|Home | About | Journals | Submit | Contact Us | Français|
We compared our institutional experience using 3D conformal radiation therapy (3DCRT) vs. IMRT (intensity-modulated radiation therapy) for anal cancer.
We performed a single-institution retrospective review of all patients with squamous cell carcinoma anal cancer treated from September 2000 through September 2011, using definitive chemoradiation with curative intent.
This study included 89 consecutive patients (37 3DCRT, 52 IMRT). Median follow-up for all patients, IMRT patients alone, and CRT patients alone was 26.5 months (range, 3.5–133.6), 20 months (range, 3.5–125.5), and 61.9 months (range, 7.6–133.6), respectively. Three-year overall survival (OS), progression-free survival (PFS), locoregional control (LRC), and colostomy-free survival (CFS) were 91.1%, 82.3%, 90.8%, and 91.3% in the IMRT cohort and 86.1%, 72.5%, 91.9%, and 93.7% in the 3DCRT group (all P > .1). More patients in the 3DCRT group required a treatment break (11 vs. 4; P = .006), although the difference in median treatment break duration was not significant (12.2 vs. 8.0 days; P = .35). Survival did not differ based on whether a treatment break was needed (all P > .1). Acute grade ≥3 nonhematologic toxicity was decreased in the IMRT cohort (21.1 vs. 59.5%; P < .0001). Acute grade ≥3 skin toxicity was worse in the 3DCRT group (P < .0001), whereas an improvement in late grade ≥3 gastrointestinal (GI) toxicity was observed in the IMRT patients (P = .012).
This study is the largest thus far to compare 3DCRT and IMRT for definitive treatment of anal cancer. Although long-term outcomes did not significantly differ based on RT technique, a marked decrease in adverse effects and the need for a treatment break was achieved with IMRT.
Definitive chemoradiation for squamous cell carcinoma (SCC) of the anal canal results in high cure rates, allowing abdominoperineal resection (APR) and thus permanent colostomy to be reserved for patients who experience a locoregional treatment failure.1–3
Unfortunately, excellent long-term outcomes using 3D conformal radiation therapy (3DCRT) are achieved at the cost of potentially severe adverse effects. For instance, the incidence of grade ≥3 gastrointestinal (GI) and skin toxicity reported in Radiation Therapy Oncology Group (RTOG) 98-11 were 34% and 48%, respectively.4 In truth, such high toxicity rates with nonconformal AP/PA(anteroposterior/posteroanterior) fields are not surprising, given the need to simultaneously cover gross disease and elective pelvic and inguinal lymph nodes while nearby sensitive normal tissues such as small bowel and genitalia may not be able to be significantly spared from higher doses.
The National Comprehensive Cancer Network (NCCN) has recognized that use of AP/PA fields for anal cancer is no longer an appropriate standard of care.5 Multifield techniques are better able to deliver a dose more conformally, thus better sparing normal tissues. Several institutions have shown that intensity-modulated radiation therapy (IMRT) for anal cancer can dramatically lower rates of acute and late severe treatment-related adverse effects while maintaining excellent rates of cure and sphincter preservation.6–12 Although there is substantial evidence that IMRT should be the standard of care for treatment of anal cancer, most of the data are retrospective in nature.7–9,11,12 We previously reviewed our institutional IMRT experience.12 With a median follow-up of 21 months, we observed 2-year locoregional control (LRC), overall survival (OS), distant-metastasis–free survival (DMFS), and colostomy-free survival (CFS) of 94.6%, 100%, 82.6%, 90%, and 94.7%, respectively. Acute grade ≥3 nonhematologic toxicity of 21.1% was comparable to that in reports from other institutions. The improvements in adverse effects seen in these retrospective data were also shown in the prospective phase II trial RTOG 0529. Although the primary end point was not met, this study showed that IMRT can reduce the incidence of severe dermatologic, GI, and hematologic adverse effects.13
IMRT appears to be superior to 3DCRT for anal cancer, but there is a lack of comparative studies.6,11 A retrospective comparative study from Stanford University showed that IMRT resulted in dramatic improvements in treatment duration, the need for treatment breaks, and the length of treatment breaks.11 Moreover, the IMRT patients had significantly higher 3-year OS, LRC, and PFS (all P < .01). Several dosimetric analyses have also described remarkably lower normal tissue doses resulting from IMRT treatment.14,15
Definitive IMRT-based chemoradiation became standard practice at our institution in 2006 for SCC of the anal canal. Because of the limited comparative data between IMRT and older RT techniques, we retrospectively evaluated our comprehensive experience treating anal cancer with these modalities.
After obtaining approval from our institutional review board (IRB), we queried a database maintained by our radiation oncology department for all patients treated for SCC of the anal canal. We then identified those patients who received concurrent chemotherapy and radiation with curative intent.
We have previously detailed our staging and IMRT treatment delivery practice.12 Briefly, initial staging included a thorough history and physical examination including digital rectal examination, routine blood chemistries, human immunodeficiency virus (HIV) status, computed tomography (CT) scan of the abdomen and pelvis, and 18-fluorodeoxyglucose positron emission tomography (PET) scan. Patients also underwent colonoscopy or sigmoidoscopy. Staging was performed according to the Cancer Staging Manual.16 Human papilloma virus (HPV) status was not routinely verified.
IMRT patients were simulated and treated, most commonly, prone on a carbon fiber belly board or in some cases supine with legs abducted in a custom mold. IMRT was delivered in consecutive daily 1.8- to 2-Gy fractions. Our IMRT treatment strategy evolved over time such that patients were initially treated with a sequential boost, and more recently this transitioned to a simultaneous integrated boost (dose painting). All patients were evaluated for a final boost to the gross tumor volume (GTV) up to approximately 60 Gy, depending on the extent of both the clinical tumor response and acute adverse effects. Normal tissue dosage constraints for the IMRT patients were established for small bowel (maximum, 45 Gy; mean <30 Gy), femoral heads (maximum, 45 Gy; mean, <30 Gy), genitalia (maximum, 50 Gy; mean <30 Gy), and bladder (mean, <45 Gy).
3DCRT patients typically underwent CT simulation and treatment in the supine position in an immobilization device. Treatment typically was delivered using AP/PA ports. En face electrons were used to boost the inguinal nodes, and the electron energies were based on nodal depth. The whole pelvis was treated with either 30.6 Gy or 45 Gy in 1.8-Gy fractions based on the treating physician's preference. Patients who received 30.6 Gy to the whole pelvis underwent a lower pelvic 14.4-Gy boost in 1.8-Gy fractions lowering the top field border was lowered to the bottom of the sacroiliac joints. A GTV boost with a 1- to 2-cm margin was delivered for T3, T4, grossly residual T2, or node-positive tumors.
Chemotherapy was delivered concurrently with radiation therapy, typically including 2 cycles of bolus 5-fluorouracil (5-FU) 1000 mg/m2 (days 1–4 and 29–32) and IV bolus mitomycin-C (MMC) 10 mg/m2 (days 1 and 29). A minority of patients was instead treated with concurrent cisplatin 75 mg/m2 (days 1 and 29) and bolus 5-FU.
Acute adverse effects were considered to have occurred within 3 months after treatment completion, whereas late toxicities occurred after at least 3 months. Treatment-related toxicity was graded according to the NCI Common Terminology Criteria for Adverse Events (CTCAE), version 4.0.
After the completion of treatment, routine follow-up consisted of physical examination and imaging at least every 3 to 6 months for 5 years and then annually. Locoregional failure (LRF) was diagnosed with biopsy after a minimum of 3 months after treatment completion. Patients with LRF were considered for APR.
Patient and radiation therapy characteristics were analyzed with the Wilcoxon Rank Sum Test by using the exact method with Monte Carlo estimation. Kaplan-Meier analysis was performed to evaluate the end points of OS, CFS, LRC, DFS, and DMFS. Each end point was calculated from the date of initial biopsy. CFS was defined as the interval until colostomy, LRC was the interval until locoregional failure diagnosis, DFS was the time of any treatment failure or death, and DMFS was the time to distant metastasis diagnosis.
Table 1 demonstrates patient and tumor characteristics. We evaluated a total of 89 patients (52 IMRT, 37 3DCRT) treated between September 2000 and September 2011. Patient and radiation therapy characteristics were similar between the treatment groups, with the exception that patients in the IMRT cohort had a higher percentage of stage III tumors (46.2 vs. 21.6%; P = .011). This difference is possibly explained by fewer 3DCRT patients being staged with PET/CT (40% vs 100%). IMRT patients received a lower median total gross tumor volume (GTV) dose (56 vs. 59.4 Gy; P = .038) and completed treatment more quickly (median, 38.5 vs. 49 days; P < .0001). Seven patients (7.9%; 6 3DCRT, 1 IMRT) received concurrent 5-FU and cisplatin, and 82 (92.1%) received concurrent 5-FU and MMC.
Median follow-up for all patients, IMRT patients alone, and 3DCRT patients alone was 26.5 months (range, 3.5–133.6), 19 months (range, 3.5–125.5), and 61.9 months (range, 7.6–133.6), respectively. Three-year OS, PFS, LRC, and CFS were 91.1%, 82.3%, 90.8%, and 91.3% for the 3DCRT patients and 86.1%, 72.5%, 91.9%, and 93.7% for the IMRT patients (all P > .1; Figure 1).
Multivariate analysis (MVA) revealed age to be the only significant variable related to OS (hazard ratio; HR) 1.060; 95% confidence interval (CI) 1.001–1.122; P = .045) whereas RT technique, sex, T stage, N stage, and treatment break were not (Table 2). None of these variables was significant with respect to LRC, DMFS, or CFS, although there was a trend toward significance for age and CFS (HR, 0.918; 95% CI, 0.832–1.012; P = .085).
Chemoradiation utilizing IMRT was also better tolerated, as evidenced by fewer patients requiring a treatment break (7.7 vs. 29.7%; P = .009). The reason for a treatment break in 4 IMRT patients included moist desquamation, perianal pain, diarrhea, and neutropenia. The IMRT patients needed shorter breaks (mean 8 vs. 12.2 days), but the difference was not statistically significant (P = .583). Three-year OS, PFS, LRC, and CFS also did not differ based on whether a break was necessary (all P > .1).
Acute grade ≥3 nonhematologic toxicity was higher in the 3DCRT group (59.5 vs. 21.1%; P < 0.0001; Table 3), primarily because of a large difference in acute skin toxicity favoring the IMRT group (64.9% vs. 11.5%; P < 0.0001). A trend was observed for improved acute grade ≥3 genitourinary (GU; P = .09) and GI (P = .06) toxicity in the IMRT patients. An improvement in late grade ≥3 GI toxicity was observed in the IMRT patients (P = .01; Table 4). We observed no difference in acute or late grade ≥3 hematologic toxicity (P > .01).
IMRT-based chemoradiation for SCC of the anal canal achieves high rates of disease control and sphincter preservation while dramatically decreasing toxicity when compared to older RT techniques.6–12 The benefit of IMRT over 3DCRT has been described dosimetrically,14 but the only previously published clinical comparison comes from Stanford.11 In that study, IMRT was superior with respect to all end points, including treatment break frequency and duration, treatment-related toxicity, survival, and disease control (all P < .01). Because of the lack of comparative data between these techniques, we reviewed our institutional experience using IMRT- and non-IMRT–based chemoradiation for SCC of the anal canal.
We evaluated 89 consecutive patients (52 IMRT, 37 3DCRT) who underwent definitive chemoradiation at our institution. There are several notable aspects of this study. First, although our IMRT follow-up was short (median, 19 months), our IMRT sample (n = 52) is second only to the sample published by Salama et al (n = 53).7 Although our median follow-up was shorter among the IMRT patients (19 vs. 61.9 months), the rates of acute and late grade ≥3 nonhematologic toxicity seen in our IMRT patients were both significantly lower than among the 3DCRT patients. Second, although our survival outcomes are similar to those in previously published data, only 4 IMRT patients in our study received a treatment break (7.7%) compared with other series in which the rate was as high as 42%. We note that the use of treatment breaks as a measure of treatment-related toxicity is not ideal because of several subjective factors, particularly the physician's individual threshold for recommending a break.11 Last, 81% of our IMRT patients were treated by using a dose-painting technique, whereas most previously published IMRT studies have been performed with a sequential boost. The only previously published dose-painting experience for anal cancer was from Kachnic et al,9 who treated elective nodes with a dose as low as a 1.5-Gy fraction and did not exceed a 1.8-Gy fraction for gross disease.9 We highlight that our dose-painting technique utilized a more aggressive fractionation schema than was used in nearly all previously published IMRT series; we delivered 1.8-Gy fractions to elective nodal regions while simultaneously delivering 2-Gy fractions to gross disease. We are not aware of any other published anal IMRT experience that used 2-Gy fractions, with the exception of the study by Salama et al .7 Furthermore, in that study, 2-Gy fractions were used in only a subset of patients. Despite our dosing schema using 2-Gy fractions, our acute grade ≥3 nonhematologic toxicity rates were similar to those in the Boston study12 (Moffitt/Boston: GI, 9.6/7%; GU, 0/7%; and skin, 11.5/9.3%). Given these promising data, future prospective dose-escalation strategies with simultaneous integrated boost for T4 or bulky nodal tumors may thus be feasible and could be considered in a cooperative group setting.
Our survival outcomes are remarkably similar to previously published IMRT studies for anal cancer, as summarized in Table 5. Our IMRT 3-year OS, LRC, and CFS rates were 86.1%, 91.9%, and 93.7%. The same respective 3-year IMRT outcomes as reported from Stanford, for example, were 88%, 92%, and 91%, respectively. When compared with historic data, the published IMRT data clearly shows a tremendous decrease in the incidence of severe treatment-related toxicity.7–12 For instance, our acute grade ≥3 nonhematologic toxicity rate, which is similar to that in other IMRT studies, was markedly lower than was reported from RTOG 98-11 (21.2% vs. 74%).4
Data have been accruing that support the use of IMRT over 3DCRT for treatment of anal cancer, yet only one retrospective study from Stanford has directly compared these RT techniques. In that study, Bazan et al11 reported a large and statistically significant survival benefit favoring the use of IMRT-based chemoradiation. Our study, which had nearly twice as many total patients (89 vs. 46), did not confirm these results. The reason for the apparent survival benefit was that the 3DCRT cohort had poorer outcomes than did the historical controls. Updated results from RTOG 98-11 (the MMC arm) included 5-year OS, LRC, and PFS of 78.2%, 80%, and 67.7%, respectively, whereas the same 3-year Stanford outcomes were 51.8%, 56.7%, 56.7%, respectively. The 3DCRT patients in the Stanford study also had much worse outcomes than were seen in our study (5-year OS, LRC, and PFS of 80.7%, 90.8%, and 75.2%), which are consistent with RTOG 98-11. Multivariate analysis from our study also showed no difference in any survival end point based on radiation technique (Table 2).
Although the exact reason that Stanford's 3DCRT outcomes were so uncharacteristically poor remains unclear, the investigators did include a higher percentage of stage III tumors (47%) than we (22%) or RTOG 98-11 (31%). Although the Stanford patients had a higher percentage of locally advanced disease, we prescribed a higher median dose (56 Gy vs. 54 Gy) for gross disease. Furthermore, a higher percentage of patients in our study were prescribed at least 55 Gy (73% vs. 29%). Last, although fewer of our 3DCRT patients had node-positive disease (19% vs. 30%), we delivered at least 45 Gy to the upper and lower pelvic nodes in 21 patients (57%) compared with only 2 patients (12%) in the Stanford experience.
Although IMRT does not appear to offer a survival benefit for patients with anal cancer, its primary benefit is clearly a reduction in the incidence of severe treatment-related toxicity, as shown in numerous studies including RTOG 0529.13 This reduction in toxicity can translate into a reduced need for a treatment break, which may potentially affect long-term outcomes.17–19 Several dosimetric analyses have also shown that IMRT results in a decrease in adverse effects.14,15 A French group generated several 3DCRT and IMRT plans for 5 patients with anal cancer to compare dose-volume parameters for normal pelvic structures.14 IMRT significantly reduced the V30 and V40 of the bladder, small bowel, and genitalia. Of interest, there was no significant dosimetric difference between the IMRT plans that used a simultaneous integrated boost (dose painting), which we utilize at our institution, in comparison with a sequential IMRT boost. Regardless, dose-painting allows treatment to be completed more quickly, which may have radiobiological implications in terms of limiting tumor repair and repopulation. The median treatment duration in our study was 38 days compared with 42 days in the Chicago study and 40 days in the Stanford study, both of which used a sequential boost. Chen et al15 also compared anal IMRT and 3DCRT plans. Although PTV coverage was comparable between the different techniques, normal tissue sparing was evident in the IMRT plans. For instance, the median doses to the external genitalia for 2 patients were 55.9% and 63.6% of the prescription dose, compared with 78.4% and 97.7%, using photons with an en face electron boost.
In conclusion, this is the largest clinical comparative study of 3DCRT vs. IMRT for anal cancer. This study did not confirm the previously suggested survival benefit for IMRT. Nonetheless, the ability of IMRT to create highly conformal plans to better avoid normal tissue does not eliminate the incidence of severe toxicity, but undoubtedly decreases its likelihood and duration. IMRT-based chemoradiation should be implemented as a standard of care for definitive treatment of anal cancer based on the strong supporting clinical and dosimetric studies that show a clear reduction in severe normal tissue toxicity compared with conventional RT techniques.
Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.