Endometrial Cancer – Rare endometrial cancers-clear cell, UPSC

Endometrial Cancer: Rare endometrial cancers—Uterine Papillary Serous Carcinoma (UPSC, USC) and Clear Cell Carcinoma (CC) of the endometrium

1. Demographics

Endometrial cancer is the most common gynecologic malignancy in the United States with approximately 60,000 new cases and 10,470 deaths per year. The majority of women with endometrial cancer (approximately 80%) will present with early-stage disease and a favorable histologic subtype, and they will enjoy excellent outcomes. However, 10% to 15% of women will present with uterine papillary serous carcinoma (UPSC) or clear cell carcinoma (CC).

While these particular histologies represent a rather small number of cases, they are responsible for a disproportionate number of the deaths (approximately 40%) related to endometrial cancer. One study showed that CC carcinoma accounts for about 3% of all endometrial cancers, but accounts for 8% of all endometrial cancer related deaths. For UPSC, these numbers are 10% and 39%, respectively. This is in part to due to the more advanced stage of disease at the time of diagnosis, the aggressive histology, and lack of effective therapy.

Uterine papillary serous carcinoma

The most common symptom of women diagnosed with UPSC, as is for women with endometrial cancer in general, is postmenopausal vaginal bleeding. This should prompt an endometrial biopsy. Often times, the diagnosis of UPSC is not made at the time of endometrial biopsy, as it may be admixed with another histologic subtype, such as a grade 3 endometrioid or clear cell carcinoma.

From a demographic perspective, women with UPSC tend to be older than women with the endometrioid histology. There also appears to be an increased risk in African American women as compared to other racial groups. Women with a history of breast cancer may also be at an increased risk for this disease and in some studies tamoxifen use has been associated with an increased risk of UPSC. However, this has not been a uniform finding in the literature. Emerging data suggests that there may be an association between BRCA1+ status and UPSC. This change in knowledge will have significant implications regarding risk reduction surgery for women who are BRCA1+.

A precursor lesion for UPSC has been identified. Endometrial intraepithelial carcinoma (EIC) has been seen in the majority of uteri with UPSC. However, in contrast with other precursor lesions, this lesion itself may be associated with metastatic disease at the time of presentation. So while it is a precursor lesion, it may not present clinicians with the opportunity to intervene with therapy prior to advanced disease.

Clear cell carcinoma

While data and studies are limited with UPSC, they are even more limited in the setting of CC carcinoma of the uterus. It often times is “associated” with UPSC, so they are often treated in a similar fashion.

The most common symptom of women diagnosed with CC, is postmenopausal vaginal bleeding. Less commonly, it may also be diagnosed subsequent to an abnormal pap smear. Endometrial biopsy or other forms of endometrial sampling typically is accurate in making the diagnosis.

As with UPSC, there is an association with a personal prior history of breast cancer. It may also occur with increased frequency in women who have undergone pelvic radiation for another condition and subsequently develop an endometrial cancer.

Diagnosis and work-up

Uterine papillary serous carcinoma/Clear cell carcinoma

The diagnosis of UPSC is typically made after a woman presents with postmenopausal vaginal bleeding. Endometrial sampling by any method (office endometrial biopsy, D&C, etc.) will typically make the diagnosis of an endometrial cancer, though UPSC may frequently be admixed with other histologies. Studies have shown that even when UPSC contributes to a small fraction of the total tumor (<10%), it may drive the presentation and prognosis for that patient. Endometrial biopsy, including office biopsy with a Pipelle, has a sensitivity of over 99%.

Ultrasound may not be as reliable for UPSC as it is for the endometrioid histologies. For type I endometrial cancers, an endometrial stripe thickness of 5 mm or less is reassuring with a sensitivity of 96% and a specificity of 61% for endometrial cancer. However, for women with type II endometrial cancers, 42% of the women who were subsequently diagnosed with an atypical histologic subtype had an endometrial stripe thickness of 5 mm or less. So caution must be exercised; in any women with postmenopausal bleeding, endometrial sampling should be obtained.

The use of preoperative CT scans has been evaluated in a study of women with endometrial cancer. In this study, preoperative imaging altered a patient’s management in only 4% of cases. However, for women with high-risk histologies, including UPSC, the CT results altered the planned management in 11% of cases. Therefore, there may be some utility in obtaining preoperative imaging, particularly in those women who have signs and symptoms of metastatic disease. At our institution, we typically do proceed with preoperative imaging in women with UPSC and CC.

CA-125 is a tumor marker that is routinely obtained in women with ovarian masses that are worrisome for carcinoma. There is some data suggesting that CA-125 may be useful in endometrial cancer as well. CA-125 may also be a useful marker in following women with UPSC for recurrence after surgery and/or adjuvant therapy. One study of women with UPSC did show that a preoperative CA-125 correlated with the disease stage at the time of surgery and was predictive of death.

Histology

Uterine papillary serous carcinoma

UPSC was first described in the 1980s. These tumors are characterized by a complex papillary architecture, broad fibrous stalks lined with a stratified epithelial lining, tumor necrosis, increased mitotic figures, and increased nuclear-to-cytoplasmic ration. Thirty percent of cases may also have psammoma bodies, which are often seen in papillary serous ovarian cancer.

Clear cell carcinoma

CC may exhibit a wider variety of histologic characteristics. The papillary form of CC is the most common. Other types include tubulocystic or solid. Common features include intraluminal mucin and intracytoplasmic vacuoles. These tumors are typically estrogen receptor and progesterone receptor negative (as is UPSC). p53 staining is less common for CC than UPSC and can be used in differing the two histologies.

3. Management

A. What therapies should you initiate immediately (i.e., emergently)?

There are no emergent management strategies for UPSC and CC carcinoma, but surgery should be the initial approach in those patients who can undergo operative interventions and whose disease appears to be resectable. Optimal resection of women with UPSC has been associated with improved outcomes. These women may be candidates for minimally-invasive surgery, with its associated improvements in complication rates.

B. What should the initial definitive therapy for the cancer be?

While UPSC and CC are clearly different histologic subtypes, they have not been exhaustively studied in a prospective fashion. This is due to their relative rarity. In addition, the majority of retrospective published work has dealt with UPSC, as it is the more common of the two atypical histologies. Often time, a woman’s tumor will also be mixed, containing variable percentages of grade 3 endometrioid, UPSC, and/or CC. For this section, the writer will combine the two histologies. While the author admits that there are differences between the two tumors, they will be presented together and only those areas that may clearly be divergent will be highlighted. We anxiously await data from two prospective Gynecologic Oncology Group trials (GOG-249; GOG -258), which will hopefully help guide future treatment for women with early (GOG-249) and advanced (GOG-258) UPSC and CC carcinomas.

Surgery

Women with both UPSC and CC carcinoma may present with extrauterine disease, even in the absence of traditional risk factors for metastatic spread, including deep myometrial invasion, lymphovascular space invasion, and cervical involvement. For women with UPSC, 20% to 25% of women will have metastatic disease even with stage IA endometrial disease. However, this is based on the older FIGO staging. Therefore, this means that even in women who have no evidence of myometrial disease, they can have metastatic disease 20% to 25% of the time if comprehensive surgical staging is undertaken.

In the one study specifically evaluating this issue for CC cancer, they found that 52% of the women who had disease “confined” to the uterus had metastatic disease at the time of surgical staging. Therefore, most would agree that these women should be comprehensively surgically staged. This would involve total hysterectomy, bilateral salpingo-oophorectomy, and pelvic and para-aortic lymph node dissection. Due to the propensity to spread through the abdominal cavity like advanced ovarian cancer, many gynecologic oncologists would also sample the omentum or perform an omentectomy. One group has shown that in the absence of gross visible omental disease, performing an omentectomy rarely altered clinical management decisions and could be left out of the staging procedures.

With the advent of minimally invasive surgery in endometrial cancer, the issue of adequate staging for these tumors has been raised. In a prospective study conducted by the Gynecologic Oncology Group (GOG), which evaluated laparoscopy versus laparotomy for the staging of women with endometrial cancer, women with UPSC and CC were included. While they represent a smaller percentage of the patients in this study due to smaller incidence or enrollment bias, there did not appear to be an obvious negative impact to these patients undergoing minimally invasive surgery.

Retrospective trials evaluating minimally invasive surgery for UPSC and CC have similarly not shown any obvious adverse effects to offering these women traditional laparoscopy or robotic assisted laparoscopy. The majority of the robotic data is from retrospective or prospective single institution studies; however, it would not seem reasonable to expect that robotic assistance would have any detrimental effects over standard laparoscopy. In fact, lymph node yields, hospital stay, estimated blood loss, and complications appear to be more favorable with the robotic platform as compared to standard laparoscopy.

In cases of gross metastatic disease at the time of surgery, there does appear to be a role for “debulking” surgery. This follows the paradigm of the surgical approach for women with advanced ovarian cancer. The majority of the retrospective studies that have been published for both UPSC and CC, do show a survival benefit for women who are maximally cytoreduced as compared to those women who have gross disease remaining at the conclusion of their surgical procedure. One large series reported that women who had no visible disease at the conclusion of their surgery had a median survival time of 51 months as compared to 14 months for those women who were “optimal” but had persistent disease and 12 months for those women who were “suboptimal.”

Early-stage (stage I/II) adjuvant therapy

There is little prospective data currently available to help guide therapy for women with this disease. Data is emerging as the Gynecologic Oncology Group (GOG 249) conducted a trial evaluating the issue of adjuvant therapy in women with high-risk early-stage disease, which is clearly the category that women with UPSC and CC fall into.

For women with Stage IA UPSC (no myometrial invasion) who do not undergo postoperative therapy, the risk of recurrence without postoperative therapy can range up to 30% depending on the study, with a cumulative risk of about 14%. With radiation this risk is 28% (range 0% to 50%); with a combination of chemotherapy and radiation, the risk is 6% (range 0% to 17%). One recent study showed no difference in 5-year overall survival for women with stage IA: 80.2% for adjuvant therapy and 79.8% for no adjuvant therapy. However, this included women with myometrial invasion to <50%. In one study with 22 women with no myometrial invasion, there was only a 5% risk of recurrence in the untreated population.

The GOG did complete a prospective trial of 21 women with early-stage UPSC who underwent whole abdominal radiation. Of the 19 women who were evaluable, 8 women died of a disease recurrence with over half of the recurrent sites occurring in the radiation fields. The authors concluded that chemotherapy should be considered in this patient population.

Based on the retrospective data that exists and acknowledging that this may not constitute more than expert opinion, the following treatment algorithm is being proposed. This algorithm assumes that the patient has been comprehensively surgically staged (Table 1).

Table 1.

There are many questions that remain and we await the data from GOG 249 to guide the management of these patients. Some of the questions that remain include which chemotherapeutic agents should we use and how many cycles should the patient receive. The default for UPSC and CC has seemed to be paclitaxel and carboplatin.

The retrospective data (and some small phase II prospective studies) have used on the average six cycles. This is extrapolated from the comfort of using six cycles of paclitaxel and carboplatin for women with ovarian cancer. How should we use radiation therapy? PORTEC and GOG 99 both could lead one to believe that the primary benefit of pelvic radiation is to reduce vaginal cuff recurrences. Therefore, we have substituted vaginal cuff brachytherapy for pelvic radiation in women with early stage disease. This would clearly decrease the time of therapy and the morbidity of therapy.

Advanced stage (III/IV) adjuvant therapy

Women with any advanced endometrial cancer, but particularly those with the atypical histologies, should be considered for eligibility in clinical trials. If there are no clinical trials available for these patients, as with the women with early-stage disease, many of the recommendations specific for UPSC and CC will be based on retrospective data. Fortunately, for women with advanced-stage endometrial cancer, there is a little more prospective cooperative group data that we can draw from.

• GOG 94 concluded that women with Stage III/IV UPSC, who were treated with whole abdominal radiation, were at risk for multisite and extrapelvic recurrences.

• GOG 122 concluded that for women with advanced stage endometrial cancer, chemotherapy (adriamycin and cisplatin) was superior to whole abdominal radiation therapy. In this trial, 83 women had UPSC and they had an increased risk of recurrence (HR1.39) and death (HR 1.56) as compared to the other endometrial histologies. In this trial, 17 women had CC and this particular histology did not have an increased risk of progression or death in this study when compared to the other histologies.

• GOG 177 compared cisplatin and adriamycin (AC) to paclitaxel, adriamycin, and cisplatin (TAP). This trial showed improved outcomes with regards to both progression free and overall survival for the paclitaxel containing regimen with a 40% reduction in the risk of recurrence. Thus, the three-arm regimen became the GOG standard regimen.

• GOG 184 compared the three-drug (adriamycin, cisplatin, paclitaxel) to the two-drug regimen (adriamycin, cisplatin) after radiation therapy for women with advanced endometrial cancer. In this study, there was no difference in the progression free or overall survival. Three-year disease-free survival rate was approximately 60%.

• GOG 209 evaluated paclitaxel and carboplatin (TC) versus the GOG standard three-drug (TAP) regimen. The results of this trial are pending. However, the interim analysis suggests that in terms of progression free and overall survival, that the paclitaxel and carboplatin (TC) combination is not inferior to the three drug (TAP) regimen.

• GOG 258 results pending. This protocol evaluates cisplatin and volume-directed radiation followed by four cycles of paclitaxel and carboplatin versus carboplatin and paclitaxel in women with optimally debulked stage III or IV endometrial cancer. We would hope that there are a significant number of women with UPSC and CC enrolled in this trial so that definitive conclusions can be drawn. This trial may address the role of sequencing chemotherapy,

Optimally debulked

There was a recent study using the National Cancer Database that evaluated the role of multimodal therapy in women with node positive (stage IIIC) UPSC. This study of 1816 women with UPSC showed that medial overall survival was 33.6 months in women who received chemotherapy alone as compared to 42.6 months (p<0.0005) in those who received chemotherapy + radiation. In this study, multivariable analysis confirmed that consolidative radiation was independently predictive of improved overall survival. GOG 258 will hopefully inform us regarding the sequence of therapy.

Based on the retrospective data that exists and acknowledging that this may not constitute more than expert opinion, the following treatment algorithm is being proposed (Table 2).

Table 2.

Suboptimally debulked

I would recommend that women with advanced stage UPSC and CC that are not optimally debulked be enrolled in clinical trials. There is currently no GOG trial evaluating this patient population.

Based on the prospective and retrospective data that exists and acknowledging that this may not constitute more than expert opinion, the following treatment algorithm is being proposed.

Suboptimally debulked Chemotherapy with paclitaxel and carboplatin

4. Complications

A. What complications could arise as a consequence of condition?

There are surgical and treatment related toxicities that can arise as part of treatment for both UPSC and CC.

B. What complications could arise as a consequence of management of chemotherapy, radiation, and/or surgery?

Surgery

Complications of surgery for UPSC and CC carcinoma depend in part on whether the surgery is performed at the time of initial diagnosis or for recurrent disease. Staging procedures for endometrial cancer are associated with both short-term and longer term sequelae. The GOG LAP-2 trial does provide us with some of the best prospective data regarding operative complications. While one cannot extrapolate from the laparoscopic results to robotic assisted laparoscopy, they should not be expected to be significantly different. Table 3 lists a select representation of the most usual complications: