Keywords

Pearls

  • 8,080 estimated new cases and 1,080 estimated deaths in the United States in 2016.

  • 75–80% are squamous cell carcinoma (SCC) ; others are adenocarcinoma or melanoma.

  • HPV: found in 85–95% and strongly associated with SCC and may be requisite for disease formation. High-grade anal intraepithelial lesions (HSIL ) are precursors. In particular HPV-16, 18 as in cervical cancer.

    • HPV vaccines in the United States: quadrivalent vaccine (HPV 6, 11, 16, and 18); 9-valent vaccine (HPV 6, 11, 16, 18, 31, 33, 45, 52, and 58); and bivalent vaccine (HPV 16 and 18).

  • 11% of untreated HSIL progress to SCC; 50% progress with extensive disease of immunosuppression; with treatment, progression is reduced to 0.4%.

  • HIV positivity increases risk, likely through an association with immunodeficiency in the setting of HPV coinfection. Increased risk if CD4 < 200.

  • Additional risk factors: >10 sexual partners, history of genital warts, receptive anal intercourse, chronic immunosuppression, and cigarette smoking.

  • Anatomy: anal canal is 3–5 cm long. Extends from anal verge to the anorectal ring. The dentate line lies within the anal canal and divides it by histology. Proximal to the dentate line is colorectal mucosa, distal to it is nonkeratinizing squamous epithelium. The dentate line contains transitional mucosa. Anal margin is a 5 cm ring of skin around the anus. Use CT to measure depth of inguinal nodes using the femoral vessels as a surrogate due to large variations.

  • Anal margin tumors : may behave like skin cancers, and can be treated as skin cancers as long as there is no involvement of the anal sphincter, tumor is <2 cm, and moderately or well-differentiated, and resected with adequate margins

  • Adenocarcinoma : higher local and distant recurrence rates with chemo-RT compared to SCC. Treatment similar to that of rectal cancer. Use 5-FU chemo-RT pre-op followed by APR.

  • Lymph node drainage : superiorly (above dentate line) along hemorrhoidal vessels to perirectal and internal iliac nodes; inferior canal (below dentate line) and anal verge to inguinal nodes.

  • Presentation: bleeding, anorectal pain/sensation of mass, altered bowel movements/rectal urgency, genital warts/condyloma, pruritus, asymptomatic.

Workup

  • H&P. Include inguinal LN evaluation. Note anal sphincter tone, pain, bleeding, HIV risk factors, inflammatory bowel disease, prior RT. For women, a comprehensive gynecological exam should be performed. On DRE, note anal sphincter tone and tumor location (clock location prone or supine position, distance from verge, circumferential involvement, size, and superior extent).

  • Labs: CBC, HIV test if any risk factors. CD4 count if HIV-positive.

  • Proctoscopy with biopsy.

  • May biopsy inguinal nodes if clinically suspicious. Only FNA, avoid open biopsy.

  • CT chest/abdomen and pelvic CT or MRI.

  • PET/CT is recommended as it is better than CT at detecting the primary tumor and is more sensitive at staging nodal disease (Winton, Br J Cancer 2009; Mistrangelo, IJROBP 2012; Cotter, IJROBP 2006; Schwarz, IJROBP 2008; Trautmann, Mol Imaging Biol 2005).

Table 23.1 Staging (AJCC 7th ed., 2010): anal canal
Full size table

Table 23.2 (AJCC 8th ed., 2017)Definition of Primary Tumor (T)

T category

T criteria

TX

Primary tumor not assessed

T0

No evidence of primary tumor

Tis

High-grade squamous intraepithelial lesion (previously termed carcinoma in situ, Bowen disease, anal intraepithelial neoplasias II—III, high-grade anal intraepithelial neoplasia)

T1

Tumor <2 cm

T2

Tumor >2 cm but <5 cm

T3

Tumor >5 cm

T4

Tumor of any size invading adjacent organ(s), such as the vagina, urethra, or bladder

Table 3
Full size table

Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois. The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

Definition of Distant Metastasis (M)

M category

M criteria

M0

No distant metastasis

M1

Distant metastasis

AJCC Prognostic Stage Groups

When T is...

And N is...

And M is...

Then the stage group is...

Tis

N0

M0

0

T1

N0

M0

I

T1

N1

M0

IIIA

T2

N0

M0

IIA

T2

N1

M0

IIIA

T3

N0

M0

IIB

T3

N1

M0

IIIC

T4

N0

M0

IIIB

T4

N1

M0

IIIC

Any T

Any N

M1

IV

TREATMENT RECOMMENDATIONS

Table 23.3 Treatment recommendations
Full size table

Trials

Chemo-RT Vs. RT

  • UKCCCR ACT I (Lancet 1996; Northover, Br J Cancer 2010): 585 pts with epidermoid cancer of anal canal or margin. RT: 45 Gy + boost (15 Gy EBRT or 25 Gy brachy) ± 5-FU + mitomycin C (MMC). 6-wk break in RT. Chemo-RT improved 3-yr LC (59% vs. 36%), but no significant change in 3-yr OS (65% vs. 58%). Poorer results with RT alone may be due to mandatory 6-wk break. 13-yr median follow-up: for every 100 pts treated with chemo-RT, 25.3 fewer pts with LRR and 12.5 fewer anal cancer deaths vs. 100 pts treated with RT alone. There was a 9.1% increase in nonanal cancer deaths in the first 5 yrs of chemo-RT, which disappeared by 10 yrs.

  • EORTC (Bartelink, JCO 1997): 110 pts. T3-4N0-3 or T1-2N1-3. RT (45 Gy + 15–20 Gy boost) + concurrent chemo (bolus 5-FU + MMC) vs. RT alone. 6-wk break in RT, prior to boost. Chemo-RT improved CR rate (80% vs. 54%), 5-yr LC (68% vs. 50%), colostomy-free survival (72% vs. 40%), and PFS (61% vs. 43%). No difference in OS (57% vs. 52%). Poorer results with RT alone may again be due to mandatory 6-wk break.

  • For pts ineligible for concurrent chemo, good results are achievable with RT alone:

    • Deniaud-Alexandre (IJROBP 2003). 305 pts treated with 45 Gy EBRT, 4–6 wk break, then boost of 20 Gy EBRT (279 pts) or brachy (17 pts). Only 19 pts received concurrent chemo. Complete response rate: T1 96%, T2 87%, T3 79%, and T4 44%. Salvage APR was used successfully for 44% of locally progressive tumors and 54% of local recurrences.

Role of Mitomycin (MMC )

  • RTOG 87-04 (Flam, JCO 1996): 291 pts treated with 45 Gy + 5FU ± MMC. Median follow-up of 36 mos. If no CR at 6 wks, gave 9 Gy boost +5-FU/cisplatin. 5-FU given as bolus × 4 day starting d1, d29 (1000 mg/m2/day). MMC given as 10 mg/m2 bolus d1, d29. MMC improved CR rate (92% vs. 85%) and decreased 4-yr colostomy rate (9% vs. 22%). No difference in 4-yr OS (75 vs. 70%).

Role of Cisplatin

  • ACT II (James, Lancet Oncol 2013): 940 pts with anal cancer [stage T1–T2 (50%), T3–T4 (43%); LN-(62%), LN+ (30%)] treated with 5-FU (1,000 mg/m2/day on d1-4 and 29–32) and RT (50.4 Gy in 28 fx), randomized to either concurrent MMC (12 mg/m2, d1) or cisplatin (60 mg/m2 on d1 and 29), and also randomized to maintenance therapy (2c of cisplatin/5-FU weeks 11 and 14) 4 wks after chemo-RT or no maintenance therapy. No significant difference in complete response at 26 wks between MMC (90.5%) and cisplatin (89.6%) groups. Similar toxicity between the MMC (71%) and cisplatin (72%) groups. No significant difference in PFS between maintenance (74%) vs. nonmaintenance (73%) groups.

  • Based on the above results and RTOG 98-11 (see below), 5-FU/MMC chemotherapy remains the standard of care.

  • It remains unclear whether the 2nd dose of MMC improves efficacy or merely increases toxicity. Some phase III trials have used 2 cycles, while others have used 1 cycle. Retrospective series suggests similar outcomes with less toxicity with only 1 cycle (Yeung, Curr Oncol 2014; White, Radiother Oncol 2015).

Role of Induction Chemo

  • No proven advantage to induction chemo exists; Results to 98-11 may indicate a disadvantage with neoadjuvant chemo.

  • RTOG 98-11 (Ajani, JAMA 2008; Gunderson, JCO 2012; Gunderson, IJROBP 2013): 644 pts, T2–T4, any N. Neoadjuvant cisplatin + 5-FU × 2 followed by concurrent cisplatin +5-FU × 2 and 45–59 Gy vs. concurrent 5-FU + mitomycin and 45–59 Gy. Worse colostomy rate in cisplatin arm (19%) vs. mitomycin arm (10%). At long-term FU, upfront RT + 5FU/MCC improved 5-yr DFS (68% vs. 58%) and OS (78% vs. 71%) vs. induction/concurrent 5FU/cisplatin + RT. T- and N-stage impacted outcomes. In 5FU/MMC arm:

    • 3-yr colostomy: T2N0 9%, T3N0 12%, T4N0 20%, T2N+ 4%, T3N+ 19%, T4N+ 28%.

    • 3-yr LRF: T2N0 10%, T3N0 22%, T4N0 27%, T2N+ 18%, T3N+ 38%, T4N+ 61%.

    • 5-yr DFS: T2N0 80%, T3N0 60%, T4N0 65%, T2N+ 68%, T3N+ 43%, T4N+ 27%.

  • ACCORD 03 Trial (Peiffert, JCO 2012): 283 pts with locally advanced anal cancer randomized to: (1) two induction chemo cycles (5-FU 800 mg/m2/d IV infusion, days 1–4 and 29–32; and cisplatin 80 mg/m2 IV, on days 1 and 29), concomitant chemo-RT (45 Gy in 25 fxs/5 wks, 5-FU and cisplatin during wks 1–5), and standard-dose boost (15 Gy); (2) two induction chemo cycles, concomitant chemo-RT and high-dose boost (20–25 Gy); (3) concomitant chemo-RT and standard dose boost; and (4) concomitant chemo-RT and high-dose boost. Induction chemo or high-dose radiation boost did not improve 5-yr colostomy-free survival rates.

Infusional 5-FU Vs. Capecitabine

  • Capecitabine is a promising alternative to 5FU for anal cancer, but phase III data are needed.

  • Phase II data with oral capecitabine concurrently with mitomycin and RT in anal cancer report overall low toxicity (Glynne-Jones, IJROBP 2008).

  • BC Cancer Agency (Peixoto, J Gastrointest Oncol 2016): retrospective single institution study of 300 pts who received either 5-FU/MMC (64.6%) vs. capecitabine/MMC (35.3%) in combination with RT for locally advanced anal cancer. No difference in disease-free survival or anal cancer-specific survival.

HIV

  • Oehler-Jänne (J Clin Oncol 2008): retrospective, multicentric cohort comparison of 40 HIV+ pts with HAART and 81 HIV- pts treated with RT or CRT. 55% of HIV+ pts had AIDS-defining clinical conditions. CR was 92% of HIV+ and 96% of HIV- cases. 5-yr OS was 61% in HIV+ and 65% in HIV- pts at a median follow-up of 36 mos. 5-yr LC worse in HIV+ pts (38%) vs. HIV- pts (87%), compromising cancer-specific survival and sphincter preservation. Increased grade 3/4 acute skin and hematological in HIV+ pts.

  • White (Am J Clin Oncol 2017): single institution retrospective cohort study of 53 consecutive HIV+ pts treated between 1987 and 2013 vs. 205 consecutive HIV- pts treated between 2003 and 2013. Median RT dose was 54 Gy (28–60 Gy), concurrent chemo was 2 cycles of 5-FU/MMC on day 1 ± day 29. 70% of the HIV+ pts were on HAART at the time of treatment, 65% of pts had an undetectable HIV viral load, and the mean CD4 count was 455. At 3 yrs, no significant difference in PFS (75% vs. 76%), colostomy-free survival (85% vs. 85%), or cancer-specific survival (79% vs. 88%, P = 0.36), respectively.

Brachytherapy

  • Not frequently used in North America due to higher complication rates, including risk of necrosis. Rates of necrosis in the range of 7–15% (Sandhu, IJROBP 1998; Gerard, Radiother Oncol 1998), 6% complication requiring surgery (Ng, IJROBP 1988).

IMRT

  • Multiple studies have reported similar LRC, DFS, colostomy rates, but comparable or lower toxicity with IMRT vs. traditional planning techniques.

  • RTOG 0529 (Kachnic, IJROBP 2013): phase II multi-institutional trial. 52 pts, 54% with stage II, 25% IIIA, and 21% IIIB. 77% experienced grade 2+ GI/GU acute AEs (vs. RTOG 9811 77%). There were significant reductions in acute adverse events (AEs): grade 2+ hematologic (73% vs. 85% in RTOG 9811), grade 3+ GI (21% vs. 36% in RTOG 9811), and grade 3+ dermatological (23% vs. 49% in RTOG 9811).

  • Call (Am J Clin Oncol 2016). Multi-institutional retrospective review of 152 pts treated with IMRT. 3-yr OS 87%, CFS T1-2 96% vs. T3-4 84%, LC T1-2 90% vs. T3-3 79%. Severe acute toxicity: skin 20%, GI 11%, and hematological 41%.

RT Dose

  • Optimal dose of RT continues to be explored.

  • Multi-institutional and retrospective analyses report improved LC for doses >54–55 Gy (e.g., Huang, World J Gastroenterol 2007; Widder, Radiother Oncol 2008).

  • ACCORD 03 trial (above) reported nonsignificant trend for improved colostomy-free survival with increased RT boost dose 20–25 Gy vs. 15 Gy (78% vs. 74%, p = 0.067).

  • Elevated dose with a treatment break does not appear to improve disease outcomes. For example, RTOG 92-08 (John, Cancer J Sci Am 1996) treated pts with 5-FU/MMC + 59.6 Gy with 2 wk planned break included and colostomy rate at 2 yrs was 30%.

Post-treatment Biopsy

  • Cummings (IJROBP 1991): no benefit to routine rebiopsy at 6 weeks post chemo-RT. Continued regression of tumor for up to 12 months, mean time to regression 3 months. ACT II trial reported optimal time point for evaluation of disease response is at 26 weeks because 72% of pts who did not show a CR at 11 weeks had achieved a CR by 26 weeks (Glynne-Jones, Lancet Oncol 2017).

  • Follow pts clinically. Biopsy for clinically suspicious lesions.

Salvage APR

  • Several studies report that salvage APR can achieve 30–77% LC after chemo-RT.

  • Ellenhorn (Ann Surg Oncol 1994): retrospective review of 38 pts treated with RT + 5-FU/MMC. 5-yr OS was 44% when salvage APR used for chemo-RT failure.

Radiation Techniques

General Points

  • IMRT is favored over 3D conformal RT to reduce toxicity. It is critical to follow detailed target volumes as used in RTOG 0529.

  • Minimize treatment breaks (try to keep under 2 weeks). Overall treatment time, but not duration of RT, has a detrimental effect on local failure and colostomy rate (Ben-Josef, JCO 2010).

  • HIV+ pts with CD4 < 50–150.

    • Consider weekly 5FU/Cisplatin.

    • Consider RT alone 4.

    • (Re)institute HAART.

  • HIV+ pts with CD4 < 150–200.

    • Personalize treatment, but consider standard of care treatment with 5FU/MMC/RT.

    • Consider cycle 2 dose reduction or omission of 2nd cycle MMC.

  • HIV+ pts : post-therapy, rigorous HIV management is needed.

Simulation and Planning

  • Simulate patient supine, frog leg in vac lock bag immobilization.

  • Anal marker to mark anal verge.

  • Consider bolus on superficial large palpable groin nodes and any exposed tumor

  • Treat with full bladder to minimize small bowel toxicity and use oral contrast 1–1.5 h before simulation. For patients who have trouble keeping a consistent full bladder, an empty bladder should be considered for reproducibility.

  • Use PET-CT findings in treatment planning.

Conventional Planning (RTOG 98-11 Technique )

  • Targets: primary tumor, grossly enlarged LN, internal/external iliac LN, inguinal LN.

  • Initial large field (all patients) treated AP/PA, energy 18 MV AP, 6 MV PA, dose 30.6 Gy at 1.8 Gy/fx.

    • Borders: superior = L5/S1. Inferior = 2.5 cm margin on anus and tumor. AP field includes lateral inguinal nodes. PA field = 2 cm lateral to greater sciatic notch (not including lateral inguinal LN).

    • Supplementary RT delivered to inguinal nodes with anterior electron fields matched with exit of PA field. Alternatively, may use modified segmental boost photon technique (Moran, IJROBP 2004).

  • Reduced field #1 (all patients) drops AP/PA superior border to inferior border of sacroiliac joints and is treated 14.4 Gy at 1.8 Gy/fx (total 45 Gy). If N0, field is also reduced off inguinal nodes after 36 Gy.

  • Reduced field #2 (for T3–T4, LN+, and T2 lesions with residual disease after 45 Gy).

    • Boost original tumor plus 2–2.5 cm margin 9–14 Gy at 1.8–2 Gy/fx (total 54–59 Gy) using either a multifield technique or laterals or a direct photon or electron perineal field.

    • Involved inguinal and/or pelvic LN should be included if small bowel can be avoided, boost 9–14 Gy (total 54–59 Gy) with electrons.

RTOG 0529 IMRT Technique

  • Follow RTOG anorectal contouring atlas (Myerson, IJROBP 2008).

  • Uses dose painting (all PTVs treated simultaneously).

  • GTVA = gross primary tumor.

  • GTVN50 = all involved nodal regions with macroscopic disease <3 cm greatest dimension.

  • GTVN54 = all nodal regions containing macroscopic disease >3 cm greatest dimension.

  • CTVA: 2.5 cm expansion around gross primary disease and anal canal.

  • CTV45, CTV50, CTV54 includes the nodal regions (respectively, uninvolved, involved with nodes <3 cm, and involved with nodes >3 cm) and a 1.0 cm expansion (except into uninvolved bone, genitourinary structures, muscles, or bowel).

  • For T2N0:

    • PTVA (primary tumor): 50.4 Gy in 28 fx of 1.8 Gy.

    • PTV42 (all nodal regions receives): 42 Gy in 28 fx of 1.5 Gy.

  • For T3-4N0:

    • PTVA: 54 Gy in 30 fx of 1.8 Gy (but for large T3 or T4 tumors, we recommend a subsequent cone-down to 55.8 to 59.4 Gy at 1.8 Gy/fx).

    • PTV45: 45 Gy in 30 fx of 1.5 Gy.

  • For N+:

    • PTVA: 54 Gy in 30 fx of 1.8 Gy.

    • PTV45 (uninvolved LN): 45 Gy in 30 fx of 1.5 Gy.

    • PTV50 (LN ≤ 3 cm): 50.4 Gy in 30 fx of 1.68 Gy.

    • PTV54 (LN > 3 cm): 54 Gy in 30 fx of 1.8 Gy.

  • For further details, see http://www.rtog.org/members/protocols/0529/0529.pdf

UCSF IMRT Doses

  • We use dose-painting (all PTVs treated simultaneously).

  • Primary tumor doses:

    • T1: 50.4–53.2 Gy/28 fx.

    • T2: 53.2 Gy/28 fx.

    • T3: 56–58.8 Gy/28 fx.

    • T4: 58.8 Gy/28 fx.

  • Involved lymph nodes:

    • 50.4 Gy/28 fx if ≤2 cm.

    • 54–58.8 Gy if >2 cm.

  • High-risk lymph nodes (perirectal, presacral, internal iliacs):

    • 47.6 Gy/28 fx.

  • Low-risk lymph nodes (external iliacs and inguinals):

    • 44.8 Gy/28 fx.

Dose Limitations

  • See RTOG 0529 constraints. UCSF constraints:

  • Small bowel: Dmax < 54 Gy, ≤ 30% volume > 45 Gy

  • Bladder: Dmax < 54 Gy; ≤ 30% volume > 45 Gy

  • Femoral Neck: Dmax < 45 Gy

  • Gluteal folds: minimize dose, < 36 Gy if possible

  • Skin (0.5 cm rind): minimize dose, < 20 Gy

Complications

  • Acute complications: skin reaction/desquamation, leukopenia, thrombocytopenia, proctitis, diarrhea, and cystitis.

  • Subacute and late complications include chronic diarrhea, rectal urgency, sterility, impotence, vaginal dryness, and vaginal fibrosis/stenosis (use vaginal dilator status post-XRT to help avoid), and possibly decreased testosterone.

  • Increased risk of late pelvic fracture, particularly among older women.

Follow-Up

  • H&P with anal & inguinal LN exam q8–12 wks until CR, then every 3–6 mos × 5 yrs. Examine more frequently if persistent disease (e.g., monthly).

  • On exam if mass increases in size, or new clinical symptoms develop (pain, bleeding, incontinence) → biopsy. If locally recurrent → salvage APR. If metastatic disease → 5-FU/cisplatin. If tumor decreasing in size, continue to follow. Median time to regression ~3 months, but tumor response can still be observed up to 6 months.

  • Anoscopy q6–12 mos × 3 yrs.

  • For T3–T4 or inguinal LN+: annual CT chest/abdomen/pelvis for 3 yrs.

  • Most recurrences occur within 2 yrs.

  • Anal pap, if available, is useful for follow-up.

  • Recommend vaginal dilator and pelvic floor physical therapy in women to help reduce stenosis/narrowing, starting at 4 weeks post-therapy.

  • Male pts may notice decrease in ejaculate; testosterone levels may be checked for sexual difficulties.