Inl. J. Radmion 0ncol~g.v Biul. Phvs.. Vol. 23, pp. 501-509 Printed m the U.S.A. All rights reserved.
Copyright
0360-3016/92 $5.04 t .oO 0 1992 Pergamon Press Ltd.
??Clinical Original Contribution
EXTENDED
FIELD IRRADIATION FOR CARCINOMA OF THE UTERINE CERVIX WITH POSITIVE PERIAORTIC NODES
ANTONIO P. VIGLIOTTI, M.D.,* B-CHEN WEN, M.D.,* DAVID H. HUSSEY, M.D.,* J. FRED DOORNBOS, M.D.,* JOHN J. STAPLES, M.D.,+ SHIRISH K. JANI, PH.D.,+ DEBORAH A. TURNER, M.D.’ AND BARRIE ANDERSON, M.D.* The University of Iowa College of Medicine, Iowa City, IA Forty-three patients were treated with extended field irradiation for periaortic metastasis from carcinoma of the uterine cervix (FIG0 stages IB-IV). Twelve patients (28%) remained continuously free of disease to the time of analysis or death from intercurrent disease, 20 (46%) had persistent cancer within the pelvis, 11 (26%) had persistent periaortic disease, and 23 (53%) developed distant metastasis. The actuarial 5-year survival rate was 32%. The results correlated well with the periaortic tumor burden at the time of irradiation. None of 19 patients (0%) with microscopic or small (< 2 cm) periaortic disease had periaortic failures, compared to 29% (4/14) of those with moderate-sized (2-5 cm) disease and 70% (7/10) of those with massive (> 5 cm) periaortic metastasis. Similarly, the 5-year survival rates were 50% (6/12) with microscopic disease, 33% (2/6) with small gross disease, 23% (3/13) with moderate-sized disease, and 0% (O/10) with massive periaortic metastases. Only 10% (l/10) of patients whose tumor extended to the Ll-2 level survived 5 years, compared with 31% (9/29) of those whose disease extended no higher than the L34 level. The periaortic failure rates correlated to some extent with the dose delivered through extended fields, although the difference was not statistically significant. Only 8% (l/13) of those who had undergone extraperitoneal lymphadenectomies developed small bowel complications, compared with 25% (7/29) of those who had had transperitoneal lymphadenectomies. The incidence of small bowel obstruction was 8% (l/13) following periaortic doses of 40004500 cGy, 10% (l/10) after 5000 cGy, and 32% (6/19) after -5500 cGy. From this, we concluded that the subset of patients who would benefit most from extended field irradiation are those in whom the residual disease in the periaortic area measures less than 2 cm in size at the time of treatment, whose disease extends no higher than L3, and whose cancer within the pelvis has a reasonable chance of control with standard radiation therapy techniques. Radiotherapy, Cervical cancer, Periaortic lymph node irradiation, Extended field irradiation.
Over the last several decades, a number of studies have been performed to evaluate the effectiveness of periaortic irradiation in patients with uterine cervix cancer (1-4, 6, 11, 13- 16, 18, 2 1). These studies were based on the observation that cancer of the uterine cervix tends to spread in an orderly fashion, first to the pelvic lymph nodes and then to the periaortic nodes, with distant metastasis occurring relatively late in the course of the disease. The rationale was that a subset of patients with metastasis to the regional lymph nodes could be cured with extended field irradiation if the disease was confined to the pelvis
and periaortic areas and small enough to be eradicated with moderate doses of radiation. This paper is a review of the University of Iowa experience with extended field irradiation for carcinoma of the uterine cervix metastatic to the periaortic lymph nodes. The principal objective was to determine whether extended field irradiation is effective in curing patients with known metastasis in the periaortic area. The specific aims were: (a) to evaluate the patterns of failure following extended field irradiation; (b) to correlate local control in the periaortic area with tumor burden, level of involvement, and radiation dose; and (c) to correlate complication rates with the surgical technique used and the radiation
Presented at the 3 1st Annual Meeting of the American Society for Therapeutic Radiology and Oncology, l-6 October 1989, San Francisco, CA. * Division of Radiation Oncology, Dept. of Radiology. + Division of Radiation Oncology, Dept. of Radiology.
* Division of Gynecologic Oncology, Dept. of OB/Gynecology. Reprint requests to: David H. Hussey, M. D., Division of Radiation Oncology, Room W 189Z-GH, University of Iowa Hospitals and Clinics, Iowa City, IA 52242-1059. Accepted for publication 6 January 1992.
INTRODUCTION
501
502
I. J. Radiation Oncology 0 Biology 0 Physics
dose delivered. The ultimate aim was to be able to select a subset of patients with a greater chance of benefiting from extended field irradiation. METHODS
AND MATERIALS
Patient population Between 1966 and 1985, 8 10 patients with carcinoma of the uterine cervix were treated with radiation therapy at the University of Iowa Hospitals and Clinics. Of these, 43 patients were treated with extended field irradiation for periaortic metastases. All but one of these patients had histologic evidence of periaortic disease. The other had clinical evidence of periaortic metastasis at laparotomy, but no biopsy was obtained. The patients ranged in age from 28 to 69 years, with a mean of 5 1 years. Thirty-eight patients had squamous carcinomas and five had adenocarcinomas. The patients were retrospectively staged on the basis of clinical findings using the 1988 FIG0 classification. Five patients presented with Stage IB, one with Stage IIA, 7 with Stage IIB, 24 with Stage IIIB, 3 with Stage IVA, and 1 with Stage IVB cancer. Two patients were treated for pelvic disease that was recurrent following a radical hysterectomy. The clinical evaluation for these patients included a history and physical examination, a chest X ray, and liver function tests. An intravenous pyelogram was performed in 42 patients, cystoscopy in 33 patients, sigmoidoscopy in 33 patients, and a barium enema in 28 patients. Six patients had pedal lymphangiograms, and four of these were positive for periaortic metastases. Abdominal CT scans were obtained in six patients, and three of these showed enlarged pelvic and periaortic lymph nodes. Twenty-three patients had negative prescalene node biopsies, and one had a positive prescalene node biopsy obtained during the course of irradiation. Peritoneal cytology was obtained in 28 patients with positive findings in 11. Forty-two patients underwent surgical staging. A transperitoneal approach was used in 29 patients and an extraperitoneal approach in 13. One patient did not have a staging lymphadenectomy, but had a positive CT-guided needle biopsy of the periaortic lymph nodes. Forty-two patients had histologic proof of periaortic nodal involvement. One patient with clinically positive periaortic nodes at laparotomy did not have a biopsy for histologic confirmation. The pelvic and periaortic nodes were palpated at laparotomy, and the number, size, and location of the positive nodes were assessed. In some cases, the suspicious nodes were resected and in others only a biopsy was performed. If the exploration was clinically negative, sampling biopsies were obtained. Only one patient had a formal lymphadenectomy. This patient had 25 nodes removed, 21 of which were involved by tumor.
Volume 23, Number 3, 1992
Radiation therapy The majority of the patients were treated with a combination of external beam irradiation and brachyradiotherapy. Twenty-five patients were treated with 6oCo gamma rays, 4 with 4 MV x-rays, and 14 with 10 MV x-rays. Anterior and posterior, parallel-opposing portals were used in 36 patients and a 4-field technique was employed in 7 patients. The treatment portals extended to the superior margins of T 12 in 12 patients, L 1 in 24 patients, and L2 in 7 patients. The dose to the periaortic nodes ranged from 3960 cGy to 6000 cGy (median = 5040 cGy). It was usually delivered at a rate of 180 cGy per day. The dose to the whole pelvis ranged from 3960 cGy in 20 fractions to 6510 cGy in 35 fractions (median = 4800 cGy in 26 fractions). The brachyradiotherapy treatment was usually delivered with radium, and the doses ranged from 1200 to 7800 mg-hr (median = 5625 mg-hr). The dose to point A, including both the external beam and intracavitary components, ranged from 5250 cGy to 13,000 cGy (median = 9510 cGy). RESULTS
The patients were treated between 1966 and 1985, and the data were analyzed in September 1989. The patients have been followed for periods ranging from 39 months to 18 years. The median follow-up time was 131 years. Survival was computed from the first day of treatment until death. Disease-free survival was computed from the first day of treatment to the first sign of failure or death. Failures were classified as occurring centrally, at the pelvic sidewall, in the periaortic area, intraperitoneally, in the supraclavicular area, and hematogenously. Patients were classified as having central failures if there was persistent ulceration or nodularity in the cervix or vagina, with or without biopsy confirmation. They were classified as having pelvic wall failures if there was progressive leg edema, hydronephrosis, or sciatic pain; or if there was a mass at the pelvic sidewall on physical examination, computerized tomography, or autopsy. Both central and pelvic wall failures were classified as pelvic failures. Patients were classified as having periaortic failures if there was clinical evidence of a persistent mass in the periaortic area or histological evidence of periaortic disease at autopsy. Patients were considered to have intraperitoneal metastases if there was ascites with a positive cytology or evidence of serosal studding at laparotomy or autopsy. Intraperitoneal, supraclavicular, and hematogenous metastases were all classified as distant metastasis. Patterns offailure The patterns of failure are listed in Table 1. Of the total group of 43 patients, 16 (37%) had progressive cancer at the primary site, 15 (35%) developed recurrent tumor at the pelvic sidewall, 11 (26%) had persistent disease in the periaortic area, 11 (26%) had intraperitoneal spread, 5
503
Extended field irradiation 0 A. P. VIGLIOTTI et al.
Table 1. Sites of failure for the total group of 43 patients with periaortic metastases*
Pelvic failures Central Pelvic wall Periaortic failures Distant metastases Intraperitoneal Supraclavicular fossa Hematogenous
Only site of failure
Total
No. (%)
No. (W)
8 (19) 5 (12) 3 (7) 1 (2) 7 (16) 0 (0) 2 (5) 5 (12)
20 (47) 16 (37) 15 (35) 11 (26) 23 (53) 11 (26) 5 (12) 16 (37)
ACTUARIAL
BY STAGE
30% 20% 10%
0%’ 0
* 12 patients (28%) had no evidence of disease (NED).
SURVIVAL
EXTENDEDFIELD CERVIX
I 12
24
33
43
30
MONTHS
( 12%) developed supraclavicular metastases, and 16 (37%)
developed hematogenous metastases. Most of these patients had progressive disease in more than one site. Twelve patients (28%) remained continuously free of disease (NED) to the time of analysis or death from intercurrent disease. The most common cause of failure in this series was distant metastasis (53%). However, a significant percentage of the patients (47%) developed progressive disease within the pelvis. Most of these also failed in the periaortic area or distantly. The relatively high local failure rate was not unexpected because the majority ofthese patients had massive pelvic disease initially. Actuarial survival for the total population is illustrated in Figure 1. Almost all of the deaths from cancer occurred within 3 years, and the actuarial 5-year survival rate was 32%. Survival rates are compared by clinical stage in Figure 2. Central failures. Sixteen patients (37%) had persistent or recurrent disease at the primary site. This correlated well with the clinical stage of the disease (Table 2). The
Fig. 2. Actuarial survival by clinical stage ( 13 patients I-II and 30 patients with stage III-IV or recurrent tumors).
central failure rate was 47% (14/30) for patients with Stage III-IV or recurrent tumors, compared to only 15% (2/l 3) for those with Stage I-II disease. The central failure rate also correlated well with the bulk of the cancer within the pelvis. Whereas 48% ( 12/25) of patients with tumors 2 6 cm in size developed central recurrences, only 22% (4/l 8) of those with tumors < 6 cm in size did (Table 2). As might be expected, the control rate at the primary site correlated well with the brachyradiotherapy dose, although it did not correlate well with the dose of external beam irradiation (Table 3). Whereas 67% (10/l 5) of patients who received < 4000 mg hr developed central recurrences, only 2 1% (6/28) of those who received > 4000 mg hr failed centrally. Similarly, the central failure rate
Table 2. Central failure rate as a function of clinical stage, tumor burden, and radiation therapy dose Central failure rate
OVERALL
ACTUARIAL
SURVIVAL
EXTENDED FIELD CERVIX 1OOII 90% '1
30% 70%
No. (%)
u*
Clinical stage Stage I-II Stage III, IV, or recurrent
2/13 (15) 14/30 (47)
.05 1
Central controlled tumor burden 9600 cGy (Table 2). Pelvic wall failures. Fifteen patients (35%) had progressive disease at the pelvic sidewall, 11 of whom also had persistent disease centrally. The failure rate at the pelvic sidewall correlated well with the extent of involvement at this site. Pelvic wall failures occurred in 43% (13/30) of patients who had tumor extension to the pelvic sidewall initially (Stage IIIB, IV, and recurrent tumors), compared with only 15% (2/ 13) of those who had no clinical evidence of pelvic wall involvement prior to treatment (Stage IB-IIB) (Fisher’s exact test; p = .075). The failure rate at the pelvic sidewall did not correlate well with either the brachytherapy or external beam dose of radiation. Periaortic failures. Eleven patients (26%) had persistent disease in the periaortic area. However, only one patient (2%) had an isolated periaortic recurrence (Table 1). In the other 10 cases, it was accompanied by progressive disease in the pelvis and/or distant metastasis. The failure rate in the periaortic area correlated well with the tumor burden in this region (Table 4). Of the total group of 43 patients, 13 had all detectable periaortic tumor resected at laparotomy and only microscopic tumor
remained. Six patients had a small gross residual disease (< 2 cm in diameter), 14 had moderate-sized residual disease (2-5 cm in diameter), and 10 had massive periaortic cancer. Massive disease was described as large matted nodes, sometimes fixed to adjacent organs or extending into the adjacent mesentery. None of 19 patients (0%) with microscopic or small gross disease had periaortic failures (Table 4). However, 29% (4/14) of those with moderate-sized disease and 70% (7/10) of those with mas-
Table 5. Correlation of 5-year survival with the highest level of periaortic involvement Superior extent of periaortic disease Ll
L2 L3 L4 Unknown Total
5 year survival O/l
I/lo* (10%) l/9 6/l 2 + 9/29* (31%) 3/ 12*I 112 1 l/41
* Fisher’s exact test; p = .189. + Excludes one patient alive NED at 39 months. * Excludes one patient alive with pelvic wall failure 4 1 months.
505
Extended field irradiation 0 A. P.VIGLIOTTI etal. Table 6. Incidence of periaortic failures and small bowel complications as a function of periaortic radiation dose Periaortic dose
Periaortic failure
Small bowel obstruction
4000-4500 cGy
31% (4/13) 20% (2/10) 21% (4/19) l/l 26% (1 l/43)
(1;;) 10% (l/10) 32% (6/19) O/l 19% (8/43)
- 5000 cGy 5500 cGy 6000 cGy Total
Table 7. Small bowel obstruction as a function of the dose of external beam irradiation and the brachyradiotherapy dose (mg-hr) mg-hrs PA dose (~GY)
Zero
~4000
40015000
50016000
>6001
Total (%)
-4500 -5000 -5500 -6000
O/3 O/l 1/3* _
O/2 O/l O/3 _
O/l O/l l/4 _
Of4 O/l 213 _
1/3* l/6 216 O/l
8 10 32 -
Total (%)
I 25
10.5
* Two patients developed partial small bowel obstructions, which were treated conservatively.
sive metastasis in the periaortic area had persistent disease at this site (Chi-square = 16.96; p = .0002). The periaortic tumor burden also correlated well with survival rates (Table 4). The 5-year survival rates were 50% (6/12) for patients with microscopic disease, 33% (2/6) for those with small gross disease, 23% (3/l 3) for those with moderate residual disease, and 0% (O/10) for those with massive periaortic disease (Chi-square = 7.17;
p = .067). Survival rates also correlated well with the level of nodal involvement, although the differences are not statistically significant (Table 5). Only 1 of 10 patients ( 10%) whose tumor extended to the Ll -2 level survived 5 years, compared with 9 of 29 patients (3 1%) whose disease extended no higher than the L3-4 level (Fisher’s exact test; p = .189). The periaortic failure rates correlated to some extent with the dose delivered to the periaortic nodes, although the differences are not statistically significant (Table 6). Doses of 4000-4500 cGy resulted in a 3 1% (4/13) periaortic failure rate, compared with 20% (2/10) with doses of - 5000 cGy and 2 1% (4/ 19) with doses of - 5500 cGy. Distant metastasis Twenty-three patients (53%) developed distant metastases. These were classified as intraperitoneal, supraclavicular, or hematogenous. The majority of these patients had involvement at more than one site. Most of the distant metastases developed within two years of treatment. The distant metastasis rate correlated well with the tumor burden in the periaortic area (Table 4). Eleven patients (26%) developed generalized intraperitoneal carcinomatosis. The incidence of intraperitoneal spread correlated well with the cytology findings at laparotomy. Whereas 55% (6/l 1) of patients with positive washings at laparotomy subsequently developed intraperitoneal carcinomatosis, none (O/ 17) of those with negative washings did (Fisher’s exact test; p = .OOl). Five patients ( 12%) developed supraclavicular metastasis, but only one became apparent prior to the development of hematogenous spread. Sixteen patients (37%) developed hematogenous metastasis to lung, bone, or liver. Complications Nine patients (2 1%) developed major complications, including eight with small bowel obstructions and one
Table 8. Clinical features of the long-term survivors Tumor burden Stage
Pelvis
PA
Superior extent
IB IIB IIB IIB IIIB IIIB IIIB IIIB IIIB IIA IB IIIB IIIB
Small Moderate Moderate Moderate Large Large Massive Moderate Massive Small Large Massive Massive
Micro. Micro. Micro. Micro. Micro. Micro. Micro. Small Small Moderate Moderate Moderate Moderate
L4 ? L4 L2 L3 L4 L3 L3 L3 L3 L3 L3 L4
Perk cytologY 0 0 + 0 0 +* 0 0 +
Pelvic dose 4500 cGy 4500 cGy 4780 cGy 4 140 cGy 4140 cGy 4980 cGy 3960 cGy 4000 cGy 3940 cGy 5040 cGy 4500 cGy 4020 cGy 5040 cGy
+ + + + + + + + + + + + +
5310 5960 6720 6240 6540 4670 1200 7760 7620 5700 2600 6360 6240
PA dose mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr
4500 4500 4780 5340 5540 5480 3960 4750 5000 5540 4500 5020 5040
cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy
Status NED @ 3$ yr NED@ 5f yr NED @ 61 yr (Vag. Net.) 4 Q 14 yr 2nd 1” (SBO)* NED @ 16 yr (SBO)* NED @ 9 yr (SBO)* NED @ 65 yr NED@ 19 yr 4 @ 7$ yr 2nd lo 4 @ 7 yr of DM (SBO)* Alive 3$ yr w/PW rec. 4 8 yr complic. (SBO)* NED@ 1Oyrs
Micro = Microscopic; NED = No evidence of disease; DM = Distant metastasis; SBO = small bowel obstruction; = Vaginal necrosis; mg-hr = Milligram hours. * Five of the long-term survivors developed small bowel obstructions.
Vag. Net.
506
I. J. Radiation Oncology 0 Biology 0 Physics
with a vaginal vault necrosis. The vaginal vault necrosis healed within 6 months, and this patient is alive with no evidence of disease 6 1 years later. Six of the eight patients with small bowel obstructions required surgical intervention, and two healed with conservative management. Seven of the eight small bowel obstructions developed within 24 months of treatment, and the other occurred 5 f years after irradiation. The incidence of small bowel obstructions correlated well with the surgical technique used. Only 1 of 13 patients (8%) who had extraperitoneal lymphadenectomies developed a small bowel obstruction, compared to 7 of 29 patients (24%) who had transperitoneal lymphadenectomies. Although an association was observed, the difference is not statistically significant (Fisher’s exact test; p = .208). The incidence of small bowel obstructions correlated well with the periaortic dose of radiation (Table 6 and 7). Only 8% (l/l 3) of the patients who received periaortic doses of 4000-4500 cGy and 10% ( l/ 10) of those who received doses of - 5000 cGy developed small bowel obstructions, compared with 32% (6/19) of those who received doses of - 5500 cGy. The only two who developed small bowel injury after external beam doses less than -5500 cGy received high brachytherapy doses (Table 7). These differences are marginally significant (- 5500 cGy vs lesser doses: p = .067; Fisher’s exact test). Long-term survivors The long-term survivors were reviewed separately to define a subset of patients who might have a greater chance of benefiting from extended field irradiation. This group included 11 patients who survived more than 5 years and 2 others who were alive at the time of analysis, less than 5 years after treatment. The clinical features of these 13 patients are summarized in Table 8. The extent of the disease within the pelvis ranged from small to massive in these long-term survivors. This shows that even large cancers of the cervix can be controlled if adequate doses of radiation are delivered. However, none of the long-term survivors had massive disease in the periaortic area, and most had limited disease in this region. Nevertheless, six of the long-term survivors were treated for gross periaortic disease, and four of them had residual disease measuring 2-5 cm in diameter. Thus, small to moderate-sized periaortic metastases can be controlled with moderate doses of radiation. Another important characteristic was the superior extent of the tumor. Almost all of the survivors had periaortic disease that extended no higher than the L3-4 level. Only one had tumor extending as high as L2. On the other hand, the presence of a positive peritoneal cytology was not as ominous as expected, since three of the longterm survivors had positive peritoneal washings. There was a relationship between the dose delivered to the periaortic area and the ultimate outcome (Table 8). Although the periaortic doses for the survivors ranged from 3960 cGy to 5540 cGy, the lower doses were effective
Volume 23, Number 3, 1992
No+ Transperitoneal Trans- and extra-peritoneal (sampling)
IB-IVA IIB-IV IB-IIIB IB-IV IB-IVB
13 29 21
of Iowa Hospitals
University
& Clinics
(15)
43
24
cGy/4-5
cGy/5-6
3940-65 10 cGy/4-7
4000-4320
4500-5000
4500 cGy/S weeks
cGy 4f-8
weeks
weeks
weeks
weeks
weeks
38% at 13-36 months 31% NED at 3 years 19% at 19-63 months 32% at 5 years
27% at 5 years
25% at 3 years 3/6 (minimum 4 years) 29% at 5 years 23% NED at 5 years 38% (minimum 2 years) 43% NED at 5 years 10% at 5 years
19%*
Survival
metastasis
19% SBO
5%
II%*
38%
45%
29%
19%
5% No SBO
116
43%*
Major complications
GOG = Gynecologic Oncology Group; SBO = Small bowel obstruction; NED = Alive with no evidence of disease. * In the M.D. Anderson series, extended fields were used if there was evidence of pelvic +/- periaortic metastases at lymphangiogram or laparotomy. The complication rate appears high because each complication was counted separately and some patients had more than one complication. + In patients not undergoing lymphadenectomy or laparotomy with biopsy, the diagnosis of periaortic nodal metastasis was made clinically, usually by lymphangiography. * In the Tufts University series, 4 patients developed major complications in 36 patients who were treated with extended fields either as part of their initial treatment (29 patients) or for periaortic recurrences (7 patients).
of Miami (22)
(5)
University
(7)
University
State University
Tufts University
Georgetown
Pennsylvania 15 100 cGy
4400-6000
Transperitoneal (variable extent) Trans- and extra-peritoneal (sampling) Transperitoneal
Instit. (14) 31
IB-IV
Roswell Park Memorial
( 19) cGy/4-6
4000-5000
Transpet-itoneal
IB-IIA
14
of Pennsylvania
University
5000 cGy/S weeks
No+
IB-IV
(12)
21
of Pennsylvania
University
weeks
4500-5 100 cGy/5-6 weeks 4320-5 130 cGy 41-6; weeks
IB-IVB IB-IV
(9)
22 18
cGy/4-7
4000-5000 cGy 6000 cGy/6 weeks
4000-6000
Transperitoneal in 83 patients* Usually transperitoneal Transperitoneal (biopsy only) Transperitoneal (sampling) Extraperitoneal
dose
proven periaortic
IB-WA
University of Tennessee UCLA (1)
257
Periaortic
or histologically
I-IV I-III
(6)
diagnosed
Periaortic lymphadenectomy
for clinically
Clinical stage
field irradiation
98 6
Hospital
No. of patients
the use of extended
GOG Study (2) Downstate Medical Center (20)
M.D. Anderson
Institution
Table 10. Studies evaluating
508
I. J. Radiation Oncology 0 Biology 0 Physics
only for patients with microscopic disease. Most of the complications occurred in patients who received high periaortic doses. All of the small bowel obstructions in the survivors occurred in patients who received periaortic doses of 5020-5540 cGy. Thus, the group that fared best following extended field irradiation had relatively limited disease in the periaortic area with the tumor extending no higher than the L3-4 level. These patients received the appropriate combination of external beam and intracavitary irradiation for their pelvic disease and doses of -4500-5000 cGy in 5-54 weeks to the periaortic area. DISCUSSION Studies to evaluate the effectiveness of periaortic irradiation in carcinoma of the uterine cervix may be classified into two broad categories: (a) those in which the periaortic lymph nodes are irradiated prophylactically, that is, in which there is no clinical evidence of periaortic metastasis, and (b) those in which the periaortic lymph nodes are irradiated for known metastatic disease. This series belongs in the second category, because periaortic metastases were documented prior to treatment. This study does not address the issue of elective irradiation of the periaortic lymph nodes in patients with no evidence of periaortic metastasis. In general, studies in which the periaortic lymph nodes have been treated prophylactically report significantly greater 5-year survival rates than those in which the periaortic nodes have been treated for known metastatic disease (Table 9 and 10). In most series, the 5-year survival rates have been in the range of 50-70% if the periaortic lymph nodes were either unevaluated or contained no evidence of metastatic disease (8, 10, 18, 19, 23). On the other hand, survival rates have been in the range of 2035% if there was clinical or histopathologic evidence of periaortic metastasis at the time of irradiation (1, 2, 5-7, 9, 12-15, 19, 20, 22). The overall results of our study are similar to those reported elsewhere, with an actuarial 5year survival rate of 32% (Fig. 1). Distant metastasis is a frequent occurrence in patients with known periaortic metastasis from carcinoma of the uterine cervix. This was the most frequent cause of treatment failure (53%) in this series. Nevertheless, half of the patients with periaortic involvement in this series did not develop distant metastasis, and the incidence of distant spread was even less if the periaortic disease was limited in extent. Therefore, the presence of periaortic metastasis is not necessarily diagnostic of distant spread. The results of this study also show that the doses that can be given with extended field irradiation can control
* The complication rates in El Senoussi’s series are high because each complication was counted separately and some patients had multiple complications.
Volume 23, Number 3, 1992
the disease in the periaortic area, particularly if the tumor burden in this region is not great. Only 26% (1 l/43) of the patients in this series had periaortic failures, and all of the patients with periaortic disease less than 2 cm in size had tumor control in this region (Table 4). Even periaortic masses measuring 2-5 cm in size were controlled in a high percentage of cases. These data suggest that a dose of -5000 cGy in 5-51 weeks is effective in controlling small to moderate sized periaortic metastases in most patients (Table 4). There is a significant risk of small bowel injury with extended fields, however, particularly if doses greater than 5000 cGy in 5-51 weeks are used (Table 6). Almost a third (6/ 19) of the patients who received periaortic doses of - 5500 cGy in this series developed small bowel complications. However, only 9% (2/23) of those who received periaortic doses of 5000 cGy or less developed small bowel complications, and this is an acceptable incidence of complications. El Senoussi et al. (6) also found a doseresponse relationship for complications following extended field irradiation. They reported complication rates of 24% following 4000-4500 cGy, 44% after 5000-5500 cGy, and 57% after 5500-6000 cGy.* They concluded that 4500 cGy in 5 weeks was the maximum dose that could be delivered safely through extended fields (Table 7). Another factor that affected the incidence of complications in this series was the surgical technique. Only 8% ( l/ 13) of patients who had extraperitoneal lymphadenectomies developed small bowel obstructions, compared to 24% (7/29) of those who had transperitoneal lymphadenectomies. These results are similar to those reported by Berman et al. (3), who found a 30% incidence of small bowel injury following transperitoneal lymphadenectomies, but only a 2.5% (l/39) incidence of bowel complications in patients who had retroperitoneal lymphadenectomies. The results of this study indicate that there is a subset of patients with documented periaortic metastasis from carcinoma of the uterine cervix who would benefit from extended field irradiation. Twenty-eight percent ( 12/43) of the patients in this series had permanent control of their disease, and the actuarial 5-year survival rate was 32%. Although this subset is small, it is large enough to warrant continued use of extended field irradiation for patients with known metastasis in the periaortic area. The patients who appear most likely to benefit from extended field irradiation are those in whom the residual disease in the periaortic area is less than 2 cm in size, whose disease extends no higher than L3, and where the primary tumor within the pelvis has a reasonable chance of control with standard radiation therapy techniques.
Extended field irradiation 0 A. P.VIGLIOTTI et al.
Extended field irradiation is probably also indicated for some patients with more extensive disease, that is, those with periaortic tumors measuring 2-5 cm in size and/or extending above L3. A reasonably high local control rate can be expected in these patients, but the survival rate
509
will be low because most of these patients already have distant metastases. On the other hand, extended field irradiation is of little value for patients with massive periaortic disease, because the periaortic control rate is poor and the incidence of distant metastasis is almost universal.
REFERENCES 1. Ballon, S. C.; Berman, M. L.; LaGasse, L. D.; Petrilli, E. S.; Castaldo, T. W. Survival after extraperitoneal pelvic and paraaortic lymphadenectomy and radiation therapy in cervical carcinoma. Gynecol. Obstet. 57( 1): 90-95; 198 1. 2. Berman, M. L.; Keys, H.; Creasman, W.; DiSaia, P.; Bundy, B.; Blessing, J. Survival and patterns of recurrence in cervical cancer metastatic to periaortic lymph nodes (A Gynecological Oncology Group Study). Gynecol. Oncol. 19: 8-l 6; 1984. 3. Berman, M. L.; LaGasse, L. D.; Waning, W. G.; Ballon, S. C.; Schlesinger, R. E.; Moore, J. G.; Donaldson, R. C. The operative evaluation of patients with cervical carcinoma by an extraperitoneal approach. Obstet. Gynecol. 50(6): 658-664; 1977. 4. Chism, S. E.; Park, R. C.; Keys, H. M. Prospects for paraaortic irradiation in treatment of cancer of the cervix. Cancer 35: 1505-1509; 1975. 5. Delgado, G.; Caglar, H.; Walker, P. Survival and complications in cervical cancer treated by pelvic and extended field radiation after paraaortic lymphadenectomy. Am. J. Roentgenol. 130: 141-143; 1978. 6. El Senoussi, M. A.; Fletcher, G. H.; Borlase, B. C. Correlation of radiation and surgical parameters in complications in the extended field technique for carcinoma of the cervix. Int. J. Radiat. Oncol. Biol. Phys. 5: 927-934; 1979. 7. Emami, B.; Waning, W. G.; Tak, W.; Anderson, B.; Piro, A. J. Para-aortic lymph node radiation in advanced cervical cancer. Int. J. Radiat. Oncol. Biol. Phys. 6: 1237-1241; 1980. 8. Haie, C.; Pejovic, M. H.; Gerbaulet, A.; Horiot, J. C.; Pourquier, H.; Delouche, J.; Heinz, J. F.; Brune, D.; Fenton, J.; Pizzi, G.; Bey, P.; Brossel, R.; Pillement, P.; Volterrani, F.; Chassagne, D. Is prophylactic para-aortic irradiation worthwhile in the treatment of advanced cervical carcinoma? Results of a controlled clinical trial of the EORTC radiotherapy group. Radiother. Oncol. 11: 101-l 12; 1988. 9. Hughes, R. R.; Brewington, K. C.; Hanjani, P.; Photopulos, G.; Dick, D.; Votava, C.; Moran, M.; Coleman, S. Extended field irradiation for cervical cancer based on surgical staging. Gynecol. Oncol. 9: 153- 16 1; 1980. 10. Inoue, T.; Morita, K. 5-year results of positive extendedfield irradiation on 76 patients with nodal metastases from cervical carcinoma stages IB-IIIB. Cancer 61: 2009-2014; 1988. 11. Jampolis, S.; Andras, E. J.; Fletcher, G. H. Analysis of sites and causes of failures of irradiation in invasive squamous
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
cell carcinoma of the intact uterine cervix. Radiology 115: 681-685; 1975. Lepanto, P.; Littman, P.; Mikuta, J.; Davis, L.; Celebre, J. Treatment of para-aortic nodes in carcinoma of the cervix. Cancer 35: 1510-1513; 1975. Perez, C. A.; Bedwinek, J. M.; Breaux, S. R. Patterns of failure after treatment of gynecologic tumors. Cancer Treat. Symp. 2: 217-231; 1983. Piver, M. S.; Barlow, J. J.; Krishnamsetty, R. Five year survival (with no evidence of disease) in patients with biopsyconfirmed aortic node metastasis from cervical carcinoma. Am. J. Obstet. Gynecol. 139: 575-578; 198 1. Podczaski, E. S.; Palombo, C.; Manetta, A.; Andrews, C.: Larson, J.; DeGeest, K.; Mortel, R. Assessment of pretreatment Iaparotomy in patients with cervical carcinoma prior to radiotherapy. Gynecol. Oncol. 33: 71-75; 1989. Potish, R. A.; Twiggs, L. B.; Prem, K. A.; Levitt, S. H.; Adcock, L. L. The impact of extraperitoneal surgical staging on morbidity and tumor recurrence following radiotherapy for cervical carcinoma. Am. J. Clin. Oncol. (CCT) 7: 245251; 1984. Rotman, M.; Choi, K.; Guze, C.; Martial, V.; Homback, N.: John, M. Prophylactic irradiation of the para-aortic lymph node chain in stage IIB and bulky stage IB carcinoma of the cervix, initial treatment results of RTOG 7920. Int. J. Radiat. Oncol. Biol. Phys. 19: 5 13-521; 1990. Rotman, M.; Moon, S.; John, M.; Choi, K.; Sail, S. Extended field para-aortic radiation in cervical carcinoma: the case for prophylactic treatment. Int. J. Radiat. Oncol. Biol. Phys. 4: 795-799; 1978. Rubin, S. C.; Brookland, R.; Mikuta, J. J.; Mangan, C.; Sutton, G.; Danoff, B. Para-aortic nodal metastases in early cervical carcinoma: long-term survival following extendedfield radiotherapy. Gynecol. Oncol. 18: 2 13-2 17; 1984. Silberstein, A. B.; Aron, B. S.; Alexander, L. L. Para-aortic lymph node irradiation in cervical carcinoma. Radiology 95: 181-184; 1970. Sotto, L. S. J.; Graham, J. B.; Pickren, J. W. Postmortem findings in cancer of the cervix. Am. J. Obstet. Gynecol. 80: 791-794; 1960. Sudarsanam, A.; Charyulu, K.; Belinson, J.; Averette, H.; Goldberg, M.; Hintz, B.; Thirumala, M.; Ford, J. Influence of exploratory celiotomy on the management of carcinoma of the cervix. Cancer 41: 1049-1053; 1978.
Copyright
0360-3016/92 $5.04 t .oO 0 1992 Pergamon Press Ltd.
??Clinical Original Contribution
EXTENDED
FIELD IRRADIATION FOR CARCINOMA OF THE UTERINE CERVIX WITH POSITIVE PERIAORTIC NODES
ANTONIO P. VIGLIOTTI, M.D.,* B-CHEN WEN, M.D.,* DAVID H. HUSSEY, M.D.,* J. FRED DOORNBOS, M.D.,* JOHN J. STAPLES, M.D.,+ SHIRISH K. JANI, PH.D.,+ DEBORAH A. TURNER, M.D.’ AND BARRIE ANDERSON, M.D.* The University of Iowa College of Medicine, Iowa City, IA Forty-three patients were treated with extended field irradiation for periaortic metastasis from carcinoma of the uterine cervix (FIG0 stages IB-IV). Twelve patients (28%) remained continuously free of disease to the time of analysis or death from intercurrent disease, 20 (46%) had persistent cancer within the pelvis, 11 (26%) had persistent periaortic disease, and 23 (53%) developed distant metastasis. The actuarial 5-year survival rate was 32%. The results correlated well with the periaortic tumor burden at the time of irradiation. None of 19 patients (0%) with microscopic or small (< 2 cm) periaortic disease had periaortic failures, compared to 29% (4/14) of those with moderate-sized (2-5 cm) disease and 70% (7/10) of those with massive (> 5 cm) periaortic metastasis. Similarly, the 5-year survival rates were 50% (6/12) with microscopic disease, 33% (2/6) with small gross disease, 23% (3/13) with moderate-sized disease, and 0% (O/10) with massive periaortic metastases. Only 10% (l/10) of patients whose tumor extended to the Ll-2 level survived 5 years, compared with 31% (9/29) of those whose disease extended no higher than the L34 level. The periaortic failure rates correlated to some extent with the dose delivered through extended fields, although the difference was not statistically significant. Only 8% (l/13) of those who had undergone extraperitoneal lymphadenectomies developed small bowel complications, compared with 25% (7/29) of those who had had transperitoneal lymphadenectomies. The incidence of small bowel obstruction was 8% (l/13) following periaortic doses of 40004500 cGy, 10% (l/10) after 5000 cGy, and 32% (6/19) after -5500 cGy. From this, we concluded that the subset of patients who would benefit most from extended field irradiation are those in whom the residual disease in the periaortic area measures less than 2 cm in size at the time of treatment, whose disease extends no higher than L3, and whose cancer within the pelvis has a reasonable chance of control with standard radiation therapy techniques. Radiotherapy, Cervical cancer, Periaortic lymph node irradiation, Extended field irradiation.
Over the last several decades, a number of studies have been performed to evaluate the effectiveness of periaortic irradiation in patients with uterine cervix cancer (1-4, 6, 11, 13- 16, 18, 2 1). These studies were based on the observation that cancer of the uterine cervix tends to spread in an orderly fashion, first to the pelvic lymph nodes and then to the periaortic nodes, with distant metastasis occurring relatively late in the course of the disease. The rationale was that a subset of patients with metastasis to the regional lymph nodes could be cured with extended field irradiation if the disease was confined to the pelvis
and periaortic areas and small enough to be eradicated with moderate doses of radiation. This paper is a review of the University of Iowa experience with extended field irradiation for carcinoma of the uterine cervix metastatic to the periaortic lymph nodes. The principal objective was to determine whether extended field irradiation is effective in curing patients with known metastasis in the periaortic area. The specific aims were: (a) to evaluate the patterns of failure following extended field irradiation; (b) to correlate local control in the periaortic area with tumor burden, level of involvement, and radiation dose; and (c) to correlate complication rates with the surgical technique used and the radiation
Presented at the 3 1st Annual Meeting of the American Society for Therapeutic Radiology and Oncology, l-6 October 1989, San Francisco, CA. * Division of Radiation Oncology, Dept. of Radiology. + Division of Radiation Oncology, Dept. of Radiology.
* Division of Gynecologic Oncology, Dept. of OB/Gynecology. Reprint requests to: David H. Hussey, M. D., Division of Radiation Oncology, Room W 189Z-GH, University of Iowa Hospitals and Clinics, Iowa City, IA 52242-1059. Accepted for publication 6 January 1992.
INTRODUCTION
501
502
I. J. Radiation Oncology 0 Biology 0 Physics
dose delivered. The ultimate aim was to be able to select a subset of patients with a greater chance of benefiting from extended field irradiation. METHODS
AND MATERIALS
Patient population Between 1966 and 1985, 8 10 patients with carcinoma of the uterine cervix were treated with radiation therapy at the University of Iowa Hospitals and Clinics. Of these, 43 patients were treated with extended field irradiation for periaortic metastases. All but one of these patients had histologic evidence of periaortic disease. The other had clinical evidence of periaortic metastasis at laparotomy, but no biopsy was obtained. The patients ranged in age from 28 to 69 years, with a mean of 5 1 years. Thirty-eight patients had squamous carcinomas and five had adenocarcinomas. The patients were retrospectively staged on the basis of clinical findings using the 1988 FIG0 classification. Five patients presented with Stage IB, one with Stage IIA, 7 with Stage IIB, 24 with Stage IIIB, 3 with Stage IVA, and 1 with Stage IVB cancer. Two patients were treated for pelvic disease that was recurrent following a radical hysterectomy. The clinical evaluation for these patients included a history and physical examination, a chest X ray, and liver function tests. An intravenous pyelogram was performed in 42 patients, cystoscopy in 33 patients, sigmoidoscopy in 33 patients, and a barium enema in 28 patients. Six patients had pedal lymphangiograms, and four of these were positive for periaortic metastases. Abdominal CT scans were obtained in six patients, and three of these showed enlarged pelvic and periaortic lymph nodes. Twenty-three patients had negative prescalene node biopsies, and one had a positive prescalene node biopsy obtained during the course of irradiation. Peritoneal cytology was obtained in 28 patients with positive findings in 11. Forty-two patients underwent surgical staging. A transperitoneal approach was used in 29 patients and an extraperitoneal approach in 13. One patient did not have a staging lymphadenectomy, but had a positive CT-guided needle biopsy of the periaortic lymph nodes. Forty-two patients had histologic proof of periaortic nodal involvement. One patient with clinically positive periaortic nodes at laparotomy did not have a biopsy for histologic confirmation. The pelvic and periaortic nodes were palpated at laparotomy, and the number, size, and location of the positive nodes were assessed. In some cases, the suspicious nodes were resected and in others only a biopsy was performed. If the exploration was clinically negative, sampling biopsies were obtained. Only one patient had a formal lymphadenectomy. This patient had 25 nodes removed, 21 of which were involved by tumor.
Volume 23, Number 3, 1992
Radiation therapy The majority of the patients were treated with a combination of external beam irradiation and brachyradiotherapy. Twenty-five patients were treated with 6oCo gamma rays, 4 with 4 MV x-rays, and 14 with 10 MV x-rays. Anterior and posterior, parallel-opposing portals were used in 36 patients and a 4-field technique was employed in 7 patients. The treatment portals extended to the superior margins of T 12 in 12 patients, L 1 in 24 patients, and L2 in 7 patients. The dose to the periaortic nodes ranged from 3960 cGy to 6000 cGy (median = 5040 cGy). It was usually delivered at a rate of 180 cGy per day. The dose to the whole pelvis ranged from 3960 cGy in 20 fractions to 6510 cGy in 35 fractions (median = 4800 cGy in 26 fractions). The brachyradiotherapy treatment was usually delivered with radium, and the doses ranged from 1200 to 7800 mg-hr (median = 5625 mg-hr). The dose to point A, including both the external beam and intracavitary components, ranged from 5250 cGy to 13,000 cGy (median = 9510 cGy). RESULTS
The patients were treated between 1966 and 1985, and the data were analyzed in September 1989. The patients have been followed for periods ranging from 39 months to 18 years. The median follow-up time was 131 years. Survival was computed from the first day of treatment until death. Disease-free survival was computed from the first day of treatment to the first sign of failure or death. Failures were classified as occurring centrally, at the pelvic sidewall, in the periaortic area, intraperitoneally, in the supraclavicular area, and hematogenously. Patients were classified as having central failures if there was persistent ulceration or nodularity in the cervix or vagina, with or without biopsy confirmation. They were classified as having pelvic wall failures if there was progressive leg edema, hydronephrosis, or sciatic pain; or if there was a mass at the pelvic sidewall on physical examination, computerized tomography, or autopsy. Both central and pelvic wall failures were classified as pelvic failures. Patients were classified as having periaortic failures if there was clinical evidence of a persistent mass in the periaortic area or histological evidence of periaortic disease at autopsy. Patients were considered to have intraperitoneal metastases if there was ascites with a positive cytology or evidence of serosal studding at laparotomy or autopsy. Intraperitoneal, supraclavicular, and hematogenous metastases were all classified as distant metastasis. Patterns offailure The patterns of failure are listed in Table 1. Of the total group of 43 patients, 16 (37%) had progressive cancer at the primary site, 15 (35%) developed recurrent tumor at the pelvic sidewall, 11 (26%) had persistent disease in the periaortic area, 11 (26%) had intraperitoneal spread, 5
503
Extended field irradiation 0 A. P. VIGLIOTTI et al.
Table 1. Sites of failure for the total group of 43 patients with periaortic metastases*
Pelvic failures Central Pelvic wall Periaortic failures Distant metastases Intraperitoneal Supraclavicular fossa Hematogenous
Only site of failure
Total
No. (%)
No. (W)
8 (19) 5 (12) 3 (7) 1 (2) 7 (16) 0 (0) 2 (5) 5 (12)
20 (47) 16 (37) 15 (35) 11 (26) 23 (53) 11 (26) 5 (12) 16 (37)
ACTUARIAL
BY STAGE
30% 20% 10%
0%’ 0
* 12 patients (28%) had no evidence of disease (NED).
SURVIVAL
EXTENDEDFIELD CERVIX
I 12
24
33
43
30
MONTHS
( 12%) developed supraclavicular metastases, and 16 (37%)
developed hematogenous metastases. Most of these patients had progressive disease in more than one site. Twelve patients (28%) remained continuously free of disease (NED) to the time of analysis or death from intercurrent disease. The most common cause of failure in this series was distant metastasis (53%). However, a significant percentage of the patients (47%) developed progressive disease within the pelvis. Most of these also failed in the periaortic area or distantly. The relatively high local failure rate was not unexpected because the majority ofthese patients had massive pelvic disease initially. Actuarial survival for the total population is illustrated in Figure 1. Almost all of the deaths from cancer occurred within 3 years, and the actuarial 5-year survival rate was 32%. Survival rates are compared by clinical stage in Figure 2. Central failures. Sixteen patients (37%) had persistent or recurrent disease at the primary site. This correlated well with the clinical stage of the disease (Table 2). The
Fig. 2. Actuarial survival by clinical stage ( 13 patients I-II and 30 patients with stage III-IV or recurrent tumors).
central failure rate was 47% (14/30) for patients with Stage III-IV or recurrent tumors, compared to only 15% (2/l 3) for those with Stage I-II disease. The central failure rate also correlated well with the bulk of the cancer within the pelvis. Whereas 48% ( 12/25) of patients with tumors 2 6 cm in size developed central recurrences, only 22% (4/l 8) of those with tumors < 6 cm in size did (Table 2). As might be expected, the control rate at the primary site correlated well with the brachyradiotherapy dose, although it did not correlate well with the dose of external beam irradiation (Table 3). Whereas 67% (10/l 5) of patients who received < 4000 mg hr developed central recurrences, only 2 1% (6/28) of those who received > 4000 mg hr failed centrally. Similarly, the central failure rate
Table 2. Central failure rate as a function of clinical stage, tumor burden, and radiation therapy dose Central failure rate
OVERALL
ACTUARIAL
SURVIVAL
EXTENDED FIELD CERVIX 1OOII 90% '1
30% 70%
No. (%)
u*
Clinical stage Stage I-II Stage III, IV, or recurrent
2/13 (15) 14/30 (47)
.05 1
Central controlled tumor burden 9600 cGy (Table 2). Pelvic wall failures. Fifteen patients (35%) had progressive disease at the pelvic sidewall, 11 of whom also had persistent disease centrally. The failure rate at the pelvic sidewall correlated well with the extent of involvement at this site. Pelvic wall failures occurred in 43% (13/30) of patients who had tumor extension to the pelvic sidewall initially (Stage IIIB, IV, and recurrent tumors), compared with only 15% (2/ 13) of those who had no clinical evidence of pelvic wall involvement prior to treatment (Stage IB-IIB) (Fisher’s exact test; p = .075). The failure rate at the pelvic sidewall did not correlate well with either the brachytherapy or external beam dose of radiation. Periaortic failures. Eleven patients (26%) had persistent disease in the periaortic area. However, only one patient (2%) had an isolated periaortic recurrence (Table 1). In the other 10 cases, it was accompanied by progressive disease in the pelvis and/or distant metastasis. The failure rate in the periaortic area correlated well with the tumor burden in this region (Table 4). Of the total group of 43 patients, 13 had all detectable periaortic tumor resected at laparotomy and only microscopic tumor
remained. Six patients had a small gross residual disease (< 2 cm in diameter), 14 had moderate-sized residual disease (2-5 cm in diameter), and 10 had massive periaortic cancer. Massive disease was described as large matted nodes, sometimes fixed to adjacent organs or extending into the adjacent mesentery. None of 19 patients (0%) with microscopic or small gross disease had periaortic failures (Table 4). However, 29% (4/14) of those with moderate-sized disease and 70% (7/10) of those with mas-
Table 5. Correlation of 5-year survival with the highest level of periaortic involvement Superior extent of periaortic disease Ll
L2 L3 L4 Unknown Total
5 year survival O/l
I/lo* (10%) l/9 6/l 2 + 9/29* (31%) 3/ 12*I 112 1 l/41
* Fisher’s exact test; p = .189. + Excludes one patient alive NED at 39 months. * Excludes one patient alive with pelvic wall failure 4 1 months.
505
Extended field irradiation 0 A. P.VIGLIOTTI etal. Table 6. Incidence of periaortic failures and small bowel complications as a function of periaortic radiation dose Periaortic dose
Periaortic failure
Small bowel obstruction
4000-4500 cGy
31% (4/13) 20% (2/10) 21% (4/19) l/l 26% (1 l/43)
(1;;) 10% (l/10) 32% (6/19) O/l 19% (8/43)
- 5000 cGy 5500 cGy 6000 cGy Total
Table 7. Small bowel obstruction as a function of the dose of external beam irradiation and the brachyradiotherapy dose (mg-hr) mg-hrs PA dose (~GY)
Zero
~4000
40015000
50016000
>6001
Total (%)
-4500 -5000 -5500 -6000
O/3 O/l 1/3* _
O/2 O/l O/3 _
O/l O/l l/4 _
Of4 O/l 213 _
1/3* l/6 216 O/l
8 10 32 -
Total (%)
I 25
10.5
* Two patients developed partial small bowel obstructions, which were treated conservatively.
sive metastasis in the periaortic area had persistent disease at this site (Chi-square = 16.96; p = .0002). The periaortic tumor burden also correlated well with survival rates (Table 4). The 5-year survival rates were 50% (6/12) for patients with microscopic disease, 33% (2/6) for those with small gross disease, 23% (3/l 3) for those with moderate residual disease, and 0% (O/10) for those with massive periaortic disease (Chi-square = 7.17;
p = .067). Survival rates also correlated well with the level of nodal involvement, although the differences are not statistically significant (Table 5). Only 1 of 10 patients ( 10%) whose tumor extended to the Ll -2 level survived 5 years, compared with 9 of 29 patients (3 1%) whose disease extended no higher than the L3-4 level (Fisher’s exact test; p = .189). The periaortic failure rates correlated to some extent with the dose delivered to the periaortic nodes, although the differences are not statistically significant (Table 6). Doses of 4000-4500 cGy resulted in a 3 1% (4/13) periaortic failure rate, compared with 20% (2/10) with doses of - 5000 cGy and 2 1% (4/ 19) with doses of - 5500 cGy. Distant metastasis Twenty-three patients (53%) developed distant metastases. These were classified as intraperitoneal, supraclavicular, or hematogenous. The majority of these patients had involvement at more than one site. Most of the distant metastases developed within two years of treatment. The distant metastasis rate correlated well with the tumor burden in the periaortic area (Table 4). Eleven patients (26%) developed generalized intraperitoneal carcinomatosis. The incidence of intraperitoneal spread correlated well with the cytology findings at laparotomy. Whereas 55% (6/l 1) of patients with positive washings at laparotomy subsequently developed intraperitoneal carcinomatosis, none (O/ 17) of those with negative washings did (Fisher’s exact test; p = .OOl). Five patients ( 12%) developed supraclavicular metastasis, but only one became apparent prior to the development of hematogenous spread. Sixteen patients (37%) developed hematogenous metastasis to lung, bone, or liver. Complications Nine patients (2 1%) developed major complications, including eight with small bowel obstructions and one
Table 8. Clinical features of the long-term survivors Tumor burden Stage
Pelvis
PA
Superior extent
IB IIB IIB IIB IIIB IIIB IIIB IIIB IIIB IIA IB IIIB IIIB
Small Moderate Moderate Moderate Large Large Massive Moderate Massive Small Large Massive Massive
Micro. Micro. Micro. Micro. Micro. Micro. Micro. Small Small Moderate Moderate Moderate Moderate
L4 ? L4 L2 L3 L4 L3 L3 L3 L3 L3 L3 L4
Perk cytologY 0 0 + 0 0 +* 0 0 +
Pelvic dose 4500 cGy 4500 cGy 4780 cGy 4 140 cGy 4140 cGy 4980 cGy 3960 cGy 4000 cGy 3940 cGy 5040 cGy 4500 cGy 4020 cGy 5040 cGy
+ + + + + + + + + + + + +
5310 5960 6720 6240 6540 4670 1200 7760 7620 5700 2600 6360 6240
PA dose mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr mg-hr
4500 4500 4780 5340 5540 5480 3960 4750 5000 5540 4500 5020 5040
cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy cGy
Status NED @ 3$ yr NED@ 5f yr NED @ 61 yr (Vag. Net.) 4 Q 14 yr 2nd 1” (SBO)* NED @ 16 yr (SBO)* NED @ 9 yr (SBO)* NED @ 65 yr NED@ 19 yr 4 @ 7$ yr 2nd lo 4 @ 7 yr of DM (SBO)* Alive 3$ yr w/PW rec. 4 8 yr complic. (SBO)* NED@ 1Oyrs
Micro = Microscopic; NED = No evidence of disease; DM = Distant metastasis; SBO = small bowel obstruction; = Vaginal necrosis; mg-hr = Milligram hours. * Five of the long-term survivors developed small bowel obstructions.
Vag. Net.
506
I. J. Radiation Oncology 0 Biology 0 Physics
with a vaginal vault necrosis. The vaginal vault necrosis healed within 6 months, and this patient is alive with no evidence of disease 6 1 years later. Six of the eight patients with small bowel obstructions required surgical intervention, and two healed with conservative management. Seven of the eight small bowel obstructions developed within 24 months of treatment, and the other occurred 5 f years after irradiation. The incidence of small bowel obstructions correlated well with the surgical technique used. Only 1 of 13 patients (8%) who had extraperitoneal lymphadenectomies developed a small bowel obstruction, compared to 7 of 29 patients (24%) who had transperitoneal lymphadenectomies. Although an association was observed, the difference is not statistically significant (Fisher’s exact test; p = .208). The incidence of small bowel obstructions correlated well with the periaortic dose of radiation (Table 6 and 7). Only 8% (l/l 3) of the patients who received periaortic doses of 4000-4500 cGy and 10% ( l/ 10) of those who received doses of - 5000 cGy developed small bowel obstructions, compared with 32% (6/19) of those who received doses of - 5500 cGy. The only two who developed small bowel injury after external beam doses less than -5500 cGy received high brachytherapy doses (Table 7). These differences are marginally significant (- 5500 cGy vs lesser doses: p = .067; Fisher’s exact test). Long-term survivors The long-term survivors were reviewed separately to define a subset of patients who might have a greater chance of benefiting from extended field irradiation. This group included 11 patients who survived more than 5 years and 2 others who were alive at the time of analysis, less than 5 years after treatment. The clinical features of these 13 patients are summarized in Table 8. The extent of the disease within the pelvis ranged from small to massive in these long-term survivors. This shows that even large cancers of the cervix can be controlled if adequate doses of radiation are delivered. However, none of the long-term survivors had massive disease in the periaortic area, and most had limited disease in this region. Nevertheless, six of the long-term survivors were treated for gross periaortic disease, and four of them had residual disease measuring 2-5 cm in diameter. Thus, small to moderate-sized periaortic metastases can be controlled with moderate doses of radiation. Another important characteristic was the superior extent of the tumor. Almost all of the survivors had periaortic disease that extended no higher than the L3-4 level. Only one had tumor extending as high as L2. On the other hand, the presence of a positive peritoneal cytology was not as ominous as expected, since three of the longterm survivors had positive peritoneal washings. There was a relationship between the dose delivered to the periaortic area and the ultimate outcome (Table 8). Although the periaortic doses for the survivors ranged from 3960 cGy to 5540 cGy, the lower doses were effective
Volume 23, Number 3, 1992
No+ Transperitoneal Trans- and extra-peritoneal (sampling)
IB-IVA IIB-IV IB-IIIB IB-IV IB-IVB
13 29 21
of Iowa Hospitals
University
& Clinics
(15)
43
24
cGy/4-5
cGy/5-6
3940-65 10 cGy/4-7
4000-4320
4500-5000
4500 cGy/S weeks
cGy 4f-8
weeks
weeks
weeks
weeks
weeks
38% at 13-36 months 31% NED at 3 years 19% at 19-63 months 32% at 5 years
27% at 5 years
25% at 3 years 3/6 (minimum 4 years) 29% at 5 years 23% NED at 5 years 38% (minimum 2 years) 43% NED at 5 years 10% at 5 years
19%*
Survival
metastasis
19% SBO
5%
II%*
38%
45%
29%
19%
5% No SBO
116
43%*
Major complications
GOG = Gynecologic Oncology Group; SBO = Small bowel obstruction; NED = Alive with no evidence of disease. * In the M.D. Anderson series, extended fields were used if there was evidence of pelvic +/- periaortic metastases at lymphangiogram or laparotomy. The complication rate appears high because each complication was counted separately and some patients had more than one complication. + In patients not undergoing lymphadenectomy or laparotomy with biopsy, the diagnosis of periaortic nodal metastasis was made clinically, usually by lymphangiography. * In the Tufts University series, 4 patients developed major complications in 36 patients who were treated with extended fields either as part of their initial treatment (29 patients) or for periaortic recurrences (7 patients).
of Miami (22)
(5)
University
(7)
University
State University
Tufts University
Georgetown
Pennsylvania 15 100 cGy
4400-6000
Transperitoneal (variable extent) Trans- and extra-peritoneal (sampling) Transperitoneal
Instit. (14) 31
IB-IV
Roswell Park Memorial
( 19) cGy/4-6
4000-5000
Transpet-itoneal
IB-IIA
14
of Pennsylvania
University
5000 cGy/S weeks
No+
IB-IV
(12)
21
of Pennsylvania
University
weeks
4500-5 100 cGy/5-6 weeks 4320-5 130 cGy 41-6; weeks
IB-IVB IB-IV
(9)
22 18
cGy/4-7
4000-5000 cGy 6000 cGy/6 weeks
4000-6000
Transperitoneal in 83 patients* Usually transperitoneal Transperitoneal (biopsy only) Transperitoneal (sampling) Extraperitoneal
dose
proven periaortic
IB-WA
University of Tennessee UCLA (1)
257
Periaortic
or histologically
I-IV I-III
(6)
diagnosed
Periaortic lymphadenectomy
for clinically
Clinical stage
field irradiation
98 6
Hospital
No. of patients
the use of extended
GOG Study (2) Downstate Medical Center (20)
M.D. Anderson
Institution
Table 10. Studies evaluating
508
I. J. Radiation Oncology 0 Biology 0 Physics
only for patients with microscopic disease. Most of the complications occurred in patients who received high periaortic doses. All of the small bowel obstructions in the survivors occurred in patients who received periaortic doses of 5020-5540 cGy. Thus, the group that fared best following extended field irradiation had relatively limited disease in the periaortic area with the tumor extending no higher than the L3-4 level. These patients received the appropriate combination of external beam and intracavitary irradiation for their pelvic disease and doses of -4500-5000 cGy in 5-54 weeks to the periaortic area. DISCUSSION Studies to evaluate the effectiveness of periaortic irradiation in carcinoma of the uterine cervix may be classified into two broad categories: (a) those in which the periaortic lymph nodes are irradiated prophylactically, that is, in which there is no clinical evidence of periaortic metastasis, and (b) those in which the periaortic lymph nodes are irradiated for known metastatic disease. This series belongs in the second category, because periaortic metastases were documented prior to treatment. This study does not address the issue of elective irradiation of the periaortic lymph nodes in patients with no evidence of periaortic metastasis. In general, studies in which the periaortic lymph nodes have been treated prophylactically report significantly greater 5-year survival rates than those in which the periaortic nodes have been treated for known metastatic disease (Table 9 and 10). In most series, the 5-year survival rates have been in the range of 50-70% if the periaortic lymph nodes were either unevaluated or contained no evidence of metastatic disease (8, 10, 18, 19, 23). On the other hand, survival rates have been in the range of 2035% if there was clinical or histopathologic evidence of periaortic metastasis at the time of irradiation (1, 2, 5-7, 9, 12-15, 19, 20, 22). The overall results of our study are similar to those reported elsewhere, with an actuarial 5year survival rate of 32% (Fig. 1). Distant metastasis is a frequent occurrence in patients with known periaortic metastasis from carcinoma of the uterine cervix. This was the most frequent cause of treatment failure (53%) in this series. Nevertheless, half of the patients with periaortic involvement in this series did not develop distant metastasis, and the incidence of distant spread was even less if the periaortic disease was limited in extent. Therefore, the presence of periaortic metastasis is not necessarily diagnostic of distant spread. The results of this study also show that the doses that can be given with extended field irradiation can control
* The complication rates in El Senoussi’s series are high because each complication was counted separately and some patients had multiple complications.
Volume 23, Number 3, 1992
the disease in the periaortic area, particularly if the tumor burden in this region is not great. Only 26% (1 l/43) of the patients in this series had periaortic failures, and all of the patients with periaortic disease less than 2 cm in size had tumor control in this region (Table 4). Even periaortic masses measuring 2-5 cm in size were controlled in a high percentage of cases. These data suggest that a dose of -5000 cGy in 5-51 weeks is effective in controlling small to moderate sized periaortic metastases in most patients (Table 4). There is a significant risk of small bowel injury with extended fields, however, particularly if doses greater than 5000 cGy in 5-51 weeks are used (Table 6). Almost a third (6/ 19) of the patients who received periaortic doses of - 5500 cGy in this series developed small bowel complications. However, only 9% (2/23) of those who received periaortic doses of 5000 cGy or less developed small bowel complications, and this is an acceptable incidence of complications. El Senoussi et al. (6) also found a doseresponse relationship for complications following extended field irradiation. They reported complication rates of 24% following 4000-4500 cGy, 44% after 5000-5500 cGy, and 57% after 5500-6000 cGy.* They concluded that 4500 cGy in 5 weeks was the maximum dose that could be delivered safely through extended fields (Table 7). Another factor that affected the incidence of complications in this series was the surgical technique. Only 8% ( l/ 13) of patients who had extraperitoneal lymphadenectomies developed small bowel obstructions, compared to 24% (7/29) of those who had transperitoneal lymphadenectomies. These results are similar to those reported by Berman et al. (3), who found a 30% incidence of small bowel injury following transperitoneal lymphadenectomies, but only a 2.5% (l/39) incidence of bowel complications in patients who had retroperitoneal lymphadenectomies. The results of this study indicate that there is a subset of patients with documented periaortic metastasis from carcinoma of the uterine cervix who would benefit from extended field irradiation. Twenty-eight percent ( 12/43) of the patients in this series had permanent control of their disease, and the actuarial 5-year survival rate was 32%. Although this subset is small, it is large enough to warrant continued use of extended field irradiation for patients with known metastasis in the periaortic area. The patients who appear most likely to benefit from extended field irradiation are those in whom the residual disease in the periaortic area is less than 2 cm in size, whose disease extends no higher than L3, and where the primary tumor within the pelvis has a reasonable chance of control with standard radiation therapy techniques.
Extended field irradiation 0 A. P.VIGLIOTTI et al.
Extended field irradiation is probably also indicated for some patients with more extensive disease, that is, those with periaortic tumors measuring 2-5 cm in size and/or extending above L3. A reasonably high local control rate can be expected in these patients, but the survival rate
509
will be low because most of these patients already have distant metastases. On the other hand, extended field irradiation is of little value for patients with massive periaortic disease, because the periaortic control rate is poor and the incidence of distant metastasis is almost universal.
REFERENCES 1. Ballon, S. C.; Berman, M. L.; LaGasse, L. D.; Petrilli, E. S.; Castaldo, T. W. Survival after extraperitoneal pelvic and paraaortic lymphadenectomy and radiation therapy in cervical carcinoma. Gynecol. Obstet. 57( 1): 90-95; 198 1. 2. Berman, M. L.; Keys, H.; Creasman, W.; DiSaia, P.; Bundy, B.; Blessing, J. Survival and patterns of recurrence in cervical cancer metastatic to periaortic lymph nodes (A Gynecological Oncology Group Study). Gynecol. Oncol. 19: 8-l 6; 1984. 3. Berman, M. L.; LaGasse, L. D.; Waning, W. G.; Ballon, S. C.; Schlesinger, R. E.; Moore, J. G.; Donaldson, R. C. The operative evaluation of patients with cervical carcinoma by an extraperitoneal approach. Obstet. Gynecol. 50(6): 658-664; 1977. 4. Chism, S. E.; Park, R. C.; Keys, H. M. Prospects for paraaortic irradiation in treatment of cancer of the cervix. Cancer 35: 1505-1509; 1975. 5. Delgado, G.; Caglar, H.; Walker, P. Survival and complications in cervical cancer treated by pelvic and extended field radiation after paraaortic lymphadenectomy. Am. J. Roentgenol. 130: 141-143; 1978. 6. El Senoussi, M. A.; Fletcher, G. H.; Borlase, B. C. Correlation of radiation and surgical parameters in complications in the extended field technique for carcinoma of the cervix. Int. J. Radiat. Oncol. Biol. Phys. 5: 927-934; 1979. 7. Emami, B.; Waning, W. G.; Tak, W.; Anderson, B.; Piro, A. J. Para-aortic lymph node radiation in advanced cervical cancer. Int. J. Radiat. Oncol. Biol. Phys. 6: 1237-1241; 1980. 8. Haie, C.; Pejovic, M. H.; Gerbaulet, A.; Horiot, J. C.; Pourquier, H.; Delouche, J.; Heinz, J. F.; Brune, D.; Fenton, J.; Pizzi, G.; Bey, P.; Brossel, R.; Pillement, P.; Volterrani, F.; Chassagne, D. Is prophylactic para-aortic irradiation worthwhile in the treatment of advanced cervical carcinoma? Results of a controlled clinical trial of the EORTC radiotherapy group. Radiother. Oncol. 11: 101-l 12; 1988. 9. Hughes, R. R.; Brewington, K. C.; Hanjani, P.; Photopulos, G.; Dick, D.; Votava, C.; Moran, M.; Coleman, S. Extended field irradiation for cervical cancer based on surgical staging. Gynecol. Oncol. 9: 153- 16 1; 1980. 10. Inoue, T.; Morita, K. 5-year results of positive extendedfield irradiation on 76 patients with nodal metastases from cervical carcinoma stages IB-IIIB. Cancer 61: 2009-2014; 1988. 11. Jampolis, S.; Andras, E. J.; Fletcher, G. H. Analysis of sites and causes of failures of irradiation in invasive squamous
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
cell carcinoma of the intact uterine cervix. Radiology 115: 681-685; 1975. Lepanto, P.; Littman, P.; Mikuta, J.; Davis, L.; Celebre, J. Treatment of para-aortic nodes in carcinoma of the cervix. Cancer 35: 1510-1513; 1975. Perez, C. A.; Bedwinek, J. M.; Breaux, S. R. Patterns of failure after treatment of gynecologic tumors. Cancer Treat. Symp. 2: 217-231; 1983. Piver, M. S.; Barlow, J. J.; Krishnamsetty, R. Five year survival (with no evidence of disease) in patients with biopsyconfirmed aortic node metastasis from cervical carcinoma. Am. J. Obstet. Gynecol. 139: 575-578; 198 1. Podczaski, E. S.; Palombo, C.; Manetta, A.; Andrews, C.: Larson, J.; DeGeest, K.; Mortel, R. Assessment of pretreatment Iaparotomy in patients with cervical carcinoma prior to radiotherapy. Gynecol. Oncol. 33: 71-75; 1989. Potish, R. A.; Twiggs, L. B.; Prem, K. A.; Levitt, S. H.; Adcock, L. L. The impact of extraperitoneal surgical staging on morbidity and tumor recurrence following radiotherapy for cervical carcinoma. Am. J. Clin. Oncol. (CCT) 7: 245251; 1984. Rotman, M.; Choi, K.; Guze, C.; Martial, V.; Homback, N.: John, M. Prophylactic irradiation of the para-aortic lymph node chain in stage IIB and bulky stage IB carcinoma of the cervix, initial treatment results of RTOG 7920. Int. J. Radiat. Oncol. Biol. Phys. 19: 5 13-521; 1990. Rotman, M.; Moon, S.; John, M.; Choi, K.; Sail, S. Extended field para-aortic radiation in cervical carcinoma: the case for prophylactic treatment. Int. J. Radiat. Oncol. Biol. Phys. 4: 795-799; 1978. Rubin, S. C.; Brookland, R.; Mikuta, J. J.; Mangan, C.; Sutton, G.; Danoff, B. Para-aortic nodal metastases in early cervical carcinoma: long-term survival following extendedfield radiotherapy. Gynecol. Oncol. 18: 2 13-2 17; 1984. Silberstein, A. B.; Aron, B. S.; Alexander, L. L. Para-aortic lymph node irradiation in cervical carcinoma. Radiology 95: 181-184; 1970. Sotto, L. S. J.; Graham, J. B.; Pickren, J. W. Postmortem findings in cancer of the cervix. Am. J. Obstet. Gynecol. 80: 791-794; 1960. Sudarsanam, A.; Charyulu, K.; Belinson, J.; Averette, H.; Goldberg, M.; Hintz, B.; Thirumala, M.; Ford, J. Influence of exploratory celiotomy on the management of carcinoma of the cervix. Cancer 41: 1049-1053; 1978.
