- Email: [email protected]
Epidermoid carcinoma of the anal region
Epidermoid
Carcinoma INGUINAL
MAUS
IV. STEARNS,
of the Anal Region
METASTASES
JR., M.D., New l’ork, New l’ork
From the Rectum and Colon Service, Memorial Center for Cancer and Allied Diseases, New York, New York.
A
LTHOUGH a considerabIe amount of literature has accumulated pertaining to epidermoid carcinoma of the anal region, relatively little data have been presented concerning the management of inguina1 node metastases occurring in patients with this neopIasm.3j4,6,7 The tendency of squamous carcinoma in this Iocation to metastasize to the inguina1 Iymph nodes is we11 recognized. Frequencies of 8 to 33 per cent have been reported.?,4,5 The opinion has been expressed that inguinal metastases are not amenabIe to cure either by surgery or by radiation although they may be prevented 63 surgery.5 The high incidence of inguinal metastasc’s and the poor prognosis associated with their presence has Ied severa authors1~2~7 to advocate that “prophylactic” groin dissections be performed as a part of the idea1 treatment. OpllosecI to the theoretic advantage of bilateral prophylactic groin dissection is the not inconsiderabIe morbidi,ty associated with this procedure. This morbldIty is due initiaIIy to poor wouncI heaIing and subsequentIy to Iymphedema of the Iower extremity. Occasionally the lymphedema is severe enough to incapacitate a patient permanently. Final evaluation of “prophyIactic” groin dissection must await publication of resuIts obtained by its proponents. In the meantime, it seems of value to anaJyze the data we now ha\.-e in regard to these metastases. X4TERIAL This review- concerns sixty-nine patients \zith epidermoid carcinoma of the periana1 skin, ana cana and rectum who were admitted either as in- or outpatients to the Rectum and Colon Service of the nlemoria1 Center for Cancer and Allied Diseases from January, 1942, to December, 1952. Some of these patients had had definitive
therapy prior to coming to Memorial HospitaI. Some were seen when their neoplasm was beyond hope of cure. The therapy rendered at Memoriai HospitaI was modified by these conditions and did not always represent idea1 treatment. The primary neopIasms were treated by a variety of methods, primarily surgery with or without irradiation in the pre- or postoperative stage. The surgical procedure varied from JocaI excision for relatively smaI1 periana1 Iesions to combined abdominoperinea1 resection of the rectum accompanied by a meticulous abdominopeJvic node dissection. The combined abdominoperineaJ resection of the rectum which is routine in this clinic for this neopIasm stresses ligation of the superior hemorrhoidaJ vessels just beIow the left colic artery, wide removaJ of the peritoneum, wide perinea1 skin excision with compIete remova of the ischiorectal fossa contents, and division of the Jevator muscles at their origin. The abdominopelvic node dissection includes removal of the tissue from about the aorta, vena cava, the iliacs (common, external and internal), the obturator space, the presacra1 and aortic bifurcation. Tissues from these individual areas are Iabeled and sent as separate specimens in order that the exact sites of node metastases may be identified. AIL of the groin dissections studied in this review were performed in the presence of clinically diagnosed groin metastases. GeneraIIy, prior to dissection the presence of node metastases was proved bv aspiration biopsy or occasionaIJy by Jocal excision. Groin dissections in this group of patients are classified as superficia1 or radical. The former dissection is limited to the inguinal area (upper thigh and Iower abdomen) superficial to the fascia of the abdominal wall and the femoral area. The radicaJ dissection in addition includes the tissues from about the externa1 and internal iliac vesseis and obturator fossa.
Stearns Clinical Factors and Incidence OJ Groin A’ode Metastases. Tables I to v show the relation of the age and sex of the patient, duration of symptoms prior to treatment, and the location and grade of the tumor to the development of groin metastases. ObviousI\- the numf>er 01 patients is too small for statistical significnncc,
Inasmuch as the purpose of the present study is to add to our understanding of the probIems of groin metastases from epidermoid carcinoma of the anal region, the analysis of end results and the method of treatment of the primarylesion are considered only in relation to this purpose. I-ABLE
1
No.... Yes......
TABLE 11 SEX OF PATIENT Groin \letastases
No.+. ...........
i
0 7
14 Jli
;;,
TABLE
1No. of \I’omen
Groin klctastases
26
14
12
‘7
Grade I
I
Gde 11
Gde III
I8 IO
I&
No. Yes.
TABLE III DL RATION OF SYMPTOMS
I I Groin Lessthan 3-6 6--12 I -2 Over z Not Llctas- ~ Stated yr. Y’. t:1st’s 3 m0. I m”. ) m”.
\,cs.
12
1
4
6
4
I
Not GMkd /
I
No.....
I,
/
No. of Men
.’
Yes .............
2 3
2 3
TABL.E
II
i7
VI
i No. Yes..
Incidence of Groin Node Metastases. Of the sixty-nine patients studied who had epidermoid carcinoma of the perianal skin, anal canal or dista1 rectum, forty failed to present groin metastases either initially or subsequently. Of the forty patients, seventeen were followed up less than two years, and twenty-three were followed up more than two years. Twenty-nine patients or 42 per cent did have groin metastases. Of the twenty-nine, sixteen had clinically diagnosed groin metastases on admission, and in thirteen groin metastases became clinicaIIy recognizabIe after the treatment of the primary neopIasm had been compIeted. Of the thirteen that developed subsequently, eleven appeared within two years, two after two years (one at nine years).
II
‘3
15
3
f,
4
but there does not appear to fle any definite relation fletwecn any of these factors and the presence of groin metastases. In Table VI the relationship hetlvcen the type of therapy for the primary tumor and the subsequent appearance of groin metastases is presented. Again the numbers in each group are few but no appreciable difference is noted. Table VII indicates the type of surgery and subsequent groin metastases. Again no appreciabIe difference is noted. Sites f~f Lymphatic Metastases. Information concerning the lymphatic metastases from this form of cancer 728
was obtained
by a study
of the
Epidermoid
Carcinoma
of AnaI
Region
rhoidai lymphatic chains studied. trve or 13 per cent showed lymphatic metastases; of sixteen pelvic Iymphadenectomies performed (external and internal iIiacs and obturator space), seven or 43 per cent showed metastases.
operative specimens and is summarized in Table VIII. It may be assumed that the specimens, grouped here according to the type of incIuded the foIIowing surgical procedure, tissues: (I) 114’ “perineal resection” a wide local TABLE \‘II TYPE OF SURGERY
Perianal. Perianal
SITES
.‘.
!p2’,
J
TABLE ~111 01-‘ LYMPHATIC METASTASES
Procrdnrr
I
I
Deep Pelvic (Iliacs, Obturator)
Superior Hcmorrhoida1 Chain
No. _ ~___ ~__
I
Negative
.__
__
Neyative
Positive
o Ii I ~ I*
3*
Positive
The Iocation of the primary tumor as related to the sites of Iymphatic metastases is presented in Table IX, which when interpreteii in conjunction with the analysis of individual specimens indicates confirmation of described routes of Iymphatic drainage based primarily on anatomic studies.1j5 One route described is from the anal plexus through the perineal Iymphatics up over the upper thigh emptying first into the superficial groin nodes. Presumably, metastases to the deep pelvic nodes may occur from the superficia1 nodes. This route most satisfactorily explains the distribution of inguinal and pelvic metastases in the following patients: Two patients had tumors confined to the perianal skin with inguina1 metastases. In one the metastases were co&red to superficial nodes while in the other esterna1 iliac and obturator node metastases were present, also. Two patients with primary Iesions involving both the periana1 skin and the anal canal had superficial groin metastases without deep pelvic node involvement. Six patients with lesions confined to the ana canal had superficial groin metastases with negative deep pelvic nodes. Four patients with Iesions involving the upper anal canal and dista1 rectum had superficial groin metastases with negative deep peIvic nodes. Another route is described as foIlowing the middIe hemorrhoida plexus to the hypogastries, obturators and then to the externa1 iliacs
._____--II Pcrinc;ll resection. Abdominoperineal resection. Abdominoperinenl resection plus pelvic node dissection * Thcsc wxc
5
i-i-I, 4,
found in radical
’
16
15
12j
x;
4
(8)
4
groin dissections.
resection of the primary with immediateIy adjacent regiona lymph nodes; (2) by the standard Miles procedure al1 of the Iymphatics aIong the superior hemorrhoidal chain up to the left colic vesseIs, those along the middle hemorrhoidal out to but not including the hypogastrics, and al1 nodes adjacent to the Ievator muscles and those in the ischiorectal fossa; (3) by combined abdominoperinea1 resection with added abdominopelvic node dissection, in addition to the aforementioned lymphatics, those about the aorta and vena cava up to the third portion of the duodenum, the sacra1 promontory and aortic bifurcation, the common, external and interna iliacs, and in the obturator spaces bilaterally. It shouId be noted at this time that none of the specimens showed metastases to the nodes at the sacra1 promontory or aortic bifurcation nor to any nodes Iocated along the levator muscIes or in the ischiorecta1 fossa aIthough direct extension was noted. To summarize TabIe VIII, of thirty-three superior hemor729
Stearns recognizabIe groin metastases were not present on admission but did appear subsequently. Of the thirteen patients, eight underwent groin dissection when the metastases became recognizabIe cIinicaIIy. Of the eight, two died of
and superficial inguina1 nodes. This route wouId expIain the folIowing distribution: Two patients whose tumors were confined to the anal cana and two patients whose primary Iesion invoIved both the anal canal and rectum TABLE OVER-ALL
-
TABLE
x SURVIVAL
GROIN
-
Three Years
Five Years
METASTASES
XI
PRESENT
ON
Treatment
ResuIt
No.
None. X-ray.. Groin dissection.
FaiIure Failure FaiIure
6
ADMISSION
Duration
Cases No.
%
No.
%
Less than 2 yr. 4 mo. Less than 3 yr. Over 5 yr.
I
7 2
Without groin metastases: Indeterminate*. FaiIures. . . Successes. . . . . With groin metastases: Indeterminate*. . Failures. Successes.
II IO
19 4 21 4
18
6;:; . . . 16.0
II II
TABLE
50.0 NO
BUT
5 21 3
GROIN
APPEARED
Tr.%tment
12.5
* Indeterminate cases are those patients who were lost track of, who died of other causes without signs of recurrent cancer, or who were treated relatively rec_htIy and couId not be observed for the designated period.
presented negative superficia1 nodes with metastases to the deep peIvic nodes. In this series no deep peIvic node dissections were performed for periana1 Iesions unIess superficia1 groin metastases were present. Therefore we cannot state that the patients did not have deep peIvic nodes without superfIcia1 metastases.
:
* Died of other static carcinoma.
ON
NO.
Failure Failure Failure Indeterminate FaiIure
cause
ADMISSION
SUBSEQUENTLY
Redt
None.. X-ray. Local excision. Groin dissection..
_.
XII
METASTASES
2 I 2 I* 2
SUCCeSS
I
SUCCeSS
2
SUCCfZSS
1
Success
I
without
evidence
Duration
Less than Less than Less than Between Less than Between Between Between Between
of recurrent
6 mo. 6 mo. 3 yr. 2-3 yr. I yr. 1-2 yr. z-3 yr. 4-3 yr. 6-7 yr. or meta-
disease, one died of other causes without recurrent or metastatic disease thirty months after groin dissection, and five are living free of recurrence two to six years. The remaining five had no curative therapy for the groin metastases, four because the primary was not controIIed and the fifth because the metastases were too buIky for compIete remova1. It couId not be shown that either the age of the patients or the Iocation of the primary tumor bore any reIation to the outcome of therapeutic groin dissection. The resuIts obtained by excising peIvic Iymph nodes containing metastatic carcinoma may be summarized as foIIows: TweIve patients underwent abdominopeIvic node dissections added to abdominoperinea1 resection of the rectum as part of the origina treatment of the primary tumor. Four of these had metastases to the iIiac or obturator nodes. None Iived more than three years. Five patients who underwent radica1 groin dissection had Sac or obturator node metastases in addition to the superficia1 groin invoIvement. Four
RESULTS
The comparative three- and five-year surviva1 rates of patients with and without groin node metastases are shown in TabIe x. Thus the three-year surviva1 of those without groin metastases was 65.5 per cent as compared with 16 per cent for those with inguina1 spread. SimiIarIy, the five-year surviva1 for those without was 30 per cent as compared to 12.5 per cent for those with groin metastases. TabIe XI is a summary of those patients presenting inguina1 node metastases on admission. Nine underwent ‘groin dissection either at the time of the treatment of the primary neoplasm or within a few weeks. AI1 patients died as a resuIt of their disease aIthough two Iived more than five years. TabIe XII summarizes the results obtained in the thirteen patients in whom cIinicaIIy 730
Epidermoid of the
Carcinoma
of’ AnaI Region Hospital may be briefly summarized as follows: Perianal lesions or small and superficial lesions very low in the anal canal are treatcad bv n-ide local excision. The groin nodes are then observed at monthly- or bimonthly intcr\,als for the first two years, at three- or four-month intervals up to five years, and at vcarlv intervals thereafter. \Vhen and if groin mctastases appear, radical groin dissection is performed on the involved side. For more extensive perianal lesions involving the anal canal and for those in other locations, the primary lesion is treated by abdominoperineal resection of the rectum with meticulous abdominopelvic node dissection up to the inguinal ligament. The superficial inguinal nodes are observed carefully at frequent intervak; and if and when they appear chnically to be diseased, superficial groin dissection is then performed.
five
patients have been free of recurrence two to six years. COMMENTS
Li;ntil more information is availabIe, the foregoing data may be used as a basis for the clinical approach to severa of the probIems presented in the consideration of proper management of inguinal metastases from epidermoid carcinoma of the ana canal. The first problem is to determine whether “prophylactic” groin dissection wouId have been of value in this group of patients. There were fifty-three patients who did not present clinically recognizable inguina1 metastases at the time the primary Iesion was treated. Inasmuch as there is no definite reIationship between the location of the primary lesion and the side to which metastases occur, fifty-three bilateral groin dissections wouId have been performed prophyIacticaIIy. Forty of this group failed to deveIop groin metastases, while thirteen or 25 per cent did. In the thirteen, the primary was uncontrohed in four, leaving nine patients in whom prophyIactic groin dissection might have been of value. ActuaIly, of the nine there were onIy three failures due to cancer. Thus fifty-three biIatera1 groin dissections would have been performed for the possible salvage of three patients or 6.0 per cent. The resuIts of therapeutic groin dissection seem paradoxic in that none of the nine dkections performed as a part of the initiaI treatment resulted in cure of the patient, while six of eight performed for metastases appearing subsequentIy apparently were effective in controlling the disease. A study of the value of excision of deep peIvic node metastases is aIso inconcIusive, as all four patients died in whom ihac or obturator node metastases were removed aIong with the primary neopIasm, whereas four of five patients who had ihac or obturator node metastases removed as part of subsequent radical groin dissection are Iiving free of cancer. It may be that malignant spread to the nodes early as shown by their presence at the time the primary Iesion is first treated indicates a more viruIent type of cancer than when metastases appear subsequentIy. However, much more cIinica1 data are necessary before any such conclusion can be drawn. Based on the foregoing considerations, the management of groin metastases fohowed by the Rectum and Colon Service at MemoriaI
SUMMAKI’ I. Sixty-nine cases of epidermoid carcinoma of the anal region seen at nlemoria1 Center between January, 1942, and December, 1952, have been reviewed. 2. Forty of the sixty-nine patients did not present clinicahy recognizabIe groin metastases. Twenty-nine or 42 per cent did show metastases to this location. 3. Sixteen of the twenty-nine patients had obvious groin metastases at the time of admission. Of the thirteen that appeared subsequently, eleven or 85 per cent appeared within the first two years of observation. 4. The study of surgical specimens in this series tends to confirm two routes of spread from the anal region: (a) by the way of the perianal Iymphatics or upper thigh directIy to the superficial groin nodes, then to the iliac and obturator; (6) by way of the internal iliac or obturator nodes and then possibIy to the superficial groin nodes. 5. The route by which malignant ceIIs are borne to the groins has IittIe relation to the exact location of the primary lesion if the anal cana is involved, although the inferior route appeared more frequent in this group. 6. Treatment of patients with groin metastases is not satisfactory as the three- and fiveyear survival rates are 16.0 and 12.5 per cent, respectively. However, it is not useless. 7. Of seventeen patients in whom groin dissections were performed, nine died of their disease in less than three years, t.wo died after
73’
Stearns sion. I assume that the primary Iesion was uncontroIIabIe in seven. Is this a correct assumption? Nine of these sixteen patients underwent groin dissection concurrently with excision of the primary Iesion or within a few weeks. AI1 died of epidermoid carcinoma, seven within three years and two after having reached the five-year period customariIy used as the yardstick for measuring survival rates. Dr. Stearns has very wiseIy refrained from condemning therapeutic groin dissection for patients who have cIinicaIIy recognizabIe groin metastasis at the time the primary Iesion is recognized and treated. What shouId we do for patients who have cIinica1 evidence of groin metastasis when first seen and whose primary Iesions appear resectable? I wonder if Dr. Stearns wouId subscribe to the folIowing: extensive abdominoperinea1 resection to incIude aortic vena cava Iymphadenectomy and biIatera1 obturator and iIiac Iymphadenectomy to the femora cana1; a wide perinea1 phase to incIude biIatera1 groin dissection to the femora1 cana1. I am raising this question, for we hope that some accord can be reached which wiI1 heIp us make some decision regarding our approach to the treatments of patients in this category. Dr. Stearns has demonstrated in TabIe XII that groin metastasis deveIoped in thirteen patients subsequent to admission. It appears that in two patients the primary Iesion was uncontroIIabIe; one patient had radiation therapy and two patients underwent IocaI excision. Eight of these patients underwent therapeutic groin dissections for metastasis that appeared subsequent to resection of the primary Iesion. It shouId be observed that the duration of freedom from cIinica1 evidence of disease was not Iong in six of these eight patients. What should we do for patients whose primary Iesions appear resectabIe and who subsequentIy show cIinica1 evidence of groin metastasis? UnIess I am mistaken, Dr. Stearns cannot concIusiveIy support the thesis that we should wait and remove these uniIateraIIy when and if groin metastasis appears. Inasmuch as I subscribe in writing, discussion and in operative procedures to extending our surgica1 attack on mahgnancy, I shouId Iike to suggest that our m.ain hope for improving the rather poor resuIts existing in treating epidermoid carcinoma rests upon an extension of our operative technic. TabIes VIII and IX demonstrate sufficient variabiIity and unpredictability in the chains and nodes invoIved to support this attitude, for even here onethird of the patients studied had positive nodes (TabIe VIII), and except for faiIure of the perianal Iesions to invoIve the superior hemorrhoida nodes (Table IX) the other Iesions metastasized in a relativeIy unpredictabIe fashion. TabIe IV indicates, without eIaboration, that one shouId perhaps
five years, one died of other causes without evidence of disease and five are Iiving free of disease two to six years. None of the patients having groin metastases present on admission is among those stiI1 aIive. 8. The presence of metastases to the deep pelvic nodes does not aIways mean a fatal outcome inasmuch as four of five patients who had obturator or iIiac node metastases removed by radical groin dissection are Iiving and free of evidence of their disease two to six years. g. If “prophyIactic” groin dissections had been the practice in this series, fifty-three biIatera1 groin dissections wouId have been performed for the possibIe saIvage of three patients who were not controIIed by therapeutic groin dissection. REFERENCES
I. BUXTON, R. W. Squamous cell anal carcinoma. Arch. Surg., 67: 821, 1953. 2. CATTELL, R. B. and WILLIAMS, A. C. Epidermoid carcinoma of the anus and the rectum. Arch. Surg., 46: 336, 1943. 3. GABRIEL. W. B. Sauamous ceil carcinoma of the * anus. Proc. Roy. Sot. Med., 34: r 39, 1941. 4. GRINNELL, R. S. An anafysis of forty-nine cases of squamous ceII carcinoma of the anus. Surg., Gynec. Ed Obst., 98: 29, 1954. 4. KEYES. E. L. Sauamous ceII carcinoma of the Iower rectum and anus. Ann. Surg., 106: 1046, 1937,
6. PACK, G. T. and REKERS, P. The management of maIignant tumors in the groin. Am. J. Surg., 56: 545, 1942. 7. PACK, G. T. and BALDWIN, J. C. Management of tumors of the anus. Am. J. Proct., 4: 297, 1953. DISCUSSION
GARNET W. AULT (Washington, D. C.): Dr. Stearns’ study is a vaIuabIe contribution to our knowIedge of epidermoid carcinoma of the anaI region, for it focuses our attention upon the incidence of initial and subsequent metastasis to the groin. It seems pertinent to the entire subject to note that groin metastasis developed in twenty-nine of a tota of sixtv-nine patients-an incidence of 42 per cent. Sixteen of these patients had groin metastasis on admission and thirteen showed evidence of metastasis subsequent to the operative remova of the primary Iesion. In eIeven patients the groin metastasis became evident within two appeared after years, an d in two the metastasis the two-year period. I shaI1 Iimit my initia1 discussion to the incidence of groin metastasis, what was done about this, and the outcome in terms of patient saIvage. Dr. Stearns has demonstrated in TabIe II that sixteen patients had groin metastasis upon admis”
I
732
Epidermoid
Carcinoma
consider altering the policy of wide local excision for perianal lesions and lesions low in the anal canal. In forty-three patients whose Iesions were so located, eighteen had groin metastasis and twenty-five did not. Dr. Stearns has very capably presented us with a report and analysis of a most important topic. Our own experience with a smaIler series of epidermoid lesions is similar to his. My critica analysis and comments are a reflection of the experience of our chnical group and of our dissatisfaction with our own rcsuIts. I am personaIIy convinced that a more adequate cancer operation is badly needed for epidermoid carcinoma of the periana1 region. ~IAUS W. STEARNS (cIosing): Thank you, Dr. Ault. I was expecting a IittIe more controversy about the question of prophylactic groin dissection because within the last year I heard two of the Ieaders in this fieId, whose only discussion condissection cerned whether prophylactic groin should be performed immediately or within two weeks after definitive surgery for the primary lesion. As to the first question regarding pohcy, we agree wholeheartedly that for the IocaIly operable lesion with .groin metastases !we iwould perform
of AnaI Region abdominoperineal resection, complete abdominopelvic node dissection, and groin dissection at the same time or within a week or two afterward. I am not quite sure that I understood the second situation which you proposed as to the patient who did not have groin mctastases on admission. Dr. Ault: What should we do for patients whose primary lesions appear resectable and who subsequently show clinical evidence of growing metastasis? Unless I am mistaken, you cannot support the thesis that we shouId wait and remove these uniIateralIy when and if groin metastasis appears. Dr. Stearns: I thought I did, for the simple reason that we had by far our best results in those cases in which we have done just that. Those patients in whom we were able to control the disease, and in whom groin metastasis subsequently became apparent, comprise the only group in which we have had success. WC have had five completely satisfactory results of eight cases in which that was done. One patient died at thirty months free of disease, so that in six of eight patients recurrent disease did not develop. I think that was my point. Dr. AL&: I wanted to re-emphasize it, vvhich is the only reason I raised the question.
733
Carcinoma INGUINAL
MAUS
IV. STEARNS,
of the Anal Region
METASTASES
JR., M.D., New l’ork, New l’ork
From the Rectum and Colon Service, Memorial Center for Cancer and Allied Diseases, New York, New York.
A
LTHOUGH a considerabIe amount of literature has accumulated pertaining to epidermoid carcinoma of the anal region, relatively little data have been presented concerning the management of inguina1 node metastases occurring in patients with this neopIasm.3j4,6,7 The tendency of squamous carcinoma in this Iocation to metastasize to the inguina1 Iymph nodes is we11 recognized. Frequencies of 8 to 33 per cent have been reported.?,4,5 The opinion has been expressed that inguinal metastases are not amenabIe to cure either by surgery or by radiation although they may be prevented 63 surgery.5 The high incidence of inguinal metastasc’s and the poor prognosis associated with their presence has Ied severa authors1~2~7 to advocate that “prophylactic” groin dissections be performed as a part of the idea1 treatment. OpllosecI to the theoretic advantage of bilateral prophylactic groin dissection is the not inconsiderabIe morbidi,ty associated with this procedure. This morbldIty is due initiaIIy to poor wouncI heaIing and subsequentIy to Iymphedema of the Iower extremity. Occasionally the lymphedema is severe enough to incapacitate a patient permanently. Final evaluation of “prophyIactic” groin dissection must await publication of resuIts obtained by its proponents. In the meantime, it seems of value to anaJyze the data we now ha\.-e in regard to these metastases. X4TERIAL This review- concerns sixty-nine patients \zith epidermoid carcinoma of the periana1 skin, ana cana and rectum who were admitted either as in- or outpatients to the Rectum and Colon Service of the nlemoria1 Center for Cancer and Allied Diseases from January, 1942, to December, 1952. Some of these patients had had definitive
therapy prior to coming to Memorial HospitaI. Some were seen when their neoplasm was beyond hope of cure. The therapy rendered at Memoriai HospitaI was modified by these conditions and did not always represent idea1 treatment. The primary neopIasms were treated by a variety of methods, primarily surgery with or without irradiation in the pre- or postoperative stage. The surgical procedure varied from JocaI excision for relatively smaI1 periana1 Iesions to combined abdominoperinea1 resection of the rectum accompanied by a meticulous abdominopeJvic node dissection. The combined abdominoperineaJ resection of the rectum which is routine in this clinic for this neopIasm stresses ligation of the superior hemorrhoidaJ vessels just beIow the left colic artery, wide removaJ of the peritoneum, wide perinea1 skin excision with compIete remova of the ischiorectal fossa contents, and division of the Jevator muscles at their origin. The abdominopelvic node dissection includes removal of the tissue from about the aorta, vena cava, the iliacs (common, external and internal), the obturator space, the presacra1 and aortic bifurcation. Tissues from these individual areas are Iabeled and sent as separate specimens in order that the exact sites of node metastases may be identified. AIL of the groin dissections studied in this review were performed in the presence of clinically diagnosed groin metastases. GeneraIIy, prior to dissection the presence of node metastases was proved bv aspiration biopsy or occasionaIJy by Jocal excision. Groin dissections in this group of patients are classified as superficia1 or radical. The former dissection is limited to the inguinal area (upper thigh and Iower abdomen) superficial to the fascia of the abdominal wall and the femoral area. The radicaJ dissection in addition includes the tissues from about the externa1 and internal iliac vesseis and obturator fossa.
Stearns Clinical Factors and Incidence OJ Groin A’ode Metastases. Tables I to v show the relation of the age and sex of the patient, duration of symptoms prior to treatment, and the location and grade of the tumor to the development of groin metastases. ObviousI\- the numf>er 01 patients is too small for statistical significnncc,
Inasmuch as the purpose of the present study is to add to our understanding of the probIems of groin metastases from epidermoid carcinoma of the anal region, the analysis of end results and the method of treatment of the primarylesion are considered only in relation to this purpose. I-ABLE
1
No.... Yes......
TABLE 11 SEX OF PATIENT Groin \letastases
No.+. ...........
i
0 7
14 Jli
;;,
TABLE
1No. of \I’omen
Groin klctastases
26
14
12
‘7
Grade I
I
Gde 11
Gde III
I8 IO
I&
No. Yes.
TABLE III DL RATION OF SYMPTOMS
I I Groin Lessthan 3-6 6--12 I -2 Over z Not Llctas- ~ Stated yr. Y’. t:1st’s 3 m0. I m”. ) m”.
\,cs.
12
1
4
6
4
I
Not GMkd /
I
No.....
I,
/
No. of Men
.’
Yes .............
2 3
2 3
TABL.E
II
i7
VI
i No. Yes..
Incidence of Groin Node Metastases. Of the sixty-nine patients studied who had epidermoid carcinoma of the perianal skin, anal canal or dista1 rectum, forty failed to present groin metastases either initially or subsequently. Of the forty patients, seventeen were followed up less than two years, and twenty-three were followed up more than two years. Twenty-nine patients or 42 per cent did have groin metastases. Of the twenty-nine, sixteen had clinically diagnosed groin metastases on admission, and in thirteen groin metastases became clinicaIIy recognizabIe after the treatment of the primary neopIasm had been compIeted. Of the thirteen that developed subsequently, eleven appeared within two years, two after two years (one at nine years).
II
‘3
15
3
f,
4
but there does not appear to fle any definite relation fletwecn any of these factors and the presence of groin metastases. In Table VI the relationship hetlvcen the type of therapy for the primary tumor and the subsequent appearance of groin metastases is presented. Again the numbers in each group are few but no appreciable difference is noted. Table VII indicates the type of surgery and subsequent groin metastases. Again no appreciabIe difference is noted. Sites f~f Lymphatic Metastases. Information concerning the lymphatic metastases from this form of cancer 728
was obtained
by a study
of the
Epidermoid
Carcinoma
of AnaI
Region
rhoidai lymphatic chains studied. trve or 13 per cent showed lymphatic metastases; of sixteen pelvic Iymphadenectomies performed (external and internal iIiacs and obturator space), seven or 43 per cent showed metastases.
operative specimens and is summarized in Table VIII. It may be assumed that the specimens, grouped here according to the type of incIuded the foIIowing surgical procedure, tissues: (I) 114’ “perineal resection” a wide local TABLE \‘II TYPE OF SURGERY
Perianal. Perianal
SITES
.‘.
!p2’,
J
TABLE ~111 01-‘ LYMPHATIC METASTASES
Procrdnrr
I
I
Deep Pelvic (Iliacs, Obturator)
Superior Hcmorrhoida1 Chain
No. _ ~___ ~__
I
Negative
.__
__
Neyative
Positive
o Ii I ~ I*
3*
Positive
The Iocation of the primary tumor as related to the sites of Iymphatic metastases is presented in Table IX, which when interpreteii in conjunction with the analysis of individual specimens indicates confirmation of described routes of Iymphatic drainage based primarily on anatomic studies.1j5 One route described is from the anal plexus through the perineal Iymphatics up over the upper thigh emptying first into the superficial groin nodes. Presumably, metastases to the deep pelvic nodes may occur from the superficia1 nodes. This route most satisfactorily explains the distribution of inguinal and pelvic metastases in the following patients: Two patients had tumors confined to the perianal skin with inguina1 metastases. In one the metastases were co&red to superficial nodes while in the other esterna1 iliac and obturator node metastases were present, also. Two patients with primary Iesions involving both the periana1 skin and the anal canal had superficial groin metastases without deep pelvic node involvement. Six patients with lesions confined to the ana canal had superficial groin metastases with negative deep pelvic nodes. Four patients with Iesions involving the upper anal canal and dista1 rectum had superficial groin metastases with negative deep peIvic nodes. Another route is described as foIlowing the middIe hemorrhoida plexus to the hypogastries, obturators and then to the externa1 iliacs
._____--II Pcrinc;ll resection. Abdominoperineal resection. Abdominoperinenl resection plus pelvic node dissection * Thcsc wxc
5
i-i-I, 4,
found in radical
’
16
15
12j
x;
4
(8)
4
groin dissections.
resection of the primary with immediateIy adjacent regiona lymph nodes; (2) by the standard Miles procedure al1 of the Iymphatics aIong the superior hemorrhoidal chain up to the left colic vesseIs, those along the middle hemorrhoidal out to but not including the hypogastrics, and al1 nodes adjacent to the Ievator muscles and those in the ischiorectal fossa; (3) by combined abdominoperinea1 resection with added abdominopelvic node dissection, in addition to the aforementioned lymphatics, those about the aorta and vena cava up to the third portion of the duodenum, the sacra1 promontory and aortic bifurcation, the common, external and interna iliacs, and in the obturator spaces bilaterally. It shouId be noted at this time that none of the specimens showed metastases to the nodes at the sacra1 promontory or aortic bifurcation nor to any nodes Iocated along the levator muscIes or in the ischiorecta1 fossa aIthough direct extension was noted. To summarize TabIe VIII, of thirty-three superior hemor729
Stearns recognizabIe groin metastases were not present on admission but did appear subsequently. Of the thirteen patients, eight underwent groin dissection when the metastases became recognizabIe cIinicaIIy. Of the eight, two died of
and superficial inguina1 nodes. This route wouId expIain the folIowing distribution: Two patients whose tumors were confined to the anal cana and two patients whose primary Iesion invoIved both the anal canal and rectum TABLE OVER-ALL
-
TABLE
x SURVIVAL
GROIN
-
Three Years
Five Years
METASTASES
XI
PRESENT
ON
Treatment
ResuIt
No.
None. X-ray.. Groin dissection.
FaiIure Failure FaiIure
6
ADMISSION
Duration
Cases No.
%
No.
%
Less than 2 yr. 4 mo. Less than 3 yr. Over 5 yr.
I
7 2
Without groin metastases: Indeterminate*. FaiIures. . . Successes. . . . . With groin metastases: Indeterminate*. . Failures. Successes.
II IO
19 4 21 4
18
6;:; . . . 16.0
II II
TABLE
50.0 NO
BUT
5 21 3
GROIN
APPEARED
Tr.%tment
12.5
* Indeterminate cases are those patients who were lost track of, who died of other causes without signs of recurrent cancer, or who were treated relatively rec_htIy and couId not be observed for the designated period.
presented negative superficia1 nodes with metastases to the deep peIvic nodes. In this series no deep peIvic node dissections were performed for periana1 Iesions unIess superficia1 groin metastases were present. Therefore we cannot state that the patients did not have deep peIvic nodes without superfIcia1 metastases.
:
* Died of other static carcinoma.
ON
NO.
Failure Failure Failure Indeterminate FaiIure
cause
ADMISSION
SUBSEQUENTLY
Redt
None.. X-ray. Local excision. Groin dissection..
_.
XII
METASTASES
2 I 2 I* 2
SUCCeSS
I
SUCCeSS
2
SUCCfZSS
1
Success
I
without
evidence
Duration
Less than Less than Less than Between Less than Between Between Between Between
of recurrent
6 mo. 6 mo. 3 yr. 2-3 yr. I yr. 1-2 yr. z-3 yr. 4-3 yr. 6-7 yr. or meta-
disease, one died of other causes without recurrent or metastatic disease thirty months after groin dissection, and five are living free of recurrence two to six years. The remaining five had no curative therapy for the groin metastases, four because the primary was not controIIed and the fifth because the metastases were too buIky for compIete remova1. It couId not be shown that either the age of the patients or the Iocation of the primary tumor bore any reIation to the outcome of therapeutic groin dissection. The resuIts obtained by excising peIvic Iymph nodes containing metastatic carcinoma may be summarized as foIIows: TweIve patients underwent abdominopeIvic node dissections added to abdominoperinea1 resection of the rectum as part of the origina treatment of the primary tumor. Four of these had metastases to the iIiac or obturator nodes. None Iived more than three years. Five patients who underwent radica1 groin dissection had Sac or obturator node metastases in addition to the superficia1 groin invoIvement. Four
RESULTS
The comparative three- and five-year surviva1 rates of patients with and without groin node metastases are shown in TabIe x. Thus the three-year surviva1 of those without groin metastases was 65.5 per cent as compared with 16 per cent for those with inguina1 spread. SimiIarIy, the five-year surviva1 for those without was 30 per cent as compared to 12.5 per cent for those with groin metastases. TabIe XI is a summary of those patients presenting inguina1 node metastases on admission. Nine underwent ‘groin dissection either at the time of the treatment of the primary neoplasm or within a few weeks. AI1 patients died as a resuIt of their disease aIthough two Iived more than five years. TabIe XII summarizes the results obtained in the thirteen patients in whom cIinicaIIy 730
Epidermoid of the
Carcinoma
of’ AnaI Region Hospital may be briefly summarized as follows: Perianal lesions or small and superficial lesions very low in the anal canal are treatcad bv n-ide local excision. The groin nodes are then observed at monthly- or bimonthly intcr\,als for the first two years, at three- or four-month intervals up to five years, and at vcarlv intervals thereafter. \Vhen and if groin mctastases appear, radical groin dissection is performed on the involved side. For more extensive perianal lesions involving the anal canal and for those in other locations, the primary lesion is treated by abdominoperineal resection of the rectum with meticulous abdominopelvic node dissection up to the inguinal ligament. The superficial inguinal nodes are observed carefully at frequent intervak; and if and when they appear chnically to be diseased, superficial groin dissection is then performed.
five
patients have been free of recurrence two to six years. COMMENTS
Li;ntil more information is availabIe, the foregoing data may be used as a basis for the clinical approach to severa of the probIems presented in the consideration of proper management of inguinal metastases from epidermoid carcinoma of the ana canal. The first problem is to determine whether “prophylactic” groin dissection wouId have been of value in this group of patients. There were fifty-three patients who did not present clinically recognizable inguina1 metastases at the time the primary Iesion was treated. Inasmuch as there is no definite reIationship between the location of the primary lesion and the side to which metastases occur, fifty-three bilateral groin dissections wouId have been performed prophyIacticaIIy. Forty of this group failed to deveIop groin metastases, while thirteen or 25 per cent did. In the thirteen, the primary was uncontrohed in four, leaving nine patients in whom prophyIactic groin dissection might have been of value. ActuaIly, of the nine there were onIy three failures due to cancer. Thus fifty-three biIatera1 groin dissections would have been performed for the possible salvage of three patients or 6.0 per cent. The resuIts of therapeutic groin dissection seem paradoxic in that none of the nine dkections performed as a part of the initiaI treatment resulted in cure of the patient, while six of eight performed for metastases appearing subsequentIy apparently were effective in controlling the disease. A study of the value of excision of deep peIvic node metastases is aIso inconcIusive, as all four patients died in whom ihac or obturator node metastases were removed aIong with the primary neopIasm, whereas four of five patients who had ihac or obturator node metastases removed as part of subsequent radical groin dissection are Iiving free of cancer. It may be that malignant spread to the nodes early as shown by their presence at the time the primary Iesion is first treated indicates a more viruIent type of cancer than when metastases appear subsequentIy. However, much more cIinica1 data are necessary before any such conclusion can be drawn. Based on the foregoing considerations, the management of groin metastases fohowed by the Rectum and Colon Service at MemoriaI
SUMMAKI’ I. Sixty-nine cases of epidermoid carcinoma of the anal region seen at nlemoria1 Center between January, 1942, and December, 1952, have been reviewed. 2. Forty of the sixty-nine patients did not present clinicahy recognizabIe groin metastases. Twenty-nine or 42 per cent did show metastases to this location. 3. Sixteen of the twenty-nine patients had obvious groin metastases at the time of admission. Of the thirteen that appeared subsequently, eleven or 85 per cent appeared within the first two years of observation. 4. The study of surgical specimens in this series tends to confirm two routes of spread from the anal region: (a) by the way of the perianal Iymphatics or upper thigh directIy to the superficial groin nodes, then to the iliac and obturator; (6) by way of the internal iliac or obturator nodes and then possibIy to the superficial groin nodes. 5. The route by which malignant ceIIs are borne to the groins has IittIe relation to the exact location of the primary lesion if the anal cana is involved, although the inferior route appeared more frequent in this group. 6. Treatment of patients with groin metastases is not satisfactory as the three- and fiveyear survival rates are 16.0 and 12.5 per cent, respectively. However, it is not useless. 7. Of seventeen patients in whom groin dissections were performed, nine died of their disease in less than three years, t.wo died after
73’
Stearns sion. I assume that the primary Iesion was uncontroIIabIe in seven. Is this a correct assumption? Nine of these sixteen patients underwent groin dissection concurrently with excision of the primary Iesion or within a few weeks. AI1 died of epidermoid carcinoma, seven within three years and two after having reached the five-year period customariIy used as the yardstick for measuring survival rates. Dr. Stearns has very wiseIy refrained from condemning therapeutic groin dissection for patients who have cIinicaIIy recognizabIe groin metastasis at the time the primary Iesion is recognized and treated. What shouId we do for patients who have cIinica1 evidence of groin metastasis when first seen and whose primary Iesions appear resectable? I wonder if Dr. Stearns wouId subscribe to the folIowing: extensive abdominoperinea1 resection to incIude aortic vena cava Iymphadenectomy and biIatera1 obturator and iIiac Iymphadenectomy to the femora cana1; a wide perinea1 phase to incIude biIatera1 groin dissection to the femora1 cana1. I am raising this question, for we hope that some accord can be reached which wiI1 heIp us make some decision regarding our approach to the treatments of patients in this category. Dr. Stearns has demonstrated in TabIe XII that groin metastasis deveIoped in thirteen patients subsequent to admission. It appears that in two patients the primary Iesion was uncontroIIabIe; one patient had radiation therapy and two patients underwent IocaI excision. Eight of these patients underwent therapeutic groin dissections for metastasis that appeared subsequent to resection of the primary Iesion. It shouId be observed that the duration of freedom from cIinica1 evidence of disease was not Iong in six of these eight patients. What should we do for patients whose primary Iesions appear resectabIe and who subsequentIy show cIinica1 evidence of groin metastasis? UnIess I am mistaken, Dr. Stearns cannot concIusiveIy support the thesis that we should wait and remove these uniIateraIIy when and if groin metastasis appears. Inasmuch as I subscribe in writing, discussion and in operative procedures to extending our surgica1 attack on mahgnancy, I shouId Iike to suggest that our m.ain hope for improving the rather poor resuIts existing in treating epidermoid carcinoma rests upon an extension of our operative technic. TabIes VIII and IX demonstrate sufficient variabiIity and unpredictability in the chains and nodes invoIved to support this attitude, for even here onethird of the patients studied had positive nodes (TabIe VIII), and except for faiIure of the perianal Iesions to invoIve the superior hemorrhoida nodes (Table IX) the other Iesions metastasized in a relativeIy unpredictabIe fashion. TabIe IV indicates, without eIaboration, that one shouId perhaps
five years, one died of other causes without evidence of disease and five are Iiving free of disease two to six years. None of the patients having groin metastases present on admission is among those stiI1 aIive. 8. The presence of metastases to the deep pelvic nodes does not aIways mean a fatal outcome inasmuch as four of five patients who had obturator or iIiac node metastases removed by radical groin dissection are Iiving and free of evidence of their disease two to six years. g. If “prophyIactic” groin dissections had been the practice in this series, fifty-three biIatera1 groin dissections wouId have been performed for the possibIe saIvage of three patients who were not controIIed by therapeutic groin dissection. REFERENCES
I. BUXTON, R. W. Squamous cell anal carcinoma. Arch. Surg., 67: 821, 1953. 2. CATTELL, R. B. and WILLIAMS, A. C. Epidermoid carcinoma of the anus and the rectum. Arch. Surg., 46: 336, 1943. 3. GABRIEL. W. B. Sauamous ceil carcinoma of the * anus. Proc. Roy. Sot. Med., 34: r 39, 1941. 4. GRINNELL, R. S. An anafysis of forty-nine cases of squamous ceII carcinoma of the anus. Surg., Gynec. Ed Obst., 98: 29, 1954. 4. KEYES. E. L. Sauamous ceII carcinoma of the Iower rectum and anus. Ann. Surg., 106: 1046, 1937,
6. PACK, G. T. and REKERS, P. The management of maIignant tumors in the groin. Am. J. Surg., 56: 545, 1942. 7. PACK, G. T. and BALDWIN, J. C. Management of tumors of the anus. Am. J. Proct., 4: 297, 1953. DISCUSSION
GARNET W. AULT (Washington, D. C.): Dr. Stearns’ study is a vaIuabIe contribution to our knowIedge of epidermoid carcinoma of the anaI region, for it focuses our attention upon the incidence of initial and subsequent metastasis to the groin. It seems pertinent to the entire subject to note that groin metastasis developed in twenty-nine of a tota of sixtv-nine patients-an incidence of 42 per cent. Sixteen of these patients had groin metastasis on admission and thirteen showed evidence of metastasis subsequent to the operative remova of the primary Iesion. In eIeven patients the groin metastasis became evident within two appeared after years, an d in two the metastasis the two-year period. I shaI1 Iimit my initia1 discussion to the incidence of groin metastasis, what was done about this, and the outcome in terms of patient saIvage. Dr. Stearns has demonstrated in TabIe II that sixteen patients had groin metastasis upon admis”
I
732
Epidermoid
Carcinoma
consider altering the policy of wide local excision for perianal lesions and lesions low in the anal canal. In forty-three patients whose Iesions were so located, eighteen had groin metastasis and twenty-five did not. Dr. Stearns has very capably presented us with a report and analysis of a most important topic. Our own experience with a smaIler series of epidermoid lesions is similar to his. My critica analysis and comments are a reflection of the experience of our chnical group and of our dissatisfaction with our own rcsuIts. I am personaIIy convinced that a more adequate cancer operation is badly needed for epidermoid carcinoma of the periana1 region. ~IAUS W. STEARNS (cIosing): Thank you, Dr. Ault. I was expecting a IittIe more controversy about the question of prophylactic groin dissection because within the last year I heard two of the Ieaders in this fieId, whose only discussion condissection cerned whether prophylactic groin should be performed immediately or within two weeks after definitive surgery for the primary lesion. As to the first question regarding pohcy, we agree wholeheartedly that for the IocaIly operable lesion with .groin metastases !we iwould perform
of AnaI Region abdominoperineal resection, complete abdominopelvic node dissection, and groin dissection at the same time or within a week or two afterward. I am not quite sure that I understood the second situation which you proposed as to the patient who did not have groin mctastases on admission. Dr. Ault: What should we do for patients whose primary lesions appear resectable and who subsequently show clinical evidence of growing metastasis? Unless I am mistaken, you cannot support the thesis that we shouId wait and remove these uniIateralIy when and if groin metastasis appears. Dr. Stearns: I thought I did, for the simple reason that we had by far our best results in those cases in which we have done just that. Those patients in whom we were able to control the disease, and in whom groin metastasis subsequently became apparent, comprise the only group in which we have had success. WC have had five completely satisfactory results of eight cases in which that was done. One patient died at thirty months free of disease, so that in six of eight patients recurrent disease did not develop. I think that was my point. Dr. AL&: I wanted to re-emphasize it, vvhich is the only reason I raised the question.
733