Whole abdominal radiotherapy in the adjuvant treatment of patients with stage III and IV endometrial cancer: A gynecologic oncology group study

Gynecologic Oncology 97 (2005) 755 – 763 www.elsevier.com/locate/ygyno

Whole abdominal radiotherapy in the adjuvant treatment of patients with stage III and IV endometrial cancer: A Gynecologic Oncology Group study Gregory Suttona,*, Janice H. Axelrodb,1, Brian N. Bundyc, Tapan Royd, Howard D. Homesleye, John H. Malfetanof,2, Borys R. Mychalczakg, Mary E. Kingh a

Division of Gynecologic Oncology, St. Vincent’s Hospital and Health Services, 2001 W. 86th Street, Indianapolis, IN 46260, USA b Gynecologic Oncology, Western Pennsylvania Hospital, Pittsburgh, PA 15224, USA c Statistics, Gynecologic Oncology Group, Roswell Park Cancer Institute, Buffalo, NY 14263, USA d Department of Radiation Oncology, St. Louis University Health Science Center, St. Louis, MO 63110, USA e Wake Forest School of Medicine, Brookview Research, Inc., Winston-Salem, NC 27103, USA f Gynecologic Oncology, Albany Medical College, Albany, NY 12208, USA g Department of Radiation Oncology, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA h Clinical Pathology, Columbia University, New York, NY 10027, USA Received 21 July 2004

Abstract Objective. To evaluate toxicity, survival, and recurrence-free interval in women with loco-regionally advanced endometrial carcinoma treated with postoperative whole abdominal radiation therapy. Methods. Whole abdominal irradiation with pelvic plus or minus para-aortic boost was initiated within 8 weeks of total abdominal hysterectomy, bilateral salpingo-oophorectomy, pelvic washings, and selective pelvic and para-aortic node sampling in eligible, consenting patients. Results. Of 180 evaluable patients entered on the study with surgically staged III and IV endometrial carcinoma maximally debulked to less than 2 cm, 77 had typical endometrial adenocarcinoma and 103 had high-risk histology, either papillary serous or clear cell carcinoma. Patients with typical endometrial adenocarcinoma were significantly younger and had significantly fewer poorly differentiated cancers. Proportionally, there were twice as many non-Whites with high-risk histologies as non-Whites with typical endometrial adenocarcinoma. Forty-five percent of patients with typical endometrial adenocarcinomas had positive pelvic nodes compared to 51% of those with high-risk histologies. Both histologic groups had similar distribution for performance status, para-aortic node positivity, site and extent of disease, and International Federation of Gynecology and Obstetrics (FIGO) stage. The frequency of severe or lifethreatening adverse effects among 174 patients evaluable for radiation toxicity included 12.6% with bone marrow depression, 15% GI, and 2.2% hepatic toxicity. The recurrence-free survival rates were 29% and 27% (at 3 years) for the typical endometrial adenocarcinoma and high-risk histologies, respectively. The survival rates were 31% and 35%, respectively. No patient with gross residual disease survived. Conclusion. Whole abdominal irradiation in maximally resected advanced endometrial carcinoma has tolerable toxicity, and it is suggested that the outcome may be improved by this adjunctive treatment in patients with completely resected disease. D 2005 Published by Elsevier Inc. Keywords: Whole abdominal radiotherapy; WAR

* Corresponding author. Fax: +1 317 415 6749. E-mail address: [email protected] (G. Sutton). 1 Current address: New Jersey Gynecologic Oncology, Little Silver, NJ 07739, USA. 2 Current address: Associates in Gynecological Care, P.C., Albany, NY 12208, USA. 0090-8258/$ - see front matter D 2005 Published by Elsevier Inc. doi:10.1016/j.ygyno.2005.03.011

756

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

Introduction Endometrial cancer is the most common malignant neoplasm arising in the female reproductive tract. Although it has the lowest death-to-case ratio of all gynecologic malignancies, 6600 women die of advanced, recurrent, or metastatic endometrial cancer every year in the United States [1]. Extirpative surgery followed in selected cases by pelvic radiation therapy is capable of controlling stage I and II disease in the majority of cases. Adjuvant therapy in stage III and localized stage IV disease has not been well defined, however, and appropriate treatment for patients with aggressive papillary serous and clear cell cancers awaits delineation. Based upon early favorable experience with ovarian cancer at the M.D. Anderson Hospital, Greer and Hamburger [2] suggested that whole abdominal radiotherapy utilizing a moving-strip technique might be beneficial in treating patients with endometrial cancer whose tumors had spread to the abdominal cavity. They reported corrected and absolute 5-year survivals of 80% and 63%, respectively, among 27 women with intraperitoneally-disseminated endometrial cancer, none with residual disease greater than 2 cm in diameter, who received whole abdominal moving-strip radiotherapy with a pelvic boost. Of patients who developed recurrent disease, three had within-field failures alone and one had simultaneous abdominal and distant relapses. Toxicity was limited to early severe enteritis and two cases of ‘‘late’’ partial bowel obstruction and a vaginal ulcer. Subsequent reports [3] of substantial small bowel toxicity among patients with ovarian cancer treated with the moving-strip technique led to the abandonment of this procedure in favor of the whole abdominal ‘‘open-field’’ procedures in use at the present time. Hendrickson et al. [4] first demonstrated that papillary serous cancers of the endometrium were associated with an extraordinary risk of relapse characterized by abdominal failures, which were fatal in the vast majority of patients. They were the first authors to suggest the use of adjuvant abdominal radiotherapy in this disease entity. The Gynecologic Oncology Group (GOG) initiated the present study to determine feasibility, survival, and progression-free interval among patients with advanced endometrial cancer of all histologic types treated with whole abdominal radiotherapy with pelvic and, in the case of paraaortic metastases, para-aortic boosts. A second major objective was to determine the progression-free interval and sites of recurrence among patients with stage I and II papillary serous and clear cell carcinomas of the endometrium treated similarly. Results for this latter group of patients will be discussed in a forthcoming publication.

Methods This report summarizes results for patients with stage III and IV endometrial cancer of all histologic subtypes. In the

present study, subjects were required to have pathologicallyconfirmed primary endometrial cancer with clinical and/or surgical stage III and IV disease without vaginal involvement, parenchymal liver metastases, lung metastases, or spread to extraperitoneal sites excluding retroperitoneal lymph nodes. Papillary serous or clear cell histologies were required to involve greater than 50% of tumor volume. Patients with para-aortic lymph node metastases were eligible only if scalene lymph node biopsy was negative. Patients were ineligible if they had received pelvic or abdominal radiation or chemotherapy, or were found to have inadequate hematologic (WBC  3000, platelets and granulocytes  1500 cells/cm), renal (creatinine > 2.0 mg%), or hepatic function (bilirubin or aspartate transaminase (AST) > 2 normal). Also ineligible were patients with GOG performance status of 4 and those with a previous or concomitant malignancy except nonmelanoma skin cancer. Written informed consent was obtained from all patients prior to study entry in accord with institutional, state, and federal regulations. Pathology review Peritoneal washings were to be obtained from the pelvis and cytologically evaluated for malignant cells. The uterus was to be evaluated for size, location of tumor, depth of myometrial invasion, histologic type, and grade of tumor. Lymph nodes and adnexa were also evaluated for the presence and location of metastases. For each patient entered, the GOG Pathology Committee reviewed appropriate slides documenting histological parameters. For patients with papillary serous and clear cell histologies, all slides were to be submitted for review. Surgery Enrollees were required to have total abdominal hysterectomy, bilateral salpingo-oophorectomy, pelvic washings, and selective para-aortic and pelvic lymph node sampling. Omentectomy was not required; however, careful inspection of the omentum was required as well as removal of sections of the omentum with gross metastases. Tumor resection to residual nodules of 2 cm or less was required. Previous therapy with hormonal agents was permitted and patients with recurrent endometrial cancer were allowed if all other protocol requirements were met. Study entry was required within 8 weeks of surgery. Radiation therapy Irradiation was to be initiated within 8 weeks of surgery; all treatments were delivered by megavoltage equipment ranging from that of cobalt-60 to maximum 25-MeV photons. Minimal source-skin distance was 80 cm and dose rates between 30 and 200 cGy/min at midplane were required. Patients were to be treated with two pairs of

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763 Table 1 Cell type distribution (N = 180) Cell type

Number

Clear cell Papillary serous Typical endometrial Adenocarcinoma Endometrioid Glassy cell Mixed epithelial Adenosquamous Villoglandular Undifferentiated

23 80 77 10 34 1 10 14 7 1

757

into study to date of reappearance or increasing parameters of disease or date of last contact. Life tables and medians were computed using the method of Kaplan and Meier [5]. Differences in recurrence-free interval or survival by patient characteristics were evaluated using the log-rank test [6]. The Pearson chi-square test was used [7] to identify correlations between the two major categories of cell type and patient and disease characteristics. The Wilcoxon rank sum test [8] was used for age at diagnosis and cell type categories.

Results parallel opposed fields (open-field technique) to the whole abdomen and pelvis. The whole abdomen was to be treated first, to a dose of 3000 cGy in 20 fractions of 150 cGy each. A decrease in the daily fraction to 125 cGy per day was allowed if gastrointestinal symptoms or leukopenia precluded use of the higher dose. After whole abdominal radiation, the pelvis was boosted to a midplane dose of 1980 cGy at 180 cGy per fraction for eleven treatments. The combined whole abdominal radiation and the total pelvic radiation required 6– 7 weeks. Patients with positive para-aortic nodes were to receive an additional boost of 1500 cGy for a total para-aortic dose of 4500 cGy. The whole abdominal field extended from 1 cm above the top of the diaphragm to the bottom of the obturator foramina. The lateral border extended 1.0 – 1.5 cm beyond the lateral peritoneal margin. Full thickness posterior kidney blocks were used throughout therapy. Blocking of the left heart above the diaphragm was to be used and portions of the lower lateral pelvic fields and femoral heads were also blocked. The pelvic field extended from the L5 – S1 interspace superiorly to the bottom of the obturator foramina inferiorly. The lateral margins were 1.5 cm lateral to the medial rim of the ilium. The para-aortic boost field was bounded by the L5 –S1 interspace inferiorly, the superior margin of the abdominal field superiorly, and the lateral extent was 8 cm wide. Interruptions in treatment exceeding 2 weeks in duration disqualified patients from protocol therapy. Radiation therapy and quality control were supervised by the Radiologic Physics Center under the sponsorship of the American Association of Physicists in Medicine. Accuracies of T3% in source output and T5% in prescribed dose delivery were required. Statistical considerations Evaluation parameters included recurrence-free survival (RFI), survival time, and frequency and severity of adverse effects. Survival was defined as observed length of life from entry into study to death, or to date of last contact. Recurrence-free interval was defined as the date from entry

A total of 274 patients were entered in this study between December 1986 and February 1994. Of these, 58 were ineligible. Inadequate surgery excluded 20 patients, 22 had wrong cell type, three had disease more advanced than protocol criteria permitted, four failed criteria for advanced stage, eight had a second primary malignancy, and one had a non-endometrial primary. Of the 214 evaluable patients, 34 had stage I or II papillary serous or clear cell cancer and are the subject of a separate report. Of the 180 patients analyzed in this study, 77 had stage III or IV typical endometrial cancer and 103 patients had stage III or IV papillary serous or clear cell carcinoma (Table 1). Characteristics of this patient population are detailed in Table 2. Patients with either papillary serous or clear cell carcinoma were significantly older (P < 0.01) than those with typical endometrioid cancers. The median age of the latter was 63 years (range: 32 –81 years) compared with median age of 68.5 years (range: 39 –85 years) for patients with papillary Table 2 Patient characteristics Characteristic

Age <50 51 – 60 61 – 70 71 – 80 81+ GOG performance status 0 1 2 3 Race White Black Other Grade 1 2 3 Unknown

Typical endometrial (N = 77)

Papillary serous/clear cell (N = 103)

No.

(%)

No.

(%)

14 15 26 20 2

(18.2) (19.5) (33.8) (26.0) (2.6)

6 17 39 36 5

(5.8) (16.5) (37.9) (35.0) (4.9)

32 41 4 0

(41.6) (53.2) (5.2) (0.0)

38 58 7 0

(36.9) (56.3) (6.8) (0.0)

68 6 3

(88.3) (7.8) (3.9)

77 21 5

(74.8) (20.4) (3.9)

20 23 34 0

(26.0) (29.9) (44.2) (0.0)

10 24 68 1

(9.7) (23.3) (66.0) (1.0)

758

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

Table 3 Extent of disease Sitea

Vagina Fallopian tube Ovarian Ligament Omentum Small bowel Colon Gutter Diaphragmb Cul-de-sac Abdominal wall Bladder Otherc

Table 5 Postoperative residual disease Typical endometrial (N = 77)

Papillary serous/clear cell (N = 103)

Residual

Typical endometrial

Papillary serous/clear cell

No.

(%)

No.

(%)

No.

No.

Microscopic Gross Total

65 12 77

(84.4) (15.6) (100.0)

91 12 103

(88.3) (11.7) (100.0)

Pos.

(%)

Pos.

(%)

4 20 28 7 15 8 12 3 1 12 2 7 12

(5.2) (26.0) (36.4) (9.1) (19.5) (10.4) (15.6) (3.9) (1.3) (15.6) (2.6) (9.1) (15.6)

3 19 36 6 22 5 11 6 8 12 2 10 13

(2.9) (18.4) (35.0) (5.8) (21.4) (4.9) (10.7) (5.8) (7.8) (11.7) (1.9) (9.7) (12.6)

Pos. = positive findings. a Patients may have two or more sites involved. b Statistically significant difference (P = 0.05). c Includes appendix, peritoneum, spleen, umbilical, liver, para-ovarian tissue, renal artery, psoas muscle, epiploic fat, mesentery, and abdominal lymph nodes.

serous cancers and median age of 71 years (range: 45 – 85 years) for patients with clear cell cancers. All groups had similar mean GOG performance status. A larger proportion of Black patients had papillary serous or clear cell cancers (21/ 27 = 78%) than did Whites (77/145 = 53%). Additionally, two-thirds of those with papillary serous or clear cell cancers had tumors of grade 3 but less than half (44%) of those with typical endometrioid tumors were grade 3. In Table 3, the extent of disease by site is enumerated. Similar frequencies of spread between the two groups were observed except for diaphragm (1.3% vs. 7.8%, P = 0.05); also, similar frequencies of lymph node spread were observed (Table 4). The frequency of gross residual disease was similar in both groups of patients (Table 5). Toxicity In Table 6, adverse events for the 174 patients evaluable for toxicity of whole abdominal irradiation are summarized. Six patients are inevaluable for adverse events because three

Pelvic nodes Negative Microscopic Gross Para-aortic nodes Negative Microscopic Gross

Table 6 Adverse events Adverse effect

Table 4 Lymph node status Node status

refused radiotherapy, one accepted only pelvic radiotherapy, and two additional patients were not treated because of declining performance status and bipolar affective disorder with schizophrenic features, respectively. Twenty-two patients had grade 3 or 4 hematologic toxicity. One patient with WBC 1900/mm3 and platelet count 29,000/mm3 complicated by an infected lymphocyst did not receive pelvic boost therapy. Grade 4 toxicity included a WBC of 900/mm3 (patient removed from study), a platelet count of 14,000/mm3, and a third patient with thrombocytopenia of 8000/mm3 1 week before therapy ended. Nausea and diarrhea were the most common acute gastrointestinal toxicities. In one patient, treatment was delayed by an episode of diverticulitis, and a second patient had a 2-day delay in radiotherapy and a subsequent recurrence while under treatment. A third patient developed a bowel obstruction secondary to progressive disease in the upper abdomen. An additional patient was taken off study after developing severe nausea and vomiting after 2 days of abdominal radiotherapy; she subsequently received pelvic radiotherapy and vaginal ovoids. Grade 4 gastrointestinal toxicity was observed in seven patients. Two had bowel obstructions associated with progressive disease at 12 and 16 months. Three patients who were NED developed obstructions requiring surgery at 10, 10, and 13 months; the first died of anastamotic breakdown and sepsis postoperatively, the second patient developed volvulus, subdiaphragmatic abscess, and sepsis at 2 months, and the third died of gastrointestinal hemorrhage, pulmonary embolus, and congestive heart failure; both were clinically NED. One patient died of a gastrointestinal bleed 5 months after therapy and a second died of hemorrhage after anticoagulation for deep venous thrombosis 60 days

Typical endometrial

Papillary serous/Clear cell

No.

(%)

No.

(%)

42 18 17

(54.5) (23.4) (22.1)

50 30 23

(48.5) (29.1) (22.3)

47 16 14

(61.0) (20.8) (18.2)

71 20 12

(68.9) (19.4) (11.7)

Hematologic Genitourinary Gastrointestinal Hepatic Pulmonary CV Neurologic Cutaneous Lymphatic Fever Other

Grade (Frequency) 0

1

63 (36%) 138 (79%) 27 (15%) 165 (95%) 156 (90%) 159 (91%) 167 (96%) 118 (68%) 164 (94%) 140 (85%) 159 (91%)

35 32 42 5 11 1 3 44 6 26 11

2 (20%) (18%) (24%) (3%) (6%) (1%) (2%) (25%) (3%) (15%) (6%)

54 4 78 0 6 2 2 10 3 6 2

3 (31%) (2%) (45%) (3%) (1%) (1%) (6%) (2%) (3%) (1%)

19 0 20 3 1 10 2 2 1 2 2

4 (11%) (11%) (2%) (1%) (6%) (1%) (1%) (1%) (1%) (1%)

3 0 7 1 0 2 0 0 0 0 0

(2%) (4%) (1%) (1%)

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

postoperatively. A last patient developed nausea, diarrhea, abdominal pain, malnutrition, and sepsis 9 months after radiation therapy while being treated for disease recurrence. She required peripheral hyperalimentation. One patient had grade 4 and three had grade 3 hepatic toxicity. The first patient developed cytologically-negative ascites immediately after radiotherapy, and biopsy showed liver necrosis. One patient developed cytologically-benign ascites 26 days after radiotherapy and a second had a liver biopsy demonstrating centrilobular veno-occlusive disease felt to be consistent with radiotherapy and not viral hepatitis. These latter two patients recovered without sequelae. An additional patient had elevated liver function studies as well as a pulmonary embolus and recurrent disease. Cardiovascular adverse events included four pulmonary emboli, two patients with congestive heart failure related to atrial fibrillation, and one with postoperative atrial flutter. One episode of severe hypertension occurred in a patient who refused antihypertensive therapy. Two episodes of hypotension occurred, one related to a postoperative epidural anesthetic. Patterns of failure and survival Table 7 lists sites of failure for patients with typical endometrial and papillary serous/clear cell cancers. Table 8 summarizes survival by residual disease. Stage III and IV papillary serous tumors Four subgroups were identified in patients with stage III and IV papillary serous tumors, separated by surgical Table 7 Site of recurrence Sites

NED Recurred Vagina Pelvis Abdomen Retroperitoneal nodes Lung Othera Lung and other Abdomen and other Pelvis and liver Vagina and liver Lung and pelvis Abdomen and pelvis Abdomen and lung Vagina and lung Retro. nodes and pelvis Unknown

Typical endometrial

Papillary serous/ Clear cell

No. (%)

No. (%)

27 (35.1) 50 (64.9) 3 7 9 0 9 10 2 2 1 0 0 1 1 0 0 5

34 (33.0) 69 (67.0) 4 3 21 1 15 6 3 2 0 1 1 4 4 1 1 2

NED, no evidence of disease. a Includes axillary, groin, and supraclavicular lymph nodes, bone, brain, spinal cord, liver, stomach, and bladder.

759

Table 8 Survival by residual disease Histology

Gross

Gross/Resected microscopic

Papillary Clear cell Endometrioid

0 (2.5 – 33.5) 25 (2.6 – 42.1) 0 (2.4 – 29)

27.5 (1.3 – 57) 12.5 (2.6 – 17.8) 47.3 (1.6 – 84)

26 (1.4 – 141.9) 16.5 (8.8 – 99) 18.4 (0.3 – 57.4)

findings and residual disease. None of the eight patients who had gross residual disease after surgery survived. Their median disease-free survival was 4.8 months (range: 2.5– 33.5 months); six had abdominal failures while one each had recurrences in lung and vagina and in an unknown site. Median survival among these patients was 11.1 months (range: 5.8 – 55.5 months). The longest survivor had pulmonary metastases. Twenty-nine patients had gross disease at the time of surgery which was felt to be completely resected. Twenty had malignant washings or ascites and nine had para-aortic metastases with negative supraclavicular biopsies. Diseasefree survival ranged from 1.3 to 57 months (median: 12.0 months), and 21 (72.4%) patients died of disease from 1.6 to 72 months after therapy. Median survival for the group was 27.4 months. Failure in the abdomen (Tpelvis) occurred in 10 patients (34.5%) and seven (24.1%) had recurrences which included the lung. Forty-three patients qualified for the study because of microscopic metastases. The most common site of spread was pelvic lymph nodes (23 patients or 53.5%) either alone or in combination with other spread. Thirteen patients had para-aortic metastases and negative supraclavicular metastases and two had para-aortic metastases without pelvic lymph node spread. Thirty-two of the 43 patients died: six of other causes, 23 of disease, 2 of unknown causes, and 1 due to treatment. Median survival and disease-free survival were 65 and 66 months (both ranged 1.4 –141.9 and 95 months), respectively. Among 26 patients with known recurrences, 9 (34.6%) included an abdominal or pelvic component and six were confined to the chest. Stage III and IV clear cell cancers Twenty-three patients had stage III (19) and IV (5) clear cell cancers of the endometrium. Four patients had from 0.1 to 2.0 cm of residual disease after surgery and all but one died of disease (one intercurrent death) between 2.6 and 42.1 months after treatment. Sites of failure included vagina, pelvis, abdomen, and axillary lymph nodes. Four of the five patients eligible for protocol because of malignant peritoneal cytology alone were alive and free of disease at 60.4 and 83 months follow-up, respectively; one patient with malignant cytology died at 6.4 months of abdominal failure. The remaining patients had either completely-resected gross disease (8) or microscopic (6) spread at the time of surgery. All but one patient with gross residual disease

760

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

Table 9 Recurrence-free interval and survival at 3 years Cell type/Stage Typical endometrial, stage Typical endometrial, stage Papillary serous/clear cell, Papillary serous/clear cell,

III IV stage III stage IV

No.

RFS (%)

Survival (%)

58 19 75 28

34.5 10.5 40.1 10.7

34.5 21.1 48.1 10.7

relapsed by 6 months and died (range of survival time 2.6– 17.8 months). Three of those with gross but resected metastases died at 1.0 –12.8 months after therapy, one died of unknown causes, and the remaining three are alive without disease at 63– 153 months follow-up. Four patients with microscopic metastases died of disease 8.8– 99 months after therapy, one died of intercurrent disease, and one was alive without cancer 64 months after therapy. Of the 10 patients failing therapy in this group, four had isolated pulmonary metastases, four had abdominal or pelvic failures, one had pelvic and retroperitoneal node metastasis, and one had axillary node metastasis. Two patients had metastases confined to the adnexa; one was alive at 63 months and the other dead of other causes. Among four patients with isolated nodal spread, all resected, two were free of disease and two died of other causes. Overall recurrence-free survival for patients with papillary serous or clear cell cancers was 32% at 3 years; patients with gross residual disease obviously did more poorly than those with microscopic residual (4.2% vs. 35.1% at 3 years, P = 0.0002). Table 9 demonstrates 3-year recurrence-free survival and survival for the two cell types and surgical stages. Stage III and IV typical endometrial (Adenocarcinomas) Seventy-seven patients with one of seven histologies (Table 1) were analyzed together. Of these, 12 patients had

gross residual disease of 0.1– 2.0 cm after surgery; none survived. Time to recurrence ranged from 0.9 to 23 months and time to death 2.4 –29 months. Five patients (41.7%) had abdominal, pelvic, or vaginal recurrences. Thirty-eight patients had gross extrauterine spread with all disease resected. Six had isolated adnexal spread and five had isolated spread to omentum, small bowel, abdominal wall, or pelvic nodes. Overall, 20 patients died of disease 1.6– 84 months after therapy; two patients died of toxicity. Recurrences were 1 – 48 months after treatment. Sites of recurrence included 7 in the abdomen/pelvis and seven with pulmonary metastases. Fourteen patients survived 59.3– 130.7 months free of disease after treatment, although one was successfully treated for an isolated groin recurrence at 19.7 months. Twenty-seven patients were found to have microscopic extrauterine disease at the time of surgery. Seven had isolated adnexal metastases; all but one also had malignant peritoneal cytology (three survived and three died of disease). Seven patients had isolated metastasis to retroperitoneal lymph nodes (2), pelvic nodes (1), serosal (1), colon (1), ligament (1), or omentum (1) had isolated metastasis to retroperitoneal lymph nodes; two had malignant cytology as well. Of these patients, two survived disease-free, one died of other causes, and four died of disease. For the 27 patients as a whole, only five were alive and free of disease at 59.3 – 88.3 months; 17 died of disease 0.3– 57.4 months after treatment. Eight had abdominal/ pelvic or vaginal recurrences and four had pulmonary metastases. Four patients died of other causes and one of toxicity. Overall, recurrence-free intervals for patients with typical endometrial adenocarcinomas and those with the ‘‘highrisk’’ cell types were similar (Fig. 1). The RFS rate was 12.6%, 22.0%, and 11.0% at 3 years for typical endometrial carcinoma, clear cell, and papillary serous cancers, respec-

Fig. 1. Recurrence-free survival by cell type.

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

761

Fig. 2. Recurrence-free survival by surgical stage.

tively. Surgical stage was the most discriminating prognostic factor identified in the study population. The risk of recurrence was significantly less (P < 0.0001) for the surgical stage III patients than those with surgical stage IV, regardless of histology (Fig. 2). Those patients with gross residual disease had a significantly higher rate of relapse than those with no visible residual or those that were completely resected (Fig. 3). Table 8 displays the 3-year rate of recurrence-free interval and survival for the two cell type categories and surgical stages. Distribution of recurrence is summarized in Table 7. The frequency of abdominal/pelvic/vaginal recurrence (‘‘infield’’ failures) in any combination was 40.0% in typical endometrial cancers and 47.1% in papillary serous and clear cell histologies. The frequencies of isolated vaginal recurrence for the two groups were 3/77 (3.9%) and 4/103 (3.9%), respectively. None of the patients with vaginal recurrences had primary vaginal involvement.

Discussion The purpose of the present study was to evaluate toxicity, survival, and progression-free survival after whole abdominal radiotherapy in women with resected stage III and IV endometrial cancer. A second objective was to determine differences in survival and progression-free survival between patients with typical endometrial adenocarcinoma and those with either papillary serous or clear cell endometrial cancer. Patient population As noted above, patients with papillary serous or clear cell cancers were significantly older than their counterparts with typical endometrial adenocarcinomas of the same advanced stages. Additionally, as has been observed by others such as Plaxe and Saltzstein [9], a larger proportion

Fig. 3. Recurrence-free survival by residual disease status.

762

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

of Black women had papillary serous cancers than did White women. The proportion of those with lymph node metastases was similar between the papillary serous or clear cell and typical endometrial adenocarcinoma groups. Toxicity Given the advanced stage and, in many cases, age of the cases treated in this series, toxicity was acceptable. Hematologic toxicity was moderate, usually self-limited, and did not interrupt or prevent therapy except in one case. It should be appreciated that the treatments reported herein occurred before the widespread availability of hematopoietic growth factors for supportive care. Acute gastrointestinal toxicity in the form of diarrhea, nausea, and vomiting, although common, rarely limited therapy. It is quite clear that patients with underlying inflammatory bowel disease such as diverticulitis tolerate abdominal radiotherapy poorly. Similar gastrointestinal toxicity has been reported by others such as Schray et al. [10] and Frank et al. [11]. Two cases of fatal gastrointestinal hemorrhage may or may not have been directly attributable to radiotherapy. The chronic serious complication rate in this study (12/165 or 7.3%) was similar to that observed by other authors. Dembo [3] reported an overall 2.2% incidence of surgical or nonsurgical bowel obstruction among 447 patients treated with moving-strip or open-field radiotherapy for ovarian cancer. In the treatment of intraperitoneally-disseminated ovarian or endometrial cancer, liver shielding produces a safe haven under the right hemidiaphragm. Omitting liver shielding clearly increases the risk of hepatic injury; in this study, four patients (2.2%) developed radiation hepatitis and one died. Fatal radiation hepatitis is rare; Ingold [12] defined radiation hepatitis as significant alteration of liver enzymes, hepatomegaly, biopsy-proven histologic changes, or ascites. He determined a safe dose to be between 3000 and 3500 cGy. Wharton et al. [13] and Schacter et al. [14] reported a total of three cases of fatal radiation hepatitis in patients receiving less than 2920 cGy using the moving-strip technique. Extensive experience at the Princess Margaret Hospital in Toronto [3] demonstrated a 0.9% incidence of radiation hepatitis among 226 patients treated with moving-strip radiotherapy and no cases of radiation hepatitis among 221 patients treated with open-field technique without liver shielding. Survival and patterns of failure It is quite apparent from this study that patients with gross residual disease after surgery for stage III or IV endometrial cancer, whether typical endometrial or papillary serous/clear cell, fare poorly despite whole abdominal radiotherapy. Death was a near universal outcome among 24 such patients regardless of histologic subtype and 15 of them suffered a recurrence within the treatment field. Radiation therapy alone cannot be advocated as a curative

measure in such patients, although those with gross residual limited to the pelvis might potentially benefit from chemosensitization plus whole abdomen radiotherapy. This approach has been piloted in a limited GOG trial [15]. On the other hand, Corn et al. [16] showed that the presence of para-aortic metastases in patients with advanced endometrioid cancers was associated with a 39% distant relapse rate, the lungs being the most common site of failure. Potish et al. [17] employed whole abdominal radiotherapy in typical endometrioid cancer without gross disease spread beyond the adnexa, but found that relapse-free survival was extremely poor in those with gross extrapelvic metastases. Other series reported by Martinez et al. [18] and Gibbons et al. [19] demonstrated that the outcome after whole abdominal therapy was better in patients with pathologic stage III endometrial cancer without gross extrapelvic spread. Among small numbers of patients with gross disease completely resected, those with clear cell histology seemed to have the best outcome, 5/9 surviving at last follow-up compared with 8/31 for papillary serous and 16/39 for typical endometrial adenocarcinomas. When groups with microscopic extrauterine spread were considered, 11/39 of patients with papillary serous, 2/6 with clear cell, and 4/25 of those with typical endometrial histologies survived. No particular advantage was observed in any subgroup except patients with typical endometrial histology and isolated adnexal metastases, where survival was 3/7. Several authors have evaluated the use of whole abdominal radiotherapy in patients with papillary serous cancers of the endometrium. Christman et al. [20] reported that three of six patients with stage I – III tumors survived after abdominal therapy (but those three were apparently stage IA). Frank et al. [11] reported a 44% crude survival in their patients, 56% of whom had abdominal or pelvic failures. Employing whole abdominal radiotherapy, Grice et al. [21] successfully treated four of six patients with stage IIIc disease, but failed in eight of nine with stage IV disease. Smith et al. [22] also interpreted their experience with whole abdominal radiotherapy in high-risk patients as favorable, but, like other reports, lacked a concurrent control arm. Martinez et al. [23] treated 132 patients with high-risk endometrial cancer with whole abdomen and pelvic radiotherapy with boosts to nodal areas or the vagina over a 20year period and reported 5- and 10-year disease-free survivals of 55% and 45%, respectively, for stage III disease. First site of failure was abdomen or pelvis in 59% of relapses and chronic grade 3 or 4 toxicity occurred in 14% of patients. Even in this single-investigator series, there is clear room for improvement. Whole abdominal radiation therapy as delivered in this study appears to offer curative therapy in a minority of patients with stage III and IV endometrial cancer, regardless of histologic type. New techniques such as conformal or intensity-modulated radiotherapy need to be utilized in future studies in this group of patients. There is also a need

G. Sutton et al. / Gynecologic Oncology 97 (2005) 755 – 763

for other adjuvant therapies in these patients if a better clinical outcome is to be realized. New treatment regimens must have the potential to reduce both abdominal and extraabdominal failures. Randomized trials involving concomitant radiochemotherapy or sequential irradiation and chemotherapy are clearly needed in this disease.

[3] [4] [5] [6]

Acknowledgments [7]

This study was supported by National Cancer Institute grants of the Gynecologic Oncology Group Administrative Office (CA 27469) and the Gynecologic Oncology Group Statistical and Data Center (CA 37517). The following Gynecologic Oncology Group institutions participated in this study; University of Alabama at Birmingham, Oregon Health Sciences University, Duke University Medical Center, Abington Memorial Hospital, University of Rochester Medical Center, Walter Reed Army Medical Center, Wayne State University School of Medicine, University of Southern California Medical Center at Los Angeles, University of Mississippi Medical Center, Colorado Foundation for Medical Care, University of California Medical Center at Los Angeles, University of Miami School of Medicine, The Milton S. Hershey School of Medicine of the Pennsylvania State University, Georgetown University Hospital, University of Cincinnati College of Medicine, University of North Carolina School of Medicine, University of Iowa Hospitals and Clinics, University of Texas Southwestern Medical Center at Dallas, Indiana University Medical Center, Wake Forest University School of Medicine, The Albany Medical College of Union University, University of California, Irvine Medical Center, Tufts New England Medical Center, Rush-Presbyterian-St. Lukes Medical Center, Stanford University Medical Center, State University of New York Downstate Medical Center, Eastern Virginia Medical School, Cleveland Clinic Foundation, The Johns Hopkins Oncology Center, State University of New York at Stony Brook, Pennsylvania Hospital, Washington University School of Medicine, Memorial Sloan-Kettering Cancer Center, Cooper Hospital University Medical Center, Columbus Cancer Council, University of Massachusetts Medical Center, Women’s Cancer Center, and University of Oklahoma Health Sciences Center.

References [1] Jemal A, Murray T, Ward E, Samuels A, Tivari RC, Ghafoor A, Feuer EF, Thun MJ. Cancer statistics, 2005. CA Cancer J Clin 2005;55:10. [2] Greer BE, Hamburger AD. Treatment of intraperitoneal metastatic adenocarcinoma of the endometrium by whole-abdominal moving

[8] [9] [10]

[11]

[12] [13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

763

strip technique and pelvic boost irradiation. Gynecol Oncol 1983;16: 365 – 73. Dembo AJ. Abdominal radiotherapy in ovarian cancer. Cancer 1985; 55:2285 – 90. Hendrickson M, Ross J, Eifel P, Kempson R, Martinez A. Uterine papillary serous carcinoma. Am J Surg Pathol 1982;6:93 – 108. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;3:457 – 81. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50: 163 – 70. Snedcor GW, Cockran WG. Statistical methods. 6th ed. Ames, Iowa’ The Iowa State University Press; 1967. Hollander M, Wolf DA. Nonparametric statistical methods. New York, NY’ John Wiley & Sons; 1999. Plaxe SC, Saltzstein SL. Impact of ethnicity on the incidence of highrisk endometrial carcinoma. Gynecol Oncol 1997;65:8 – 12. Schray MF, Martinez A, Howes AE. Toxicity of open field whole abdominal irradiation as primary postoperative treatment in gynecologic malignancy. Int J Radiat Oncol, Biol, Phys 1989;16: 397 – 403. Frank AH, Tseng PC, Haffty BG, Papadopoulos DP, Kacinski BM, Danling SW, et al. Adjuvant whole abdominal radiation therapy in uterine papillary serous carcinoma. Cancer 1991;68:1516 – 9. Ingold JA, Reed GB, Kaplan HS, Bagshaw MA. Radiation hepatitis. Am J Roentgenol 1965;93:200 – 8. Wharton JT, Delclos L, Gallager S, Smith JP. Radiation hepatitis induced by abdominal irradiation with cobalt 60 moving strip technique. Am J Roentgenol 1973;117:73 – 80. Schacter L, Crum E, Spitzer T, Maksem J, Diwan V, Kolli S. Fatal radiation hepatitis: a case report and review of the literature. Gynecol Oncol 1986;24:373 – 80. Reisinger RA. Phase I study of cisplatin and whole abdominal radiotherapy for the management of stage III and IV endometrial cancer. Gynecol Oncol 1996;63:299 – 303. Corn BW, Lanciano RM, Greven KM, Schultz DK, Reisinger SA, Stafford PM, et al. Endometrial cancer with para-aortic adenopathy: patterns of failure and opportunities for cure. Int J Radiat Oncol, Biol, Phys 1992;24:223 – 7. Potish RA, Twiggs LB, Adcock LL, Prem KA. Role of whole abdominal radiation therapy in the management of endometrial cancer: prognostic importance of factors indicating peritoneal metastases. Gynecol Oncol 1985;21:80 – 6. Martinez A, Schray M, Podratz K, Stanhope R, Malkasian G. Postoperative whole abdomino-pelvic irradiation for patients with high risk endometrial cancer. Int J Radiat Biol Phys 1989;17:371 – 7. Gibbons S, Martinez A, Schray M. Adjuvant whole abdominal pelvic irradiation for high risk endometrial carcinoma. Int J Radiat Oncol, Biol, Phys 1991;21:1019 – 25. Christman JE, Kapp DS, Hendrickson MR, Howes AE, Ballon SC. Therapeutic approaches to uterine papillary serous carcinoma: a preliminary report. Gynecol Oncol 1987;26:228 – 35. Grice J, Ek M, Greer B, Koh WJ, Muntz JG, Cain J, et al. Uterine papillary serous carcinoma: evaluation of long-term survival in surgically-staged patients. Gynecol Oncol 1998;69:69 – 73. Smith RS, Kapp DS, Chen Q, Teng NN. Treatment of high-risk uterine cancer with whole abdominal radiation therapy. Int J Radiat Oncol, Biol, Phys 2000;48:767 – 8. Martinez AA, Weiner S, Podratz K, Armin A-R, Stromberg JS, Stanhope R, et al. Improved outcome at 10 years for serouspapillary/clear cell or high-risk endometrial cancer patients treated with adjuvant high-dose abdomino-pelvic irradiation. Gynecol Oncol 2003;90:537 – 46.