The risk of malignancy index (RMI) scoring system is used to predict whether a pelvic mass is malignant (Box 13.1). Women with an RMI of >200 should be referred to a specialist centre for further management and surgery. An RMI of >200 has a positive predictive value of 87% and a sensitivity of 88% for diagnosing malignant disease.

Box 13.1
Risk of malignancy index (RMI)

Feature	RMI score
Ultrasound features: multilocular cyst solid areas bilateral lesions ascites intra-abdominal metastases	0 = none 1 = one abnormality 3 = two or more abnormalities
Premenopausal Postmenopausal	1 3
CA-125	U/ml

RMI score = ultrasound score × menopausal score × CA-125 level in U/ml.

To assess operability patients should have a CT of the abdomen and pelvis (Figure 13.1), and a CXR. In patients with a pleural effusion, cytological diagnosis is required to determine if the effusion is malignant.

Figure 13.1 A&B, CT scan in ovarian cancer. Contrast enhanced CT in a patient with advanced ovarian cancer demonstrates bilateral solid (black arrows) and cystic (white arrows) adnexal masses and omental cake (white arrow heads).

Courtesy of Dr. E Sala, University of Cambridge.

Patients with ascites without a mass on CT, should have cytological and immunohistochemical analysis, in addition to CA-125, CEA and CA19-9 (p. 287). The immunohistochemical profile of an ovarian tumour is typically CK 7 positive and CK 20 negative. A serum CA125:CEA ratio >25 is strongly suggestive of an ovarian rather than a GI primary.

Box 13.2
FIGO staging of ovarian cancer

Stage IV – distant metastases to the liver or outside the peritoneal cavity

Management (Figure 13.2)

Primary surgery

Surgery involves total abdominal hysterectomy (TAH), bilateral salpingo-oophorectomy (BSO), infra-colic omentectomy, peritoneal washings, biopsy of any lesions or adhesions, blind peritoneal biopsies of the bladder, cul de sac, paracolic gutter, hemidiaphragm and pelvic sidewall and pelvic and para-aortic node sampling. Surgical staging is important to guide treatment with chemotherapy and estimate prognosis.

Figure 13.2 Management of ovarian cancer.

Early stage disease (stage I and II)

Surgery

Initial management in early stage disease is maximal cytoreductive surgery and surgical staging. Around one-third of patients with apparent stage I–IIA disease will have a higher stage of disease after full staging, mostly stage III (77%). The frequency of lymph node involvement for apparent stage I disease is 20%. There is no evidence that total lymphadenectomy improves survival. Patients with mucinous tumours should have an appendicectomy, as this may be the site of the primary.

Fertility preserving surgery

In younger patients with stage IA tumours and favourable histology, unilateral salpingo-oophorectomy and staging may be carried out, although data on fertility preserving surgery is limited. Endometrial biopsy should be carried out as a synchronous primary is present in 10% of patients. Wedge biopsy of the contralateral ovary should only be taken if it appears abnormal, as the probability of involvement of a macroscopically normal ovary is low (2.5%).

Chemotherapy

For patients with stage IA/B G1 non-clear cell cancer (low risk) 5-year disease free survival is >90% without chemotherapy if optimal surgery has been performed. For patients with high risk features (G3, clear cell, stage IIA disease), 5-year recurrence rates are 25–40% and adjuvant chemotherapy is recommended. In these patients, chemotherapy improves 5-year disease free survival (DFS) by 11% (from 65% to 76%), and 5-year overall survival (OS) by 8% (75% to 82%). Chemotherapy with 6 cycles of carboplatin and paclitaxel is recommended, although 6 cycles of carboplatin is also acceptable (Box 13.3). For stage I G2 tumours the role of chemotherapy is not clear but 6 cycles of carboplatin may be offered.

Advanced stage disease (stage III and IV)

Patients with advanced disease may develop medical complications as a result of their cancer, notably thrombosis which has consequences for further management (Box 13.4).

Box 13.4
Thrombosis in ovarian cancer

Patients with ovarian cancer have a relatively high risk of venous thromboembolism. Approximately 5% of ovarian cancer will have either a DVT or PE within in the first 2 years after diagnosis, with 30% of these occurring within the first 3 months. Rates of thrombosis are highest in those with metastatic disease, medical co-morbidities and clear cell histology. Management is with low molecular weight heparin. Vena-caval filters may be considered prior to surgery to prevent pulmonary emboli if the clot extends proximally, is non-resolving, or if the patient develops further emboli despite optimal anticoagulation.

Surgery

The extent of cytoreductive surgery is the most important prognostic factor after stage. The aim of surgery is to achieve a complete macroscopic debulking, or failing this, an ‘optimal debulking’. The definition of ‘optimal debulking’ is visible residual tumour of <1 cm. Studies have shown that patients with residual >2 cm show no improvement in survival over patients without debulking, and are considered incurable. Outcomes according to cytoreductive surgery are shown in Box 13.5. However, even after a complete response 25–50% will recur later.

Box 13.5
Stage-wise survival according to surgical cytoreduction at primary surgery

Meta-analyses suggest that patients with stage IV disease obtain a survival benefit after optimal debulking surgery. Even in patients with liver metastases, studies indicate an improved survival for those in whom optimal extra-hepatic cytoreduction is achieved (27 vs. 8 months). However, more extensive surgery such as hepatic metastatectomies does not appear to improve the survival of patients. Patients with chemotherapy resistant disease have a poor prognosis, even if optimal extra-hepatic cytoreduction is achieved.

Assessment of operability

Surgery is the primary treatment of choice where a complete debulking is considered possible. Patients with bulky disease or a poor performance status (25%) are treated with neo-adjuvant chemotherapy then reassessed for interval debulking surgery (IDS). Surgery should only be attempted where optimal debulking is possible. In general, debulking is less likely to be successful in patients with bulky upper abdominal disease or fixed tumour. Even if debulking surgery is not possible, palliative bowel surgery may be appropriate to relieve impending or actual bowel obstruction.

Neoadjuvant chemotherapy

A randomized trial (EORTC 55971) has shown that neo-adjuvant chemotherapy has equivalent efficacy with lower morbidity than primary debulking surgery in patients with stage bulky III/IV disease. Neo-adjuvant chemotherapy usually comprises 3 cycles of carboplatin and paclitaxel (or carboplatin alone) followed by interval debulking surgery for patients responding to chemotherapy, then a further three cycles of adjuvant carboplatin and paclitaxel. Approximately 10% of patients will progress on chemotherapy (platinum refractory) and the prognosis for these patients is poor and not improved by surgery. For this group further treatment is with a change of chemotherapy or best supportive care.

Adjuvant chemotherapy

Adjuvant chemotherapy is aimed at improving overall survival. There are no studies comparing best supportive care with chemotherapy. Meta-analyses suggest that adjuvant treatment with platinum compared with non-platinum monotherapy achieves a 30% relative risk reduction in mortality implying some benefit from platinum based chemotherapy. Four phase III trials have investigated the benefit of adding taxanes to platinum. Two trials (GOG111 and OV10) showed a mean survival benefit of 10–14 months whist other two (ICON3 and GOG132) did not. A meta-analysis of the four trials demonstrates a small survival advantage with the addition of paclitaxel to platinum. Many expert consider the combination of carboplatin and paclitaxel as the standard adjuvant chemotherapy.

The choice between single agent carboplatin and carboplatin with paclitaxel as first line treatment should be made after discussion of the risks and benefits with the patient. Since platinum exerts the major effect, and paclitaxel increases toxicity, single agent carboplatin will be the treatment of choice for patients with poor performance status or co-morbidities.

Intraperitoneal (IP) chemotherapy

A number of studies have investigated intraperitoneal chemotherapy. Although some have shown a small improvement in PFS and OS (GOG-172), the toxicities were high (p. 38). In the UK, IP chemotherapy is not currently offered outside a clinical trial.

Prognosis

Prognosis according to stage is shown in Table 13.1. The most important prognostic variables after stage are in order; degree of differentiation, cyst rupture, substage of disease and age.

Table 13.1 Prognosis of ovarian cancer

FIGO stage	5-year survival (%)	5-year disease-free survival
I	80–90	70–85
II	65–80	55–65
IIIa	50	45
IIIb	40	25
IIIc	30	20
IV	15	10

Relapse

Despite 70% of patients with advanced disease having a complete clinical response at the end of surgery and chemotherapy (no detectable disease on CT scan and/or normal CA-125), 70% of these will relapse, with a median progression free survival from diagnosis of 16 months. The time from first relapse to death is now around 2 years. There are no studies comparing best supportive care with chemotherapy in relapsed patients, but the recent studies suggest that chemotherapy improves survival.

Presentation of relapse

55–70% of patients present with an asymptomatic rise in CA-125. The median time from this to onset of symptoms is 3 months. 30–45% of patients present with symptoms, of whom 30% relapse in the abdomen, 36% in the pelvis, 14% in abdomen and pelvis, and the remainder have distant metastases (frequently liver metastases).

Relapse may be defined by conventional RECIST criteria (p. 44) or by the gynaecological cancer inter-group (GCIG) CA-125 criteria. CA-125 relapse for those who had elevated pre-treatment value which normalized after first line treatment (60% of all new patients) is defined as ≥2× upper limit of normal on two occasions not less than 1 week apart. Patients in whom CA-125 was elevated but never normalized (30% of all new patients): relapse is defined as a CA-125 value ≥2× nadir value on two occasions. The definition of relapse in patients with normal CA-125 prior to initial treatment (10% of all new patients) is similar to that of patients with elevated CA-125 which normalized after first line treatment.

Treatment (Figure 13.3)

Chemotherapy for relapsed disease is palliative and aims to prevent or treat symptoms in order to improve quality of life and possibly extend survival. In selected cases with a long disease free interval (exceeding 12 or 18 months) and localized relapse, surgery may be an option and long term disease free survival may be possible.

Figure 13.3 Treatment of relapsed ovarian cancer.

The probability of response to re-treatment with platinum chemotherapy depends upon the interval between the completion of platinum-based first line therapy to the time of relapse, known as the platinum free interval (PFI). A longer PFI is associated with a higher response rate, longer PFS and improved survival.

Disease relapsing >6 months after platinum treatment is considered ‘platinum sensitive’, however this group is subdivided into ‘relatively’ platinum sensitive (PFI 6–12 months) and fully platinum sensitive (PFI >12 months).

Although most of the data about the PFI and its relation to response is taken from first relapse studies, this finding has been extended to second and subsequent relapses. Standard practice is continue re-treating with platinum (or platinum combination) until the patient becomes platinum resistant, regardless of the line of treatment. It has been suggested that the response rate to platinum can be improved by ‘artificially’ extending the platinum free interval by using non-platinum agents, but this has not been proven.

Platinum sensitive disease

Patients with a PFI of >18 months show response rates of 60–95% to second line therapy, with a median time to progression of 10–12 months. Those with a PFI of 12–18 months show response rates of 50–60%, and a time to progression of approximately 9 months. Patients with a PFI of 6–12 months show response rates of 25–35% to platinum chemotherapy and a median time to progression of <6 months.

ICON4 showed an improvement in response rate for carboplatin and paclitaxel compared with carboplatin alone (57% vs. 50%), progression free survival (12 months vs. 9 months), and 2-year survival (57% vs. 50%) and median survival (29 vs. 24 months). In relapsed platinum sensitive disease, combination treatment is therefore recommended unless there is significant residual neuropathy. Gemcitabine and carboplatin is sometimes used as combination therapy, as phase III trial results show similar improvements in response rates and progression free survival to carboplatin and paclitaxel.

In patients with a PFI of 6–12 months, the benefits of combination chemotherapy are reduced. Single agent carboplatin is more commonly offered in this group. An alternative that may be considered in this group is peglyated doxorubicin (caelyx), since response rates from non-comparative phase III trials are similar.

Platinum resistant disease

Approximately 25% of patients will progress during, or within 6 months of completion of first line treatment. These patients have a low probability of response to any chemotherapy and should be considered for clinical trials if possible. The response to platinum agents is <10%. The overall median time-to-progression (TTP) is 2–3 months from starting second line treatment. The median OS from time of relapse is 8–10 months, and 2-year survival is 17–21%.

Caelyx, paclitaxel and topotecan are useful; all with response rates are around 6–13%. Rates of stable disease 28–43%, PFS is 2–3 months and OS is 8–10 months. Caelyx is most commonly chosen as the initial treatment because of its favourable toxicity profile. Paclitaxel tends to be used only when patients have not received paclitaxel with 1st line therapy. Recently some centres in the UK have used a weekly carboplatin and paclitaxel regime which has shown favourable response rates. However this regime awaits validation in a prospective RCT. Other agents with activity are oral etoposide, gemcitabine, vinorelbine, ifosfamide and tamoxifen. In this group of patients there is no evidence that combination treatment is superior to single agent.

Hormonal treatment

A proportion of ovarian cancers express the oestrogen receptor (ER) in a percentage of the tumour population. After the use of tamoxifen and aromatase inhibitors in breast cancers, several groups have used these agents in relapsed ER positive ovarian cancers. Compared with breast cancer a lower proportion of tumour cells express ER but the response to aromatase inhibitors has been shown to reflect the percentage of ER positivity. In a phase II study 17% had a response and 26% had stable disease by CA-125 criteria when letrozole was used in patients with significant ER positivity in their tumour at the time of relapse. By RECIST criteria 9% had a PR and 42% achieved stable disease. The responses were greatest amongst those with the highest ER scores.

Future directions

Examples of agents under investigation include anti-angiogenics such as VEGF inhibition with bevacizumab, and PARP inhibitors. Poly (ADP-ribose) polymerase (PARP) is an enzyme involved in the repair of single strand DNA breaks. The involvement of the PARP pathway in ovarian cancer was suggested by the relationship of BRCA positive patients who have an increased incidence of ovarian cancer and a deficiency in homologous repair. In addition 50% of sporadic serous papillary ovarian cancers also have impaired PARP expression. PARP inhibitors have shown some benefit in the prevention of further relapse when used in a maintenance setting after response to chemotherapy.

Palliative care

The most common palliative symptoms in advanced ovarian cancer are ascites, pleural effusion and small bowel obstruction. Permanent ascitic drains can be placed for those with frequent recurrent ascites. The most debilitating symptom is of small bowel obstruction which is frequently recurrent. Treatment of small bowel obstruction consists of minimal oral intake and symptomatic control with anti-emetics (haloperidol or cyclizine), antispasmodics (e.g. buscopan) and steroids (dexamethasone) if not responding to the initial treatment. Pain relief should be added to the syringe driver (e.g. diamorphine). If there is colicky abdominal pain metoclopramide should be avoided as this will exacerbate the symptoms. If secretions remain high, octreotide given via a separate syringe driver can be useful. Intravenous fluids should be given but the use of parenteral nutrition should be reserved for patients with a meaningful chance of treatment to remission, e.g. with bowel obstruction at presentation or at first relapse after a long disease free interval.

Follow-up

Most (approximately 95%) recurrences will occur in the first 3 years and relapse after 5 years is rare. Follow-up is directed at detecting localized relapse which is potentially curable with surgery or radiotherapy, e.g. local vaginal vault relapse. Follow-up should include clinical examination, CA-125 (if elevated at presentation) and intermittent imaging if CA-125 was not informative or with symptoms. A typical follow-up would be 3-monthly for 2–3 years then 6-monthly up to 5 years. Patients with an informative CA-125 can become very focused on their blood result at each appointment so a clear plan of action with a raised CA-125 is recommended as there is no evidence that treatment with an asymptomatic elevated CA-125 improves outcome.

Endometrial cancer

Introduction

Endometrial cancer is the fifth most common female cancer in developed countries. In the UK 6800 new patients are diagnosed every year. The incidence rises steeply to a peak at 64–74, with 75% of cases diagnosed in post-menopausal women. In the UK, women have a lifetime risk of 0.9% of developing endometrial cancer. Endometrial cancer is responsible for 2% of all cancer deaths in women in the UK. Most cases are diagnosed at an early stage and overall 5-year survival is 75%.

Aetiology

Table 13.2 shows the risk factors for endometrial cancer. Less than 5% of endometrial cancers are hereditary, and most of these arise in women with hereditary non-polyposis coli (HNPCC) or Lynch syndrome II (p. 51).

Table 13.2 Risk factors for endometrial cancer

Risk factor	Relative risk
Increased age	–
Unopposed oestrogen	2–10
Late menopause (after age 55)	2
Nulliparity	2
Polycystic ovary syndrome	3
Obesity	2–4
Diabetes mellitus	2
Hereditary non-polyposis colorectal cancer	22 to 50% lifetime risk
Tamoxifen	2/1000