- Email: [email protected]
Analysis of factors contributing to treatment failures in Stages IB and IIA carcinoma of the cervix
Analysis of factors contributing to treatment failures in Stages IB and IIA carcinoma of the cervix C. K. CHUNG, M.D. WILLIAM A. NAHHAS, M.D. JOHN A. STRYKER, M.D. STEPHEN L. CURRY, M.D. ARTHUR B. ABT, M.D. RODRIGUE MORTEL, M.D.
Hershey, Pennsylvania Between April, 1971, and September, 1977, 98 patients with Stages 18 and IIA cervical cancer who underwent surgical exploration prior to treatment at Hershey Medical Center were studied. Those who had bulky primary tumor (2=4 em) had a higher incidence of nodal metastases (80% vs. 16%), local recurrences (40% vs. 5%), and distant metastases (40% vs. 1%). Patients who had positive nodes had more local recurrences (24% vs. 6%) and distant metastases (28% vs. 0%). Those with grossly positive nodes had more distant metastases (60% vs. 7%) than those with microscopically positive nodes. Those who had positive nodes, vascular invasion, and/or deep invasion of the cervix (2: 70% of thickness) in the radical hysterectomy specimen had more nodal metastases and local recurrences. Postoperative radiation seemed to prevent local recurrences (40% vs. 6%) and improve the 2-year tumor-free survival rate (94% vs. 55%). Patients who had bulky primary tumors and/or grossly positive nodes at laparotomy may require systemic therapy in view of the high incidence Of distant failures. (AM. J. 0BSTET. GYNECOL. 1313:550, 1980.)
Su CCESSFU Ltreatment of Stages IB and IIA cancer of the cervix requires control of the central tumor and eradication of metastatic cancer foci when present in the regional nodes. The purpose of this investigation was to determine the prognostic significance of pelvic node metastasis, size and depth of invasion of the primary lesion, and the presence of vascular invasion in the radical hysterectomy specimen. The value of postoperative radio-
From the Divisions of Radiation Oncology and Gynecologic Oncology, the Department of Pathology, and The Specialiud Cancer Research Center, Milton S. Hershey Medical Center, College of Medicine. The Pennsylvania State University. Presented at the Sixty-fifth Annual Meeting of the Radiologic Society of North America, Atlanta, Georgia, November 29, 1979. Receivedfor publication February 13, 1980. Revised june 27, 1980. Accepted july 21, 1980. Reprint requests: Rodrigue Mortel, M.D., Division of Gynecologic Oncology, M. S. Hershey Medical Center, The Pennsylvania State University, Hershey, Pennsylvania 17033.
550
therapy for those m the high-risk group was also evaluated.
Methods One hundred forty patients with previously untreated FIGO (International Federation of Gynaecology and Obstetrics) Stages IB and IIA carcinoma of the cervix were treated at the Milton S. Hershey Medical Center of The Pennsylvania State University between April, 1971, and September, 1977. Forty-two of these patients who were medically ineligible or had barrelshaped cervix underwent no surgical exploration and were not included in the study, thus leaving as eligible a total of 98 patients who underwent laparotomy. The patients were divided into three groups according to the treatment plan (Table 1). At laparotomy, sampling of aortic nodes and multiple biopsies of grossly enlarged pelvic nodes were done. If nodes were positive and/or the primary tumor was found to be more advanced, the condition was regarded as nonresectable and the patient was treated with radiotherapy alone. If nodes were negative, radical hysterectomy with bilateral pelvic lymphadenectomy was performed. Postoperative radiation was usually administered to 0002·9378/80/210550+07$00.70/0
©
1980 The C. V. Mosby Co
Volume 138 ~umber
Treatment failures in cervical carcinoma
551
.'i
Table I. Ninety-eight patients with Stages IB and IIA cervical cancer who underwent laparotomy No. of patients Treatment
Operation* Operation plus
I Average age (yr)
65
8
II
I
45 46
Totalt
9
85
4
13
47
*Radical hysterectomy with bilateral pelvic lymphadenectomy. tRadical hysterectomy with bilateral pelvic lymphadenectomy plus postoperative radiotherapy. those in whom positive nodes were found on radical hysterectomy and pelvic lymphadenectomy specimen. A totai of 73 patients underwent operation aione (radical hysterectomy with bilateral pelvic lymphadenectomy). Tweive patients underwent operation pius postoperative radiotherapy: 10 patients with positive nodes. one patient with positive vaginal resectJon margin, and one patient in whom the tumor penetrated through the wall on the radical hysterectomy specimen. (The other five patients with positive nodes did not receive postoperative radiotherapy because of surgical complications.) Thirteen patients underwent radiotherapy alone: nine patients with gross nodal metastases and four patients with more advanced disease at laparotomv (the tumor extended into the parametrium, 1, pelvic wall, 1, and bladder, 2). Radiotherapy was administered in a fairly standard fashion and it consisted of external-beam radiotherapy to the whole pelvis with t.he use of a 10-mev linear accelerator followed by intracavitary brachytherapy. External-beam radiation therapy delivered a minimum whole-pelvic dose of 5,100 rads through anteriorposterior opposed or four-field technique. All fields were treated every day, 5 days a week with a daily tumor dose of 170 rads. The minimum treatment volume was maintained as follows: superiorly, the upper margin of L5; inferiorly, the upper third of the obturator foramen, or lower if indicated; laterally, at least 1 em beyond the lateral margin of the bony pelvis at the widest plane of the pelvis. Additional para-aortic node irradiation was given to three patients with positive para-aortic nodes. Intracavitary brachytherapy consisted of the Fletcher-Suit afterloading system with cesium-137. Two 48-hour applications with a 2-week interval were used for primary radiotherapy; the first insertion of cesium delivered an average of 2,714 (2,587 to 3,226) mg hr, with 2,234 rads (2,088 to 2,640) to point A. and 486 rads (336 to 576) to the lateral
Left
Lymph node group
External iliac Obturator J-J vnno-!:!!drir ..... ,r .... e-...............
postoperative
radiotherapyt Radiotherapy Total
Table II. Sites of lymph node metastases* (Stages iB and IIA cervical cancer)
Parametrial Total pelvic Common iliac Para-aortic
5 2
4
4
I
15
2
%
ll
36
.,
!6
19
fj
3
0 0
I
()
:)
9
II
2 2
25
80
:I
10 10
2
2 12 2
No.
1 2
14
0
:~
31
2
100
*Number of lymph nodes dissected: eight to 76 nodes (average, 24.3). tNumber of sites of nodal metastases. Table III. Incidence of lymph node metastases (+)Sodes Charack~.stics
Stage: IB
IIA
Histologic features:
Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Size of primary tumor*: ~4cm
<4cm Depth of invasiont:
~70% thickness <70% thickness
Vascular invasiont:
+
1Vo. C\1 /0!'!
% 0"
~ lfO~
6;;.}
4/13
31
23/88 2/8 0/2
26 25
12/15 13/83
80t 16
6/15 9170
40t 13
5/8
63t 13
10177
()
icClinical measurement.
tP < 0.05 (chi-square test). hysterectomy specimen findings.
~Radical
pelvic wall (lateral margin of the bony pelvis at the widest plane through the pelvis). The second insertion delivered an average of 2,964 (2, 145 to 3, 144) mg hr with 2,382 rads (1,914 to 2,640) to point A and 434 rads (384 to 528) to the lateral pelvic wall. Therefore, a total of 5,679 mg hr with 4,616 rads to point A and 920 rads to the lateral pelvic wall was delivered in two insertions. For the postoperative radiotherapy group, seven patients underwent external pelvic irradiation only (5, 100 rads in 6 weeks), and five patients underwent, in addition, one 48-hour insertion of a vaginal applicator to deliver an average of 1,566 (960 to 1,920) mg hr with 4,864 (4,080 to 5,760) rads to the vaginal surface. The size of the primary tumor, as measured at the time of initial clinical staging, was divided into two
552 Chung et al. .\m.
.\!ovember L I'IKII Obstet. Gmeml.
J.
Table IV. Sites of failure by treatment modalities Local failures No. of patients
Treatment Operation Operation plus postoperative radiotherapy Radiotherapy · · · Total
73 12 13
98
% 3 8 13
2 1 1
3 0
3
4
4
4 0 23
6
0 1 6
()
8
46
7
6
7
Table V. Treatment results Local recurrence No. of patients Stage IB IIA Histologic features: Souamous-cell carcinoma Adenocarcinoma Adenosquamous carcinoma Size of prirr~ry turr..or ~4cm
<4cm Depth of invasion ~70% thickness <70% thickness Vascular invasion
+
No.
I
Distant metastasis
%
No.
I
2-Year survival* (NED)t %
No.
I
%
85 13
9
11
8
5 2
6 15
72 11
85 85
88 8 2
8 1 1
9 13 50
7 0 0
8 0 0
75 7 1
85 88 50
IS 83
6 4
40 5
6 1
40 1
7 76
47:j: 92
15 70
4 3
27 4
2 0
13 0
10 65
67:j: 93
8 77
2 4
25 5
1 0
13 0
4 72
SO:j: 94
*Absolute recurrence-free survival. tNo evidence of disease at the follow-up examination immediately after the cut-off time at 24 months after treatment. :j:P < 0.05.
groups: 4 em or lllore vs. less than 4 em on clinical measurement. The hysterectomy specimen was examined in a fresh state and after formalin fixation. Fullthickness sections through the tumor (minimum of three) were taken from areas grossly showing the greatest depth of penetration. These were stained with hematoxylin and eosin. The maximum depth of tumor invasion was determined as a percentage of thickness of the cervical-endocervical segment. Vascular invasion (lymphatic and venous) w,as determined from the slides stained with hematoxylin and eosin. Sections of all lymph nodes were studied. When nodes were large, a single representative section was taken. In specimens in which nodes were not grossly detectable, a sample of tissue was subjected to search for foci of lymphatic tissue. The incidence of nodal metastasis was presented according to size of primary tumor, degree of depth of invasion. and nresence of vascular invasion. Treatment results were analyzed as local recurrence, distant metastases, and 2-year absolute recurrence-free survival (NED), by size of primary tumor, status of -
-
'-
lymph node metastases, hysterectomy specimen findings (vascular invasion and depth of invasion), and treatment modalities. Particularly, treatment outcome was compared between operation alone and operation plus postoperative radiotherapy in those in the highrisk 0 1oup with lymph node metastases, deep penetration of the tumor, and vascular invasion. Finally, comorder to find out whether the combined-therapy group had increased complications. Follow-up care was carried out in the Gynecologic Oncology Clinic. Absolute survival was calculated from the initial date of treatment. All patients were followed until death or for a minimum of 2 years from the date of first treatment. Three patients were lost to follow-up and counted as dead. Statistical evaluation was carried out according to the chi-square method.
l
Results
Of a total of 98 patients, 10 ( 10%) had gross lymph node metastases at the time of laparotomy; 15 of 85
Volume 13ii Number 5
Treatment failures in cervical carcinoma
553
Table VI. Status of nodal m.etastases and prognoses Nodal site
Status
of metastasis
Pelvic
Gross*(+) Microt (+)
Common iliac
Gross(+) Micro(+)
Para-aortic
0/6
Oil
Total
1/3 (33%) 2/2 II 1 3/3 (100%) 6/10 (60%) 1/15 (7%) 7/25 (28%) 0173 (0%)
Oil 2/3 (67%) 3/10 (30%) 3/15 (20%) 6/25 (24%) 4173 (6%)
I~·
'flO
9/13 13/19 (68%) 1/2 1/1 2/3 (67%1 0/2
Oil
1/3 (33%) 2/2
Gross(+) Micro(+) Node(+) Node(-)
,.
3/6
0/13 3/19 (16%) 112
31l3 3/19 (16%) 112
Gross(+) Micro(+)
2-Yrar survival (NED)
Distant metastases
Local recurrence
Ill
1/3 {:{:l 0/c) 5/10 (50%)~ 11/15 (73%) 16/25 (64%)§ 67173 (92%)
*Gross nodal metastases at laparotomy, which were all removed. snecimen. tMicrosconicallv ,--r--rnositive -- nocles ---- on --- racliral --------- hvsterertomv r *Not significant. §P < 0.05. -~
1
1
--,~--
Table VII. Retrospective comparison of treatment results in the high-risk group: Operation alone vs. operation plus postoperative radiotherapy Local recurrence Operation* Pathologic specimen findings
No.
Lymph node metastasis ( +) Depth of invasion (~70% thickness) Vascuiar invasion (+) Total
2/5:j: 4/10 2/5 8/20
I
2-Year survival (NED!
Operation* + RTt
%
No.
40 40 40 40
1110 0/5 0/3
1118
I
Operation*
%
No.
10
2/5 71l0
0 0 6
2/5
11120
l
Operation + RTt
%
No.
40 70 40 55
9/10 5/5 3/3
171l8§
I
% 90 100 iOO 94
*Radical hysterectomy. tRadical hysterectomy plus postoperative radiotherapy. :j:Five patients who had microscopic positive nodes did not receive postoperative radiotherapy because of surgical complications. §P = 0.05. patients (18%) who underwent radical hysterectomy with bilateral pelvic lymphadenectomy were found to have unexpected lymph node metastases microscopically in the pathologic specimen. Of the 85 patients with Stage IB carcinoma, 21 had positive nodes (25%); 19 had pelvic only, one had pelvic and para-aortic, and one had para-aortic only. Of the 13 patients with Stage IIA carcinoma, 4 patients had nodal metastasis (31 %), three had pelvic only, and one had para-aortic oniy. Table II shows the distribution of lymph node metastases; the external iliac node was the most common site of involvement. Only two patients had lymph node metastases bilaterally. The incidences of lymph node metastasis, by stage, histologic features, size of primary tumor, as well as depth of invasion and vascular invasion, are shown in T-:JhlP TTJ ThP _.. ...... .....,,. ...................... ,. ... ..._
~t':llrrP
fiR
u~::
TJ A\
seem to influence lymph node metastasis; however, patients with bulky primary tumors (2:4 ern) had a significantly higher incidence of nodal metastasis than did those with smaller tumors (80% vs. 16%). Those who had vascular invasion and/or deep invasion of the cervix (2:70% of thickness) on hysterectomy specimen had more lymph node metastases. In this series, 88% of recurrence occurred within 24 months, and aii within 30 months of treatment. Therefore, the 2-year NED rate is reliable for the purpose of comparing the various treatment results. Ten patients experienced local failure in the pelvis (four central and six lateral), and seven patients developed distant metastases (four patients in the bones, two patients in the supraclavicular nodes, and one patient in the mediastinum). The sites of failure by treatment modality are
~nrl h1ct-£'\ln.~r .f'..::..-:~h1r£»co
'-'OAA.t)"'- , .... ...., ..,..,, .._.._,.._, u,..L.o.'-" .L.L.Lo;:,IL..OJ.L'-'E).a"- .L"-U..IL..U.L'-'"
(squamous-cell carcinoma vs. adenocarcinoma) do not
vorable conditions received radiotherapy alone, the
554 Chung et al. Am.
November I, 1980 Gvnecol.
J. Obstet.
Table VIII. Complications by treatment modalities Complications
Fistulas Urinary retention and/or incontinence Lymphocyst Pelvic abscess ~-\cute
1
Operation (7 3 patients)
Operation + postoperative radiotherapy ( 12 patients)
Radiotherap)' ( 13 patients)
7 (9%)* I3
0 (0%)
I (7%}t
2 2
pyelonephritis
Surgical injury to bladder or rectum Severe hemorrhage Small bowel obstruction Rectosigmoiditis Severe complication*
0 0 0
0 0 0 0 0 0 I I
I (8%)
3 (23%)
1
0 0 0
3
I
I I
0
17 (23%)
*Seven patients with 10 fistulas (five ureterovaginal, three vesicovaginal, one rectovaginal, and one vesicocutaneous) underwent surgical correction within 2 to IO months after radical hysterectomy, with good primary results in all cases. tRectovaginal fistula. :j:Required second surgical intervention.
Tabie IX. Two-year absolute survival (NED) Stage IB cancer
Stage JIA cancer
OveraU survival (%)
Operation Operation plus ....,....,.,r:,.,.....,...,. .. ..,..: .. ..,. ..,...,.,li,....foho'l"<:~or..'l.r yv;;n. vp'-• A"l ,..._ a aouu. n.a•.._1 "'t'!
Radiotherapy Positive nodes Negative nodes Total
NED*(%)
OveraUsurvival(%)
58/65 IO/II
89 9I
95 9I
7/8 Ill
88 100
100 100
4/9 14/2I 58/64 72/83
44 67 9I 85
67 8I 95 92
3/4 2/4 9/9 IIII3
75 50 IOO 85
75 75 100 92
*Recurrence-free survival. treatment results were worse in that group. The incidences of local recurrence, distant metastasis, and absolute 2-year survival (NED) rate by stage, histologic findings, and size of primary tumor, as well as depth of invasion and vascular invasion, are shown in Table V. The stage and histologic features did not seem to influence the prognosis, but the patients who had bulky primary tumors had a higher incidence of local recurrence and distant metastases and a lower 2-year NED rate, and those who had vascular invasion and/ or deep invasion of the cervix had a higher incidence of local recurrence and lower 2-year NED rate. The correlations of the status of nodal metastases with prognosis are shown in Table VI. The patients who had positive nodes had more local recurrences, distant metastases, and lower 2-year NED rates. The patients with grossly positive nodes had more distant metastases than those with only microscopically positive nodes. All three patients with. para-aortic node metastases developed distant metastases. A retrospective analysis comparing the treatment results in the high-risk group for patients who received postoperative radiotherapy and for those who did not
revealed that postoperative radiotherapy seemed to prevent local recurrence and improve the 2-year NED rate (Table VII). The complications by treatment modality are shown in Table VIII. The complication rate in those patients with the combined approach was not increased in this small series; one patient had occasional urinary incontinence, the other patient sustained surgical injury lo the bladder which was repaired. The severe complications which required a second surgical intervention in each treatment group were comparable. The 2-year absolute survival (NED) rate by stage, treatment, and nodal status is shown in Table IX. Surgical procedures for resectable lesions achieved an 89% 2-year NED rate, and postoperative radiotherapy a 92% rate, whereas primary radiotherapy for nonresectable lesions achieved a 54% 2-year NED rate. The 2-year NED rates for patients who had positive nodes and for those who had negative nodes were 64% and 92%, respectively. Of the five patients who experienced surgical failure, four underwent radiotherapy, and success was achieved in three. One patient with local recurrence in
Volume I:lH !'.:umber 5
the postoperative radiotherapy group underwent chemotherapy which was not successful. Of four patients who experienced radiotherapy failure, one underwent a successful operation (total pelvic exenteration) and three received chemotherapy that failed.
Comment It is evident from this retrospective analysis that in
patients with Stages IB and IIA cervical cancer the various parameters associated with the prognosis are the ~ize of the cervical lesion, lymph node metastases, vascular invasion, and depth of invasion of the tumor. Piver and ChungS and Freidell and Graham 3 have confirmed that there is good correlation between the size of the cervical tumor, lymph node metastases, and survival. The patients with bulky primary tumors (~4 em) had a higher incidence of failure (40% with both local recurrence and distant metastasis) in our series. Those patients with barrel-shaped cervices did not undergo laparotomy and were treated with either primary radiotherapy or preoperative radiotherapy followed bv simple hysterectomy. Therefore, these patients were not included in this study. The patients with vascular invasion had more nodal metastases and local recurrences than did those without vascular invasion. In Van Nagel! and associates 7 series of I 00 patients with Stage lB cervical cancer who underwent radical hysterectomy, vascular invasion was associated with a significant increase in nodal metastases and tumor recurrences. No reports have previously been made of the prognostic significance of depth of invasion of cervical cancer. In the present series. the patients with deep invasion of the tumor (~70% of thickness) had three times as many nodal metastases and seven times as many local recurrences as those who did not have deep invasion. Boyce and associate~' also noticed that there was good correlation between the depth of invasion, vascular invasion, and prognosis in Stage IB cervical carcinoma. In treatment bv radiation or surgical procedures, the lymph nodes are more of a problem than is the central disease. Several authors have reported 5-year survival rates as high as 61 %" and as low as 38%2 in patients with Stage lB carcinoma who at the time of radical hysterectomv and bilateral pelvic lymphadenectomy
Treatment failures in cervical carcinoma 555
had pelvic nodes containing metastatit cancer. There are a few reports of survival of patients with operation-documented positive nodes who were treated by radiotherapy. Of 13 patients with positive nodes surgically documented prior to radiation. Lagasse and associates4 reported that only three sunived:) vears (23%). Wharton and associates" reported that II oUit patients with positive nodes were surviving 1~) to >IH months (17%). In the present series, the :2-year '.JED rate hlr patients with positive nodes was 6-1%: those who had lymph nodes that contained microscopic deposits of cancer achieved only a 73% 2-year :-.JED rate. compared to a 50c7c rate for those with gro;,sly positive nodes at laparotomy. The latter group had nine times more distant metastases than the former group. Also, the site of lymph node metastases seemed to affect the incidence of distant metastases, since ti79f of those with common iliac and/or para-aortic node imulvement developed distant metastases. whereas onh Jfi<>(. of those with only pelvir node metastases de,·eloped distant metastasis. The role of postoperative radiotherapv in the management of cervical carcinoma has not been extensively studied. ln this series, postoperative radiotherapy for those with vascular invasion and/or deep invasion of the cenix and for those with microsmpicallv involved nodes seemed to prevent local recurrence and improve the 2-vear tumor-h·ee survival rate. The :2-vear NED rates for operation alone and for operation plus postoperative radiotherapy in the high-risk group were 55% and 94c7c, respective!}. The- difference was significant at P = 0.05. Since the complication rate in those patients treated with the combined approach was not increased. our early observations suggest, therefore, that postoperative radiotherapv is indicated in patients who have microscopically positiw nodes, vascular invasion, and/or deep penetration of tumor in the radical hysterectomy specimen. Thi~ remains to be proved by randomized prospective studv lt1 addition, if further studies confirm the risk of disseminated disease in those patients with gross metastase' to lymph nodes at laparotomy. adjuvant chemothe-rapy or im· munotherapy in addition to local irradiation should be considered in the future.
REFERENCES I. Boyce, ]., and Rotman, M.: Cervical cancer refresher
course at the Twenty-first Annual Meeting of the Ameri· can Society of Therapeutic Radiologists, New Orleans, Louisiana, October, 1979. 2. Brunschwig, A., and Daniel, W. W.: The surgical treat-
ment of cancer of the cervix uteri. AM. .J. OasTET. GYNECOL. 75:875, 1958. 3. Friedell, G. H., and Graham, J. B.: Regional lymph node involvement in small carcinoma of the tervix. Surg. GynecoL Obstet. 108:513, 1959.
556 Chung et al.
:--Jovernher
Am.
4. Lagasse, L. D., Smith, M. L., Moore,]. G., et al.: The effect of radiation theranv on nelvic lvmnh node involvement in Stage I carcinoma' ~f th~ cervix: A~. J. 0BSTET. GYNECOL. 119:328, 1974. 5. Liu, \V .. and ~v1eigs,j. V.: Radical hysterecton1y and pelvic lymphadenectomy, AM. J. 0BSTET. GYNECOL. 69:1, 1955. 6. Piver, M.S., and Chung, W. S.: Prognostic significance of cervical lesion, size, and pelvic node metastasis in cervical carcinoma, Obstet. Gynecol. 46:507, 1975.
J.
I. I ~IKII
Ohsret. Gvnccoi.
7. Van Nageli,J. R., Donaldson, E. S., Wood, E. G .. et al.: The significance of vascular invasion and lymphocytir infiltration in invasive cervical cancer, Cancer 41:228, I97S. 8. Wharton,]. T.,Jones, H. W., Day, T. G., eta!.: Preirradiation celiotomy and extended field irradiation for invasive carcinoma of the cervix, Obstet. Gynecol. 49:333, 1977.
Hershey, Pennsylvania Between April, 1971, and September, 1977, 98 patients with Stages 18 and IIA cervical cancer who underwent surgical exploration prior to treatment at Hershey Medical Center were studied. Those who had bulky primary tumor (2=4 em) had a higher incidence of nodal metastases (80% vs. 16%), local recurrences (40% vs. 5%), and distant metastases (40% vs. 1%). Patients who had positive nodes had more local recurrences (24% vs. 6%) and distant metastases (28% vs. 0%). Those with grossly positive nodes had more distant metastases (60% vs. 7%) than those with microscopically positive nodes. Those who had positive nodes, vascular invasion, and/or deep invasion of the cervix (2: 70% of thickness) in the radical hysterectomy specimen had more nodal metastases and local recurrences. Postoperative radiation seemed to prevent local recurrences (40% vs. 6%) and improve the 2-year tumor-free survival rate (94% vs. 55%). Patients who had bulky primary tumors and/or grossly positive nodes at laparotomy may require systemic therapy in view of the high incidence Of distant failures. (AM. J. 0BSTET. GYNECOL. 1313:550, 1980.)
Su CCESSFU Ltreatment of Stages IB and IIA cancer of the cervix requires control of the central tumor and eradication of metastatic cancer foci when present in the regional nodes. The purpose of this investigation was to determine the prognostic significance of pelvic node metastasis, size and depth of invasion of the primary lesion, and the presence of vascular invasion in the radical hysterectomy specimen. The value of postoperative radio-
From the Divisions of Radiation Oncology and Gynecologic Oncology, the Department of Pathology, and The Specialiud Cancer Research Center, Milton S. Hershey Medical Center, College of Medicine. The Pennsylvania State University. Presented at the Sixty-fifth Annual Meeting of the Radiologic Society of North America, Atlanta, Georgia, November 29, 1979. Receivedfor publication February 13, 1980. Revised june 27, 1980. Accepted july 21, 1980. Reprint requests: Rodrigue Mortel, M.D., Division of Gynecologic Oncology, M. S. Hershey Medical Center, The Pennsylvania State University, Hershey, Pennsylvania 17033.
550
therapy for those m the high-risk group was also evaluated.
Methods One hundred forty patients with previously untreated FIGO (International Federation of Gynaecology and Obstetrics) Stages IB and IIA carcinoma of the cervix were treated at the Milton S. Hershey Medical Center of The Pennsylvania State University between April, 1971, and September, 1977. Forty-two of these patients who were medically ineligible or had barrelshaped cervix underwent no surgical exploration and were not included in the study, thus leaving as eligible a total of 98 patients who underwent laparotomy. The patients were divided into three groups according to the treatment plan (Table 1). At laparotomy, sampling of aortic nodes and multiple biopsies of grossly enlarged pelvic nodes were done. If nodes were positive and/or the primary tumor was found to be more advanced, the condition was regarded as nonresectable and the patient was treated with radiotherapy alone. If nodes were negative, radical hysterectomy with bilateral pelvic lymphadenectomy was performed. Postoperative radiation was usually administered to 0002·9378/80/210550+07$00.70/0
©
1980 The C. V. Mosby Co
Volume 138 ~umber
Treatment failures in cervical carcinoma
551
.'i
Table I. Ninety-eight patients with Stages IB and IIA cervical cancer who underwent laparotomy No. of patients Treatment
Operation* Operation plus
I Average age (yr)
65
8
II
I
45 46
Totalt
9
85
4
13
47
*Radical hysterectomy with bilateral pelvic lymphadenectomy. tRadical hysterectomy with bilateral pelvic lymphadenectomy plus postoperative radiotherapy. those in whom positive nodes were found on radical hysterectomy and pelvic lymphadenectomy specimen. A totai of 73 patients underwent operation aione (radical hysterectomy with bilateral pelvic lymphadenectomy). Tweive patients underwent operation pius postoperative radiotherapy: 10 patients with positive nodes. one patient with positive vaginal resectJon margin, and one patient in whom the tumor penetrated through the wall on the radical hysterectomy specimen. (The other five patients with positive nodes did not receive postoperative radiotherapy because of surgical complications.) Thirteen patients underwent radiotherapy alone: nine patients with gross nodal metastases and four patients with more advanced disease at laparotomv (the tumor extended into the parametrium, 1, pelvic wall, 1, and bladder, 2). Radiotherapy was administered in a fairly standard fashion and it consisted of external-beam radiotherapy to the whole pelvis with t.he use of a 10-mev linear accelerator followed by intracavitary brachytherapy. External-beam radiation therapy delivered a minimum whole-pelvic dose of 5,100 rads through anteriorposterior opposed or four-field technique. All fields were treated every day, 5 days a week with a daily tumor dose of 170 rads. The minimum treatment volume was maintained as follows: superiorly, the upper margin of L5; inferiorly, the upper third of the obturator foramen, or lower if indicated; laterally, at least 1 em beyond the lateral margin of the bony pelvis at the widest plane of the pelvis. Additional para-aortic node irradiation was given to three patients with positive para-aortic nodes. Intracavitary brachytherapy consisted of the Fletcher-Suit afterloading system with cesium-137. Two 48-hour applications with a 2-week interval were used for primary radiotherapy; the first insertion of cesium delivered an average of 2,714 (2,587 to 3,226) mg hr, with 2,234 rads (2,088 to 2,640) to point A. and 486 rads (336 to 576) to the lateral
Left
Lymph node group
External iliac Obturator J-J vnno-!:!!drir ..... ,r .... e-...............
postoperative
radiotherapyt Radiotherapy Total
Table II. Sites of lymph node metastases* (Stages iB and IIA cervical cancer)
Parametrial Total pelvic Common iliac Para-aortic
5 2
4
4
I
15
2
%
ll
36
.,
!6
19
fj
3
0 0
I
()
:)
9
II
2 2
25
80
:I
10 10
2
2 12 2
No.
1 2
14
0
:~
31
2
100
*Number of lymph nodes dissected: eight to 76 nodes (average, 24.3). tNumber of sites of nodal metastases. Table III. Incidence of lymph node metastases (+)Sodes Charack~.stics
Stage: IB
IIA
Histologic features:
Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Size of primary tumor*: ~4cm
<4cm Depth of invasiont:
~70% thickness <70% thickness
Vascular invasiont:
+
1Vo. C\1 /0!'!
% 0"
~ lfO~
6;;.}
4/13
31
23/88 2/8 0/2
26 25
12/15 13/83
80t 16
6/15 9170
40t 13
5/8
63t 13
10177
()
icClinical measurement.
tP < 0.05 (chi-square test). hysterectomy specimen findings.
~Radical
pelvic wall (lateral margin of the bony pelvis at the widest plane through the pelvis). The second insertion delivered an average of 2,964 (2, 145 to 3, 144) mg hr with 2,382 rads (1,914 to 2,640) to point A and 434 rads (384 to 528) to the lateral pelvic wall. Therefore, a total of 5,679 mg hr with 4,616 rads to point A and 920 rads to the lateral pelvic wall was delivered in two insertions. For the postoperative radiotherapy group, seven patients underwent external pelvic irradiation only (5, 100 rads in 6 weeks), and five patients underwent, in addition, one 48-hour insertion of a vaginal applicator to deliver an average of 1,566 (960 to 1,920) mg hr with 4,864 (4,080 to 5,760) rads to the vaginal surface. The size of the primary tumor, as measured at the time of initial clinical staging, was divided into two
552 Chung et al. .\m.
.\!ovember L I'IKII Obstet. Gmeml.
J.
Table IV. Sites of failure by treatment modalities Local failures No. of patients
Treatment Operation Operation plus postoperative radiotherapy Radiotherapy · · · Total
73 12 13
98
% 3 8 13
2 1 1
3 0
3
4
4
4 0 23
6
0 1 6
()
8
46
7
6
7
Table V. Treatment results Local recurrence No. of patients Stage IB IIA Histologic features: Souamous-cell carcinoma Adenocarcinoma Adenosquamous carcinoma Size of prirr~ry turr..or ~4cm
<4cm Depth of invasion ~70% thickness <70% thickness Vascular invasion
+
No.
I
Distant metastasis
%
No.
I
2-Year survival* (NED)t %
No.
I
%
85 13
9
11
8
5 2
6 15
72 11
85 85
88 8 2
8 1 1
9 13 50
7 0 0
8 0 0
75 7 1
85 88 50
IS 83
6 4
40 5
6 1
40 1
7 76
47:j: 92
15 70
4 3
27 4
2 0
13 0
10 65
67:j: 93
8 77
2 4
25 5
1 0
13 0
4 72
SO:j: 94
*Absolute recurrence-free survival. tNo evidence of disease at the follow-up examination immediately after the cut-off time at 24 months after treatment. :j:P < 0.05.
groups: 4 em or lllore vs. less than 4 em on clinical measurement. The hysterectomy specimen was examined in a fresh state and after formalin fixation. Fullthickness sections through the tumor (minimum of three) were taken from areas grossly showing the greatest depth of penetration. These were stained with hematoxylin and eosin. The maximum depth of tumor invasion was determined as a percentage of thickness of the cervical-endocervical segment. Vascular invasion (lymphatic and venous) w,as determined from the slides stained with hematoxylin and eosin. Sections of all lymph nodes were studied. When nodes were large, a single representative section was taken. In specimens in which nodes were not grossly detectable, a sample of tissue was subjected to search for foci of lymphatic tissue. The incidence of nodal metastasis was presented according to size of primary tumor, degree of depth of invasion. and nresence of vascular invasion. Treatment results were analyzed as local recurrence, distant metastases, and 2-year absolute recurrence-free survival (NED), by size of primary tumor, status of -
-
'-
lymph node metastases, hysterectomy specimen findings (vascular invasion and depth of invasion), and treatment modalities. Particularly, treatment outcome was compared between operation alone and operation plus postoperative radiotherapy in those in the highrisk 0 1oup with lymph node metastases, deep penetration of the tumor, and vascular invasion. Finally, comorder to find out whether the combined-therapy group had increased complications. Follow-up care was carried out in the Gynecologic Oncology Clinic. Absolute survival was calculated from the initial date of treatment. All patients were followed until death or for a minimum of 2 years from the date of first treatment. Three patients were lost to follow-up and counted as dead. Statistical evaluation was carried out according to the chi-square method.
l
Results
Of a total of 98 patients, 10 ( 10%) had gross lymph node metastases at the time of laparotomy; 15 of 85
Volume 13ii Number 5
Treatment failures in cervical carcinoma
553
Table VI. Status of nodal m.etastases and prognoses Nodal site
Status
of metastasis
Pelvic
Gross*(+) Microt (+)
Common iliac
Gross(+) Micro(+)
Para-aortic
0/6
Oil
Total
1/3 (33%) 2/2 II 1 3/3 (100%) 6/10 (60%) 1/15 (7%) 7/25 (28%) 0173 (0%)
Oil 2/3 (67%) 3/10 (30%) 3/15 (20%) 6/25 (24%) 4173 (6%)
I~·
'flO
9/13 13/19 (68%) 1/2 1/1 2/3 (67%1 0/2
Oil
1/3 (33%) 2/2
Gross(+) Micro(+) Node(+) Node(-)
,.
3/6
0/13 3/19 (16%) 112
31l3 3/19 (16%) 112
Gross(+) Micro(+)
2-Yrar survival (NED)
Distant metastases
Local recurrence
Ill
1/3 {:{:l 0/c) 5/10 (50%)~ 11/15 (73%) 16/25 (64%)§ 67173 (92%)
*Gross nodal metastases at laparotomy, which were all removed. snecimen. tMicrosconicallv ,--r--rnositive -- nocles ---- on --- racliral --------- hvsterertomv r *Not significant. §P < 0.05. -~
1
1
--,~--
Table VII. Retrospective comparison of treatment results in the high-risk group: Operation alone vs. operation plus postoperative radiotherapy Local recurrence Operation* Pathologic specimen findings
No.
Lymph node metastasis ( +) Depth of invasion (~70% thickness) Vascuiar invasion (+) Total
2/5:j: 4/10 2/5 8/20
I
2-Year survival (NED!
Operation* + RTt
%
No.
40 40 40 40
1110 0/5 0/3
1118
I
Operation*
%
No.
10
2/5 71l0
0 0 6
2/5
11120
l
Operation + RTt
%
No.
40 70 40 55
9/10 5/5 3/3
171l8§
I
% 90 100 iOO 94
*Radical hysterectomy. tRadical hysterectomy plus postoperative radiotherapy. :j:Five patients who had microscopic positive nodes did not receive postoperative radiotherapy because of surgical complications. §P = 0.05. patients (18%) who underwent radical hysterectomy with bilateral pelvic lymphadenectomy were found to have unexpected lymph node metastases microscopically in the pathologic specimen. Of the 85 patients with Stage IB carcinoma, 21 had positive nodes (25%); 19 had pelvic only, one had pelvic and para-aortic, and one had para-aortic only. Of the 13 patients with Stage IIA carcinoma, 4 patients had nodal metastasis (31 %), three had pelvic only, and one had para-aortic oniy. Table II shows the distribution of lymph node metastases; the external iliac node was the most common site of involvement. Only two patients had lymph node metastases bilaterally. The incidences of lymph node metastasis, by stage, histologic features, size of primary tumor, as well as depth of invasion and vascular invasion, are shown in T-:JhlP TTJ ThP _.. ...... .....,,. ...................... ,. ... ..._
~t':llrrP
fiR
u~::
TJ A\
seem to influence lymph node metastasis; however, patients with bulky primary tumors (2:4 ern) had a significantly higher incidence of nodal metastasis than did those with smaller tumors (80% vs. 16%). Those who had vascular invasion and/or deep invasion of the cervix (2:70% of thickness) on hysterectomy specimen had more lymph node metastases. In this series, 88% of recurrence occurred within 24 months, and aii within 30 months of treatment. Therefore, the 2-year NED rate is reliable for the purpose of comparing the various treatment results. Ten patients experienced local failure in the pelvis (four central and six lateral), and seven patients developed distant metastases (four patients in the bones, two patients in the supraclavicular nodes, and one patient in the mediastinum). The sites of failure by treatment modality are
~nrl h1ct-£'\ln.~r .f'..::..-:~h1r£»co
'-'OAA.t)"'- , .... ...., ..,..,, .._.._,.._, u,..L.o.'-" .L.L.Lo;:,IL..OJ.L'-'E).a"- .L"-U..IL..U.L'-'"
(squamous-cell carcinoma vs. adenocarcinoma) do not
vorable conditions received radiotherapy alone, the
554 Chung et al. Am.
November I, 1980 Gvnecol.
J. Obstet.
Table VIII. Complications by treatment modalities Complications
Fistulas Urinary retention and/or incontinence Lymphocyst Pelvic abscess ~-\cute
1
Operation (7 3 patients)
Operation + postoperative radiotherapy ( 12 patients)
Radiotherap)' ( 13 patients)
7 (9%)* I3
0 (0%)
I (7%}t
2 2
pyelonephritis
Surgical injury to bladder or rectum Severe hemorrhage Small bowel obstruction Rectosigmoiditis Severe complication*
0 0 0
0 0 0 0 0 0 I I
I (8%)
3 (23%)
1
0 0 0
3
I
I I
0
17 (23%)
*Seven patients with 10 fistulas (five ureterovaginal, three vesicovaginal, one rectovaginal, and one vesicocutaneous) underwent surgical correction within 2 to IO months after radical hysterectomy, with good primary results in all cases. tRectovaginal fistula. :j:Required second surgical intervention.
Tabie IX. Two-year absolute survival (NED) Stage IB cancer
Stage JIA cancer
OveraU survival (%)
Operation Operation plus ....,....,.,r:,.,.....,...,. .. ..,..: .. ..,. ..,...,.,li,....foho'l"<:~or..'l.r yv;;n. vp'-• A"l ,..._ a aouu. n.a•.._1 "'t'!
Radiotherapy Positive nodes Negative nodes Total
NED*(%)
OveraUsurvival(%)
58/65 IO/II
89 9I
95 9I
7/8 Ill
88 100
100 100
4/9 14/2I 58/64 72/83
44 67 9I 85
67 8I 95 92
3/4 2/4 9/9 IIII3
75 50 IOO 85
75 75 100 92
*Recurrence-free survival. treatment results were worse in that group. The incidences of local recurrence, distant metastasis, and absolute 2-year survival (NED) rate by stage, histologic findings, and size of primary tumor, as well as depth of invasion and vascular invasion, are shown in Table V. The stage and histologic features did not seem to influence the prognosis, but the patients who had bulky primary tumors had a higher incidence of local recurrence and distant metastases and a lower 2-year NED rate, and those who had vascular invasion and/ or deep invasion of the cervix had a higher incidence of local recurrence and lower 2-year NED rate. The correlations of the status of nodal metastases with prognosis are shown in Table VI. The patients who had positive nodes had more local recurrences, distant metastases, and lower 2-year NED rates. The patients with grossly positive nodes had more distant metastases than those with only microscopically positive nodes. All three patients with. para-aortic node metastases developed distant metastases. A retrospective analysis comparing the treatment results in the high-risk group for patients who received postoperative radiotherapy and for those who did not
revealed that postoperative radiotherapy seemed to prevent local recurrence and improve the 2-year NED rate (Table VII). The complications by treatment modality are shown in Table VIII. The complication rate in those patients with the combined approach was not increased in this small series; one patient had occasional urinary incontinence, the other patient sustained surgical injury lo the bladder which was repaired. The severe complications which required a second surgical intervention in each treatment group were comparable. The 2-year absolute survival (NED) rate by stage, treatment, and nodal status is shown in Table IX. Surgical procedures for resectable lesions achieved an 89% 2-year NED rate, and postoperative radiotherapy a 92% rate, whereas primary radiotherapy for nonresectable lesions achieved a 54% 2-year NED rate. The 2-year NED rates for patients who had positive nodes and for those who had negative nodes were 64% and 92%, respectively. Of the five patients who experienced surgical failure, four underwent radiotherapy, and success was achieved in three. One patient with local recurrence in
Volume I:lH !'.:umber 5
the postoperative radiotherapy group underwent chemotherapy which was not successful. Of four patients who experienced radiotherapy failure, one underwent a successful operation (total pelvic exenteration) and three received chemotherapy that failed.
Comment It is evident from this retrospective analysis that in
patients with Stages IB and IIA cervical cancer the various parameters associated with the prognosis are the ~ize of the cervical lesion, lymph node metastases, vascular invasion, and depth of invasion of the tumor. Piver and ChungS and Freidell and Graham 3 have confirmed that there is good correlation between the size of the cervical tumor, lymph node metastases, and survival. The patients with bulky primary tumors (~4 em) had a higher incidence of failure (40% with both local recurrence and distant metastasis) in our series. Those patients with barrel-shaped cervices did not undergo laparotomy and were treated with either primary radiotherapy or preoperative radiotherapy followed bv simple hysterectomy. Therefore, these patients were not included in this study. The patients with vascular invasion had more nodal metastases and local recurrences than did those without vascular invasion. In Van Nagel! and associates 7 series of I 00 patients with Stage lB cervical cancer who underwent radical hysterectomy, vascular invasion was associated with a significant increase in nodal metastases and tumor recurrences. No reports have previously been made of the prognostic significance of depth of invasion of cervical cancer. In the present series. the patients with deep invasion of the tumor (~70% of thickness) had three times as many nodal metastases and seven times as many local recurrences as those who did not have deep invasion. Boyce and associate~' also noticed that there was good correlation between the depth of invasion, vascular invasion, and prognosis in Stage IB cervical carcinoma. In treatment bv radiation or surgical procedures, the lymph nodes are more of a problem than is the central disease. Several authors have reported 5-year survival rates as high as 61 %" and as low as 38%2 in patients with Stage lB carcinoma who at the time of radical hysterectomv and bilateral pelvic lymphadenectomy
Treatment failures in cervical carcinoma 555
had pelvic nodes containing metastatit cancer. There are a few reports of survival of patients with operation-documented positive nodes who were treated by radiotherapy. Of 13 patients with positive nodes surgically documented prior to radiation. Lagasse and associates4 reported that only three sunived:) vears (23%). Wharton and associates" reported that II oUit patients with positive nodes were surviving 1~) to >IH months (17%). In the present series, the :2-year '.JED rate hlr patients with positive nodes was 6-1%: those who had lymph nodes that contained microscopic deposits of cancer achieved only a 73% 2-year :-.JED rate. compared to a 50c7c rate for those with gro;,sly positive nodes at laparotomy. The latter group had nine times more distant metastases than the former group. Also, the site of lymph node metastases seemed to affect the incidence of distant metastases, since ti79f of those with common iliac and/or para-aortic node imulvement developed distant metastases. whereas onh Jfi<>(. of those with only pelvir node metastases de,·eloped distant metastasis. The role of postoperative radiotherapv in the management of cervical carcinoma has not been extensively studied. ln this series, postoperative radiotherapy for those with vascular invasion and/or deep invasion of the cenix and for those with microsmpicallv involved nodes seemed to prevent local recurrence and improve the 2-vear tumor-h·ee survival rate. The :2-vear NED rates for operation alone and for operation plus postoperative radiotherapy in the high-risk group were 55% and 94c7c, respective!}. The- difference was significant at P = 0.05. Since the complication rate in those patients treated with the combined approach was not increased. our early observations suggest, therefore, that postoperative radiotherapv is indicated in patients who have microscopically positiw nodes, vascular invasion, and/or deep penetration of tumor in the radical hysterectomy specimen. Thi~ remains to be proved by randomized prospective studv lt1 addition, if further studies confirm the risk of disseminated disease in those patients with gross metastase' to lymph nodes at laparotomy. adjuvant chemothe-rapy or im· munotherapy in addition to local irradiation should be considered in the future.
REFERENCES I. Boyce, ]., and Rotman, M.: Cervical cancer refresher
course at the Twenty-first Annual Meeting of the Ameri· can Society of Therapeutic Radiologists, New Orleans, Louisiana, October, 1979. 2. Brunschwig, A., and Daniel, W. W.: The surgical treat-
ment of cancer of the cervix uteri. AM. .J. OasTET. GYNECOL. 75:875, 1958. 3. Friedell, G. H., and Graham, J. B.: Regional lymph node involvement in small carcinoma of the tervix. Surg. GynecoL Obstet. 108:513, 1959.
556 Chung et al.
:--Jovernher
Am.
4. Lagasse, L. D., Smith, M. L., Moore,]. G., et al.: The effect of radiation theranv on nelvic lvmnh node involvement in Stage I carcinoma' ~f th~ cervix: A~. J. 0BSTET. GYNECOL. 119:328, 1974. 5. Liu, \V .. and ~v1eigs,j. V.: Radical hysterecton1y and pelvic lymphadenectomy, AM. J. 0BSTET. GYNECOL. 69:1, 1955. 6. Piver, M.S., and Chung, W. S.: Prognostic significance of cervical lesion, size, and pelvic node metastasis in cervical carcinoma, Obstet. Gynecol. 46:507, 1975.
J.
I. I ~IKII
Ohsret. Gvnccoi.
7. Van Nageli,J. R., Donaldson, E. S., Wood, E. G .. et al.: The significance of vascular invasion and lymphocytir infiltration in invasive cervical cancer, Cancer 41:228, I97S. 8. Wharton,]. T.,Jones, H. W., Day, T. G., eta!.: Preirradiation celiotomy and extended field irradiation for invasive carcinoma of the cervix, Obstet. Gynecol. 49:333, 1977.