Weekly or daily cisplatin and 5-fluorouracil (5-FU)-based chemoradiation was evaluated for patients with locally advanced squamous cell carcinoma (SCC) of the anal canal treated at a single institution over a 20-year period.
Weekly or daily cisplatin and 5-fluorouracil (5-FU)-based chemoradiation was evaluated for patients with locally advanced squamous cell carcinoma (SCC) of the anal canal treated at a single institution over a 20-year period.
A retrospective, single-institution analysis was conducted of patients receiving concurrent 5-FU/cisplatin and radiotherapy for locally advanced SCC from 1989 to 2009. Endpoints included clinical complete response rate, local recurrence rate, colostomy-free survival, disease-free survival (DFS), overall survival, and treatment-related toxicity.
A total of 197 patients were evaluable. The majority had American Joint Committee on Cancer stage II (41%) or stage III (46%) disease; most were T2 (44%) or T3 (27%); bulky nodal disease (N2-N3) was noted in 24% of patients. Patients received weekly (20 mg/m2) or daily (4 mg/m2) cisplatin during radiotherapy. Median radiation dose was 55 Gy. Clinical complete response was observed in 185 patients (94%). After a median follow-up of 8.6 years, local recurrence rate was 11%. Sixteen patients (8%) developed distant metastases. The 5-year DFS was 81%, the 5-year overall survival was 86%, and the 5-year colostomy-free survival was 88%. By univariate analysis, N-stage was a poor prognostic indicator for 5-year DFS (P = .02, 95% confidence interval = 1.17-2.01) and distant metastases (P = .046, 95% confidence interval = 1.09-2.13). Increased T-stage correlated with the necessity for salvage surgery (P = .01).
The combination of weekly/daily cisplatin and 5-FU–based chemotherapy with concurrent radiotherapy is an effective regimen, and our long-term results indicate that cisplatin is an alternative to mitomycin C and may be considered for the treatment of locally advanced SCC of the anal canal. Cancer 2013;119:3769–3775. © 2013 American Cancer Society.
It is estimated squamous cell carcinoma (SCC) of the anal canal will be diagnosed in 7060 individuals in 2013, compared with 6230 in 2012. More than 85% of patients will present with locally advanced disease, and the majority will be cured with the use of chemotherapy with concurrent radiation therapy. As a result, few treatment modifications have developed over the past 30 years, with the commonly accepted standard of care being 5-fluorouracil (5-FU) plus mitomycin C (MMC) as the accepted doublet for radiation sensitization. Historically, MMC has been provided at variable doses of 10 mg/m2 (day 1 and 29) or 12 to 15 mg/m2 (day 1).
Mitomycin C is commonly associated with significant myelosuppression, which may limit its use in patients at risk for developing anal carcinoma, primarily the elderly, the immunocompromised, and the immunosuppressed. An alternative radiation sensitizer, cisplatin, has been previously investigated in the treatment of locally advanced disease.
RTOG 98-11 was a phase 3 trial designed to evaluate the role of 5-FU in combination with cisplatin or MMC. Based on the potential of induction chemotherapy to improve outcome, 2 cycles of induction chemotherapy were administered, but only in the investigational cisplatin arm. The primary endpoint was 5-year disease-free survival (DFS) with secondary endpoints of overall survival (OS), colostomy-free survival (CFS), and time to relapse. Patients on the control arm received MMC, (10 mg/m2 on days 1 and 29), and 5-FU, (1000 mg/m2 per day on days 1 to 4 and 29 to 32) with concurrent radiation therapy. In the investigational arm, patients received induction chemotherapy with 5-FU (1000 mg/m2 per day on days 1 to 4, days 29 to 32, days 57 to 60, and days 85 to 88 [days 57 and 85 corresponding to days 1 and 29 of radiotherapy]) + cisplatin (75 mg/m2 on days 1 and 29 and repeated on days 57 and 85). All patients received a minimum dose of 45 Gy, in 25 fractions of 1.8 Gy, over 5 weeks to the primary tumor. After a median follow-up of more than 5 years, an improvement in 5-year DFS (67.8% versus 57.8%, P = .006) and OS (78.3% versus 70.7%; P = .026) was noted for the control arm of MMC when compared to the induction plus cisplatin arm; trends toward statistical significance in locoregional failure (LRF, P = .087), colostomy failure (CF, P = .074), and CFS (P = .05) were noted. The investigators concluded that the cisplatin arm was inferior and 5-FU plus MMC should remain as the standard of care. However, the use of induction chemotherapy in the cisplatin arm and the associated 8-week delay until radiation therapy was initiated confounded a true comparison of the final results.
In contrast, the largest phase 3 trial, the UK ACT II trial, was a direct comparison of cisplatin versus MMC and also examined the role of maintenance (adjuvant) treatment. The ACT II trial enrolled 940 patients in a 2 × 2 randomized trial with a control arm of 5-FU (1000 mg/m2, days 1-4, 29-32) + MMC (12 mg/m2, day 1) versus an investigational arm of 5-FU (1000 mg/m2, days 1-4, 29-32) + cisplatin (60 mg/m2, day 1 and 29) with a primary endpoint of 6-month response rate. Patients were then randomized to no maintenance therapy versus maintenance therapy of 5-FU (1000 mg/m2, days 1-4, 29-32) + cisplatin (60 mg/m2, day 1 and 29) to be initiated 4 weeks after the completion of radiation therapy. The primary endpoint was recurrence-free survival. The 6-month complete response rate was equivalent (95%) in both arms. Greater hematologic toxicity was noted in the MMC arm during radiation therapy (24% versus 13%, P < .01). After a median follow-up of 5 years, no difference in DFS, OS, and CFS was observed and no benefit was seen with the use of maintenance chemotherapy (P = .21, hazard ratio [HR] = 0.81, 95% confidence interval [CI] = 0.57-1.13).
RTOG 98-11 and ACT II provided chemotherapy as a radiation sensitizer during weeks 1 and 5 but used different treatment doses for both MMC and cisplatin. Both studies also excluded chronically immunosuppressed patients.
At our institution, we have also previously reported favorable results using continuous daily infusional 5-FU and cisplatin. We have continued to use this same cisplatin doublet regimen but have made modifications to the treatment, including weekly bolus cisplatin, a higher radiation dose in patients with T3/T4 disease (58-59 Gy at standard fractionation), and since 2007 the inclusion of intensity-modulated radiotherapy (IMRT). With the reported results of RTOG 89-11 and ACT II as well as our own earlier experience with cisplatin, we have decided to review our long-term results with cisplatin-based chemoradiation in SCC of the anal canal. Our objective is to review our 20-year experience with weekly or daily cisplatin-based chemoradiation for locally advanced SCC of the anal canal and to assess the limitations of treatment compared to historical controls.
Data collection and analysis were approved by the institutional review board at The University of Texas MD Anderson Cancer Center (MDACC), and a waiver of informed consent was obtained. Consecutive patients with American Joint Committee on Cancer (AJCC) stage I through III SCC of the anal canal treated with concurrent 5-FU plus cisplatin and radiation therapy only at MDACC during 1989 through 2009 were included in this analysis. Medical records including scanned documents and electronic medical records were reviewed for patient demographics, AJCC stage, histological confirmation of SCC of the anal canal, radiation oncology dosimetry records, and clinical outcomes including baseline history of sexually transmitted diseases, chronic immunosuppression, or immunocompromised state. Human immunodeficiency virus (HIV)-positive patients were not excluded from this analysis. All patients were seen by the multidisciplinary treatment team (medical oncology, radiation oncology, and surgical oncology) prior to treatment with a complete multidisciplinary team evaluation at least 8 to 12 weeks following the completion of chemoradiation therapy. Weekly laboratory values were required to be obtained prior to each chemotherapy administration as per institutional requirements. Patients are also required to be evaluated weekly for clinical toxicity by the radiation oncologist as per institutional requirements and were seen in every 1 to 2 weeks by the medical oncologist to determine tolerance of chemotherapy administration.
Acute treatment-related toxicities were documented (Common Terminology Criteria for Adverse Events, version 4.0) as well as clinical complete response (cCR), partial response, local and distant recurrence, CFS, salvage surgery, and OS. The cCR identified at any time point during physician follow-up was sufficient for the purposes of this study. Patients were followed on surveillance with a physical examination including a digital rectal examination every 3 to 4 months for the first 2 years and then biannually for years 3 through 5, laboratory work, flexible sigmoidoscopy or proctoscopy and an annual chest x-ray or computed tomography (CT) scan of the chest as well as CT scan or magnetic resonance imaging (MRI) scan of the abdomen and pelvis.
The chemotherapy regimens provided at the discretion of the treating physician included: 1) intravenous (IV) 5-FU (300 mg/m2) + IV cisplatin (4 mg/m2 per day); 2) IV 5-FU (300 mg/m2) + cisplatin (20 mg/m2, day 1 only of each week of radiation therapy); 3) capecitabine (825 mg/m2 twice daily [bid]) + cisplatin (4 mg/m2 per day); or 4) capecitabine (825 mg/m2 bid) + cisplatin (20 mg/m2, day 1 only of each week) (Table 1). All chemotherapy was provided Monday through Friday and only on days of radiation therapy.
|Chemotherapy||N = 201 (%)||Regimen|
|5-FU/cisplatin||177 (88)||IV 5-FU (300 mg/m2) + cisplatin 4 mg/m2 per day, Mon-Fri or IV 5-FU (300 mg/m2) + cisplatin 20 mg/m2, day 1 only of each week, Mon-Fri|
|Capecitabine/cisplatin||24 (12)||Capecitabine (825 mg/m2 BID) + cisplatin 4 mg/m2 per day, Mon-Fri or Capecitabine (825 mg/m2 BID) + cisplatin 20 mg/m2, day 1 only of each week, Mon-Fri|
Patients were treated with a 3-dimensional conformal technique until 2007; following that, all patients were treated with a standardized IMRT technique; cases involving T1N0 tumors received prophylactic inguinal irradiation. Patients were treated initially with anterior and posterior pelvic fields with a dose of 30.6 Gy in 1.8-Gy fractions. The superior border of the pelvic field was placed at the bottom of the sacroiliac joints prior to 1999, and at the L5/S1 interspace from 1999 onward. The lateral borders covered the medial inguinal nodes, and the inferior border was placed at least 3 cm below the inferior border of the tumor or the anal verge. Patients were then treated with a 3-field technique, with posterior, right lateral, and left lateral fields, with a cumulative dose of either 5040 cGy (prior to 2003) or 45 Gy (from 2003 onward), in 1.8-Gy fractions. In this 3-field portion of the treatment, the superior border was placed at the bottom of the sacroiliac joints and the inferior border was unchanged from the initial field. For the posterior field, the lateral borders were placed 1.5 to 2 cm outside the pelvic brim. For the lateral fields, the posterior border was placed behind the sacrum, and the anterior border was placed 0 to 2 cm behind the pubic symphysis.
Finally, a boost using a reduced 3-field technique was delivered to the primary tumor with a 2- to 3-cm margin, typically to a cumulative dose of 55 Gy. After 2003, the boost dose was individualized based on the T stage. Patients with Tx or T1 tumors received 50.4 Gy, those with T2 tumors received 55 Gy, and those with T3 or T4 tumors received 59 Gy. The involved inguinal nodes were treated to a cumulative dose of 55 Gy, using supplemental appositional electron fields, whereas uninvolved inguinal regions received a dose of 30.6 Gy from the initial pelvic field. After 2003, the dose to involved nodes was individualized based on nodal size. Nodes less than 2 cm received 50 Gy and nodes greater than 5 cm received 59 Gy.
Starting in March of 2007, a simultaneous integrated boost IMRT technique was used, with dose based on the T and N stage. The gross primary tumor volume (GTV) was delineated based on clinical examination findings, as well as diagnostic imaging with CT scan, and/or MRI. The clinical target volume, CTV 1, was defined as the GTV with a 1-cm margin. The CTV 2 encompassed the regional lymph nodes, including the mesorectal, presacral, bilateral inguinal, external iliac, and internal iliac nodes. The CTV volumes were expanded by 0.5 to 0.8 cm to obtain the corresponding planning target volumes (PTVs). Within the CTV 2, enlarged nodes were treated with a 5 mm PTV margin to a dose between 50 Gy (< 2 cm), 54 Gy (between 2 cm and 5 cm), or 58 Gy (> 5 cm) based on their size.
Patients with Tx or T1 tumors were treated in 25 fractions with 1.72 Gy and 2 Gy per fraction to the elective nodal volume and gross disease, for total doses of 43 Gy and 50 Gy, respectively. Patients with T2 tumors were treated with 27 fractions, using a dose of 1.67 Gy and 2 Gy to the elective volume and gross disease, for total doses of 45 Gy and 54 Gy, respectively. Patients with T3 or T4 tumors were treated with 29 fractions, using a dose of 1.62 Gy and 2 Gy to the elective volume and gross disease, for total doses of 47 Gy and 58 Gy, respectively. Enlarged nodes were treated with a 5-mm PTV margin to a dose based on their size, as noted above. Generally, plans were created using at least five 6-MV photon beams.
Three endpoints were examined from the time of initiation of treatment. DFS was the time to either local or distant disease recurrence or death as a result of anal cancer; OS was the time to death; CFS was the time to colostomy. OS, DFS, and CFS were determined using the Kaplan-Meier method. The log-rank test was used for univariate comparisons. Cox proportional hazards models were used to determine whether there is an association between prognostic factors with the above outcome endpoints. All computations were carried out using Stata MP, version 10. A P value < .05 was regarded as statistically significant.
A total of 201 consecutive patients were identified in the squamous cell anal carcinoma database who were treated with concurrent chemoradiation therapy with cisplatin at the MDACC (Table 2). One hundred forty-seven of the patients were women (73%). Documentation of known chronic immunosuppression was present in 7 patients. Nodal involvement was present in 38% of patients, and a small proportion of patients had T4 disease (11%). Histologically, 38% of patients had moderate to moderately-poorly differentiated histology; 27% of patients had poorly differentiated disease. Undifferentiated squamous cell histology was noted in 25% of patients; basaloid/cloacogenic histology was reported in 5 of these patients. A small proportion of patients had AJCC stage I disease, with 87% of patients having AJCC stage II-III disease. Of these patients, 177 (88%) were treated with IV 5-FU and either daily cisplatin (4 mg/m2 per day) or a bolus dose of 20 mg/m2 (day 1 of each week) on Monday through Friday; a minority of patients (12%) received the same dose of cisplatin but received capecitabine at 825 mg/m2 bid, on Monday through Friday, on days of radiation therapy only (Table 1). Four patients were lost to follow-up with treatment outcome unknown. Twenty patients (5%) received IMRT. The median radiation dose received was 55 Gy in 30 fractions (range, 34-69.9 Gy).
|Characteristic||Value N = 201 (%)|
|Age, y (mean)||56|
|Sex: male/female||54 (27) / 147 (73)|
|History of STD/immunosuppression|
|Hepatitis C||1 (<1)|
|AJCC stage (7th edition)|
|Stage I||26 (13)|
|Stage II||82 (41)|
|Stage IIIA||41 (20)|
|Stage IIIB||52 (26)|
|AJCC primary tumor|
|AJCC lymph nodes|
|Well to moderate||6 (3)|
|Moderate to Poor||8 (4)|
Of the 197 patients evaluable for response, 185 (94%) patients had a cCR (Table 3) after a median time to evaluation of clinical response of 18 weeks. Twelve (6%) patients had a partial response (6 with residual tumor, 2 with residual nodal disease, and 4 with both); salvage abdominoperineal resection (APR) was pursued in 6 of the 12 patients; 3 patients were considered surgically unresectable due to development of distant disease before salvage surgery could be performed, 1 was determined to be a poor surgical candidate due to existing comorbidities, and 2 patients were lost to follow-up. Notably, increased T-stage correlated with need for pursuit of salvage surgery (P = .01). No statistical difference in clinical response was associated with AJCC stage, histological differentiation, T stage, or N stage. No difference in clinical response was noted when intravenous 5-FU was provided versus capecitabine.
|CCR N (%)||PR N (%)||Pa|
|Total||185 (94)||12 (6)|
|Male||48 (94)||3 (6)|
|Female||137 (94)||9 (6)|
|Stage I||26 (100)||0 (0)|
|Stage II||77 (96)||3 (4)|
|Stage IIIA||35 (88)||5 (12)|
|Stage IIIB||47 (92)||4 (8)|
|AJCC primary tumor||.459|
|T1||31 (97)||1 (3)|
|T2||81 (94)||5 (6)|
|T3||50 (94)||3 (6)|
|T4||19 (86)||3 (14)|
|Unknown||4 (100)||0 (0)|
|AJCC lymph nodes||.070|
|N0||120 (96)||5 (4)|
|N1||24 (89)||3 (11)|
|N2||28 (97)||1 (3)|
|N3||13 (81)||3 (17)|
|Well||14 (100)||0 (0)|
|Moderate to well||5 (83)||1 (17)|
|Moderate||62 (93)||4 (7)|
|Moderate to poor||6 (75)||2 (25)|
|Poor||49 (92)||4 (8)|
|Undifferentiated||49 (98)||1 (2)|
|5-FU/cisplatin||165 (95)||9 (5)|
|Capecitabine/cisplatin||20 (87)||3 (13)|
A total of 197 patients were evaluable for acute and chronic treatment toxicity. Treatment delays were rare (N = 11) and only accounted for 6% of all patients treated. Only 1 patient developed grade 3 neutropenia; grade 4 myelosuppression was not noted. Grade 3 radiation dermatitis was not unexpected, and developed in 56% of patients. Other common grade 3 acute toxicities included 12% diarrhea, 14% nausea/vomiting, and 4% mucositis. Rare instances of grade 4 acute toxicities occurred in 3% of all patients including diarrhea (1%), nausea/vomiting (0.5%), gastrointestinal bleeding due to gastric ulcer (1%), and radiation dermatitis (0.5%).
The development of chronic irreversible toxicities were noted in a minority of patients (N = 8, 4%); 1 additional elderly patient with a known history of osteoporosis developed bilateral mild insufficiency fractures above the acetabulum within 6 months after completing treatment, but these were noted to be healing within the year.
A voluntary APR was completed in 1 patient for palliation of symptoms attributed to chronic stool incontinence 9 months after completing their treatment. One patient died of renal failure secondary to chronic cystitis; however, that patient had a history of urethral strictures and chronic cystitis before treatment for anal carcinoma, which may have contributed to this outcome. Fistulas or radionecrosis developed in 2 patients, but one was in the setting of a local recurrence. Grade 2 and 3 dyspareunia resulted in persistent sexual dysfunction in 2 patients. One patient developed late pain due to radiation fibrosis of the inguinal region 4 years after treatment attributed to radiation therapy and continues to be followed by the Department of Pain Management. One patient developed chronic cisplatin-induced magnesium wasting syndrome and continues under the care of a nephrologist for more than 5 years.
Eighty-five percent of patients were followed on surveillance for a minimum of 2 years after assessment for complete versus partial treatment response; 80% of patients were followed for a minimum of 3 years; 65% of patients were followed for a minimum of 5 years. Standard surveillance included a physical examination including a digital rectal examination every 3 to 4 months, laboratory work, flexible sigmoidoscopy or proctoscopy, and an annual chest x-ray or computed tomography (CT) scan of the chest as well as CT scan or MRI of the abdomen and pelvis.
After a median follow-up of 8.6 years, locoregional recurrent disease (defined as nodal or primary recurrence) developed in 22 (11%) patients and distant metastasis developed in 16 (8%) patients. Salvage surgery with APR was pursued in 18 of the 22 locoregionally recurrent patients, whereas 4 of the 16 patients underwent surgical resection of their metastatic site of disease. Sites of metastatic disease included the liver, lungs, pelvis, distant lymph nodes, and bone. The estimated 5-year DFS was 81% (95% CI = 74%-86%); the 5-year OS was 86% (95% CI = 79%-90%); and the 5-year CFS was 88% (95% CI = 82%-92%). By univariate analysis, N-stage was a poor prognostic indicator for 5-year DFS (P = .02, HR = 1.53, 95% CI = 1.17-2.01) and development of distant metastasis (P = .04, HR = 1.52, 95% CI = 1.09-2.13), but not for OS (P = .59, HR = 1.15, 95% CI = 0.88-1.51) or CFS (P = .17, HR = 1.24, 95% CI = 0.86-1.79) (Figs. 1-3).
Over the past decade, the incidence of anal carcinoma continues to rise 2% annually, with the majority of cases in the US being Caucasian females > 65 years old who are likely to have multiple existing comorbidities. Furthermore, despite the use of antiretroviral therapy, an immunocompromised HIV-positive patient is still at risk of developing SCC of the anal canal but will commonly present at a younger age, in their mid-40s.[11, 12] Hence, chemotherapy options other than MMC are of value in patients receiving chemoradiation with a goal of curative intent. Patients with locally advanced SCC of the anal canal have been traditionally treated with concurrent 5-FU plus MMC and radiation therapy, but the use of MMC and the potential risk of treatment-related myelosuppresion is not an optimal choice for all patients, especially for those more prone to developing SCC of the anal canal such as the elderly, the immunosuppressed, and the immunocompromised patient.
Our retrospective analysis indicates the doublet of 5-FU plus weekly/daily cisplatin with concurrent radiation therapy in the treatment of locally advanced SCC of the anal canal is an effective and well-tolerated regimen providing additional support for the use of 5-FU and cisplatin as an acceptable alternative to 5-FU plus MMC without apparent compromise of efficacy. As noted in our analysis, grade 3 neutropenia was infrequent. The weekly dosing or daily dosing of cisplatin may also be more advantageous in allowing the treating oncologist to have greater control in addressing toxicities, which may be helpful in those individuals most susceptible to anal cancer, namely, the immunosuppressed patient. The strengths of this study are the long median follow-up, the large sample size, and the relative homogeneity of the treatment. In addition, we believe this is the only completed analysis reporting on the efficacy of weekly/daily cisplatin during radiation therapy in the treatment of SCC of the anal canal.
Limitations of this analysis include that it is a single-institution, retrospective study, in which not all late or chronic assessments of toxicities can uniformly be captured due to variable follow-up patterns, and not fully consistently documented, ie, the necessity of IV hydration during treatment. Another limitation is that we are unable to assess the minor degree of sphincter dysfunction; however, because significant fecal incontinence and sphincter dysfunction require surgical intervention, we presume our estimate of severe chronic toxicity requiring a palliative colostomy to be fairly accurate when using surgical intervention as an index.
The role of cisplatin as a radiation sensitizer in locally advanced SCC has been previously explored as indicated in RTOG 98-11 and ACT II, where the role of induction and maintenance chemotherapy was explored, respectively. Although cisplatin is an accepted radiation sensitizer in other malignancies, its role in the treatment of SCC of the anal canal continues to be debated based on the trial design of RTOG 98-11 and the endpoints of superiority versus noninferiority. The final results of the ACT II phase 3 trial indicate that cisplatin as a radiation sensitizer is not inferior to MMC for local recurrence, OS, and CFS. There is clearly no role for cisplatin-based induction or postradiation maintenance (adjuvant therapy) in the treatment of locally advanced SCC based on the final results of RTOG 98-11 and ACT II studies.
The role of cisplatin continues to be investigated in the treatment of SCC of the anal canal. The doublet regimen of 5-FU plus cisplatin currently serves as the chemotherapy backbone for two phase 2 trials in combination with the chimeric monoclonal antibody against the epidermal growth factor receptor (EGFR), cetuximab, in both HIV-positive (AMC045) and HIV-negative (ECOG 3205) patients with locally advanced SCC. AMC045 is the first prospective study to include HIV-positive patients. The investigators recently reported that the combination is feasible, tolerable, and promising in both HIV-positive and HIV-negative patient populations; final results are to be reported at a later date.
With the continued rising incidence of SCC, it is likely we will continue to see the creation of more innovative clinical trials inclusive of HIV+ and other immunosuppressed patients, and also using combination therapy with biomarker-driven targeted agents as promising radiation sensitizers. Currently, for an off-protocol patient, the doublet of 5-FU/cisplatin is a reasonable, tolerable, and potential alternative to 5-FU/mitomycin C for patients being treated for locally advanced SCC of the anal canal.
This research was funded in part by the Farrah Fawcett Foundation, the E.B. Anal Cancer Fund, and an anonymous philanthropic donation. No other specific funding was used.
The authors made no disclosure.