MRI-guided adaptive radiotherapy in locally advanced cervical cancer from a Nordic perspective

Figures & data

Table II. Treatment characteristics for the IGABT and NOCEAC cohorts of patients with locally advanced cervical cancer treated at Aarhus University Hospital.

Cohort	IGABT	NOCECA	p-value
EBRT target
Pelvis	122 (87%)	99 (100%)	< 0.001
Pelvis and para-aortic region	18 (13%)	0
EBRT technique
3D conformal	86 (61%)	99 (100%)	< 0.001
IMRT	54 (39%)	0
EBRT dose and fractionation
Type 1: 50/45 Gy in 25 fx	44 (31%)	47 (48%)	< 0.001
Type 2: 50/45 Gy in 25 fx + 10 Gy/5 fx	10 (7%)	42 (42%)
Type 3: 50/56 Gy in 28 fx + 24 Gy/12 fx	0	10 (10%)
45 Gy/25 fx (no nodes)	40 (29%)	0
50 Gy/30 fx (nodal SIB 60 Gy/30 fx)	46 (33%)	0
EBRT dose
Whole pelvis, Gy (mean, SD)	46 ± 2	45 ± 4	0.11
Central pelvis, Gy (mean, SD)	49 ± 4	56 ± 9	< 0.001
BT prescription and imaging
Point A, x-ray	0	89 (100%)*	< 0.001
HR CTV, CT	3 (2%)	0
HR CTV, MRI	137 (98%)	0
BT source and applicator
Caesium tube and box, IC	0	43 (48%)	< 0.001
Ir192 (PDR), IC	80 (57%)	46 (52%)
Ir192 (PDR), IC/IS	60 (43%)	0
Chemotherapy
Concomitant	110 (79%)	0	< 0.001
Neoadjuvant	10 (7%)	0	0.007
Overall treatment time
EBRT+ BT, days (mean, range)	47 (36–70)	44 (34–63)	< 0.001

*10 NOCECA patients did not receive BT (EBRT Type 3).

Table I. Patient and tumour characteristics for two cohorts treated with either MRI-guided IGABT or x-ray-based BT (NOCECA) at Aarhus University Hospital for locally advanced cervical cancer.

Cohort	IGABT	NOCECA	p-value
Number of patients	140	99
Follow-up, years (mean, range)	3.0 (0.5–6.5)	2.9 (0.1–5.4)	0.59
Age, years (mean, range)	56 (27–84)	61 (28–80)	0.34
FIGO (%)			< 0.001
IB	14 (10%)	0
IIA	6 (4%)	0
IIB	79 (56%)	55 (56%)
IIIA	8 (6%)	1 (1%)
IIIB	27 (20%)	42 (42%)
IVA	6 (4%)	1 (1%)
Tumour size, cm (mean, range)	6 (2–11)	7 (2–12)	< 0.001
Histology (%)			0.35
Squamous cell	116 (83%)	88 (89%)
Adenocarcinoma	17 (12%)	9 (9%)
Adeno-squamous	7 (5%)	2 (2%)
Pelvic nodes	58 (41%)	NA	–
Para-aortic nodes	12 (9%)	0	0.003

Figure 1. Cumulative dose (EQD2) of EBRT and IGABT in terms of D2 cm³ to the sigmoid plotted as a function of the D90 to the HR CTV. DVH parameters for a standard point A-based plan for the first 70 consecutive patients (2005–2008) is shown in panel A. Optimised DVH parameters actually used for treatment in all 140 patients are given in panel B divided into the first 70 patients treated 2005–2008 (grey circles) and the last 70 consecutive patients treated 2009–2011 (white circles). The planning aim of 85 Gy to D90 of HR CTV and the DVH constraints of first 75 and later 70 Gy are indicated by dotted lines.

Table III. DVH parameters for EBRT and BT calculated in EQD2 (mean, range) for 140 consecutive patients treated with IGABT in the period 2005–2011 at Aarhus University Hospital. The DVH parameters for the 70 first and 70 last consecutively treated patients are given in separate columns. In addition, standard point A-based DVH parameters are given for the first 70 patients treated 2005–2008. Number of patients fulfilling the planning aim (HR CTV D90 > 85 Gy, D2 cm³ bladder < 90 Gy and D2 cm³ rectum and sigmoid < 75 Gy) are also indicated.

Treatment period	2005–2008	2005–2008	2009–2011	p-value
	standard	optimised	optimised	first vs. last
HR CTV D90 (Gy)	92 (62–132)	91 (72–107)	91 (69–102)	0.82
IR CTV D90 (Gy)	–	69 (60–78)	67 (60–73)	0.037
Bladder ICRU point (Gy)	–	67 (49–91)	65 (48–91)	0.22
Bladder D2 cm^3 (Gy)	79 (55–177)	73 (56–89)	69 (52–82)	< 0.001
Rectum ICRU point (Gy)	–	67 (55–90)	64 (52–83)	0.011
Rectum D2 cm^3 (Gy)	68 (53–109)	66 (53–77)	62 (51–75)	< 0.001
Sigmoid D2 cm^3 (Gy)	73 (52–107)	69 (53–78)	62 (49–74)	< 0.001
Planning aim obtained (%)	15/70 (21%)	56/70 (80%)	65/70 (93%)	0.026

Figure 2. D90 of HR CTV shown as a function of HR CTV volume in 140 consecutive patients treated with MRI-based IGABT for locally advanced cervical cancer. Local failure was observed in 11 patients (black spot). The planning aim for D90 of HR CTV and the median D90 and volume of HR CTV are indicated by dotted lines.

Figure 3. Actuarial local control, overall survival and ≥ grade 3 combined urological-gastrointestinal morbidity in 140 patients treated with IGABT (black lines). For comparison the curves for overall survival and morbidity in 99 patients treated with 2D x-ray-based brachytherapy (NOCECA) are indicated (grey lines). Patient number at risk for overall survival is indicated below the x-axis.

Table IV. Outcome in terms of actuarial three-year disease control, survival and morbidity in 140 consecutive patients treated with MRI-guided IGABT compared with 99 patients treated with x-ray-based BT (NOCECA).

Cohort	IGABT	NOCECA	p-value	Hazard ratio
Local control	91%	–	–	–
Pelvic control	85%	76%	0.12	1.6 (0.9–2.8)
Cancer-specific survival	87%	68%	0.001	5.0 (2.8–8.9)
Overall survival	79%	63%	0.005	1.8 (1.2–2.8)
Overall survival IIB–IV	77%	63%	0.01	1.7 (1.1–2.6)
G2 ≥ urological	17%	18%	0.29	1.4 (0.7–2.7)
G2 ≥ gastrointestinal	18%	35%	< 0.001	2.9 (1.6–5.2)
G2 ≥ vaginal	33%	87%	< 0.001	4.8 (2.7–8.4)
G2 ≥ overall	55%	90%	< 0.001	4.3 (2.9–6.4)
G3 ≥ urological	1%	2%	0.23	3.6 (0.5–26.0)
G3 ≥ gastrointestinal	3%	8%	0.08	3.8 (0.9–15.5)
G3 ≥ vaginal	4%	9%	0.08	2.8 (0.9–8.7)
G3 ≥ overall	7%	15%	0.02	3.0 (1.2–7.3)