Dosimetric comparison of different radiation techniques (IMRT vs. 3-dimensional) of the “true” (deep) ano-inguinal lymphatic drainage of anal cancer patients

We selected ten patients with a diagnosis of anal cancer and clinical not involved inguinal nodes who were already treated with primary chemoradiation protocol between 2012 and 2017. We prospectively generated a standardized elective clinical target volume for anal cancer as recommended by RTOG [5] on the original planning CT scan with 3 mm slice thickness for each patient. All patients were in prone position. Subsequently, we calculated two plans for each individual, one for volumetric arc therapy (VMAT) and one 3-dimensional (3D)-plan. All plans were created for a Varian Clinac® DHX linear accelerator (Varian Medical Systems, Palo Alto, CA, USA). Dose prescription for both radiation techniques was 36 Gy (1.8 Gy single dose) to PTV1, which included the primary tumor region (PTR), the elective pelvic lymph nodes and the inguinal nodes and subsequently 14.4 Gy (1.8 Gy single dose) to PTV2, which includes the PTR and the elective pelvic nodes without inguinal lymph nodes. Aim was, to cover the primary tumor site and the pelvic lymph nodes with a total dose of 50.4 Gy and the inguinal lymph nodes with a total dose of 36 Gy (single dose 1.8 Gy). Dose constraints for organs at risk (OAR) (rectum, sigmoid, small bowel, femoral head left & right, penis/scrotum or vagina/vulva, skin of AILD, urinary bladder) orientated on Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC) [21].

For VMAT, regularly 3 arcs in the main plan (PTV1) and 2 arcs for the boost plan (PTV2) (6 or 15 MV) were used. The dose was prescribed to the median of the PTV (ICRU83).

For 3D-radiation-plans, we utilized 6–12 fields with wedges (15–45°). Main fields were planned from posterior and lateral. Segment fields were used to improve dose coverage and dose homogeneity. The 3D planning was done according to the technique used in the past. Analogous to VMAT, the dose was prescribed to the median of the PTV (ICRU83). For both techniques we used Eclipse 13.0 Treatment Planning System (Varian Medical Systems, Palo Alto, CA, USA) for contouring and dose comparison.

We defined the AILD on each patient almost identical to our previous AILD-fluorescence-study (Fig. 1a) [8]. The cranial border of the AILD (the upper ischioanal fossa) was the origin of the levator ani muscle. The caudal demarcation was defined 3 cm below the lower end of the anal canal. On inguinal site, the ventral and medial demarcation was the skin of the medial thigh; dorsal was the connection line of the dorsal edge of the gluteal muscles. The lateral demarcations were the adductor muscles (anal) or the medial edge of the sartorius or iliopsoas adductor muscles (inguinal).

In our previous study we demonstrated, that especially the caudal parts of the AILD were insufficiently covered by the 30 Gy isodose in IMRT-treated patients [8]. In contrast, the cranial parts of the AILD (ischioanal fossa) were adjacent to the anal canal and usually covered by higher doses. The interesting part of the AILD begins where the first subcutaneous connective tissue connection from the area of the anus to the inguinal lymph nodes below the pubic bone exists. This connection is not included into the CTV in various established contouring guidelines [5, 22, 23]. Therefore, we also compared the dose of different subvolumes of the AILD: We divided the AILD into a cranial part (AILDcranial = ischiorectal fossa) and a caudal part (AILDcaudal = first 3 cm below the anal verge). Furthermore, we divided the AILDcaudal into three different levels, each of 1 cm longitudinal extension (Fig. 1b).

We compared absolute (Dmean, Dmedian, D98%, D2%) and relative dose parameters (V10-V50) of the AILD for both radiation techniques. In addition, we analyzed different dose parameters to the AILDcranial, AILDcaudal and Level1, Level2 and Level3, as represented in the dose-volume histogram.

The dose parameters of the AILD hardly differed between the two irradiation techniques (Table 1). Both, the Dmean and the Dmedian of the AILD were significant lower using IMRT compared to 3D-RT (p = 0.008; 9 = 0.017), even if the absolute differences were just 1.8 Gy for Dmean and 3.3 Gy for Dmedian. Ninety-eight percent of this volume was covered by 34 Gy (IMRT) and 37 Gy (3D) (p = 0,075). There were no relevant differences in D98% however, there was a small but significant difference in the D2% (VMAT: 51.8 Gy, 3D-RT: 53.2).

With a subdivision into different regions of interest, the AILD can be examined more precisely (Table 2). The cranial aspect of the AILD (AILDcranial), which largely corresponds to the ischio-rectal fossa, encloses the anal canal and thus the PTV. Unsurprisingly, here the Dmean and Dmedian almost reached the prescription dose of 50.4 Gy regardless of the radiotherapy technique, and 98 % of the cranial volume was covered by 34 Gy (IMRT) and 37 Gy (3D) (p = 0,075). Also the clinical relevant V30Gy reached almost 100 % of the volume.

We chose the V30Gy as an important dose parameter because RTOG98–11 used 30.6 Gy for elective irradiation of the groin in patients with uninvolved inguinal nodes and therefore should be considered as sufficient for treatment of micrometastases. The V30Gy of the AILD was significant higher for 3D-RT (79%) than for IMRT-treated patients (73%). However, in some critical areas of the caudal AILD (Level2), the differences reached more than 10 % (73.4% versus 62.6%).

The direct drainage from the anus to inguinal site is on ventral site of the perineal pouch. Below the pubic symphysis, branches of the pudendal vessels and the first fat tissue components of the ischioanal fossa reach the external genitalia following the adductor muscles (lateral) and the skin (medial) [24]. We chose this area as the first level of our anatomically detailed breakdown of dose distribution (Level1). The anus and inguinal site at this level were covered by the previous PTV and received adequate dose with both radiation techniques. For this critical level, the V30Gy was 85 % with 3D-RT and only 4 % lower (81%) with IMRT. The absolute differences are indeed small and it would be daring to assume that these small differences increase the risk of inguinal recurrences.

Around 1 cm below this level we defined Level2. The differences in dose distribution between the two techniques increased. Figure 2 shows the 30 Gy isodose (color wash) in a selected patient using 3D-RT (2a) or IMRT (2b). The caudal parts of the AILD were properly covered with 3D-RT, whereas in IMRT the V30Gy was more conformal surrounding the PTV, and less volume of the AILD was covered. Ultimately, this region could be the cause of regional recurrences in some patients with high risk tumors treated with IMRT. With immune-fluorescence methods we were previously able to show the AILD to fall about 3 cm below the level of the anal verge. This level (Level3) is the most caudal potential lymphatic drainage of the anal verge and unlikely to contain micrometastases in low risk tumors of the anal canal. Since inguinal recurrences rarely occur 3 cm below the primary tumor, we do not expect this level to be responsible for inguinal relapse in the vast majority of cases. Nevertheless, in high risk anal cancer, especially for big tumors (≥ T3) of the anal verge with a high number of positive inguinal and pelvic lymph nodes at diagnose, this part of the AILD could be of clinical relevance. The change from 3D-RT to IMRT might have an impact regarding inguinal recurrence in those patients.

The most common grade 3 or 4 toxicity in the two big prospective trials (2D- or 3D-techniques) was radio dermatitis. In both trials (RTOG98–11 and ART II) grade ≥ 3 skin toxicity was 48 % in the mitomycin-based arm [1, 2]. Scher et al. summarized 8 IMRT-studies which demonstrated skin toxicity after radiation of anal cancer patients (n = 39–78) [25]. Dose to the inguinal lymph nodes was mainly 45 Gy. Grade ≥ 3 skin toxicity ranged from zero to 42 % (mean: 23%). With IMRT-techniques ≥3, skin toxicity was significantly reduced in RTOG-0529 (23%), whereas in other IMRT-studies it could reach 69 % [15, 26, 27]. Because of the fact that the critical part of the ano-inguinal lymphatic drainage is located subcutaneously on the medial thigh, we would expect an increase in this relevant toxicity. Furthermore, the genito-urinary side effects might increase [28].

In order to ascertain the clinical benefit of inclusion of the AILD into the CTV, the number and pattern of recurrence are important. Overall, the data on IMRT-treated patients and explicit inguinal recurrences and patterns of spread are less well explored. In the RTOG-0529-study, loco regional failure was 13 % after 4 years. The Study did not show any detailed differentiation of the site of failure, though [17]. Only Tomasoa et al. provided detailed data regarding the number and patterns of inguinal recurrence. In a retrospective analysis of 106 patients treated by SIB-IMRT, four patients had inguinal relapse (4%) [15]. Three out of these four patients had initially involved inguinal nodes. A total dose to lymph nodes of 49.5 à 1.5 Gy was given, which means a higher biological dose than in conventional techniques. The authors (Tomasoa et al.) compared this to similar results of Wright et al. (3-field technique) and mentioned that more conformal techniques (IMRT) with higher total dose might not reduce inguinal relapse [15, 29]. This indicates that microscopic disease in the AILD but not the lymph nodes themselves might be responsible for inguinal relapse. Another important finding of these studies was that none of inguinal recurrences occurred below approximately 1 cm of the level of the anal canal. Wright et al. however, could show that inguinal recurrences arise up to 3 cm below the anal verge (conventional technique) [29]. Up to this date, insufficient data concerning inguinal recurrence in anal carcinoma is available. However, there are a few cases of inguinal relapse even in inguinal-treated patients. To balance the relevant skin toxicity with the risk of inguinal relapse in high risk tumors (T3-T4, inguinal involvement at diagnosis, anal verge), an inclusion of the AILD into the target volume could be useful.