Managing advanced-stage cervical cancer

Best Practice & Research Clinical Obstetrics and Gynaecology Vol. 19, No. 4, pp. 591–609, 2005 doi:10.1016/j.bpobgyn.2005.03.001 available online at http://www.sciencedirect.com

9 Managing advanced-stage cervical cancer D. Allen*

MBChB, MMed, PhD, FCOG(SA), FRANZCOG, CGO

Associate Professor Department of Gynaecological Oncology, Mercy Hospital for Women, University of Melbourne, Clarendon Street, East Melbourne, Vic. 3002, Australia

K. Narayan

MBBS, MD, PhD, FRANZCR

Associate Professor Peter MacCallum Cancer Centre and University of Melbourne, Vic., Australia

The current staging methods and the definition of advanced cervical cancer are discussed. The clinical International Federation of Gynaecology and Obstetrics (FIGO) staging system has been found to be inaccurate and this limits treatment planning. More accurate management could be based on surgicopathological features of the tumour. However, the latest imaging techniques have the potential to give us this information in a non-invasive way. To deliver optimal treatment in advanced cervical cancer, we need to optimize the way we categorize the prognostic groups. Accurately delineating the extent of the disease will potentially minimize treatment morbidity and improve survival. The techniques of chemoradiation are also discussed in detail. Key words: cervical cancer; advanced stage; imaging; management.

The performance of regular Pap smears has played a major role in decreasing the incidence of cervical cancer worldwide. Advanced-stage cervical cancers are mainly diagnosed in women who remain largely unscreened. The survival of women with cervical cancer is directly determined by the extent of disease at the time of diagnosis. In most developed countries, the incidence of and mortality from cervical cancer has decreased significantly over the last 10 years. The lifetime risk of cervical cancer in 1994 in Victoria, Australia was 1:80, and this had decreased to 1:250 by 2002. The incidence and mortality rates for cervical cancer in Victoria, Australia are shown in Table 1.1 The incidence in women over the age of 75 years is higher than in the other age groups, again because of reduced screening. Figure 1 shows the incidence of cervical cancer by

* Corresponding author. Tel.: C61 3 927 02896; Fax: C61 3 927 02448. E-mail address: [email protected] (D. Allen).

1521-6934/$ - see front matter Q 2005 Elsevier Ltd. All rights reserved.

592 D. Allen and K. Narayan

Table 1. The incidence and mortality rates of cervical cancer in Victoria, Australia.

Total Crude incidence rate per 100 000 Mortality Crude mortality rate per 100 000

1994

1995

1996

1997

1998

1999

2000

2001

2002

304 13.5

245 10.7

226 9.8

194 8.3

214 9.1

172 7.2

150 6.2

150 6.2

156 6.3

76 3.4

76 3.3

65 2.8

72 3.1

56 2.4

48 2.0

52 2.2

63 2.6

44 1.8

40 35

1994

30

2002

25 20 15 10 5 85+

80-84

75-79

70-74

65-69

60-64

55-59

50-54

45-49

40-44

35-39

30-34

25-29

20-24

15-19

0

Figure 1. Incidence of cervical cancer by age group in Victoria, Australia in 1994 and 2002. Source: Canstat. Cancer Council of Victoria 2002. 2004.

age group in 1994 and 2002. The 10 most common female cancers in Victoria, Australia are shown in order of frequency in Table 2. Prognostic factors for cervical cancer include stage, tumour volume and depth of invasion, histologic type, lymph node metastases and lymphovascular invasion. The size of the primary tumour is an important prognostic factor and influences the choice of primary therapy.

DEFINITION OF ADVANCED CERVICAL CANCER Traditionally, FIGO stages IA, IB and IIA cervical cancers have been regarded as early disease, and FIGO stages IIB, IIIA, IIIB, IVA and IVB have been regarded as advanced Table 2. The 10 most common sites of female cancers in Victoria, Australia (in order of frequency). 1994: Breast, bowel, melanoma, lung, lymphoma, endometrium, cervix, ovary, leukaemia, bladder 1997: Breast, bowel, melanoma, lung, lymphoma, endometrium, ovary, head and neck, leukaemia, pancreas 2002: Breast, bowel, melanoma, lung, lymphoma, endometrium, ovary, bladder, head and neck, leukaemia

Managing advanced-stage cervical cancer 593

disease. This allowed for different prognostic groupings, as well as a convenient way of deciding and applying uniform treatment policies for the different stages. Generally, surgery was the preferred method of treatment for early-stage cervical cancer and radiotherapy for advanced-stage cervical cancer. FIGO staging is clinical and based on resectability of the tumour, and hence on compartmental extent of the disease and not on the size of the lesion. Realizing the poor prognosis of large, voluminous, barrel-shape stage IB cervical cancer and the difficulty in patient selection for surgery in stage IB, the concept of tumour size was introduced in 1995.2 Stage IB is divided into IB1 and IB2. Stage IB1 consists of tumour with a clinical diameter of %4 cm, and tumours larger than 4 cm are staged as IB2. Gradually, stage IB2 and stage IIA O4 cm have been perceived as advanced cervical cancers. Although arbitrary in its conception, this division is helpful in treatment selection. In the last 15 years, several studies have been published which suggested that the definition of advanced cervical cancer required further consideration.

ARGUMENTS FOR CHANGING THE CURRENT DEFINITION The arbitrary nature of 4 cm as a basis of dividing stage IB was shown by Finan et al.3 Two hundred and twenty-nine patients with stage IB cervical cancer treated with radical hysterectomy were assigned to the new FIGO IB1 (nZ181) and IB2 (nZ48) stages based on clinical tumour diameter. Stage IB2 patients had a significantly worse 5-year survival (72.8%) compared with stage IB1 patients (90.0%) (PZ0.0265). Multivariable analysis suggested that the new staging system did not have an independent impact on survival. Stage acts through nodal status in its impact on survival. In other words, patients with larger tumours but negative nodes did better than those with small tumours and positive nodes. In addition, 38% of the patients with stage IB1 disease and 72% of the patients with stage IB2 disease also received adjuvant, postoperative pelvic radiotherapy. As pointed out in an editorial by Grigsby4, 46% of all patients with stage IB disease in the Finan series received radical surgery and radiotherapy. Patients with stage IB carcinoma of the cervix treated with radiotherapy alone have shown similar diseasefree survival5 as reported in the Finan series. The consequences for patients receiving both surgery and radiotherapy can be severe. This was amply demonstrated by Landoni et al.6 in a landmark phase III clinical trial. Newly diagnosed stage IB and IIA cervical cancer patients were randomized to primary surgery or radiotherapy. Adjuvant radiotherapy was delivered after surgery in women with surgical stage pT2b or greater, less than a 3-mm margin of cervical stroma, cut-through tumour or positive nodes. The primary outcome measures were 5-year survival and the incidence of complications. There were 170 patients in the surgery group and 167 in the radiotherapy group included in the intention-to-treat analysis. After a median follow-up of 87 (range 57–120) months, the 5-year overall and diseasefree survivals were identical in the surgery and radiotherapy groups (83 and 74%, respectively, for both groups). Fifty-four percent of patients with stage IB1 and 85% of patients with stage IB2 cervical cancer received adjuvant radiotherapy. Forty-eight (28%) patients from the surgery group had severe morbidity compared with 19 (12%) patients from the radiotherapy group (PZ0.0004). Clearly, the clinical size of the tumour alone was not sufficient to separate patients such that the need for adjuvant radiotherapy could be predicted pre-operatively. Recent advances (discussed below)

594 D. Allen and K. Narayan

may enable better patient selection for either surgery or chemoradiotherapy as the principal modality of treatment.

DEFINITIONS USED IN THIS ARTICLE We have used the following categories of cancer as advanced cervical cancer: † all cervical cancers staged as FIGO IB2 or higher; † all FIGO stage I cancers with a clinical tumour diameter of !4 cm and which are known to have nodal and metastatic disease, including tumours involving the uterine corpus; and † all cancers requiring adjuvant postoperative irradiation.

STAGING OF ADVANCED CERVICAL CANCER Assessing the size of the primary tumour and the extent of spread is an important step before deciding on a management plan. This information can be obtained by clinical (FIGO) staging, surgical staging (laparotomy or laparoscopy) and imaging [magnetic resonance imaging (MRI), positron emission tomography (PET) and computer tomography (CT) scans]. Clinical staging: examination under anaesthesia One of the major obstacles in prognosticating for carcinoma of the cervix is the inability to clinically delineate the extent of the disease accurately. The FIGO staging system, which is widely used, remains inappropriate to this end. In terms of clinical FIGO staging, Lagasse et al.7 have shown that examination under anaesthesia (EUA) revealed 25% inaccuracy in stage I and 50% inaccuracy in stage II. This is reflected in 5-year survival rates reported from around the world. The survival range in stage I varies from 52 to 90%, stage II from 38 to 68%, stage III from 22 to 61%, and stage IV from 0 to 10%.8 FIGO staging rules are shown in Table 3.8 Surgical staging The idea of disease mapping in cervical cancer is not new. It was proposed in the 1970s by investigators who performed staging laparotomy in the hope of tailoring subsequent radiotherapy. Much useful information about the pattern of nodal disease and radiation Table 3. FIGO staging rules for cervical cancer. Examination under anaesthetic: May include inspection, palpation, colposcopy, endocervical curettage, hysteroscopy, cystoscopy, proctoscopy Investigations: Chest X-ray, intravenous pyelogram, bone scan. Full blood examination, liver and renal function tests can be performed Other rules: If there is doubt as to the stage, the earlier stage is chosen. Once clinically decided, the stage cannot be changed. The carcinoma should be confirmed by histological examination

Managing advanced-stage cervical cancer 595

tolerance in the pelvis and para-aortic region has been obtained. The incidence of pelvic nodal metastases (from surgical series) in FIGO stages IB and IIA was 25% in tumours !4 cm and 31% in tumours O4 cm.6 In FIGO stage III, 55% of patients had positive nodes.9 The incidence of para-aortic nodal metastases was 16% (13–20%) in stage II and 28% (16–38%) in stage III.10 The pattern of nodal metastases was also described by Renolleau et al.11 One hundred and eighty-two cases of cervical cancer with lymph node invasion were treated at the Institut Curie between 1960 and 1988 by colpohysterectomy and lymph node dissection. External iliac (and obturator) lymph node invasion was found in 95% of cases, situated in the middle and internal chains. Common iliac lymph node invasion was found in 24% of cases, but was only exceptionally isolated (3%). Para-aortic lymph node invasion was found in 8.2% of cases. Similarly, Winter et al.12 observed in 140 operable cervical cancer patients with positive nodes that the obturator group was involved in 76% of cases, the external iliac group in 63% of cases and the presacral and common iliac nodes in only 4% of cases. Para-aortic nodes between the inferior mesenteric artery (IMA) and the renal artery are usually not involved if nodes below the IMA are negative. Of the 20 histological positive node samples, 15 patients had positive nodes inferior to the IMA and five patients had positive nodes above the IMA. All five also had concurrent positive nodes inferior to the IMA. There were no patients with negative nodes below the IMA and positive nodes above the IMA.13 Laparoscopic lymph node dissection In the absence of CT-identifiable disease outside the pelvis, para-aortic lymph node involvement can only be identified following surgical intervention. There is evidence that this is worthwhile since patients with locoregionally advanced cervical cancer that are surgically staged have a better outcome than those who are not.13 Nevertheless, the surgical staging introduces significant cost and morbidity into the staging paradigm, and the optimum method remains controversial. Although the most accurate technique, transperitoneal staging laparotomy is associated with high morbidity.14,15 Retroperitoneal lymph node dissection is less morbid16,17 and has similar accuracy to the open technique. Laparoscopic node dissection is more accurate than either and has similar morbidity to retroperitoneal node dissection but is still invasive. Magnetic resonance imaging Clinical estimation of tumour diameter is a poor correlate for the actual tumour volume.18 The advent of CT scans raised considerable hope of estimating tumour volume, but lack of distinction between tumour and normal tissue in CT images rendered this modality useless.19 One of the first reports of MRI of the uterus was published in 198320 wherein authors could not only distinguish uterus from the surrounding soft tissues, but could also see the transitional zone which separated the corpus uteri and cervix. MRI can clearly distinguish between gross cervical tumour and surrounding normal tissue.21 The extent of cervical tumour in MRI images and subsequent whole mount histopathological sections showed a good approximation.22 MRI had been used to determine tumour volume in cervical cancer patients.23 A comparison of the tumour volume calculated from MRI and pathology specimens was made by Hawnaur et al.24 Clinical staging with EUA can only determine the axial dimensions of the tumour. Even within the axial dimensions, it is not possible to estimate the contribution of normal

596 D. Allen and K. Narayan

cervical tissue. Narayan et al.25 found that 24 of 31 (77%) patients with a measurable cervical lesion had craniocaudal or longitudinal axis as the principal axis. Craniocaudal diameter can only be detected through MRI or histological examination. Craniocaudal dimension has prognostic significance. Hayashi and Kato, using a 4-cm cut-off value, noted a 5-year disease-free survival of 70% in tumours !4 cm long and 37% in tumours O4 cm long26, whereas the survival figures for tumours with a transverse diameter of !4 cm and O4 cm were 63 and 50%, respectively. Positron emission tomography Staging lymphadenectomy was the only means of accurately diagnosing nodal disease in inoperable cervical cancer. Such pretherapeutic lymphadenectomy was associated with considerable morbidity.16 Selective diagnostic lymphadenectomy has been shown to improve survival in patients treated by radical radiotherapy. The information about nodal disease is incorporated in defining the upper level of the radiotherapy field and areas requiring an additional radiotherapy boost.27 It has now become possible to detect lymph node metastases non-surgically using PET28,29 in cervical cancer patients planned for radical radiotherapy (see Figure 2).

Figure 2. Positron emission tomography/computer tomography (PET/CT). This patient diagnosed as a poorly differentiated adenocarcinoma, FIGO stage IB cervical cancer, with a 6-cm clinical tumour diameter. Pretreatment MRI (inset) showed this to be a 4.6!7.6!4.5 cm3 tumour expanding cervix and lower uterine corpus (it could be an endometrial cancer). PET/CT revealed two right and one left pelvic nodes. Note that metabolically active primary tumour tissue corresponds exactly to the tumour outline in MRI.

Managing advanced-stage cervical cancer 597

DIAGNOSTIC WORK-UP Once the diagnosis of cervical cancer is established by histology, the patient should be staged using the FIGO rules (see Table 3). If the examination reveals a tumour larger than 4 cm or a FIGO stage higher than IB2, the patient should be investigated with PET and MRI scans. Smaller tumours thought to have a large endocervical component should also be investigated with MRI and PET scans. Patients with positive nodes on PET scan, tumour reaching uterine corpus on MRI scan or those in whom the entire cervix is invaded by tumour leaving !3 mm intact cervical stroma should be regarded as locally advanced cervical cancer and treated by concurrent chemoradiotherapy. If treated primarily by surgery, these patients almost always require postoperative radiotherapy.

CONCURRENT CHEMORADIOTHERAPY Concurrent chemoradiotherapy is the current standard of practice for patients with advanced-stage cervical cancer. The chemotherapy agent of choice is cisplatin that is commenced during the first week of radiotherapy and given weekly. Carboplatin can be used in patients where cisplatin is unsuitable. Concurrent chemoradiotherapy is used both for primary treatment and for posthysterectomy women with high risk factors. Evidence for chemoradiotherapy An overall survival advantage and a 3-year absolute survival advantage of 12% for cisplatin-based therapy given concurrently with radiotherapy has been demonstrated by randomized phase III trials.30–35 These studies have included women treated with primary radiation therapy and women found to have poor prognostic factors at the time of primary surgery. Administration of chemotherapy The chemotherapy is started in the first week of radiotherapy and continued weekly for four to five cycles. Prehydration fluid and mannitol is given intravenously prior to infusing cisplatin 40 mg/m2. Posthydration fluid follows the chemotherapy. A 5-HT3 antagonist and dexamethasone are given prior to starting the cisplatin (see Table 4 for the schema). Cisplatin is generally well tolerated with very little nausea experienced. Weekly carboplatin at area under curve (AUC) 2 can be used as an alternative agent if cisplatin is contraindicated.

RADIOTHERAPY External beam radiotherapy Where possible, patients should be treated on a linear accelerator to avoid an acute radiation reaction in the gluteal folds. The ideal is 12–18 MV photons. Treatments should be given daily Monday to Friday, or five fractions per week.

598 D. Allen and K. Narayan

Table 4. Chemotherapy regime for concurrent chemoradiotherapy. Before chemotherapy: Prehydration 1 L 5% dextrose 0.9% sodium chloride over 1 hour Mannitol 20% 200 mL over 30 minutes Urine output of 400 mL before starting cisplatin infusion Dolasetronw 200 mg orally (Maxolonw 10 mg as necessary may be included) Dexamethasone 8 mg IV Chemotherapy: Cisplatin 40 mg/m2 in 500 mL 0.9% sodium chloride IV infusion over 30–60 minutes Total 1000 mL output by time cisplatin completed, otherwise Lasixw IV After chemotherapy: Posthydration 1 L 4% dextrose 0.18% sodium chloride over 2 hours Maxolonw 10 mg orally as required over the next 2 days

Node-negative patients should receive pelvic radiotherapy. Those with positive pelvic nodes should have their pelvis treated up to the level of the L3–L4 vertebral disc. Those with common iliac or para-aortic disease should have extended field radiotherapy (EFRT) up to the T12-L1 vertebral disc. The EFRT field should be simulated and treated in the supine position. Pelvic radiotherapy Patients should be treated in the prone position on a belly board to exclude the small bowel as far as possible. During simulation, oral radiological contrast should be used to delineate the small bowel. This information may be useful in shielding the small bowel in lateral pelvic portals where the anterosuperior portion of the radiation field can be blocked without shielding the lymphatic drainage areas. Barium mixed with lubricant should be injected into the vagina with a canula. A tampon inserted in the vagina prevents the barium mixture from seeping out. The barium mixture coats the vagina, delineating any vaginal tumour. This information should be used to adequately cover the tumour with a 3-cm margin in the postero-inferior aspect of the lateral field. This is particularly useful in deciding the placement of a rectal shield. Planning and dose prescription Patients should receive 45–50.4 Gy external beam radiotherapy delivered homogeneously to the pelvis in 25–28 fractions. Forty Gy in 20 fractions is also acceptable. In rare circumstances when intracavitary radiotherapy cannot be performed, a shrinking field technique should be used to bring gross tumour volume with adequate margins to a minimum of 65 Gy. An attempt should be made to exclude all small bowel from the treatment field after 50.4 Gy. A four-field box technique with parallel opposed anteroposterior-posteroanterior (AP-PA) and two opposing lateral fields should be used. The external radiotherapy target volume should encompass, with adequate margins, the primary cervical tumour, its gross extension, any grossly involved pelvic lymph nodes and possible microscopic extension to pelvic lymph nodes and the uterus.

Managing advanced-stage cervical cancer 599

AP-PA pelvic fields The superior border should be through the L5-S1 interspace unless the target volume (e.g. involved common iliac lymph nodes) would not be encompassed adequately in a cephalad direction. In the latter case, the superior border should be through the L3–L4 interspace. The lateral border will be 2 cm beyond the lateral margins of the bony pelvis. The inferior border will be inferior to the obturator foramina or the lowest extension of disease with at least a 3-cm margin. The inferior extent of cervical cancer or vaginal extension should be marked so that the inferior border of disease can be documented. Uninvolved normal tissues may be blocked, although the position of the uterus should be drawn on at least one anterior and lateral field to ensure adequate coverage by transfer of the volume from CT scan or MRI of the pelvis onto the simulation films. Any significant nodal disease as seen on CT scan or PET should be documented on the simulation films. Lateral pelvic fields The anterior border should be a horizontal line drawn just anterior to the symphysis pubis, and the posterior border should be a vertical line at the posterior border of the sacrum intersecting the third and fourth sacral space. If MRI is available, the posterior border should include a 3-cm margin at the posterior extent of the tumour or uterus, whichever is greater. Superior and inferior borders will be the same as for the anterior and posterior fields. If clips are present from the lymph node dissection to document the position of the lymph nodes, these should be used as a guide when anterior blocks are designed to shield the small bowel. At least 3 cm should not be blocked anterior to the vertebral body. The outer table of the sacrum should be blocked to protect the sacral plexus on lateral fields. Parametrial boost fields The superior border should be reduced to include the true pelvis alone. The upper border of the true pelvis field is defined as 1 cm above the inferior aspect of the sacroiliac joint. The parametrial block should be a minimum of 4–5 cm wide if a bilateral parametrial boost is used, and may be shaped to the point A isodose. A unilateral parametrial boost should be used if only unilateral involvement was noted. Nodal boost will depend on the position of the nodes but should have a 3-cm margin around the gross nodal disease. Where required, parametrial and nodal boost should be 6–10 Gy in 2-Gy fractions. The prescription point should be at the center of the unblocked portion of the field. The parametrial boost should be delivered, if possible, between implants 1 and 2 in case of low dose rate brachytherapy (LDR) or immediately after implant. In the case of high dose rate brachytherapy (HDR), the boost should be completed before the last HDR treatment. Extended field radiotherapy Patients should be simulated and treated in the supine position. Oral radiological contrast and vaginal contrast should be used to place rectal and small bowel shielding as described above (see Figure 3). Planning CT scans are recommended to help calculate the dose volume histogram of the kidneys.

600 D. Allen and K. Narayan

Figure 3. Simulation portals. Extended field radiotherapy simulation portals. Colour densities are coregistered positron emission tomography details from the same patient, showing metabolically active primary tumour and lymph node metastases. (A) AP-PA treatment portal, (B) parallel and opposed lateral treatment portal. Inset indicates external beam nodal boost treatment field. Primary was boosted with intracavitary brachytherapy.

AP-PA pelvic fields The superior border should be through T12-L1. The lateral border will be 2 cm beyond the lateral margins of the bony pelvis. The lateral field from the upper border to the interspace between L3 and L4 should have a 2-cm margin lateral to the vertebral profile; to cover para-aortic lymphatics and nodes adequately at the same time, it should shield the kidneys. Enlarged nodes in this region may require individual field modification. Lateral margins below L3–L4 should diverge progressively to merge with lateral pelvic margins to accommodate the diverging course of common iliac nodes as they course towards the lateral pelvic wall to become the external iliac group of nodes. The inferior border will be inferior to the obturator foramina or the lowest extension of disease with at least a 3-cm margin. Uninvolved normal tissues (as described above in pelvic field) may be blocked, although the position of the uterus should be drawn on at least one anterior and lateral field to ensure adequate coverage by transfer of the volume from CT scan or MRI of the pelvis onto the simulation films. Any significant nodal disease as seen on CT scans or PET should be documented on the simulation films.

Managing advanced-stage cervical cancer 601

Lateral pelvic fields The anterior border should be a horizontal line drawn just anterior to the symphysis pubis. Most of the anterior abdominal field, except 3 cm anterior to the vertebral body, should be blocked. This block should expand below the level of L5-S1 to merge with anterior small bowel shielding of the lateral pelvic field in keeping with the forward course of the external iliac group of nodes. The posterior border should be a vertical line at the posterior border of the sacrum intersecting the third and fourth sacral spaces. However, the spinal cord and nerve roots should be shielded by placing the posterior margin of the field to exclude the posterior 6–8 mm of vertebral bodies with the help of customized shielding, which should merge below with the sacral shielding as described above. If MRI is available, the posterior border should be placed with a 3-cm margin at the posterior extent of the tumour or uterus, whichever is greater. Superior and inferior borders will be the same as for the anterior and posterior fields. If clips are present from the lymph node dissection to document the position of the lymph nodes, these should be used as a guide when anterior blocks are designed to shield the abdominal content and small bowel. Parametrial and nodal boost fields would be the same as described above. Dose consideration Patients who received O52 Gy of EBRT to the central pelvis had disease specific survival (DSS) rates of 27–34%, compared with 53% for patients treated with lower doses of EBRT to the central pelvis and more intensive ICRT (P!0.0001). At 5 years, the actuarial risk of major complications is also greater for patients treated with O52 Gy of EBRT to the central pelvis (57–68%), compared with those who had 48–52 Gy (28%) and those who had %47 Gy (15%) (P!0.0001). The highest DSS (51%) and lowest actuarial complication rate (17%) were achieved when modest doses of EBRT were combined with relatively intensive intracavitary radiotherapy (ICRT) (P!0.01 for both comparisons).36 Whole pelvic EBRT involves irradiating the whole of the rectum and bladder, thus limiting the dose to point A from ICRT. Similarly, the radiation dose to the para-aortic region (in a concurrent chemoradiotherapy regimen) should be limited to 45–50.4 Gy given in 1.8-Gy fractions. The 2-year progression-free survival based solely on para-aortic lymph node status in 101 cervical cancer patients treated by chemoradiotherapy was 64% in CT-negative and PET-negative patients, 18% in CT-negative and PET-positive patients, and 14% in CT-positive and PET-positive patients (P!0.0001).37 There were no para-aortic failures in 24 patients who had PET-positive nodes, 3 cm or less, and who received a mean dose of 43.3 Gy to the para-aortic region.38 Treatment morbidity in patients requiring fields higher than T12-L1 is very high and these patients should be treated with palliative intent. In our experience, there have been no survivors when gross nodal disease was present above the renal vessels.

BRACHYTHERAPY Low dose rate brachytherapy Following the completion of external beam radiotherapy, the patient should receive 35–45 Gy to point A by intracavitary implant with radium or its equivalent. The patient may receive this in two applications. A minimum of 80 Gy total dose should be delivered to point A. If possible, the first insertion should be performed within 1 week upon

602 D. Allen and K. Narayan

completion of external beam radiation therapy. If two implants are contemplated, the second implant should be completed such that the overall treatment time remains within 8 weeks. High dose rate brachytherapy HDR brachytherapy should start at the end of external beam radiotherapy. Any parametrial or nodal boost can be given during the first 2 weeks of brachytherapy. All treatment should be completed within 8 weeks. Each HDR brachytherapy fraction should aim to give 6.0 Gy to point A. Each patient should be treated with five insertions of 6.0 Gy/fraction to deliver an LDR equivalent dose of 40 Gy. If MRI-compatible applicators are available, MRI should be done with the applicator in the treatment position following a standard treatment of 6 Gy to point A. This planning MRI should be used to optimize isodoses to cover the observable and expected residual tumour. This information should be used for the remaining four insertions. Since, in the latter case, treatment is optimized to the residual tumour and not to point A, the dose prescription likewise should relate to the residual tumour and suspected region of tumour infiltration in the uterine corpus or target volume (as seen on pretreatment MRI) and not point A (see Figure 4). The resulting dose to point A and tumour dose should be recorded separately. It is strongly recommended that tandem and ovoids be used for HDR brachytherapy. A tandem and cylinder is appropriate only for patients where tandem and ovoid application is not possible due to the extent of disease. Determination of normal tissue tolerance In order to stay below an LDR equivalent of 120–130 Gy to the vaginal surface at the vault and 70 Gy to the rectum for five HDR insertions, including the 45-Gy

Figure 4. Target volume for brachytherapy. (A) Pretreatment cervical tumour expanding the cervix and extending into the lower part of the corpus uteri. The tumour/normal tissue interface is clearly visible. (B) The same tumour as in (A) after 40 Gy external beam chemoradiotherapy at first brachytherapy insertion. Tumour has responded well and is much reduced in size. However, the tumour/normal tissue interface is blurred. The target for brachytherapy now consists of residual abnormal tissue at the cervix and the entire uterine corpus, save a narrow rind of uterine tissue at the fundus. This target is covered by 6 Gy (100%) optimized isodose line as indicated by ‘^’.

Managing advanced-stage cervical cancer 603

contribution from the external beam radiation, the vagina should receive 8–9 Gy and the rectum should receive less than 4.1 Gy for each HDR fraction of 6 Gy (70% of the prescribed dose to point A). The dose to the bladder should be less than 4.6 Gy per HDR fraction of 6 Gy (77% of the prescribed dose to point A). As in LDR brachytherapy, every attempt should be made to deliver tumoricidal doses, even if the late responding tissues receive a slightly higher dose. Importance of brachytherapy Brachytherapy is an essential part of radical radiotherapy for cervical cancer. In a review of 1096 patients treated for FIGO stage IIIB cervical cancer, the survival rate was significantly better for patients whose treatment included intracavitary radiotherapy.36 Similarly, higher doses of irradiation delivered to the medial and lateral parametrium with external beam irradiation and intracavitary insertions correlated with a lower incidence of parametrial failures in all stages, except in FIGO stage IB.39 Use of transabdominal ultrasound in tandem placement One of the well-known complications associated with intra-uterine brachytherapy is the risk of creating a false passage within the myometrium and subsequent uterine perforation. Several investigators have used ultrasound to assist in tandem placement and to avoid uterine perforation.40–42 Uneven tumour shrinkage from prior external beam chemoradiotherapy can also result in eccentric placement of tandem in the uterine cavity without causing any false passage. The presence of fibroids can also result in the less than ideal placement of tandem. We believe that even in the absence of uterine perforation and with the tandem in the ideal position within the pelvis (as verified by orthogonal radiography), and with the bladder and rectal doses within tolerances, it may not be possible to avoid radiation-associated enteric complications. Such (unexplained) radiation injury to the small intestine has recently been described following the use of HDR brachytherapy for cervical cancer.43 Ideally, we would recommend the use of transabdominal ultrasound while placing the tandem in the uterus to avoid perforation, and subsequent MRI-assisted dosimetry to avoid radiation toxicity to normal tissues. Figure 5 shows a pretreatment cervical tumour and altered uterine anatomy following EBRT and tandem resting in a false passage within the uterine isthmus. Nevertheless, with the MRI-assisted optimized treatment plan, the patient did receive brachytherapy and remains disease free.

TREATMENT-RELATED TOXICITY Almost all patients feel tired while undergoing concurrent chemoradiotherapy. Occasionally, tiredness may be due to reduced haemoglobin levels following bone marrow suppression secondary to chemotherapy in patients receiving pelvic radiotherapy. In those receiving EFRT, the risk of bone marrow suppression would be much greater. All patients should have a full blood count done once a week and if found to have a haemoglobin level of !100 g/L, they should receive a blood transfusion. Many centres prefer keeping haemoglobin levels O120 g/L. The role of erythropoietin in the management of radiation-induced anaemia is controversial.44,45 These patients should also be monitored for signs of infection secondary to leucopenia. Most patients

604 D. Allen and K. Narayan

Figure 5. False passage. This patient diagnosed in late 2001 had a poorly differentiated squamous cell carcinoma, FIGO stage IIIB cervical cancer, with a 7-cm clinical tumour diameter, reaching the right pelvic side wall. (A) Pretreatment magnetic resonance imaging (MRI) showed this to be a 6!6!5 cm3 tumour expanding and distorting the cervix and the lower part of uterine corpus. (B) A good tumour response following external beam radiotherapy and concurrent chemotherapy. This MRI was done at the time of first brachytherapy insertion and since it was known that the tandem was resting in a false passage, treatment was optimized to exclude surrounding normal tissue without attempting to include the entire uterus in the treatment volume. In subsequent fractions, anatomy improved and the patient was replanned to include the entire uterus. (C) Currently, the patient remains free of treatment-related complications and is disease free. Note that the cervix and the lower part of the uterus have been replaced with fibrosis. 1 denotes body of uterus.

will develop diarrhoea in the third week of treatment that may progressively get worse. Initial management of diarrhoea consists of following a low residue diet. This can be supplemented by loperamide hydrochloride as required. Some patients may require the use of diphenoxylate hydrochloride. Those receiving EFRT may suffer severe nausea and vomiting which is not adequately controlled by metoclopramide. These patients are best treated by 5-HT3 antagonists taken in the form of a wafer that quickly dissolves in the mouth. Adequate oral hydration should be maintained. Occasionally, a few patients may develop burning micturition. If bladder infection has been ruled out, this can usually be treated with a urinary alkalinizer taken orally. Most patients would recover from acute side effects of concurrent chemoradiotherapy within 6–8 weeks of completing treatment.

POST-TREATMENT FOLLOW-UP All patients should be encouraged to use a vaginal cylinder and oestrogen cream following completion of radiotherapy treatment to prevent vaginal adhesions and mucosal atrophy. We recommend daily use of oestrogen application for 2–3 months after radiotherapy and twice weekly thereafter. Most centres follow patients for 5–10 years post treatment. Patients have a pelvic examination every 3–4 months in the first 2 years after completing their treatment and 6 monthly thereafter. Yearly follow-up is appropriate after 5 years following treatment. The majority of recurrences occur during the first 2 years after treatment, and 84% of these patients are symptomatic at the time of recurrence.46 The median time from recurrence to death is 2 years.47 It appears that surveillance for recurrence in low-risk asymptomatic patients could be

Managing advanced-stage cervical cancer 605

less intense. In a recent study by Grigsby et al.48, PET scans were used to evaluate recurrence after radical radiotherapy. The 5-year overall survival rates were 92% for those with no abnormal fluorodeoxy-D-glucose (FDG) uptake and 46% for those with persistent fluorodeoxy-D-glucose (FDG) uptake in the treated field, whereas there were no survivors for those with new FDG uptake outside the treated field (P!0.001).

RESULTS OF CONCURRENT CHEMORADIOTHERAPY STUDY AT PETER MACCALLUM CANCER CENTRE, EAST MELBOURNE The prognostic significance of FIGO stage, histology and MRI parameters (tumour volume and involvement of the uterine corpus) in advanced cervical cancer has been studied. One hundred and seventy-nine cervical cancer patients with FIGO stages IB to IVA who had pretreatment MRI and were treated radically according to the methods described above were analysed for survival. The patient characteristics are shown in Table 5. The prognostic factors and survival figures are shown in Table 6.

Table 5. Baseline characteristics of advanced cervical cancer patients treated at Peter MacCallum Cancer Centre, Melbourne, Australia. Variable All patients FIGO stage

Clinical diameter (cm) At EUA Histology

Age (years) Corpus invasion Tumour volume (mL)

Treatment

Level Ib IIa IIb IIIa IIIb IVa Median (range) Tertiles SCC Clear cell Endometriod/mucinous Serous Median (range) No Yes Geometric mean (% cv) Median (range) Quartiles Surgery alone Chemoradiotherapy Surgery and radiotherapy

Number of patients 179 60 29 49 5 29 7 4.0 (0.0–8.0) 4.0, 5.0 151 4 23 1 60 (22–94) 72 107 27 (120%) 33 (0.1–200) 14, 33, 62 6 153 20

% 100 34 16 27 3 16 4

84 2 13 1 40 60

3 85 11

cv, coefficient of variation; EUA, examination under anaesthesia; SCC, squamous cell carcinoma.

606 D. Allen and K. Narayan

Table 6. Five-year overall survival rates for patient subgroups. Factor

Level

All patients FIGO stage

Clinical diameter

Histology

Age (years) Corpus invasion Tumour volume (mL)

1 2 3 4 !4.0 4.0–!5.0 R5.0 Squamous cell carcinoma Adenocarcinoma !60 R60 No Yes !14 mL 14–!33 mL 33–!62 mL R62 mL

n

Deaths

5-year OS rate

% se

179 60 78 34 7 59 57 63 151

74 21 24 18 6 17 24 33 62

55 60 60 43 14a 70 54 45 56

4 8 6 9 13 6 7 7 4

28 89 90 72 107 40 49 45 45

12 40 34 16 58 8 17 18 31

46 50 60 77 41 82 62 58 25

13 6 6 5 5 6 8 8 7

OS, overall survival; se, standard error. There were 74 deaths experienced by the 179 patients. Kaplan-Meier estimated 5-year survival rates for various patient subgroups are shown above. a OS rate at longest follow-up, 4.7 years.

PALLIATIVE RADIOTHERAPY A number of palliative radiotherapy regimens are available. Our preferred regimen was the one studied from 1985 to 1989 by the Radiation Therapy Oncology Group 8502.49 The dose used was 44.40 Gy in 12 fractions, 3.7 Gy twice daily with a rest of 4 weeks after 14.80 Gy and 29.60 Gy. A total of 290 cases were analysable for late effects. The primary site consisted of gynaecologic (40%), colorectal (28%), genito-urinary (25%) and miscellaneous (7%). The extent of tumour consisted of pelvis only (62%) and additional tumour outside the pelvis (38%). None of the patients with !30 Gy (less than three courses) developed late toxicity. The crude late complications rate was 6%. This schedule has significant logistic benefits and has been shown to produce good tumour regression and excellent palliation of symptoms.49

CONCLUSION Advanced-stage cervical cancers are mainly diagnosed in women who remain largely unscreened. FIGO staging has traditionally defined both the prognostic and treatment modalities in the management of cervical cancer. The size of the primary tumour is an important prognostic factor and influences the choice of primary therapy.

Managing advanced-stage cervical cancer 607

The following categories of cervical cancer can be regarded as advanced: (1) FIGO stage IB2 or higher; (2) nodal or metastatic disease, including tumours involving the uterine corpus; and (3) cancers requiring adjuvant postoperative irradiation. Clinical estimation of tumour diameter is a poor correlate for the actual tumour volume. Unlike CT scans, MRI can clearly distinguish between gross cervical tumour and surrounding normal tissue. Patient staging, disease mapping and defining treatment portals using MRI and PET for advanced cervical cancer patients have been described. Better non-invasive assessment of patients can prevent the use of dual treatment toxicities when surgery and radiotherapy are used. Concurrent chemoradiation is the current standard of practice. Patients should be treated using EBRT on a linear accelerator of 12–18 MV photons. EBRT and pelvic radiotherapy should be prescribed. Patients should receive 45–50.4 Gy EBRT delivered homogeneously to the pelvis in 25–28 fractions. Forty Gy in 20 fractions is also acceptable. An attempt should be made to exclude all small bowel from the treatment field after 50.4 Gy. A four-field box technique with parallel opposed AP-PA and two opposing lateral fields should be used.

Practice points † the use of combined chemoradiation enhances survival in advanced-stage cervical cancers † imaging with MRI and PET scans assists treatment planning

Research agenda † further assess the use of MRI and PET scans in the planning process

ACKNOWLEDGEMENTS We acknowledge the assistance of Ms Sylvia van Dyk with the preparation of the radiotherapy planning section.

REFERENCES 1. Canstat Cancer in Victoria 2002. Annual Victorian Cancer Registry Statistical Report. The Cancer Council Victoria Epidemiology Centre 2004. 2. Creasman WT. New gynecologic cancer staging. Gynecol Oncol 1995; 58: 157–158. 3. Finan MA, DeCesare S, Fiorica JV et al. Radical hysterectomy for stage IB1 vs IB2 carcinoma of the cervix: does the new staging system predict morbidity and survival? Gynecol Oncol 1996; 62: 139–147. 4. Grigsby PW. Stage IB1 vs IB2 carcinoma of the cervix: should the new FIGO staging system define therapy? Gynecol Oncol 1996; 62: 135–136. 5. Grigsby PW. Primary radiotherapy for stage IB or IIA cervical cancer. J Natl Cancer Inst Monogr 1996; 21: 61–64.

608 D. Allen and K. Narayan *6. Landoni F, Maneo A, Colombo A et al. Randomised study of radical surgery versus radiotherapy for stage Ib-IIa cervical cancer. Lancet 1997; 350: 535–540. 7. Lagasse LD, Creasman W, Shingleton HM et al. Results and complications of operative staging in cervical cancer: experience of the Gynecologic Oncology Group. Gynecol Oncol 1980; 9: 90–98. 8. Morrow CP Curtin JP Townsend DE. Tunours of cervix. In synopsis of Gynecologic Oncology. New York: Churchill Livingstone, 1993 pp. 111–152. 9. Lapolla JP, Schlaerth JB, Gaddis O & Morrow CP. The influence of surgical staging on the evaluation and treatment of patients with cervical carcinoma. Gynecol Oncol 1986; 24: 194–206. 10. Hacker NF. Cervical cancer. In Berek JS & Hacker NF (eds.) Practical Gynecologic Oncology. Philadelphia: Lippincott Williams and Wilkins, 2004, pp. 345–405. 11. Renolleau C, Laroussinie MP, Mosseri V et al. Pelvic lymph node invasion in cancer of the uterine cervix. Bull Cancer 1995; 82: 377–383. 12. Winter R, Petru E & Haas J. Pelvic and para-aortic lymphadenectomy in cervical cancer. Baillieres Clin Obstet Gynaecol 1988; 2: 857–866. 13. Altintas A, Vardar MA, Evruke C & Aridogan N. Is it essential to perform complete paraaortic lymph node dissection if no metastases have been shown in the lower part of the aorta? Eur J Gynaecol Oncol 1995; 16: 120–122. 14. Nelson Jr. JH, Boyce J, Macasaet M et al. Incidence, significance, and follow-up of para-aortic lymph node metastases in late invasive carcinoma of the cervix. Am J Obstet Gynecol 1977; 128: 336–340. 15. Tewfik HH, Buchsbaum HJ, Latourette HB et al. Para-aortic lymph node irradiation in carcinoma of the cervix after exploratory laparotomy and biopsy-proven positive aortic nodes. Int J Radiat Oncol Biol Phys 1982; 8: 13–18. *16. Fine BA, Hempling RE, Piver MS et al. Severe radiation morbidity in carcinoma of the cervix: impact of pretherapy surgical staging and previous surgery. Int J Radiat Oncol Biol Phys 1995; 31: 717–723. 17. Bolla M, Salvat J, Sarrazin R et al. Feasibility of retroperitoneal pelvic lymph node exploration in cervical carcinoma. Assessment of morbidity in 33 cases treated by combined radio-surgery or definitive radiotherapy. Radiother Oncol 1996; 40: 233–239. *18. Burghardt E, Baltzer J, Tulusan AH & Haas J. Results of surgical treatment of 1028 cervical cancers studied with volumetry. Cancer 1992; 70: 648–655. 19. Kim SH, Choi BI, Han JK et al. Preoperative staging of uterine cervical carcinoma: comparison of CT and MRI in 99 patients. J Comput Assist Tomogr 1993; 17: 633–640. 20. Hricak H, Alpers C, Crooks LE & Sheldon PE. Magnetic resonance imaging of the female pelvis: initial experience. AJR Am J Roentgenol 1983; 141: 1119–1128. 21. Subak LL, Hricak H, Powell CB et al. Cervical carcinoma: computed tomography and magnetic resonance imaging for preoperative staging. Obstet Gynecol 1995; 86: 43–50. 22. Burghardt E, Hofmann HM, Ebner F et al. Magnetic resonance imaging in cervical cancer: a basis for objective classification. Gynecol Oncol 1989; 33: 61–67. 23. De Souza NM, McIndoe GA, Soutter WP et al. Value of magnetic resonance imaging with an endovaginal receiver coil in the pre-operative assessment of stage I and IIa cervical neoplasia. Br J Obstet Gynaecol 1998; 105: 500–507. 24. Hawnaur JM, Johnson RJ, Buckley CH et al. Staging, volume estimation and assessment of nodal status in carcinoma of the cervix: comparison of magnetic resonance imaging with surgical findings. Clin Radiol 1994; 49: 443–452. 25. Narayan K, McKenzie A, Fisher R et al. Estimation of tumor volume in cervical cancer by magnetic resonance imaging. Am J Clin Oncol 2003; 26: e163–e168. 26. Hayashi T & Kato T. Usefulness of tumor size on MR imaging in assessing the prognosis of uterine cervical cancer treated with radiation. Nippon Igaku Hoshasen Gakkai Zasshi 1999; 59: 250–255. 27. Goff BA, Muntz HG, Paley PJ et al. Impact of surgical staging in women with locally advanced cervical cancer. Gynecol Oncol 1999; 74: 436–442. 28. Rose PG, Adler LP, Rodriguez M et al. Positron emission tomography for evaluating para-aortic nodal metastasis in locally advanced cervical cancer before surgical staging: a surgicopathologic study. J Clin Oncol 1999; 17: 41–45. *29. Narayan K, McKenzie AF, Hicks RJ et al. Relation between FIGO stage, primary tumor volume, and presence of lymph node metastases in cervical cancer patients referred for radiotherapy. Int J Gynecol Cancer 2003; 13: 657–663.

Managing advanced-stage cervical cancer 609 30. Whitney CW, Sause W, Bundy BN et al. Randomized comparison of fluorouracil plus cisplatin versus hydroxyurea as an adjunct to radiation therapy in stage IIB-IVA carcinoma of the cervix with negative para-aortic lymph nodes: a Gynecologic Oncology Group and Southwest Oncology Group study. J Clin Oncol 1999; 17: 1339–1348. *31. Eifel PJ, Winter K, Morris M et al. Pelvic irradiation with concurrent chemotherapy versus pelvic and paraaortic irradiation for high-risk cervical cancer: an update of Radiation Therapy Oncology Group Trial (RTOG) 90-01. J Clin Oncol 2004; 22: 872–880. 32. Rose PG, Bundy BN, Watkins EB et al. Concurrent cisplatin-based radiotherapy and chemotherapy for locally advanced cervical cancer (published erratum appears in N Engl J Med 1999; 341: 708). N Engl J Med 1999; 340: 1144–1153. *33. Rose PG, Bundy BN, Watkins EB et al. Concurrent cisplatin-based radiotherapy and chemotherapy for locally advanced cervical cancer. N Engl J Med 1999; 340: 1144–1153. 34. Peters III. WA, Liu PY, Barrett RJ et al. Concurrent chemotherapy and pelvic radiation therapy compared with pelvic radiation therapy alone as adjuvant therapy after radical surgery in high-risk early-stage cancer of the cervix. J Clin Oncol 2000; 18: 1606–1613. *35. Thomas GM. Improved treatment for cervical cancer—concurrent chemotherapy and radiotherapy. N Engl J Med 1999; 340: 1198–1200. 36. Logsdon MD & Eifel PJ. Figo IIIB squamous cell carcinoma of the cervix: an analysis of prognostic factors emphasizing the balance between external beam and intracavitary radiation therapy. Int J Radiat Oncol Biol Phys 1999; 43: 763–775. *37. Grigsby PW, Siegel BA & Dehdashti F. Lymph node staging by positron emission tomography in patients with carcinoma of the cervix. J Clin Oncol 2001; 19: 3745–3749. 38. Grigsby PW, Singh AK, Siegel BA et al. Lymph node control in cervical cancer. Int J Radiat Oncol Biol Phys 2004; 59: 706–712. 39. Perez CA, Breaux S, Madoc-Jones H et al. Radiation therapy alone in the treatment of carcinoma of uterine cervix. I. Analysis of tumor recurrence. Cancer 1983; 51: 1393–1402. 40. Granai CO, Doherty F, Allee P et al. Ultrasound for diagnosing and preventing malplacement of intrauterine tandems. Obstet Gynecol 1990; 75: 110–113. 41. Wong F & Bhimji S. The usefulness of ultrasonography in intracavitary radiotherapy using selectron applicators. Int J Radiat Oncol Biol Phys 1990; 19: 477–482. 42. Tanaka S, Hishikawa Y, Kamikonya N & Miura T. Sonographic positioning of endouterine applicator. Radiat Med 1987; 5: 92–93. 43. Lertsanguansinchai P, Lertbutsayanukul C, Shotelersuk K et al. Phase III randomized trial comparing LDR and HDR brachytherapy in treatment of cervical carcinoma. Int J Radiat Oncol Biol Phys 2004; 59: 1424–1431. 44. Henke M, Laszig R, Rube C et al. Erythropoietin to treat head and neck cancer patients with anaemia undergoing radiotherapy: randomised, double-blind, placebo-controlled trial. Lancet 2003; 362: 1255–1260. 45. Kavanagh BD, Fischer BA, Segreti EM et al. Cost analysis of erythropoietin versus blood transfusions for cervical cancer patients receiving chemoradiotherapy. Int J Radiat Oncol Biol Phys 2001; 51: 435–441. 46. Gerdin E, Cnattingius S, Johnson P & Pettersson B. Prognostic factors and relapse patterns in early-stage cervical carcinoma after brachytherapy and radical hysterectomy. Gynecol Oncol 1994; 53: 314–319. *47. Wilailak S, Srisupandit S, Tangtrakul S et al. Recurrence of cervical carcinoma following radical hysterectomy and pelvic node dissection. J Med Assoc Thai 1993; 76(supplement 1): 69–73. *48. Grigsby PW, Siegel BA, Dehdashti F et al. Posttherapy [18F] fluorodeoxyglucose positron emission tomography in carcinoma of the cervix: response and outcome. J Clin Oncol 2004; 22: 2167–2171. 49. Spanos Jr. WJ, Clery M, Perez CA et al. Late effect of multiple daily fraction palliation schedule for advanced pelvic malignancies (RTOG 8502). Int J Radiat Oncol Biol Phys 1994; 29: 961–967.