Changes in hepatic blood flow during whole body hyperthermia

Abstract

Purpose: Changes in blood flow distribution are important for heat dispersion and for supportive therapeutic strategies such as simultaneous whole body hyperthermia (WBH) and administration of chemotherapy. The aim of this clinical study was to determine changes in hepatic blood flow during WBH for the treatment of metastatic cancer.

Materials and methods: This observational clinical study was part of a phase I/II feasibility study of WBH. WBH was induced using a radiant heat device. Hepatic blood flow was estimated using indocyanine green clearance measurements. The plasma disappearance rate of indocyanine green (PDR-ICG) was recorded in percent/min. We used an invasive thermo-dye-dilution technique to estimate hepatic blood flow, cardiac output, and volume status. Mean arterial blood pressure was also measured invasively. To determine the effects of hyperthermia the measurements were performed at defined temperature points.

Results: In 10 of 22 treatments the PDR-ICG fell below normal values during hyperthermia, which represented a significant fall in hepatic blood flow. Cardiac output, volume status, and mean arterial blood pressure did not differ between patients whose liver blood flow was reduced and those whose liver blood flow remained unchanged.

Conclusions: We observed distinct reductions in hepatic blood flow during WBH, which suggested a significant redistribution of blood flow away from the core during WBH. This was not mirrored by global circulatory parameters.

• whole body hyperthermia
• plasma disappearance rate of indocyanine green
• blood flow distribution
• hepatic blood flow

Introduction

Induced whole body hyperthermia of 41.8°–42.2°C for at least one hour may be an effective component of a multimodal approach to the treatment of inoperable, locally advanced, or recurrent solid tumours [1–3]. Blood flow distribution has consequences for heat dispersion [4–7]. Alterations in hepatic blood flow in particular can have significant consequences for the metabolism of chemotherapeutic agents. For instance 5-FU and ifosfamid are two examples of anti-cancer drugs whose clearance depends on liver blood flow [8].

Various aspects of tumour perfusion during hyperthermia have been investigated in the context of experimental studies of hyperthermia: intratumoural blood flow has been shown to be related to factors such as intratumoural pH and intratumoural partial pressure of oxygen, and vice versa [4], [5], [9]; and it has been established that tumour angiogenesis generally produces disordered vessel architecture. Vasodilation and vasoconstriction of these disordered capillaries is for the most part independent of normal external controls because the base of local regulation–the muscle layer–is often either missing or dysfunctional [10], [11]. Nevertheless a fraction of vessels in the tumour will remain subject to external regulatory mechanisms.

Experimental studies have shown that during local hyperthermia, hypoperfusion occurs within the tumour while the surrounding tissue becomes hyperperfused [12]. To some extent the tumour blood vessels can be thought of as being the victims of a ‘steal’ phenomenon [5].

In healthy subjects the liver receives about one third of cardiac output. About one third of the liver blood flow is well oxygenated high-pressure blood from the hepatic artery and about two thirds is poorly oxygenated blood from the portal vein system (which originates from the gastrointestinal system). As part of the normal adaptive response to cardiovascular challenges, the intestinal vascular bed acts as a primary blood volume reserve. In general, hypovolaemia and low cardiac output as well as local mediators generated in the hepatic sinusoids (such as adenosine, nitric oxide and prostaglandins) have significant effects on hepatic blood flow [13–15].

The aim of this clinical study was to determine the changes in hepatic blood flow induced by radiant heating in the context of WBH treatment of metastatic cancer in humans. Our secondary aim was to determine any accompanying changes in global circulatory parameters (cardiac output, volume status, and mean arterial blood pressure) that are known to be associated with liver blood flow.

Materials and methods

Patients

This observational study was part of a phase I/II feasibility study of whole body hyperthermia for the treatment of metastatic cancer [16]. Each patient fulfilling inclusion criteria underwent WBH on one to three occasions at intervals of four weeks. The inclusion criteria were disseminated advanced malignancies and a life expectancy greater than 12 weeks. Patients underwent pre-treatment evaluation by an experienced anaesthesiologist. This anaesthesiological assessment consisted of history taking, electrocardiography, transthoracic echocardiography, chest X-ray, cranial computed tomography, lung function tests, and any further laboratory investigations deemed appropriate by the anaesthesiologist. Exclusion criteria were: standard contraindications to chemotherapy, a history of coronary heart disease, symptomatic cardiac insufficiency (NYHA Class II or above), brain metastasis, hyperthyroidism, pulmonary vital capacity less than 50% predicted, forced expiratory volume in one second less than 50% predicted, or a greater than 2.5-fold increase above normal levels in laboratory parameters such as alanine aminotransferase (ALAT), aspartate aminotransferase (ASAT), bilirubin, or γ-glutamyltransferase (γ-GT). Chemotherapeutic drugs were administered during WBH. The regimen naturally varied between cancers: folinic acid/ 5-fluorouracil plus mitomycin for colorectal cancer, ifosfamide (IFO) plus carboplatin (CBCDA) for ovarian cancer and IFO plus CBCDA plus etoposide for others (Table I).

Table I.  Patient characteristics according to treatment groups with a reduction of hepatic blood flow below normal value (reduced liver blood flow treatment) and in treatments without a reduction of hepatic blood flow (unchanged liver blood flow treatment). Data are presented as mean ± standard error of mean, n = number of patients.

	Reduced liver blood flow	Unchanged liver blood flow
Age (years)	43.9 ± 3.4	44.7 ± 3
Gender (male/female) (n)	7/3	9/3
Weight (kg)	88.6 ± 7.3	82.6 ± 6.1
Body-surface (m^2)	2 ± 0.1	2 ± 0.1
Height (cm)	179 ± 4	177 ± 3
Diagnoses (n)
Colorectal cancer	5	5
Germ cell tumour	5	4
Cervical carcinoma	0	1
Ovarian cancer	0	2
Hepatic metastases (n)	8	11

For WBH, all patients underwent total intravenous anaesthesia and mechanical ventilation. Patients were deliberately hyperoxygenated with the aim of sensitizing cancer cells to chemotherapy [17]. The anaesthesiological management, including volume management, was in accordance with our previously published algorithm. The algorithm emphasises the primacy of adequate volume status for circulatory stability and global organ perfusion [16].

Whole body hyperthermia

WBH was applied using the IRATHERM 2000 system (von Ardenne Institut, Germany), which is an ‘open chamber’ radiant heat applicator device. Water-filtered infrared waves penetrate to between 3–5 mm under the cutis to reach blood vessels in the subcutis and immediately induce a drastic vasodilation as described in detail elsewhere [1]. Heating was started after induction of anaesthesia and was continued up to the high temperature. A core body temperature of 41.8° to 42.2°C was generally achieved after a heating period of between 2 to 3 h, and was maintained for 60 min further. Patients were cooled passively. The study was approved by the local Review Board for Human Subject Research and written informed consent was obtained from all patients prior to treatment.

Estimation of hepatic blood flow

Indocyanine green is a water-soluble compound which is strictly intravascular and selectively excreted unchanged exclusively by hepatocytes without undergoing enterohepatic recirculation. The clearance of indocyanine green corresponds to hepatic blood flow and is called the plasma disappearance rate of indocyanine green (PDR-ICG) [18], [19]. Levels of PDR-ICG of less than 18% per minute have been reported to be subnormal [20]. An aortic catheter (Pulsiocath 4F, PV 2024L; Pulsion Medical Systems, Munich, Germany) was inserted via a femoral artery sheath to record the indocyanine green concentrations from which the washout curve was plotted (COLD-Z021 system, Pulsion Medical Systems, Munich, Germany). Cooled indocyanine green solution (2 mg/mL in glucose 5%, at less than 5°C) was administered via a catheter in the internal jugular vein. The thermo-dilution curves were plotted in real time and were used to extrapolate cardiac output and intravascular volume parameters (intrathoracic blood volume index, ITBVI, normal range 900–1100 mL/kg) [21].

PDR-ICG, arterial blood pressure, cardiac output, and volume status were recorded at the following standardised points during WHB: upon starting the heating at 37°C (baseline), 30 min after reaching the target temperature of 41.8°–42.2°C, and at 39°C during the cooling phase.

The liver function assays: alanine aminotransferase (ALAT), aspartate aminotransferase (ASAT), bilirubin, and γ-glutamyltransferase (γ-GT), were measured within 24 hours of WHB and 2–4 days after WBH. These liver function tests are global laboratory parameters used to monitor hepatocyte integrity and function.

Statistical analyses

Where a pre-existing reduction in PDR-ICG of less than 18%/min was observed at baseline, the data were excluded from the statistical analysis. The reason for this was that it would not have been possible to distinguish reduced hepatic blood flow due to WBH in the context of pre-existing liver dysfunction. In a number of cases, data collection was abandoned when PDR-ICG measurement failed at baseline (37°C body temperature). Repeating the baseline PDR-ICG would have required interruption of the treatment, which could not ethically be justified.

When the PDR-ICG level fell below 18%/min during WBH, the treatment was classified as a ‘reduced liver blood flow treatment’; while when the PDR-ICG remained greater than or equal to 18%/min during WBH, the treatment was classified as an ‘unchanged liver blood flow treatment’. In those patients who received more than one WBH there were a number of cases of patients who had different classifications between treatments.

All data were processed using the Statistical Package for the Social Sciences, Version 14.0 (SPSS). Results are presented as mean ± the standard error of mean (SEM). Differences between PDR-ICG groups were tested with the Mann-Whitney U-test (corrected for repeated measurements) for ordinal scaled parameters and Fisher's exact test was used to compare nominal scaled parameters. Significance was assumed at p < 0.05 in all tests.

Results

In all cases (41 WBH treatments), the baseline body temperature (bladder temperature measured via a urinary catheter placed immediately after induction of general anaesthesia) was below 37°C. During ten of these treatments the baseline PDR-ICG measurement failed within the first 30 min. As stated above, these treatments were excluded from the statistical analysis because the measurement could not be repeated without delaying or interrupting the WBH. A further nine WBH treatments were excluded from the statistical analysis because the clearance of ICG was below the minimum normal value of 18%/min at baseline.

Accordingly 22 WBH treatments from 14 patients were included in the statistical analysis. Ten of these 22 were classified as reduced liver blood flow treatments (Table I, Figure 1) and the remaining 12 treatments were classified as unchanged liver blood flow treatments. As mentioned above, three patients had one classification during their first WBH but a different classification during a subsequent WBH.

Figure 1. PDR-ICG at different temperature points. Plasma disappearance rate of indocyanine green (PDR-ICG) represents hepatic blood flow. Reduced liver blood flow treatment: PDR-ICG decreased <18%/min during WBH, unchanged liver blood flow: PDR-ICG remained ≥18%/min during WBH. Data are presented as mean ± standard error of mean. *Significant differences between both groups in Mann-Whitney U-test.

There were 16 male and 6 female patients. The average body mass was 85.5 ± 4.7 kg; the average height was 178 ± 3 cm; and the average body surface was 2 ± 0.1 m². The primary malignancies were as follows: colorectal cancer (10 patients), germ cell tumour (9 patients), cervical cancer (1 patient), ovarian cancer (2 patients). Hepatic metastases were demonstrated using computer tomography in 19 of the 22 included patients. There were no differences between the two PDR-ICG treatment groups (reduced liver blood flow treatment and unchanged liver blood flow treatment) concerning age, gender, body weight, body surface area, height, underlying diseases, and existence of hepatic metastases (Table I). Liver metastases were present in 11 of the 12 unchanged liver blood flow treatments and in 8 of the 10 reduced liver blood flow treatments (Table I). At high temperature (41.8°C–42.2°C) the cardiac index (8.03 ± 0.24 L/min/m²) and volume status was higher than normal (intrathoracic blood volume index; 1116 ± 22 mL/m²). Mean arterial blood pressure was decreased at high temperature (53.6 ± 1.4 mmHg). There were no differences with respect to these parameters between the reduced liver blood flow treatment and the unchanged liver blood flow treatment groups at 37°C (baseline), high temperature (41.8°C–42.2°C) or at 39°C (cooling phase) (Table II).

Table II.  Circulatory parameters in treatments with a reduction of hepatic blood flow below normal value (reduced liver blood flow treatment) and in treatments without a reduction of hepatic blood flow (unchanged liver blood flow treatment) at standardised temperatures.

	Reduced liver blood flow			Unchanged liver blood flow
	37°C	41.8–42.2°C	39°C	37°C	41.8–42.2°C	39°C
MAP (mmHg)	63.7 ± 3.3	54.6 ± 1.5	58.2 ± 3	65.3 ± 3.7	52.9 ± 1.5	59.1 ± 2.2
CVP (mmHg)	9.4 ± 1.3	10.6 ± 1.5	9.30 ± 1.34	8.58 ± 0.74	10.3 ± 0.7	9.17 ± 0.68
CI (l/min/m^2)	5.37 ± 0.33	8.51 ± 0.22	6.58 ± 0.16	5.16 ± 0.34	7.82 ± 0.32	5.93 ± 0.31
ITBVI (ml/m^2)	1168 ± 114	1129 ± 39	1010 ± 31	1013 ± 31	1107 ± 26	990 ± 30

Values of liver enzymes 24 h and 2–4 days after WBH did not differ between the two groups (Table III).

Table III.  Laboratory parameters within 24 h after treatment and within 2–4 days after treatments. Treatments with a reduction of hepatic blood flow below normal value (reduced liver blood flow treatment) compared with treatments without a reduction of hepatic blood flow (unchanged liver blood flow treatment). Data are presented as mean ± standard error of the mean. ASAT, aspartate aminotransferase; ALAT, alanine aminotransferase; γGT, γglutamyltransferase.

	Within 24 hours after treatment		2–4 days after treatment
	Reduced liver blood flow	Unchanged liver blood flow	Reduced liver blood flow	Unchanged liver blood flow
ASAT (U/L)	19.1 ± 3.3	14 ± 1.2	36.3 ± 13.7	26.9 ± 6.3
ALAT (U/L)	28.8 ± 6.5	17.8 ± 2.7	35.2 ± 13.2	32.1 ± 8.1
γGT (U/L)	84.3 ± 37.9	43.8 ± 17.3	126 ± 34	58.4 ± 14.8
Bilirubin (mg/dL)	0.64 ± 0.09	0.58 ± 0.08	1.16 ± 0.37	0.97 ± 0.2

Discussion

We made a number of observations during this clinical study. Firstly, hepatic blood flow was reduced during high temperature in a significant number of treatments irrespective of baseline liver function. Secondly, the cardiac output, volume status and mean arterial pressure measurements were not the cause of the reduction in hepatic blood flow that we observed. Thirdly, patients who demonstrated a reduction in hepatic blood flow during one WBH did not necessarily demonstrate reductions in hepatic blood flow during subsequent WBHs.

Baseline hepatic blood flow was high in both treatment groups (Figure 1), i.e. the baseline clearance of ICG was excellent in all included treatments. The marked changes during whole body hyperthermia that we observed in circulatory parameters have previously been described in detail [16], [22], [23]. Hepatic perfusion, however, has not yet been described in humans undergoing WBH treatment in use of radiant heat device and chemotherapy. Similar experimental studies using ICG were performed in the 1970s in healthy subjects [15].

Overall we observed more than a doubling of cardiac output during hyperthermia. However, we found no differences with respect to cardiac output, volume status or mean arterial pressure when we compared reduced liver blood flow and unchanged liver blood flow WBH treatments (Table II). Our findings suggest that during WBH hepatic blood flow may be independent of known systemic circulatory determinants of hepatic blood flow such as cardiac output and intravascular volume. We observed high cardiac output and a reduced mean arterial blood pressure. This observation may be explained by vasodilation somewhere in the circulation. Intuitively, cutaneous vasodilation would be the most likely reason for reduced mean arterial blood pressure despite increased cardiac output. It is assumed that the application of external heat from radiant heat devices results in an intense vasodilation in the subcutis. Indeed, substantial cutaneous vasodilation might cause shifting of perfusion away from core areas such as the intestinal tract and liver. We speculate that the liver may suffer as a result of selective hyperperfusion of other areas such as the cutis at high temperature during WBH (cutaneous steal phenomenon). Measuring cutaneous perfusion was beyond the scope of this study and we are unable to confirm this hypothesis. Cutaneous hyperperfusion is maximal during radiant heating and cutaneous vasodilation is an established autoregulatory response to hyperthermia [24]. A reduction of liver blood flow at high temperature during WBH might reduce the exposure of the liver and/or liver metastases to chemotherapy that time. These observed changes in blood flow, highlighted the importance of the correct timing of drug delivery during WBH. Drug delivery at the high temperature phase of WBH should be avoided, when targeting liver tumours with chemotherapy.

Experimental studies have demonstrated interdependency between tumour perfusion, intratumoural oxygen partial pressure and/or pH levels [5]. In order to examine the effects of hepatic metastasis on hepatic blood flow, hepatic metastases were sought using computed tomography. In this clinical study, WBH was applied to patients with advanced cancers including cancers with liver metastases. Due to the high incidence of hepatic metastases in our patient sample, it was not possible to determine the effects of hepatic metastasis on hepatic blood flow, and further studies will be necessary.

The underlying cancer or choice of chemotherapeutic agents might influence liver blood flow during WBH. We observed that a number of patients had one classification (unchanged liver blood flow treatment or reduced liver blood flow treatment) during their first treatment but a different classification during a subsequent treatment with the same chemotherapeutic agents. We also found no differences between the groups concerning the primary tumours, which may be attributable to the fact that each primary tumour was treated in a unique manner. The study, however, was insufficiently powered to make definitive statements in this respect.

Liver blood flow depends on the portal venous system and the arterial hepatic blood flow. Usually the portal venous system and the arterial blood supply accurately compensate for changes in one another. This compensation cannot however be measured using the methods that we used. The reduction in liver blood flow might be attributable to, for instance, reduced blood flow in the small intestine and consequent reduced portal venous flow. Alternatively, it might simply be attributable to reduced hepatic arterial blood flow. Nevertheless, reduced liver blood flow and liver excretion rates were measured and this may have clinical significance with respect to the metabolism of chemotherapeutics [6–8].

Acute hepatic cellular destruction as a possible further cause of ICG clearance reduction seems implausible, because liver enzyme levels did not indicate hepatic cellular destruction.

In conclusion, in this clinical study a distinct reduction in hepatic blood flow was observed during WBH and confirmed drastic changes in blood flow distribution away from the core at high temperature. The reduction in hepatic blood flow was not, however, suggested by values of common circulatory parameters measured during WBH.

Acknowledgements

M.D. and O.A. contributed equally to this study. None of the authors are involved in any organisation with a direct financial, intellectual, or other interest in the subject of the manuscript. The study was supported by the Deutsche Krebshilfe, Deutsche Forschungsgemeinschaft (SFB 273, Grako 331).

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