Incomplete thermal ablation stimulates proliferation of residual renal carcinoma cells in a translational murine model

Authors


Judith Jans, Department of Urology, University Medical Center Utrecht, Stratenum, STR02.118, Postbox 80030, 3508 TA Utrecht, the Netherlands. e-mail: j.j.m.jans@umcutrecht.nl

Abstract

What's known on the subject? and What does the study add?

Thermal ablation influences the local tissue microenvironment. Several studies have reported that residual tumour cells may exhibit a more aggressive phenotype.

This study shows that incomplete CA and RFA cause an increased proliferation and decreased apptosis of residual renal tumour cells. This may be caused by stimulatory factors such as hypoxia, HSPs and inflammatory cells.

OBJECTIVE

  • • To compare the effect of incomplete thermal ablation vs partial nephrectomy (PN) on growth stimulation and cellular survival in renal tumours.

MATERIALS AND METHODS

  • • Renca renal tumours were transplanted under the renal capsule of mice (four to six mice/group) after which incomplete radiofrequency ablation (RFA), cryoablation (CA) or PN was performed.
  • • At several time points after treatment, presence of cell proliferation, apoptosis, hypoxic areas, inflammatory factors and the heat-shock proteins (HSPs) 70 and 90 were evaluated using immunohistochemistry.

RESULTS

  • • At 2 h after thermal ablation residual tumour cells showed increased proliferation. This hyperproliferation was significantly stronger after RFA than CA (P < 0.05) and not present after PN.
  • • Residual cells showed increased apoptosis after 2 h and decreased apoptosis from 2 days after thermal ablation. Apoptotic cells were significantly less evident at 3 days after RFA (P < 0.001).
  • • Hypoxic areas and HSPs were increasingly present from 2 h up to 7 days after thermal ablation (P < 0.001).
  • • Inflammatory cells infiltrated mainly the necrotic areas after thermal ablation, and their abundance peaked at 1 week after ablation (P < 0.05).
  • • The increased cell growth was preceded by hypoxia and presence of HSPs.

CONCLUSIONS

  • • CA and RFA result in an increased proliferation and decreased apoptosis of residual renal tumour cells.
  • • This hyperproliferation may be caused by stimulatory factors, e.g. hypoxia, HSPs and inflammatory cells, and could facilitate recurrences of renal tumours after thermal ablation.
  • • This study highlights the importance of achieving complete tumour destruction.
Abbreviations
CA

cryoablation

HSP

heat-shock protein

PN

partial nephrectomy

RFA

radiofrequency ablation

INTRODUCTION

The ‘gold standard’ of treatment of small renal masses (<4 cm) is partial nephrectomy (PN) [1]. Triggered by the increasing incidence of small renal masses and in search of options for patients unfit for open or laparoscopic surgery, minimally invasive nephron-sparing therapies have been developed. Both radiofrequency ablation (RFA) and cryoablation (CA) induce necrosis and apoptosis of tumour cells. Successful ablation is currently defined as the absence of enhancement on follow-up imaging with or without a negative biopsy of the ablated lesion, which does not increase in size over time [2]. Local recurrences after ablation continue to be a problem, particularly in larger tumours. Optimal temperatures for RFA are between 50 and 100 °C and for CA < −20 °C [3]. The thermal energy is conducted from the tip of a probe to the surrounding tissue. Thermal energy conduction is inversely related to distance; therefore, cells at the edge of the lesion may undergo sub-lethal injury. RFA seems to require more re-ablations to achieve treatment success, although a selection bias may be present [3].

Thermal ablation influences the local tissue microenvironment. Several studies have reported that residual tumour cells may exhibit a more aggressive phenotype [4–8]. Factors associated with this phenomenon are local expression of cytokines, heat-shock proteins (HSPs), hypoxia and influx of inflammatory cells. Hypoxia at areas surrounding the ablated region appears to be an important consequence of thermal ablation [8], as it leads to activation of signalling pathways that allow tumour cells to undergo adaptive changes promoting survival [9].

Although one aims for complete ablation, this is not always achieved. Therefore, clarifying the cellular effects of thermal ablation vs PN on residual tumour cells is important and may improve the techniques of CA, RFA and PN to prevent an escape from cell death. The aim of the present study was to examine renal tumour cell proliferation, apoptosis, and factors important for cellular survival after CA, RFA and PN, in an orthotopic murine model for renal cancer.

MATERIALS AND METHODS

For the cell cultures, Balb/C renal carcinoma cells (Renca, National Cancer Institute, USA) were maintained at 37 °C and 5% CO2 in Dulbecco's Modified Eagle's Medium (Lonza, the Netherlands) supplemented with 10% fetal calf serum, 1% penicillin/streptomycin and 1% glutamine.

All experiments were conducted in agreement with the Netherlands Experiments on Animals Act and European convention guidelines. Surgical procedures were performed under isoflurane anaesthesia. Buprenorfine was used s.c. for analgesia before, and 24 h after surgery. In 103 male Balb/C mice (aged 9–11 weeks, Charles River, the Netherlands), a Renca tumour cube with a diameter of 2 mm was transplanted under the renal capsule [10]. After 1 week incomplete RFA, CA or PN was performed (Fig. 1). Access to the left kidney was obtained through a flank incision. For RFA treatment a 19-G non-cooled bipolar RFA probe with an active length of 1 cm (CelonProSurge micro, Celon AG, Germany) was inserted in the renal tumour. RFA was performed at 2 W, 120 J (60 s), using the CELON Power System (Celon AG, Germany).

Figure 1.

Schematic drawing of the preclinical murine renal tumour model. A, Incomplete CA and RFA of renal tumours. B, Incomplete PN of renal tumours.

CA was performed with a 17-G probe with a built-in thermocouple (IceSeed, Galil Medical, Israel). Tumour treatment consisted of two cycles of 20 s freeze followed by 30 s thaw [11]. During CA and RFA the tumour was exteriorized while the abdominal cavity and skin were protected with gauze. For PN the kidney was mobilised, the tumour was partly excised to a level of near complete resection, followed by a 1-min tamponade for haemostasis. Control mice were sham-operated by inserting a probe in the tumour without performing ablation (sham), or received no treatment (control). To enable detection of hypoxia, pimonidazole hydrochloride (Hypoxyprobe-1, 90201, HPI Inc., USA) was injected i.v. at a dose of 60 mg/kg 1 h before termination. At several time points (2 h and 1, 3, 7 and 14 days) after treatment mice were killed humanely and the kidneys were harvested.

For immunohistochemical analysis, formalin-fixed paraffin-embedded tissue was cut in 5-µm sections. Antigen retrieval was achieved by boiling in citrate buffer (pH 6). Immunohistochemistry was performed to detect proliferation (Ki67 clone SP6, Neomarkers, USA), apoptosis (Casp3 clone C92-605, BD Pharmingen, USA), hypoxia (pimonidazole adducts, Pab2627, HPI Inc, USA), macrophages (F4/80, Serotec, USA), leukocytes (anti-CD45, Abcam, USA) and HSP70 and HSP90 (Santa Cruz, USA). Brightvision (Immunologic, the Netherlands) was used as secondary antibody, followed by diaminobenzidine/H2O2. For Ki67, Casp3, CD45 and F4/80, 5–13 fields at the direct border of the lesion were randomly selected at x20. Subsequently, positive cells were scored by one observer ‘blinded’ to treatment using a grid overlay. Hypoxia, HSP70 and HSP90 were analysed using a semi-quantitative method as previously described [12]. Images of all fields directly surrounding the lesion in one representative slide per tumour (6–13 fields/slide) were selected at x10. The percentage of positive areas in each field was calculated with ImageJ software (National Institute of Health, USA).

RESULTS

CA AND RFA STIMULATE TUMOUR CELL PROLIFERATION AND DECREASE APOPTOSIS

In untreated tumours the mean (sem) proliferation rate was 28.3 (6.3)% (Fig. 2). Ablation resulted in a two-fold higher proliferation rate of cells at the border of the ablated region from 2 h after thermal ablation (P < 0.001), peaking at 3 days (P < 0.001). Proliferation rates remained high up to 2 weeks after thermal ablation, at a mean (sem) of 40.4 (1.2)% after RFA and 48.2 (1.1) after CA (P < 0.001). Importantly, RFA induced significantly more proliferation than CA, at a mean (sem) of 66.3 (1.2)% after RFA vs 57.5 (1.2)% after CA (P < 0.001). PN showed a slight increase in proliferation at the resection site 1 day after treatment (P= 0.003), after which proliferation levels decreased to baseline levels. Sham treatment did not affect proliferation.

Figure 2.

A, Cell proliferation at the border of thermally ablated or partially resected renal tumours (Ki67, mean ±sem, n= 4–6 mice/group). Cell proliferation at this region significantly increased from 2 h after thermal ablation (P < 0.001), with a two-fold increased proliferation rate at 3 days (P < 0.001). RFA induced significantly more proliferation than CA (P < 0.001). Proliferation rates remained 1.5-fold increased 14 days after treatment. B, Representative images of Ki67 staining at the peak of proliferation rates 3 days after treatment (×20). Brown nuclear staining indicates proliferation. C, Apoptotic cells at the border of thermally ablated or partially resected renal tumours (Casp3, mean ±sem, n= 4–6 mice/group). Apoptosis at this region showed a trend of decrease after 24 h, only RFA showed a significant decrease of apoptotic cells 3 days after ablation (P < 0.001). Within the ablated areas apoptotic cells were frequently present until 7 days after thermal ablation. D, Representative images of Casp3 staining at 3 days after treatment (×20). Brown nuclear and cytoplasmatic staining indicates apoptosis. Ctrl, control. scale bars 204 µm.

Apoptotic cells were not abundantly present in untreated tumours with a mean (sem) of 2.4 (0.35)% (Fig. 2). At 2 h after thermal ablation there was a slight increase in apoptotic cells at the border of the ablated region (P < 0.001). However, at 2 days apoptosis at this area was decreasing. After 3 days there was a significant decrease (P < 0.01) in apoptotic cells surrounding the RFA-treated area. While CA also showed a trend towards decreased presence of apoptotic cells, this difference was not significant. Up to 7 days after thermal ablation there was a plenitude of apoptotic cells within the ablated area (Fig. 2D).

CA AND RFA INDUCE TISSUE HYPOXIA

In control (untreated) and sham-operated renal tumours, there was minor hypoxia surrounding necrotic areas throughout the tumour. All thermally ablated tumours invariably revealed profound hypoxia in the viable tumour at the border of the ablated area (Fig. 3). This hypoxic region was most pronounced 2 h after ablation in both CA and RFA, at a mean (sem) of 27.3 (1.0)% after RFA and 27.8 (1.3)% after CA, and remained present for up to 1 week. CA-treated tumours showed a more rapid decrease in hypoxia than RFA-treated tumours; hypoxia was significantly more pronounced from day 3 in RFA compared with CA (P < 0.001). PN and sham treatment did not detectably increase hypoxia at the border of the (resected) tumour.

Figure 3.

A, Hypoxia at the border of thermally ablated and partially resected renal tumours, detected by pimonidazole staining (mean ±sem, n= 4–6 mice/group). Hypoxia was significantly increased in both CA and RFA treated tumours 2 h after treatment (P < 0.001), and was present up to 7 days. B, Representative images of pimonidazole staining 2 h after treatment. Intense brown staining indicates hypoxia (×10). RFA and CA strongly induced hypoxia at the border of the ablated area (VT, viable tissue). PN induced slight hypoxia at the border of the resected area (arrow). CTRL, CONTROL; Scale bar 102 µm.

INFLAMMATORY CELLS AFTER CA AND RFA

CA resulted in an increased number of macrophages at the border of ablated tumours after 3 days with a mean (sem) of 24.9 (6.8)% (P= 0.014), reaching a maximum after 1 week, at 30.2 (6.6)% (P < 0.001; Fig. 4A). Macrophages were less abundant after RFA, and were only significantly increased 14 days after RFA with a mean (sem) of 26.4 (7.8)% (P= 0.003). There were significantly more leukocytes (CD45 + cells) in the tumour tissue surrounding the ablated area 3 days after CA and RFA treatment, at a mean (sem) of 10.5 (0.7)% after CA and 10.5 (0.5)% after RFA (P < 0.05), and peaked after 1 week at 12.3 (0.9)% after CA and 11.1 (0.9)% after RFA (Fig. 4B). After 2 weeks leukocytes had returned to baseline levels in the tumour tissue surrounding the ablated area, but not within the ablated area. Within the ablated area leukocytes and macrophages showed an increased influx from day 1 after treatment, which was higher after CA than RFA (P < 0.05). In both treatments the peak presence of inflammatory cells was found several days later than the observed peak in hyperproliferation. PN and sham treatment did not result in a significant influx of macrophages or leukocytes.

Figure 4.

A, Presence of macrophages at the border of the thermally ablated or resected renal tumours (mean ±sem, n= 4–6 mice/group). There was a significant influx of macrophages 3 days after CA (P < 0.05) and 14 days after RFA (P < 0.05). B, Representative images of F4/80 staining for macrophages at the border of thermally ablated or resected renal tumours 14 days after treatment (×20). C, Presence of leukocytes at the border of the thermally ablated or resected renal tumours (mean ±sem, n= 4–6 mice/group). There was a significant influx of leukocytes from 3 days after CA and RFA (P < 0.05). D, Representative images of CD45-staining for leukocytes at the border of thermally ablated or resected renal tumours 7 days after treatment (×20). Ctrl, control; scale bars 204 µm.

HSPS ARE INCREASED AFTER RFA, CA AND PN

HSP90 was present in untreated tumours, but was only increased at the border of RFA-treated tumours, where it showed a marked increase after 2 h with a mean (sem) of 4.9 (3.7)% compared with the 0.4 (0.9)% baseline level (P < 0.05), and a peak after 3 days of 15.8 (8.2)% (P < 0.001; Fig. 5B). HSP70 was observed in untreated tumours, and was increasingly present in all treatment groups after 2 h (P < 0.05; Fig. 5A) and peaked after 1 day. RFA resulted in a three-fold higher presence of HSP70 than CA, at a mean (sem) of 6.9 (1.3) vs 2.3 (0.7)% (P < 0.001). From day 3 HSP70 could not be seen in the PN treatment group and showed a strong decrease in CA- and RFA-treated tumours compared with earlier time points. Sham treatment did not show any significant differences in HSP70 and HSP90.

Figure 5.

A, Expression of HSP70 at the border of thermally ablated and partially resected renal tumours (mean ±sem, n= 4–6 mice/group). HSP70 was significantly increased in both CA, RFA and PN treated tumours 2 h after treatment (P < 0.001). Increased levels of HSP70 remained present up to 7 days for RFA and CA, and 1 day after PN. B, Representative images of HSP70 staining 2 h after treatment. Intense brown staining indicates increased expression of HSP70 above the baseline levels of HSP70 (lighter brown staining; ×10). Especially induced a strong HSP70 expression at the border of the ablated area (VT, viable tissue). After PN there were small areas of increased HSP70 levels at the border of the resected area (arrow). C, Increased expression of HSP90 was only seen at the border of RFA-treated tumours, up to 3 days after treatment (P < 0.001). D, Representative images of HSP90 staining 3 days after treatment (×10). All tumours showed a slight physiological presence of HSP90 (light brown staining). An increased expression in RFA-treated tumours was seen as dark brown staining. Ctrl, control; scale bar 102 µm.

DISCUSSION

The success of CA and RFA depends on several factors, including location of the tumour within the kidney, tumour size, tumour biology and the influence of tumour micro-environment. Local recurrence rates after RFA and CA remain higher than after PN [13]. In the present translational study, we showed that incomplete ablation induced changes that stimulated growth of surviving renal tumour cells. These results are in accordance with various studies of thermal ablation in the liver, showing aggressive local regrowth after treatment [4–6,8,14]. The mechanisms stimulating hyperproliferation and apoptosis include hypoxia, immune modulation and regulation of cell survival by HSPs. Therefore, we determined which factors were present at the border of incompletely ablated renal tumours.

We found that both CA and RFA caused profound hypoxia directly adjacent to the ablated area. The presence of hypoxia has been reported in several solid tumours and is related to tumour aggressiveness [15,16]. While prolonged periods of hypoxia are damaging to normal cells, cancer cells can undergo adaptive changes that allow growth and survival under hypoxic conditions. The increased levels of hypoxia induced by thermal ablation are sufficient to provide an additional growth-stimulating microenvironment for the surviving tumour cells. In addition to hypoxia, HSPs showed an early overexpression at the border of the ablated area. HSPs are overexpressed under conditions of cellular stress. It is known that together with heat activation, factors such as inflammation, hypoxia and tissue trauma play a major role [17]. Indeed, RFA resulted in overexpression of both HSP70 and HSP90 at the border of the ablated area 2 h after treatment, while CA and PN caused upregulation of HSP70 that was less pronounced. Although some conflicting results on the influence of HSPs on damaged cells are reported, there is strong evidence that HSP overexpression plays an important role in increased cell survival through recovery of damaged cells, cell proliferation and inhibition of apoptosis. Increased expression of HSPs in surviving tumour cells at the border of thermally ablated tumours may lead to the development of more aggressive tumour cells [17–19].

Another factor that may be involved in hyperproliferation and cell survival is the immune response. The process of inflammation after thermal ablation involves a complex network of chemical signals and cell interactions in response to tissue damage. Surviving tumour cells at the border, as well as dead tumour cells within the ablated area release chemokines that attract inflammatory cells [20]. As described by others, leukocytes were present first upon thermally induced necrosis, followed by infiltration of macrophages [21]. The high baseline level of macrophages in renal tumours can be explained by the observation that often tumour-derived chemical signals stimulate a continuous migration of monocytes into tumour tissue [20]. Macrophages can be differentiated into two types: tumour-supportive and tumour-suppressive macrophages. Tumour-supportive macrophages and other inflammatory cells provide growth factors and cytokines that stimulate cell proliferation, angiogenesis and metastasis formation. Hypoxia is one of the major stimulators of the tumour-supportive macrophages [20,22]. However, in the present study inflammatory cells are unlikely to be the major factor involved in the hyperproliferation of tumour cells in the early phase after thermal ablation, as the influx of macrophages occurred several days after tumour cell hyperproliferation. However, because at 2 weeks after treatment tumour cells were still proliferating 1.5-fold more compared with the control tissue, macrophages could play a stimulatory role in this later phase.

Although it has been described that surgical resection may affect tumorigenesis [23–25], there was only a slightly increased cell proliferation rate and no significant difference in apoptosis seen after PN in the present study. This is in accordance with clinical studies of positive surgical margins after PN that indicate that recurrence rates are only slightly increased in these patients, and do not affect survival [26–28]. This suggests that a growth stimulatory effect on residual tumour cells is limited in RCC and the observed proliferation after CA and RFA can be ascribed to the thermal ablation.

We compared the two most commonly used thermally ablative procedures for small renal masses, to PN. With the described cellular effects we do not intend to question thermal ablation as a method. CA and RFA initiate cellular effects that may have secondary consequences: a number of growth stimulatory factors involving hypoxia, HSPs and immune responses may interact with surviving tumour cells. Based on the observation that RFA induces significantly more hyperproliferation, it might be recommended to use RFA for smaller lesions or employ larger safety margins around the tumour, although further clinical testing would be required to examine this. Although it is technically easier to perform multiple consecutive RFA procedures compared with CA [29], the present results stress the importance of complete ablation. Whenever possible, PN remains the procedure of choice, as it induces limited cellular proliferation, it is likely to cause the most complete tumour resection and has the best clinical results. Adjuvant therapies inhibiting hypoxia- or HSP-related responses after thermal ablation may be an appealing strategy to prevent escape of viable cells from lethal injury [8,17].

In conclusion, after incomplete CA and RFA we showed an increased proliferation and decreased apoptosis of surviving renal tumour cells. This hyperproliferation may be caused by stimulatory factors, e.g. hypoxia, HSPs and inflammatory cells, and could facilitate recurrences of renal tumours after thermal ablation. This underlines the importance of complete tumour destruction.

ACKNOWLEDGMENTS

The authors would like to thank M. van Amersfoort, J.T.C. Beekhuis, J.A.S. van Ginkel, M.A.J. van Osch and N.D. ter Hoeve for their assistance, and L.G.M. Daenen and R. Giles for critically reading the manuscript. We thank A.G. Celon for providing the radiofrequency equipment. We thank Galil for providing the cryo equipment.

CONFLICT OF INTEREST

None declared.

Ancillary