H. Tumors of hematopoietic system. In: Atlas of Tumor Pathology. Section III, Fascicle 8. Armed Forces Institute of Pathology, 1966. 20. RIDER, W. D. Epithelial cancer of tonsillar area. Radiology, 1962, 78, 754-760. 22. SCANLON, P. W., DEVINE, K. D., WOOLNER, L. B.,andMCBEAN,J. B. Cancer of tonsil: 131 patients treated in II year period 1950 through 17.

Downloaded from www.ajronline.org by 37.44.207.21 on 01/24/17 from IP address 37.44.207.21. Copyright ARRS. For personal use only; all rights reserved

Ackerman,

1960.

SCHULZ,

treatment 1965, 24.

AM.

NUCLEAR

STAPLE,

M.

J.

ROENTGENOL.,

MED.,

1967,

RAD.

zoo,

75, 958967. T. W.,

HOLTZ,

TAPLEY,

26.

TELOH, Surg.,

27.

S., OGURA,

J.,

and

TERZ,

H.

125, 28.

A.

1952,

J. J.,

tonsillar

JANUARY,

1972

results of for corntreatment.

Cancer of tonsil. 65, 693-701.

and fossa.

FARR,

Surg.,

Arch.

A.M.A.

H. W. Carcinoma Gynec. & Obst.,

of 1967,

58 1-590.

TERZ, J. sarcoma

J., and

FARR,

of tonsil.

H. W. Primary Surgery, 1969, 65,

lympho772-776.

R. H. Oxygen therapy: biological considerations. In: Modern Trends in Radiotherapy. Edited by T. J. Deeley and C. A. P. Wood. Appleton-Century-Crofts, Inc., New York, 1967, pp. 52-71.

29.

THOMLINSON,

30.

WALKER, of tonsil: in group

THERAPY

arch cancerLaryngoscope,

Powers

N. DuV., EVANS, R. A., KLIGERMAN, M. N., and JACOX, H. W. Carcinoma of tonsillar area: factors influencing results of treatment. AM. J. ROENTGENOL., RAD. THERAPY & NuCLEAR MED., 1959, 82, 626-633.

25.

894-903.

D. Tonsil and palatine by radiotherapy.

and

POWERS, W. E. Carcinoma oftonsil: radiation therapy and consideration bined radiation and surgical Missouri Med., 1965, 62, 909-91 I.

MARTIN,

& 23.

Ogura

162-168.

J. H.,

and SCHULZ, M. D. Carcinoma review of treatment and its results of ninety cases. Radiology, 1947, 49,

Suggest Documents