Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

During the directorship of Professor Gerhard Thews, research at the Institute of Physiology, University of Mainz, traditionally focussed on oxygen transport in blood, lung, brain, and heart. Joining his research team in 1970 as a postdoctoral research fellow, I was asked to “investigate the oxygen transport and respiratory gas exchange in other clinically relevant tissues and organs to expand the scope of research of the Institute” (G. Thews). After a careful and time-consuming literature search, I finally decided to study oxygen transport to the spleen (murine, rabbit, human) and to solid, malignant tumors, since reliable oxygenation data for these tissues were not available at that time, especially in terms of translation of the preclinical data to the clinical setting. In the following, chronology-oriented chapters are presented, and selected data obtained over the last 40 years are described, clearly showing the progress in relevant information.

The oxygenation status and data on the respiratory gas exchange of the spleen have been described earlier [1–3].

2 The Past

Since 1970, in vivo investigations have been carried out on isotransplanted rat tumors after the development and implementation of a “tissue-isolated” tumor model in the rat kidney involving a single artery feeding the tumor and a single vein draining the tumor, thus enabling the measurement of total blood flow and of the biologically relevant arteriovenous concentration differences of the substrates and catabolites of interest [4–6]. Key results using this tumor preparation are as follows (for details, see [4–12]): (a) tumor blood flow (TBF) and oxygen availability exhibit pronounced intra-tumor and inter-tumor heterogeneities; (b) tumor oxygenation is distinctly poorer than in normal tissue and shows similar heterogeneities to those found for TBF; (c) increasing oxygen availability through increasing TBF, arterial oxygen content, and hemoglobin concentration (cHb) can increase oxygen uptake and can improve tissue oxygenation; (d) oxygen availability is the major determinant of the oxygen consumption rate of cancers in situ; (e) oxygen consumption rate of cancers in situ is thus a function of TBF and arterial oxygen content; (f) weight-related TBF and tissue oxygenation generally decrease with increasing tumor size (not necessarily applicable to the clinical setting, see below); and (g) contrary to conventional belief, there is no evidence for a general mitochondrial dysfunction, speaking against a principal role of the Warburg effect in its original concept [6, 13].

Modulation of the tumor oxygenation status has been described as a result of therapeutic measures (irradiation [14], localized hyperthermia [15–19], photodynamic therapy [20], normobaric and hyperbaric hyperoxia [6, 21, 22], improvement of perfusion [23, 24], and correction of anemia using erythropoietin [25]).

Between 1977 and 1985, HbO2 saturation of single red blood cells (RBCs) in tumor microvessels was registered in experimental rat tumors [26, 27], in cancers of the oral cavity and of the rectum, and in primary and metastatic bone tumors [28–30]. In accordance with the studies on experimental murine tumors, the oxygenation status in human cancers was poorer than in the normal tissue, exhibited pronounced intra-tumor and inter-tumor heterogeneities, and was positively correlated with the vascular density. In contrast to the experimental situation, the O2 status in human tumors showed no size dependency.

In 1985, investigations were started to assess the oxygen status of orthotopically xenografted human breast cancers in immune-deficient rnu/rnu rats. In order to allow measurements of TBF and the relevant arteriovenous concentration differences, a novel “tissue-isolated tumor” model was implemented [31, 32]. Experiments using different tumor histologies showed that – comparable to experimental murine tumors – the oxygen consumption and the median tissue pO2 both were a function of TBF and the oxygen availability, respectively [33]. Theoretical analysis of the oxygen supply conditions in these xenografted human tumors led to the conclusion that oxygen seems to be the limiting substrate for unlimited proliferation and glucose for tumor cell survival [34, 35].

Using 31P-NMR-spectroscopy studies starting in 1987, correlations between the bioenergetic status, the tissue oxygenation, and the intracellular pH (pHi) were evaluated. In experimental murine tumors, pHi was neutral to alkaline whereas the extracellular pH (pHe) was acidic [36]. Intracellular pH was found to be alkaline as long as the median tissue pO2 was above 10 mmHg. Below this threshold, pHi became acidic [37–41] and the gradient between the intracellular pH and extracellular pH flattened.

In 1989 systematic investigations on gynecological patient cancers (cervix, breast, vulva) were initiated. In these patients, the pretherapeutic oxygenation status of primary and recurrent tumors was assessed using the pO2 histography system [42–46]. Publication of these data had a tremendous impact in defining the role of tumor hypoxia in malignant progression and therapeutic resistance [47–52]. Key findings were as follows: (a) approx. 60 % of pretreatment cervical cancers were hypoxic; (b) cancer oxygenation was distinctly poorer than that of the normal tissues at the site of tumor growth; (c) the extent of hypoxia was independent of clinical size, stage, histology, grade, lymph node status, and various patient demographics; (d) hypoxia was aggravated in anemic patients; (e) hypoxia was less pronounced on transgression of stage IVA cervical cancers into the bladder wall; (f) recurrent tumors were more hypoxic than their primaries; and (g) there was no typical topological distribution of hypoxic areas within tumors (periphery vs. center).

Since 1990, investigations on hypoxia-driven malignant progression followed, based on the observations that in multivariate analysis, hypoxia was found to be a strong, independent, and adverse prognostic factor for overall and disease-free survival in cervical cancer patients [53–58].

In the last 10 years, the recognition of tumor hypoxia as a pivotal factor driving the development of a highly malignant phenotype – in which the HIF system, genetic instability, and clonal selection play a central role – has encouraged attempts to correlate the expression of “endogenous” hypoxia markers (HIFs, GLUT-1, CA IX) with the oxygenation status in identical, non-necrotic tumor microareas. Our results clearly showed that there is no correlation between the protein expression of these markers and pO2 data measured with O2 microsensors [59–62]. This supports the hypothesis that the HIF system can be stabilized even under normoxic conditions (e.g., by oncogenic growth factors, certain cytokines, glucose deprivation, acidosis, and gene mutations). From this it can be concluded that HIF-1α and its target genes cannot be considered as strict hypoxia markers, but instead should be considered to be markers of hypoxia-associated malignant progression.

3 The Present

Since 2008 our research focus is on the classification and quantification of hypoxia subtypes in xenografted human squamous cell carcinomas of the head and neck. In these experiments, hypoxia subtypes are categorized as follows: (a) continuous (chronic) hypoxia due to diffusion limitations or sustained microvascular flow stop by disturbed Starling forces, (b) intermittent (acute) hypoxia due to temporary obstructions of tumor microvessels or distinct fluctuations of RBC fluxes, and (c) hypoxemic hypoxia due to patient anemia or plasma flow in microvessels only [63]. Using tumor cryosections and (immuno-) fluorescence, detection and quantification of these subtypes showed that chronic hypoxia is the dominating subtype in vital tumor tissue, followed by acute and hypoxemic hypoxia. Analyses using microcirculatory supply units yielded pronounced (tumor size-dependent) intra-tumor heterogeneity and distinct variability between different tumor lines [64–67].

4 The Future

The extent of hypoxia subtypes, their respective fractions of total hypoxia, their time frames, and biological and therapeutic consequences will be investigated in the near future. Furthermore, detection and reliable quantification of hypoxia subtypes in the clinical setting are urgently needed, especially for critical judgment of fractionation schedules in radio(chemo-)therapy.