PERSPECTIVES OPINION
Squamous-cell carcinoma of the anus: progress in radiotherapy treatment Rob Glynne-Jones, David Tan, Robert Hughes and Peter Hoskin
Abstract | Chemoradiotherapy is the standard-of‑care treatment of squamous-cell carcinoma of the anus (SCCA), and this has not changed in decades. Radiation doses of 50–60 Gy, as used in many phase III trials, result in substantial late morbidities and fail to control larger and node-positive tumours. Technological advances in radiation therapy are improving patient outcomes and quality of life, and should be applied to patients with SCCA. Modern techniques such as intensitymodulated radiotherapy (IMRT), rotational IMRT, image-guided radiotherapy using cone-beam CT, and stereotactic techniques have enabled smaller margins and highly conformal plans, resulting in decreased radiation doses to the organs at risk and ensuring a shorter overall treatment time. In this Perspectives article, the use of novel approaches to target delineation, optimized radiotherapy techniques, adaptive radiotherapy, dose-escalation with external-beam radiotherapy (EBRT) or brachytherapy, and the potential for modified fractionation are discussed in the context of SCCA. Squamous-cell carcinoma of the anus (SCCA) is a rare cancer with an incidence of 1–2 cases per 100,000 people, although this incidence is increasing1,2. Distant metastases appear late in the course of disease and local failure occurs usually at the primary site, according to the results of retrospective patient series3,4, population series2 and randomized trials5. The primary aim of treatment is to achieve locoregional control of the cancer and preserve anal function, while maintaining the best possible quality of life. The accepted standard treatment of SCCA typically consists of concurrent chemoradiotherapy using fluoropyrimidines and mitomycin C (MMC). Conventional doses of radiation in patients with SCCA are typically between 50 and 60 Gy in standard fractionated schedules using 1.8–2.0 Gy per fraction (TABLE 1, BOX 1). With the increasing use of intensitymodulated radiotherapy (IMRT), the optimal radiation dose, overall treatment time (OTT), and treatment schedule remain a matter of debate. In this Perspectives, we examine the current challenges in the treatment of patients with SCCA using radiation.
Radiotherapy for SCCA Treatment of patients with SCCA using radiotherapy alone has, in the past, been difficult and continues to pose a challenge to radiation oncologists owing to the large irregular pelvic target volumes, tumour and organ motion, and the many organs at risk (OAR) within close proximity of these target volumes; all of these considerations make striking the optimal balance between efficacy and toxicity a challenge. Thus, a ‘one-sizefits-all’ approach to delivery of radiotherapy is no longer sustainable in this patient population. Individualization of treatment requires the compilation of primary data into large datasets, in order to create evidence-based prediction models, which enable clinical, radiological and molecular information to be integrated with sufficient statistical power to be successfully validated in an independent patient cohort6. The occurrence of late treatmentassociated morbidities is particularly relevant to patients with SCCA, owing to the fact that prolonged overall survival is expected for many patients. In several randomized studies,
NATURE REVIEWS | CLINICAL ONCOLOGY
radiotherapy was delivered using 2D or 3D techniques7–9, with large volumes of normal tissue present within the target volume. The use of approaches such as anterior–posterior and/or posterior–anterior (APPA) radiation fields with concurrent chemotherapy has been associated with incidences of grade 3/4 acute toxicities of >70%, with 15% of patients requiring treatment breaks or early termination of radiotherapy; moreover grade ≥3 late pelvic toxicities occurred in 10% of patients9,10. Of note, findings from Nordic population studies indicate that anal continence is significantly worse than that of the general population many years after receiving chemoradiotherapy as a treatment of SCCA11. Findings of clinical trials conducted in the 1990s confirmed that chemotherapy combined with radiotherapy is a more successful treatment strategy than radiation alone, and MMC is an essential component of chemotherapy regimens for patients with SCCA7,8,12. Findings of subsequent phase III trials failed to show any additional improvement in patient outcomes by escalating the radiotherapy dose13, replacing MMC with cisplatin during chemoradiotherapy10, the use of induction cisplatin-based chemotherapy9,13, or applying maintenance chemotherapy after chemoradiotherapy10. Thus, concurrent chemoradiotherapy using 5‑fluorouracil (5‑FU) and MMC remains the standard of care for patients with SCCA14,15. Locoregional control is usually achieved with this approach, resulting in a 5‑year survival rate of 80–90% in patients with early stage cancers16–19. In patients with more advanced stage (T3/T4) cancers, 5‑year relapse-free survival (RFS) and colostomy-free survival (CFS) rates are in the region of 50–60%; hence the use of higher doses of radiation is advocated by some clinicians, although concerns remain that the sphincter mechanism will be compromised by exposure to doses of radiation >54–56 Gy11. The anatomic proximity of radiosensitive pelvic structures, particularly the gastrointestinal tract, exposes normal tissues to high doses of radiation with a consequent risk of late toxicities and a substantial negative effect on patients’ quality of life. Acute haematological and gastrointestinal ADVANCE ONLINE PUBLICATION | 1
© 2016 Macmillan Publishers Limited. All rights reserved
PERSPECTIVES Table 1 | Randomized phase III trials and treatment characteristics of CRT in patients with anal cancer Trial characteristics
Stage included
Treatments
NACT
CRT schedule
Consolidation
Gap
ACT 1 UKCCR (1996)7 n = 585
Any except T1 for local excision
RT versus 5-FU/ MMC/RT
No
45 Gy/20–25 F +/- 5-FU 1,000 mg/m2 on days 1–4 and 29–32, MMC 10 mg/m2 on day 1
No
6 weeks after 45 Gy
EORTC 22861 Bartelink et al. (1997)12 n = 110
T3/4N0‑3 and T1‑2N1‑3
RT versus 5-FU/ MMC/RT
No
45 Gy/25 F +/- 5-FU 750 mg/m2 on days 1–5 and 29–33, MMC 15 mg/m2 on day 1
No
6 weeks after 45 Gy
No
4–6 weeks after 45–50.4 Gy if biopsy tumour positive
No
Max 10 day gap
No
3 weeks after 45 Gy
2 cycles of 5-FU 1,000 mg/m2 on days 1–4,
No gap
RTOG 87‑04/ ECOG Any T, or N Flam et al. (1996)8 stage n = 291
RTOG 98–11 Ajani et al. (2008)9 n = 644
ACCORD‑03 Peiffet et al. (2012)13 n = 307
T2–4, any N stage
T1/T2, N0 excluded
6 week gap followed by a 15 Gy (CR) or 20 Gy boost (PR)
No
5-FU/RT +/- MMC 9 Gy Salvage CRT for biopsy proven residual post CRT
CRT with 5-FU/ MMC versus NACT and 5-FU/cisplatin
2 × 2 factorial NACT and CRT (5-FU/cisplatin) +/- HDRT
ACT II James et al. (2013)10 n = 940
Any T, or N stage
2 × 2 factorial CRT 5-FU/MMC versus 5-FU/ cisplatin +/- maintenance 5-FU/cisplatin
45–50.4 Gy, 5-FU 1,000 mg/m2 on days 1–4 and 29–32 +/- MMC 10 mg/m2 on days 1 and 29 9 Gy boost + 5-FU / cisplatin 100 mg/m2 if residual disease
5-FU 1,000 mg/ m2 on days 1–4, 29–32
45–59 Gy, 5-FU 1,000 mg/m2 on days 1–4 and 29–32, MMC 10 mg/m2 on days 1 and 29
Cisplatin 75 mg/ m2 on days 1 & 29
Or
5-FU 800 mg/m2 per day on days 1–4, 29–32
45 Gy/25 F, 5-FU 800 mg/m2 on days 1–4 and 29–32, cisplatin 80 mg/m2 on days 1 and 29
Cisplatin 80 mg/m2 on days 1 & 29
Or
No
50.4 Gy/28 F, 5-FU 1,000 mg/m2 on days 1–4 and 29–32, MMC 12 mg/m2 on day 1
5-FU 1,000 mg/m2 on days 57–60 and 85–88, cisplatin 75 mg/m2 on days 57 and 85
5-FU 800 mg/m2 on days 57–60 and 85–88, cisplatin 80 mg/m2 on days 57 and 85 and 15 or 20–25 Gy (HDRT)
Or 5-FU 1,000 mg/m2 on days 1–4 and 29–32, cisplatin 60 mg/m2 on days 1 and 29
Cisplatin 60 mg/m2 on day 1 and every 3 weeks
5-FU, 5‑fluorouracil; CR, complete response; CRT, chemoradiation; F, fraction; HDRT, high-dose radiotherapy; MMC, Mitomycin C; NACT, neoadjuvant chemotherapy; PR, partial response; RT, Radiotherapy.
toxicities, dysuria and skin reactions, can occur during and after chemoradiotherapy, in addition to a substantial risk of late adverse effects on bowel, urinary, and sexual function, which evolve over 1–3 years following chemoradiotherapy, and can substantially reduce the quality of life of patients with SCCA2,20,21. Modern radiotherapy approaches In 2014, authors of a review of clinical trials investigating chemoradiotherapy in patients with SCCA22 identified statistically significant associations between locoregional failure and decreased overall survival, which reflects several factors. The following factors can all influence outcomes: the overall ability of the patient to tolerate treatment, which is influenced by sex, age, HIV status, and presence of co‑morbidities; human papilloma virus (HPV) status, which might
be modified by past or present smoking; tumour size; nodal involvement; and, finally, the accuracy of radiation delivery. Compliance to chemoradiotherapy is variable; some patients do not complete planned courses of chemoradiotherapy, or only do so following a gap in treatment. Investigators consistently report that patients with HIV-associated SCCA who are receiving highly active antiretroviral therapy (HAART) — and therefore have similar CD4 counts to those of patients with SCCA who are HIV-negative — have similar responses to chemoradiotherapy in terms of long-term tumour control, although some researchers have expressed concerns regarding acute toxicities in this patient population23–25. Before treatment of patients with HIV-associated SCCA, performance status, viral load, presenting CD4 count, presence of comorbidities, and size and
2 | ADVANCE ONLINE PUBLICATION
stage of the tumour should all be taken into account. Owing to the relative rarity of HIV-associated SCCA outside of some large cities, we recommend that clinicians should decide on the approach for the management of patients who are HIV-positive on a case‑by‑case basis, in collaboration with clinicians who specialize in treatment and management of the underlying infectious disease. The use of IMRT might be of particular benefit to this subset of patients as it reduces the skin dose and volume of normal tissue within the high-dose treated target. More-sophisticated radiotherapy techniques are currently available, including: conformal radiotherapy with sequential boost; static IMRT with inverse planning and simultaneous integrated boost (SIB); rotational IMRT; and helical tomography IMRT applying continuous modulation www.nature.com/nrclinonc
© 2016 Macmillan Publishers Limited. All rights reserved
PERSPECTIVES of the multileaf collimator, dose rate, and gantry rotational speed. Use of rotational techniques reduces the need for monitor units, and reduces the daily treatment duration compared with use of standard IMRT techniques. The implementation of IMRT and image-guided radiotherapy (IGRT) with daily imaging and adaptive planning might enable decreased exposure of OAR to radiation while also preserving locoregional control. Findings from a prospective phase II study conducted by the Radiation Therapy Oncology Group (RTOG)26, as well as two retrospective series examining the effects of IMRT27,28, suggest a reduction in acute toxicity can be obtained with use of these techniques compared with 2D or 3D techniques. Therefore, use of this approach might reduce the need for treatment breaks, which can have an adverse effect on local control of the disease. Use of IMRT might also enable dose-escalation or dose-intensification strategies to be used in patients who are at a high risk of local recurrence (such as those with SCCA of stages T4N0 and T3–4N+). Use of MRI compared with CT‑based planning offers improved spatial resolution and enables better definition of tumour size, local extent and spread, and invasion of adjacent organs, as well as the extent of nodal involvement29. In addition, most anal carcinomas are 18F-fluorodeoxyglucose (FDG)-avid; therefore, metabolically active sites in both the primary tumour and regional nodes can be identified to facilitate optimal radiotherapy planning. Tumour shrinkage during radiotherapy might affect the extent of planning target volume (PTV) dose coverage and OAR sparing. These anatomical changes can be imaged with kV or MV cone-beam CT (CBCT), enabling replanning — either routinely during treatment, or selectively if triggered by preset limits, a process described as image-guided IMRT. Adjustments to the PTV, based on findings of such CT‑based approaches, can be undertaken in order to deliver patientspecific treatment plans with tight tumour margins and the expectation of reduced toxicities compared with conventionally planned radiotherapy. Use of brachytherapy or stereotactic ablative-radiotherapy (SBRT) to apply a highly focused, high-dose radiation ‘boost’ to the primary-tumour area is feasible in selected patients especially those with more locally advanced disease (T3/T4) that might benefit from dose escalation; proton therapy is a further option to explore in the future.
Box 1 | Literature review strategy A literature search was used to examine relevant English language publications from 1974 to September 2014 using the PubMed, MEDLINE and Cancerlit online databases, supplemented by hand-searching of abstracts from recent international meetings. We employed the keywords “anal cancer”, “squamous cell carcinoma”, “local recurrence”, “survival”, “concurrent irradiation”, “chemotherapy”, “radiotherapy”, “chemoradiation”, “radiochemotherapy” and “combined modality”. Studies were considered eligible for inclusion if patients had been randomly allocated, or treatment had been prospectively determined. We read the full text of 347 articles likely to offer relevant information on the defined endpoints of complete clinical response, local control, disease-free survival, relapse-free survival, colostomy-free survival, and overall survival. We retrieved 230 retrospective/unspecified studies, and 65 reviews and/or guidelines, but no published systematic reviews or individual patient meta-analyses. The PARADAC meta-analysis project is reported to have collected individual patient data from phase II and all the randomized phase III trials to examine radiation parameters, but is currently not published. We identified 28 trials with a total of 3,842 patients comprising 6 randomized phase III chemoradiation trials; 4 phase I; 3 phase I/II and 14 other phase II trials (TABLES 1,2,3 AND 4).
Conventional chemoradiation Historically, in randomized phase III trials designed to investigate the efficacy of chemoradiotherapy for patients with anal cancer, treating physicians used doses of 1.8 Gy per day, utilizing a 2D planned shrinking-field technique over the course of treatment, to achieve a total radiation dose of 45–65 Gy (TABLE 2)7–10,12. In these trials, generous radiation-field sizes were used to avoid a geometric miss; the pelvic bones were used as reference points for the position of the pelvic nodes, based on data from lymphangiograms30,31, CT scans, and early radical surgical series32–34. Variations exist in the radiation-field sizes used in different studies, particularly in the superior extent. The routine inclusion of the common iliac nodes, and full coverage of the entire internal iliac nodal system within the radiation field, by setting the upper clinical target volume (CTV) border at or above the sacral promontory3,4, remains controversial. In the ACT II trial35, the superior aspect of the initial APPA field was defined as 2 cm above the inferior aspect of the sacroiliac joints, usually at the S1–S2 interface, and with beam divergence, the estimated dose delivered to the common iliac nodes was small. Very few isolated tumour recurrences were observed above this field in patients whose outcomes were included in the ACT II dataset5. In the Norwegian National Population Cohort2, the recommended superior border was at the level of the lower border of the sacroiliac joints. Only if the primary cancer extended into the rectal mucosa, or if the pelvic nodes were considered to be involved, was the radiation-field extended up to the sacral promontory2. Despite the fact that approximately 50% of patients received radiation therapy using a field that only extended to the lower end of the sacroiliac
NATURE REVIEWS | CLINICAL ONCOLOGY
joint, no recurrences were observed above the upper limits of this treatment field2. Hence, the routine inclusion of common iliac lymph nodes within the radiotherapy field might not be justified. The doses of radiation delivered to the primary tumour and elective nodal groups also varied substantially between trials. In the RTOG 9811 trial9, a minimum radiation dose of 45 Gy was delivered to the primary tumour and perirectal nodes in 25 daily fractions, with an additional reduced-field boost of 10–14 Gy in 2 Gy fractions for patients with more-advanced disease. A treatment break of up to 10 days was permitted for patients with adverse skin reactions9. For patients in the elective nodal groups, a conventionally fractionated initial dose of 30.6 Gy was delivered to the upper pelvis, followed by 36 Gy to the lower pelvic and node-negative inguinal nodes9. In the ACCORD‑03 trial13, investigators explored the use of an adapted radiotherapy boost dose according to the degree of response. The radiotherapy technique consisted of either a standard four-field box or an APPA conformal technique extending from L5–S1 to the perianal region with a dose of up to 45 Gy in 25 fractions using 6–25 MV X‑rays, or electrons to the inguinal region. In the ACT II trial10, a standard dose of 50.4 Gy was delivered to the primary tumour over 28 daily fractions using a two-phase technique, with no gap between the phases. In phase one, a radiation dose of 30.6 Gy was delivered in 17 daily fractions using nonconformal APPA fields10. In phase two, radiation therapy was delivered as a 3D‑conformal CT‑planned boost delineated before the start of phase one to deliver 19.8 Gy to the International Commission on Radiation Units and Measurements (ICRU) intersection point in 11 daily fractions over ADVANCE ONLINE PUBLICATION | 3
© 2016 Macmillan Publishers Limited. All rights reserved
PERSPECTIVES Table 2 | Comparison of results from randomized phase III trials of CRT in patients with anal cancer Studies
Primary Endpoint
Complete response
DFS
Local failure rate
OS
ACT 1 UKCCCR (1996)7 n = 585
Local failure
30% with RT versus 39% with CRT at 6 weeks
NR
61% with RT versus 39% with CRT at 3 years; (P 12 Gy/h, is the most frequently used technique for remote afterloading, where the radiotherapy source is inserted following the brachytherapy implant and planning83. Interstitial implantation of radioisotopes is an invasive procedure, which requires local or general anaesthesia, and demands skill and experience from the operator. Findings from two French studies84,85 in a cohort of 438 patients with SCCA suggest that delivery of EBRT plus a brachytherapy boost resulted in better local control than use of EBRT alone; however, the currently available data are insufficient to enable direct comparisons of acute and late toxicities following EBRT with brachytherapy. Most brachytherapy techniques described in the literature are based on use of low-dose-rate or pulsed-dose-rate approaches86. Typically, in the treatment of patients with SCCA, a minimum EBRT dose of 45 Gy is followed by a 15–20 Gy dose of radiation, delivered using a perineal boost or interstitial brachytherapy13,87. In one study, investigators used a combination of external irradiation, and 5‑FU and MMC, followed by an interstitial 192 Ir implant after a 2‑month treatment gap, and reported a 5‑year survival rate of 65% and an anal preservation rate of 61%; thus, normal sphincter function was retained in more than 90% of surviving patients88. For optimal preservation of sphincter function, we recommend restricting the use of brachytherapy to patients with tumours that extend less than half the circumference of the anal canal, are less than 5–10 mm in thickness, and are 37.5 days, local control is typically inferior compared with that observed in patients with an interval of 10% improvement in local control. Nevertheless, any gains in the effectiveness provided by use of escalated-dose treatment must be weighed up against the increased risk of late toxicities that might affect the function of the internal and external sphincter muscles and/or small bowel. Concerns remain that doses of radiation in excess of 56 Gy might lead to faecal incontinence and the need for a defunctioning colostomy. At such doses, damage to both the anal-sphincter mechanism and the lamina propria of the rectum exist, potentially leading to stenosis, stricture formation, and reduced rectal capacity, is a substantial risk. Studies of anorectal function using anorectal manometry have been performed in patients receiving pelvic radiotherapy, and have demonstrated a decrease in maximum anal canal resting pressure, rectal threshold volume, and maximum tolerated rectal volume in patients exposed to increasing doses of radiation110. In the ACT II trial10, excellent compliance with radiotherapy meant that only 37 of 472 patients (8%) in the MMC group and 44 of 468 patients (9%) in the cisplatin group failed to complete the planned dose, and
Squamous-cell carcinoma of the anus: progress in radiotherapy treatment Rob Glynne-Jones, David Tan, Robert Hughes and Peter Hoskin
Abstract | Chemoradiotherapy is the standard-of‑care treatment of squamous-cell carcinoma of the anus (SCCA), and this has not changed in decades. Radiation doses of 50–60 Gy, as used in many phase III trials, result in substantial late morbidities and fail to control larger and node-positive tumours. Technological advances in radiation therapy are improving patient outcomes and quality of life, and should be applied to patients with SCCA. Modern techniques such as intensitymodulated radiotherapy (IMRT), rotational IMRT, image-guided radiotherapy using cone-beam CT, and stereotactic techniques have enabled smaller margins and highly conformal plans, resulting in decreased radiation doses to the organs at risk and ensuring a shorter overall treatment time. In this Perspectives article, the use of novel approaches to target delineation, optimized radiotherapy techniques, adaptive radiotherapy, dose-escalation with external-beam radiotherapy (EBRT) or brachytherapy, and the potential for modified fractionation are discussed in the context of SCCA. Squamous-cell carcinoma of the anus (SCCA) is a rare cancer with an incidence of 1–2 cases per 100,000 people, although this incidence is increasing1,2. Distant metastases appear late in the course of disease and local failure occurs usually at the primary site, according to the results of retrospective patient series3,4, population series2 and randomized trials5. The primary aim of treatment is to achieve locoregional control of the cancer and preserve anal function, while maintaining the best possible quality of life. The accepted standard treatment of SCCA typically consists of concurrent chemoradiotherapy using fluoropyrimidines and mitomycin C (MMC). Conventional doses of radiation in patients with SCCA are typically between 50 and 60 Gy in standard fractionated schedules using 1.8–2.0 Gy per fraction (TABLE 1, BOX 1). With the increasing use of intensitymodulated radiotherapy (IMRT), the optimal radiation dose, overall treatment time (OTT), and treatment schedule remain a matter of debate. In this Perspectives, we examine the current challenges in the treatment of patients with SCCA using radiation.
Radiotherapy for SCCA Treatment of patients with SCCA using radiotherapy alone has, in the past, been difficult and continues to pose a challenge to radiation oncologists owing to the large irregular pelvic target volumes, tumour and organ motion, and the many organs at risk (OAR) within close proximity of these target volumes; all of these considerations make striking the optimal balance between efficacy and toxicity a challenge. Thus, a ‘one-sizefits-all’ approach to delivery of radiotherapy is no longer sustainable in this patient population. Individualization of treatment requires the compilation of primary data into large datasets, in order to create evidence-based prediction models, which enable clinical, radiological and molecular information to be integrated with sufficient statistical power to be successfully validated in an independent patient cohort6. The occurrence of late treatmentassociated morbidities is particularly relevant to patients with SCCA, owing to the fact that prolonged overall survival is expected for many patients. In several randomized studies,
NATURE REVIEWS | CLINICAL ONCOLOGY
radiotherapy was delivered using 2D or 3D techniques7–9, with large volumes of normal tissue present within the target volume. The use of approaches such as anterior–posterior and/or posterior–anterior (APPA) radiation fields with concurrent chemotherapy has been associated with incidences of grade 3/4 acute toxicities of >70%, with 15% of patients requiring treatment breaks or early termination of radiotherapy; moreover grade ≥3 late pelvic toxicities occurred in 10% of patients9,10. Of note, findings from Nordic population studies indicate that anal continence is significantly worse than that of the general population many years after receiving chemoradiotherapy as a treatment of SCCA11. Findings of clinical trials conducted in the 1990s confirmed that chemotherapy combined with radiotherapy is a more successful treatment strategy than radiation alone, and MMC is an essential component of chemotherapy regimens for patients with SCCA7,8,12. Findings of subsequent phase III trials failed to show any additional improvement in patient outcomes by escalating the radiotherapy dose13, replacing MMC with cisplatin during chemoradiotherapy10, the use of induction cisplatin-based chemotherapy9,13, or applying maintenance chemotherapy after chemoradiotherapy10. Thus, concurrent chemoradiotherapy using 5‑fluorouracil (5‑FU) and MMC remains the standard of care for patients with SCCA14,15. Locoregional control is usually achieved with this approach, resulting in a 5‑year survival rate of 80–90% in patients with early stage cancers16–19. In patients with more advanced stage (T3/T4) cancers, 5‑year relapse-free survival (RFS) and colostomy-free survival (CFS) rates are in the region of 50–60%; hence the use of higher doses of radiation is advocated by some clinicians, although concerns remain that the sphincter mechanism will be compromised by exposure to doses of radiation >54–56 Gy11. The anatomic proximity of radiosensitive pelvic structures, particularly the gastrointestinal tract, exposes normal tissues to high doses of radiation with a consequent risk of late toxicities and a substantial negative effect on patients’ quality of life. Acute haematological and gastrointestinal ADVANCE ONLINE PUBLICATION | 1
© 2016 Macmillan Publishers Limited. All rights reserved
PERSPECTIVES Table 1 | Randomized phase III trials and treatment characteristics of CRT in patients with anal cancer Trial characteristics
Stage included
Treatments
NACT
CRT schedule
Consolidation
Gap
ACT 1 UKCCR (1996)7 n = 585
Any except T1 for local excision
RT versus 5-FU/ MMC/RT
No
45 Gy/20–25 F +/- 5-FU 1,000 mg/m2 on days 1–4 and 29–32, MMC 10 mg/m2 on day 1
No
6 weeks after 45 Gy
EORTC 22861 Bartelink et al. (1997)12 n = 110
T3/4N0‑3 and T1‑2N1‑3
RT versus 5-FU/ MMC/RT
No
45 Gy/25 F +/- 5-FU 750 mg/m2 on days 1–5 and 29–33, MMC 15 mg/m2 on day 1
No
6 weeks after 45 Gy
No
4–6 weeks after 45–50.4 Gy if biopsy tumour positive
No
Max 10 day gap
No
3 weeks after 45 Gy
2 cycles of 5-FU 1,000 mg/m2 on days 1–4,
No gap
RTOG 87‑04/ ECOG Any T, or N Flam et al. (1996)8 stage n = 291
RTOG 98–11 Ajani et al. (2008)9 n = 644
ACCORD‑03 Peiffet et al. (2012)13 n = 307
T2–4, any N stage
T1/T2, N0 excluded
6 week gap followed by a 15 Gy (CR) or 20 Gy boost (PR)
No
5-FU/RT +/- MMC 9 Gy Salvage CRT for biopsy proven residual post CRT
CRT with 5-FU/ MMC versus NACT and 5-FU/cisplatin
2 × 2 factorial NACT and CRT (5-FU/cisplatin) +/- HDRT
ACT II James et al. (2013)10 n = 940
Any T, or N stage
2 × 2 factorial CRT 5-FU/MMC versus 5-FU/ cisplatin +/- maintenance 5-FU/cisplatin
45–50.4 Gy, 5-FU 1,000 mg/m2 on days 1–4 and 29–32 +/- MMC 10 mg/m2 on days 1 and 29 9 Gy boost + 5-FU / cisplatin 100 mg/m2 if residual disease
5-FU 1,000 mg/ m2 on days 1–4, 29–32
45–59 Gy, 5-FU 1,000 mg/m2 on days 1–4 and 29–32, MMC 10 mg/m2 on days 1 and 29
Cisplatin 75 mg/ m2 on days 1 & 29
Or
5-FU 800 mg/m2 per day on days 1–4, 29–32
45 Gy/25 F, 5-FU 800 mg/m2 on days 1–4 and 29–32, cisplatin 80 mg/m2 on days 1 and 29
Cisplatin 80 mg/m2 on days 1 & 29
Or
No
50.4 Gy/28 F, 5-FU 1,000 mg/m2 on days 1–4 and 29–32, MMC 12 mg/m2 on day 1
5-FU 1,000 mg/m2 on days 57–60 and 85–88, cisplatin 75 mg/m2 on days 57 and 85
5-FU 800 mg/m2 on days 57–60 and 85–88, cisplatin 80 mg/m2 on days 57 and 85 and 15 or 20–25 Gy (HDRT)
Or 5-FU 1,000 mg/m2 on days 1–4 and 29–32, cisplatin 60 mg/m2 on days 1 and 29
Cisplatin 60 mg/m2 on day 1 and every 3 weeks
5-FU, 5‑fluorouracil; CR, complete response; CRT, chemoradiation; F, fraction; HDRT, high-dose radiotherapy; MMC, Mitomycin C; NACT, neoadjuvant chemotherapy; PR, partial response; RT, Radiotherapy.
toxicities, dysuria and skin reactions, can occur during and after chemoradiotherapy, in addition to a substantial risk of late adverse effects on bowel, urinary, and sexual function, which evolve over 1–3 years following chemoradiotherapy, and can substantially reduce the quality of life of patients with SCCA2,20,21. Modern radiotherapy approaches In 2014, authors of a review of clinical trials investigating chemoradiotherapy in patients with SCCA22 identified statistically significant associations between locoregional failure and decreased overall survival, which reflects several factors. The following factors can all influence outcomes: the overall ability of the patient to tolerate treatment, which is influenced by sex, age, HIV status, and presence of co‑morbidities; human papilloma virus (HPV) status, which might
be modified by past or present smoking; tumour size; nodal involvement; and, finally, the accuracy of radiation delivery. Compliance to chemoradiotherapy is variable; some patients do not complete planned courses of chemoradiotherapy, or only do so following a gap in treatment. Investigators consistently report that patients with HIV-associated SCCA who are receiving highly active antiretroviral therapy (HAART) — and therefore have similar CD4 counts to those of patients with SCCA who are HIV-negative — have similar responses to chemoradiotherapy in terms of long-term tumour control, although some researchers have expressed concerns regarding acute toxicities in this patient population23–25. Before treatment of patients with HIV-associated SCCA, performance status, viral load, presenting CD4 count, presence of comorbidities, and size and
2 | ADVANCE ONLINE PUBLICATION
stage of the tumour should all be taken into account. Owing to the relative rarity of HIV-associated SCCA outside of some large cities, we recommend that clinicians should decide on the approach for the management of patients who are HIV-positive on a case‑by‑case basis, in collaboration with clinicians who specialize in treatment and management of the underlying infectious disease. The use of IMRT might be of particular benefit to this subset of patients as it reduces the skin dose and volume of normal tissue within the high-dose treated target. More-sophisticated radiotherapy techniques are currently available, including: conformal radiotherapy with sequential boost; static IMRT with inverse planning and simultaneous integrated boost (SIB); rotational IMRT; and helical tomography IMRT applying continuous modulation www.nature.com/nrclinonc
© 2016 Macmillan Publishers Limited. All rights reserved
PERSPECTIVES of the multileaf collimator, dose rate, and gantry rotational speed. Use of rotational techniques reduces the need for monitor units, and reduces the daily treatment duration compared with use of standard IMRT techniques. The implementation of IMRT and image-guided radiotherapy (IGRT) with daily imaging and adaptive planning might enable decreased exposure of OAR to radiation while also preserving locoregional control. Findings from a prospective phase II study conducted by the Radiation Therapy Oncology Group (RTOG)26, as well as two retrospective series examining the effects of IMRT27,28, suggest a reduction in acute toxicity can be obtained with use of these techniques compared with 2D or 3D techniques. Therefore, use of this approach might reduce the need for treatment breaks, which can have an adverse effect on local control of the disease. Use of IMRT might also enable dose-escalation or dose-intensification strategies to be used in patients who are at a high risk of local recurrence (such as those with SCCA of stages T4N0 and T3–4N+). Use of MRI compared with CT‑based planning offers improved spatial resolution and enables better definition of tumour size, local extent and spread, and invasion of adjacent organs, as well as the extent of nodal involvement29. In addition, most anal carcinomas are 18F-fluorodeoxyglucose (FDG)-avid; therefore, metabolically active sites in both the primary tumour and regional nodes can be identified to facilitate optimal radiotherapy planning. Tumour shrinkage during radiotherapy might affect the extent of planning target volume (PTV) dose coverage and OAR sparing. These anatomical changes can be imaged with kV or MV cone-beam CT (CBCT), enabling replanning — either routinely during treatment, or selectively if triggered by preset limits, a process described as image-guided IMRT. Adjustments to the PTV, based on findings of such CT‑based approaches, can be undertaken in order to deliver patientspecific treatment plans with tight tumour margins and the expectation of reduced toxicities compared with conventionally planned radiotherapy. Use of brachytherapy or stereotactic ablative-radiotherapy (SBRT) to apply a highly focused, high-dose radiation ‘boost’ to the primary-tumour area is feasible in selected patients especially those with more locally advanced disease (T3/T4) that might benefit from dose escalation; proton therapy is a further option to explore in the future.
Box 1 | Literature review strategy A literature search was used to examine relevant English language publications from 1974 to September 2014 using the PubMed, MEDLINE and Cancerlit online databases, supplemented by hand-searching of abstracts from recent international meetings. We employed the keywords “anal cancer”, “squamous cell carcinoma”, “local recurrence”, “survival”, “concurrent irradiation”, “chemotherapy”, “radiotherapy”, “chemoradiation”, “radiochemotherapy” and “combined modality”. Studies were considered eligible for inclusion if patients had been randomly allocated, or treatment had been prospectively determined. We read the full text of 347 articles likely to offer relevant information on the defined endpoints of complete clinical response, local control, disease-free survival, relapse-free survival, colostomy-free survival, and overall survival. We retrieved 230 retrospective/unspecified studies, and 65 reviews and/or guidelines, but no published systematic reviews or individual patient meta-analyses. The PARADAC meta-analysis project is reported to have collected individual patient data from phase II and all the randomized phase III trials to examine radiation parameters, but is currently not published. We identified 28 trials with a total of 3,842 patients comprising 6 randomized phase III chemoradiation trials; 4 phase I; 3 phase I/II and 14 other phase II trials (TABLES 1,2,3 AND 4).
Conventional chemoradiation Historically, in randomized phase III trials designed to investigate the efficacy of chemoradiotherapy for patients with anal cancer, treating physicians used doses of 1.8 Gy per day, utilizing a 2D planned shrinking-field technique over the course of treatment, to achieve a total radiation dose of 45–65 Gy (TABLE 2)7–10,12. In these trials, generous radiation-field sizes were used to avoid a geometric miss; the pelvic bones were used as reference points for the position of the pelvic nodes, based on data from lymphangiograms30,31, CT scans, and early radical surgical series32–34. Variations exist in the radiation-field sizes used in different studies, particularly in the superior extent. The routine inclusion of the common iliac nodes, and full coverage of the entire internal iliac nodal system within the radiation field, by setting the upper clinical target volume (CTV) border at or above the sacral promontory3,4, remains controversial. In the ACT II trial35, the superior aspect of the initial APPA field was defined as 2 cm above the inferior aspect of the sacroiliac joints, usually at the S1–S2 interface, and with beam divergence, the estimated dose delivered to the common iliac nodes was small. Very few isolated tumour recurrences were observed above this field in patients whose outcomes were included in the ACT II dataset5. In the Norwegian National Population Cohort2, the recommended superior border was at the level of the lower border of the sacroiliac joints. Only if the primary cancer extended into the rectal mucosa, or if the pelvic nodes were considered to be involved, was the radiation-field extended up to the sacral promontory2. Despite the fact that approximately 50% of patients received radiation therapy using a field that only extended to the lower end of the sacroiliac
NATURE REVIEWS | CLINICAL ONCOLOGY
joint, no recurrences were observed above the upper limits of this treatment field2. Hence, the routine inclusion of common iliac lymph nodes within the radiotherapy field might not be justified. The doses of radiation delivered to the primary tumour and elective nodal groups also varied substantially between trials. In the RTOG 9811 trial9, a minimum radiation dose of 45 Gy was delivered to the primary tumour and perirectal nodes in 25 daily fractions, with an additional reduced-field boost of 10–14 Gy in 2 Gy fractions for patients with more-advanced disease. A treatment break of up to 10 days was permitted for patients with adverse skin reactions9. For patients in the elective nodal groups, a conventionally fractionated initial dose of 30.6 Gy was delivered to the upper pelvis, followed by 36 Gy to the lower pelvic and node-negative inguinal nodes9. In the ACCORD‑03 trial13, investigators explored the use of an adapted radiotherapy boost dose according to the degree of response. The radiotherapy technique consisted of either a standard four-field box or an APPA conformal technique extending from L5–S1 to the perianal region with a dose of up to 45 Gy in 25 fractions using 6–25 MV X‑rays, or electrons to the inguinal region. In the ACT II trial10, a standard dose of 50.4 Gy was delivered to the primary tumour over 28 daily fractions using a two-phase technique, with no gap between the phases. In phase one, a radiation dose of 30.6 Gy was delivered in 17 daily fractions using nonconformal APPA fields10. In phase two, radiation therapy was delivered as a 3D‑conformal CT‑planned boost delineated before the start of phase one to deliver 19.8 Gy to the International Commission on Radiation Units and Measurements (ICRU) intersection point in 11 daily fractions over ADVANCE ONLINE PUBLICATION | 3
© 2016 Macmillan Publishers Limited. All rights reserved
PERSPECTIVES Table 2 | Comparison of results from randomized phase III trials of CRT in patients with anal cancer Studies
Primary Endpoint
Complete response
DFS
Local failure rate
OS
ACT 1 UKCCCR (1996)7 n = 585
Local failure
30% with RT versus 39% with CRT at 6 weeks
NR
61% with RT versus 39% with CRT at 3 years; (P 12 Gy/h, is the most frequently used technique for remote afterloading, where the radiotherapy source is inserted following the brachytherapy implant and planning83. Interstitial implantation of radioisotopes is an invasive procedure, which requires local or general anaesthesia, and demands skill and experience from the operator. Findings from two French studies84,85 in a cohort of 438 patients with SCCA suggest that delivery of EBRT plus a brachytherapy boost resulted in better local control than use of EBRT alone; however, the currently available data are insufficient to enable direct comparisons of acute and late toxicities following EBRT with brachytherapy. Most brachytherapy techniques described in the literature are based on use of low-dose-rate or pulsed-dose-rate approaches86. Typically, in the treatment of patients with SCCA, a minimum EBRT dose of 45 Gy is followed by a 15–20 Gy dose of radiation, delivered using a perineal boost or interstitial brachytherapy13,87. In one study, investigators used a combination of external irradiation, and 5‑FU and MMC, followed by an interstitial 192 Ir implant after a 2‑month treatment gap, and reported a 5‑year survival rate of 65% and an anal preservation rate of 61%; thus, normal sphincter function was retained in more than 90% of surviving patients88. For optimal preservation of sphincter function, we recommend restricting the use of brachytherapy to patients with tumours that extend less than half the circumference of the anal canal, are less than 5–10 mm in thickness, and are 37.5 days, local control is typically inferior compared with that observed in patients with an interval of 10% improvement in local control. Nevertheless, any gains in the effectiveness provided by use of escalated-dose treatment must be weighed up against the increased risk of late toxicities that might affect the function of the internal and external sphincter muscles and/or small bowel. Concerns remain that doses of radiation in excess of 56 Gy might lead to faecal incontinence and the need for a defunctioning colostomy. At such doses, damage to both the anal-sphincter mechanism and the lamina propria of the rectum exist, potentially leading to stenosis, stricture formation, and reduced rectal capacity, is a substantial risk. Studies of anorectal function using anorectal manometry have been performed in patients receiving pelvic radiotherapy, and have demonstrated a decrease in maximum anal canal resting pressure, rectal threshold volume, and maximum tolerated rectal volume in patients exposed to increasing doses of radiation110. In the ACT II trial10, excellent compliance with radiotherapy meant that only 37 of 472 patients (8%) in the MMC group and 44 of 468 patients (9%) in the cisplatin group failed to complete the planned dose, and
