Long-term local control and survival after preoperative radiochemotherapy in combination with deep regional hyperthermia in locally advanced rectal cancer

Abstract

Purpose: The aim of this study was to evaluate the impact of deep regional hyperthermia on long-term local control and survival in locally advanced non-metastatic rectal cancer. Methods: In total 103 patients with locally advanced non-metastatic rectal cancer were treated preoperatively with either neoadjuvant radiochemotherapy alone (n = 43) or the same treatment with additional deep regional hyperthermia (n = 60). The two groups were compared with respect to local control, overall survival (OS), disease-free survival (DFS), and distant metastases-free survival (DMFS). Results: Patients receiving additional hyperthermia had excellent long-term local control with a 5-year Kaplan-Meier estimate of 98% compared with 87% in the radiochemotherapy only group (p = 0.09). Five-year rates for OS (88% versus 76%, p = 0.08), DFS (77% versus 73%, p = n.s.) and DMFS (75% versus 77%, p = n.s.) were not statistically different between the two groups. Conclusion: Radiochemotherapy combined with hyperthermia results in excellent long-term local control.

• Chemoradiotherapy
• disease-free survival
• hyperthermia
• local control
• overall survival
• radiochemotherapy
• rectal cancer

Introduction

Preoperative radiochemotherapy (RCT) is standard of care for locally advanced non-metastatic rectal cancer. The combination of total mesorectal excision in combination with preoperative RCT yields local control rates higher than 90% [1,2]. Yet low lying tumours, distant failures and local recurrences still remain a challenge in the multimodal treatment of rectal cancer. Deep regional or superficial hyperthermia has been shown to improve treatment outcomes in a variety of malignant diseases when combined with radiotherapy [3–5]. For instance, a randomised controlled phase III study showed a significant benefit in overall survival (OS) for the combination of hyperthermia with radiotherapy compared with radiotherapy alone in the treatment of cervical cancer [4]. Several mechanisms of action by which hyperthermia augments irradiation have been investigated. These mechanisms include radiosensitising effects by inhibition of DNA repair, immuno-stimulatory effects, increased perfusion, and therefore improved oxygenation of hypoxic tissues and direct cell death [6–9]. Yet clinical data for the combination of hyperthermia with RCT in rectal cancer is rare [10–12]. In a previous report we have shown higher rates of pathological complete responses and sphincter-sparing surgeries in patients who were treated with preoperative radiochemotherapy plus hyperthermia (HRCT) [13]. We now report follow-up data on patients treated with this quadrimodal treatment regimen consisting of radiochemotherapy, hyperthermia and total mesorectal excision. Furthermore this intensified treatment regimen is compared to the current standard of care (RCT).

Methods

Patients

In this non-randomised retrospective study a total of 103 patients with UICC Stage II or III adenocarcinoma of the middle or lower rectum were treated with preoperative RCT either with or without concomitant regional hyperthermia at the Department of Radiation Oncology, University Hospital of Tübingen, Germany, between 2007 and 2010. Patient and tumour characteristics are summarised in Table 1. Standard pre-therapeutic staging procedures included rectoscopy, abdominal CT and/or MRI and chest imaging (CT or conventional X-ray).

Table 1. Patient characteristics.

	RCT group	HRCT group	p
Number of patients	43	60
Age			0.052
Mean (years)	68.56	64.33
Range (years)	43–87	38–89
Sex			n.s.
Male	29 (67.4%)	34 (56.7%)
Female	14 (32.6%)	26 (43.3%)
Tumour distance from anal verge			0.007
Mean (cm)	7.62	5.89
0–5 cm	13 (30.2%)	28 (46.7%)	0.04
6–10 cm	23 (53.5%)	30 (50%)
11–15 cm	7 (16.3%)	2 (3.3%)
Clinical stage			n.s.
cT3 cN0 cM0	7 (16.3%)	10 (16.7%)
cT3 cN+ cM0	29 (67.4%)	45 (75%)
cT4 cN0 cM0	1 (2.3%)	1 (1.7%)
cT4 cN+ cM0	6 (14%)	4 (6.7%)
Tumour grade			n.s.
G1	3 (7%)	8 (13.3%)
G2	36 (79.1%)	46 (76.7%)
G3	6 (14%)	6 (10%)

RCT, radiochemotherapy; HRCT, radiochemotherapy plus hyperthermia; n.s., not significant.

Hyperthermia was contraindicated in the presence of metal implants, high grade cardiac co-morbidities, or actively bleeding tumours.

Radiochemotherapy

Linac-based normo-fractionated RCT was applied in 1.8 Gy fractions to a cumulative dose of 50.4 Gy. Radiotherapy was delivered as conventional three-dimensional three-field radiotherapy or intensity-modulated radiotherapy (IMRT). 5-Fluorouracil (1000 mg/m² body surface/day) was given as a continuous venous infusion over 120 h during the first and fifth week of treatment. Four cycles of post-operative 5-fluorouracil-based chemotherapy were recommended to all patients after complete resection.

Hyperthermia

Hyperthermia was delivered once or twice weekly with a Sigma Eye or Sigma-60 applicator (BSD 2000/3D, BSD Medical Systems, Salt Lake City, UT, USA). European Society for Hyperthermic Oncology guidelines for quality assurance were followed. Continuous temperature measurements in the bladder, rectum and vagina were performed with endoluminal catheters (Bowman Probes, BSD Medical Systems). Target temperature was 40.5 °C for at least 60 min and the maximum treatment duration per hyperthermia session was 90 min. T90 and CEMT43 were calculated as described before [13]. A water bolus was used for cooling and energy coupling.

End points and statistical analysis

This retrospective study was performed to evaluate the impact of hyperthermia on long-term local control, OS, disease-free survival (DFS), and distant metastases-free survival (DMFS). Follow-up data was obtained from departmental patient records. Statistical analysis was performed using commercial software (SPSS 19, IBM, Armonk, NY). Estimated survival was determined according the Kaplan-Meier method, survival curves were compared using the log-rank test. A multivariate Cox regression was used to determine independent prognosticators for survival parameters.

Results

Patient characteristics

Pretreatment patient characteristics are shown in Table 1. Mean age at diagnosis for the entire cohort was 66 years. Forty-three patients received multimodal treatment with RCT, 60 patients were treated with the same regimen plus deep regional hyperthermia. No significant difference existed in patient characteristics with regard to age, gender, pretreatment tumour stage or grade. In the HRCT group tumours were located closer to the anal verge.

Treatment related parameters

Hyperthermia-related parameters are shown in Table 2. Mean T90 and CEMT43 were 39.3 °C and 1.1 min respectively. A median of four hyperthermia treatments was applied. All 103 patients received radiotherapy as planned; four of 43 patients (9.3%) in the RCT group did not receive the second cycle of concomitant chemotherapy, while three of 60 patients (5%) in the HRCT group required dose reduction of the second cycle to 75%.

Table 2. Hyperthermia related parameters.

	Min	Max	Median
CEMT43 (min)	0.0	9.2	1.1
T90 (°C)	37.1	40.6	39.3
Number of treatments	1.0	9.0	4.0
Treatment time (min)	9.0	636.0	247.0

Survival data

Median follow-up for the entire cohort was 4.9 years and 5.1 years for patients alive. For the entire cohort estimated 5-year rates for OS, DFS, local control and DMFS were 83%, 75%, 93% and 76% respectively. Patients in the HRCT group had a trend towards longer OS (88% versus 76% after 5 years, p = 0.08) and local recurrence-free survival (98% versus 87% at 5 years, p = 0.09). No impact of hyperthermia on DFS and DMFS was observed (Figure 1). A total of five local recurrences occurred during follow-up, one in the HRCT group and four in the RCT group. Four of the five patients with local recurrences were poor responders to preoperative radiochemotherapy (ypT3 or ypT4) or had synchronous distant metastases. In a univariate analysis post-operative nodal stage (ypN0 versus ypN+) was a significant prognosticator for OS (89% versus 68% after 5 years, p = 0.03), DFS (84% versus 51% after 5 years, p < 0.01) and DMFS (84% versus 54% after 5 years, p < 0.01). Post-operative T-stage (ypT0/ypT1 versus ypT2/ypT3/ypT4) was prognostic for longer DFS (95% versus 69% after 5 years, p = 0.02) and DMFS (95% versus 70% after 5 years, p = 0.03). Patients younger than 60 years had significantly worse local control (83% versus 98% after 5 years, p = 0.01). Local control after five years was 100% for patients who had achieved a pathological complete response. Hyperthermia-related parameters including the number of hyperthermia sessions were not predictive for OS, DFS, LC, or DMFS, Table 3. In a multivariate analysis post-operative nodal stage remained a significant prognosticator for OS, DFS and DMFS (Table 4).

Figure 1. (a) Overall survival, (b) disease-free survival, (c) local control, and (d) distant metastases-free survival after preoperative radiochemotherapy alone (dashed line) or in combination with deep regional hyperthermia (solid line).

Table 3. Five-year survival estimates and univariate analysis.

	Total N	5 years OS	p	5 years DFS	p	5 years LC	p	5 years DMFS	p
Treatment
RCT	43	0.76	0.17	0.73	0.95	0.87	0.22	0.75	0.98
HRCT (1–3 treatments)	60	0.94		0.74		1.00		0.74
HRCT (4 or more treatments)	60	0.85		0.78		0.97		0.78
Sex
Female	40	0.86	0.54	0.73	0.54	0.94	0.97	0.75	0.74
Male	63	0.82		0.76		0.96		0.76
Age
<60 years	28	0.85	0.75	0.65	0.16	0.83	0.01	0.67	0.29
>60 years	75	0.83		0.80		0.98		0.80
<70 years	55	0.90	0.04	0.75	0.96	0.91	0.29	0.77	0.80
>70 years	48	0.75		0.75		0.97		0.75
Location of tumour
0–5 cm	41	0.84	0.90	0.75	0.77	0.93	0.78	0.77	0.79
6–10 cm	53	0.83		0.77		0.95		0.77
11–15 cm	9	0.78		0.67		0.89		0.67
T-category
T3	91	0.85	0.37	0.77	0.15	0.94	0.35	0.77	0.40
T4	12	0.70		0.62		0.89		0.70
N-category
cN0	19	0.72	0.10	0.79	0.61	0.92	0.85	0.79	0.69
cN+	84	0.86		0.74		0.94		0.75
Post-operative T-category
ypT0 or ypT1	22	0.95	0.13	0.95	0.02	1.00	0.18	0.95	0.03
ypT2 or ypT2 or ypT3	81	0.80		0.69		0.91		0.70
Post-operative N-category
ypN0	75	0.89	0.03	0.84	<0.01	0.97	0.09	0.84	<0.01
ypN+	28	0.68		0.51		0.84		0.54
pCR
No	90	0.82	0.43	0.72	0.17	0.92	0.34	0.73	0.18
Yes	13	0.92		0.92		1.00		0.92
Tumour grade
G1	12	0.81	0.88	0.61	0.30	1.00	0.39	0.61	0.26
G2 or G3	91	0.83		0.77		0.93		0.78
CEM43
<1 or no hyperthermia	88	0.83	0.79	0.74	0.45	0.92	0.35	0.75	0.49
≥1	15	0.86		0.84		1.00		0.84
T90
<39 °C or no hyperthermia	58	0.79	0.12	0.71	0.43	0.90	0.26	0.73	0.54
≥39 °C	45	0.89		0.80		0.97		0.80
<40 °C or no hyperthermia	94	0.82	0.22	0.74	0.49	0.93	0.48	0.75	0.52
≥40 °C	9	1.00		0.89		1.00		0.89

OS, overall survival; DFS, disease-free survival; LC local control; DMFS, distant metastases-free survival.

Table 4. Multivariate analysis.

	HRCT for OS	p	HRCT for DFS	p	HRCT for LC	p	HRCT for DMFS	p
Age
<60 years vs >60 years					7.19	0.09
<70 years vs >70 years	0.32	0.04
Post-operative T-category
ypT0/ypT1 vs ypT2/ypT3/ypT4			0.15	0.06			0.15	0.07
Post-operative N-category
ypN0 vs ypN+	0.33	0.03	0.27	<0.01			0.29	0.01

HRCT, hyperthermia plus radio radiochemotherapy; OS, overall survival; DFS, disease-free survival; LC local control; DMFS, distant metastases-free survival.

Discussion

Our group had previously reported local response rates and toxicities after trimodal preoperative treatment consisting of radiochemotherapy and deep regional hyperthermia for locally advanced rectal cancer [13]. In that report, pathological complete response rates were significantly higher in patients who had received at least four hyperthermia treatments compared with patients treated with radiochemotherapy alone or less than four hyperthermia sessions. As there is still a debate about the predictive value of pCR we sought to evaluate long-term oncological parameters and patterns of recurrence in our cohort [14,15]. Five-year estimates for OS and DFS of 83% and 75% are slightly superior to previously published data from prospective trials that did not use hyperthermia [1,16]. However, on subgroup analyses we did not see a statistical effect of hyperthermia on OS, DFS, DMFS, or local control. Yet there was a trend towards a higher local control rate in the hyperthermia group with only one local recurrence (1.6%). Due to the small number of events the study was likely underpowered for detecting a difference at a statistically significant level. On the other hand we cannot rule out the possibility that asymptomatic local or distant recurrences remained undetected since routine follow-up was performed according to the German colorectal cancer guidelines which do not recommend cross-sectional imaging on a regular basis. In our previous report we had observed a correlation between the number of hyperthermia treatments and the pCR rate while in the current study a higher number of hyperthermia treatments was not prognostic for superior outcome regarding OS, DFS, DMFS and LC. Again, the low number of local recurrences does not allow the statistical detection of prognosticators, particularly in the subset of patients treated with hyperthermia. Similar results have been reported in another prospective study of combined radiochemotherapy and hyperthermia. In that study of 37 patients with locally advanced primary rectal cancer thermal parameters correlated with response to treatment but not with improved OS or DFS [17]. Our data is also in line with a report from a Korean group that failed to show any impact on long-term outcome parameters for rectal cancer from hyperthermia added to preoperative radiochemotherapy [18].

A recent prospective phase II study applied a quadrimodal preoperative regimen for locally advanced fixed (T4) rectal cancers. Treatment consisted of hypofractionated radiotherapy with concomitant capecitabine and oxaliplatin, rectal metronidazole and capacitative hyperthermia with a target temperature of 41–45 °C. The authors report a complete resection rate of 92% and a local failure rate of 13.6% after a median follow-up of 24.9 months [19]. These numbers are superior to data reported before for this population with a very high risk for local failure [20]. Unfortunately the concomitant use of several potential radiosensitisers with an experimental hypofractionation scheme does not allow the impact of hyperthermia to be determined in that trial. Studies comparing survival end points after radiochemotherapy and resection for rectal cancer with or without hyperthermia are rare. A meta-analysis by the Cochrane Collaboration included six randomised controlled trials that compared preoperative radiotherapy with preoperative radiotherapy (without chemotherapy) combined with deep regional hyperthermia in locally advanced or recurrent rectal cancer [21]. Besides a significantly higher pathological response rate the authors report a benefit in OS after two years.

However, this survival was no longer present after three, four and five years. The circumstance that until now no survival benefit can be detected by the addition of hyperthermia to radiochemotherapy in locally advanced rectal cancer is likely to be explained by the fact that most treatment failures occur outside the pelvis.

In our cohort, post-operative nodal stage was the strongest independent predictor of DFS, a finding consistent with the literature [22]. As with all retrospective studies our study is prone to biases. For instance, the trend towards longer OS seen in the hyperthermia group is possibly linked with the younger age in this group and the non-randomised treatment allocation.

Despite missing evidence for improvements in DFS and DMFS by the addition of hyperthermia, further investigations in this field are warranted: First, and as shown before, hyperthermia can lead to higher tumour regression and downsizing which in turn may allow for sphincter preservation in patients with low lying tumours [13,21]. Second, in our cohort the intensified local treatment translated into excellent long-term local control. Third, additional hyperthermia is a promising approach in recurrent rectal cancer when re-irradiation doses are limited by previous radiotherapy [23,24]. The German HT01 trial is a single institution prospective phase II trial applying ten sessions of deep regional hyperthermia in combination with 5-FU-based radiochemotherapy according to a previously published protocol [25]. Recently results of the German CAO/AIO/ARO-04 trial were presented. For the first time a higher rate of 3-year DFS could be shown by the addition of oxaliplatin to preoperative radiochemotherapy in rectal cancer in a randomised controlled phase III trial [26]. The concept of incorporating oxaliplatin into preoperative radiochemotherapy regimens has been adopted by the HYREC trial. In this prospective multicentre phase II trial a preoperative 5-fluorouracil and oxaliplatin-based radiochemotherapy regimen similar to the one in the experimental arm of the CAO/AIO/ARO-04 trial is combined with 10 sessions of deep regional hyperthermia for patients with locally advanced or recurrent rectal cancer [27]. So far there is no predictive marker that could select patients who are likely to benefit from hyperthermia. One of the possible mechanisms by which hyperthermia enhances the effects of radiochemotherapy is increased perfusion, and therefore improved oxygenation of hypoxic tissues [28]. Hypoxia is one of the major mechanisms of radioresistance and can be identified by PET tracers, e.g. F-MISO [29]. Therefore hyperthermia might be particularly effective in tumours with such hypoxia areas. However, this hypothesis needs has to be investigated in a prospective manner first.

Conclusion

Deep regional hyperthermia in combination with preoperative radiochemotherapy results in excellent long-term local control. Clinical trials to confirm this observation in a prospective manner are underway.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.