Poor outcome in hypoxic endometrial carcinoma is related to vascular density.

BACKGROUND: Identification of endometrial carcinoma (EC) patients at high risk of recurrence is lacking. In this study, the prognostic role of hypoxia and angiogenesis was investigated in EC patients. METHODS: Tumour slides from EC patients were stained by immunofluorescence for carbonic anhydrase IX (CAIX) as hypoxic marker and CD34 for assessment of microvessel density (MVD). CAIX expression was determined in epithelial tumour cells, with a cut-off of 1%. MVD was assessed according to the Weidner method. Correlations with disease-specific survival (DSS), disease-free survival (DFS) and distant disease-free survival (DDFS) were calculated using Kaplan-Meier curves and Cox regression analysis. RESULTS: Sixty-three (16.4%) of 385 ECs showed positive CAIX expression with high vascular density. These ECs had a reduced DSS compared to tumours with either hypoxia or high vascular density (log-rank p = 0.002). Multivariable analysis showed that hypoxic tumours with high vascular density had a reduced DSS (hazard ratio [HR] 3.71, p = 0.002), DDFS (HR 2.68, p = 0.009) and a trend for reduced DFS (HR 1.87, p = 0.054). CONCLUSIONS: This study has shown that adverse outcome in hypoxic ECs is seen in the presence of high vascular density, suggesting an important role of angiogenesis in the metastatic process of hypoxic EC. Differential adjuvant treatment might be indicated for these patients.

Prognostic value of the proliferation marker Ki-67 in laryngeal carcinoma: results of the accelerated radiotherapy with carbogen breathing and nicotinamide phase III randomized trial.

BACKGROUND: The prognostic and predictive value of the proliferation marker Ki-67 was investigated in a randomized trial comparing accelerated radiotherapy with carbogen breathing and nicotinamide (ARCON) to accelerated radiotherapy in laryngeal carcinoma. METHODS: Labeling index of Ki-67 (Li Ki-67) in immunohistochemically stained biopsies and the colocalization with carbonic anhydrase IX (CAIX) were related to tumor control and patient survival. RESULTS: On average, node-positive patients had a higher Li Ki-67 (median 14% vs 8%; p < .01). In patients with a high Li Ki-67, the 5-year regional control and metastases-free survival were 79% versus 96% (p < .01) and 71% versus 88% (p = .05) for accelerated radiotherapy and ARCON, respectively. The 5-year local control and disease-specific survival were not significantly different. Patients with low Ki-67 expression had an excellent outcome with accelerated radiotherapy alone. CONCLUSION: Patients with laryngeal carcinomas with high proliferative activity are at increased risk of regional and distant metastases formation. This risk can be reduced by treatment with ARCON.

Dasatinib Inhibits DNA Repair after Radiotherapy Specifically in pSFK-Expressing Tumor Areas in Head and Neck Xenograft Tumors.

Src family kinases (SFKs) have been implicated in resistance to both radiation and epidermal growth factor receptor (EGFR) inhibition. Therefore, we investigated whether inhibition of SFK through dasatinib (DSB) can enhance the effect of radiotherapy in two in vivo human head and neck squamous cell carcinoma (HNSCC) models. Response to DSB and/or radiotherapy was assessed with tumor growth delay assays in two HNSCC xenograft models, SCCNij153 and SCCNij202. Effects on EGFR signaling were evaluated with Western blot analysis, and effects on DNA repair, hypoxia, and proliferation were investigated with immunohistochemistry. DSB and radiotherapy induced a significant growth delay in both HNSCC xenograft models, although to a lesser extent in SCCNij202. DSB did not inhibit phosphorylated protein kinase B (pAKT) or phosphorylated extracellular signal-regulated kinase 1/2 (pERK1/2) but did inhibit (phosphorylated) DNA-dependent protein kinase. Moreover, DSB reduced repair of radiation-induced DNA double-strand breaks as shown by an increase of p53-binding protein 1 (53BP1) staining 24 hours after radiation. This effect on DNA repair was only observed in the cell compartment where phosphorylated SFK (pSFK) was expressed: for SCCNij153 tumors in both normoxic and hypoxic areas and for SCCNij202 tumors only in hypoxic areas. No consistent effects of DSB on hypoxia or proliferation were observed. In conclusion, DSB enhances the effect of radiotherapy in vivo by inhibition of radiation-induced DNA repair and is a promising way to improve outcome in HNSCC patients.

Pattern of CAIX expression is prognostic for outcome and predicts response to ARCON in patients with laryngeal cancer treated in a phase III randomized trial.

BACKGROUND AND PURPOSE: In a phase III trial in patients with advanced stage laryngeal carcinoma comparing ARCON (accelerated radiotherapy with carbogen breathing and nicotinamide) to accelerated radiotherapy alone (AR) the prognostic and predictive value of CAIX, a hypoxia-associated protein, was investigated. MATERIAL AND METHODS: 261 Paraffin embedded tumor biopsies and 79 fresh frozen biopsies from patients entered in the trial were immunohistochemically stained for CAIX. CAIX-fraction and CAIX expression pattern were related to tumor control and patient survival. RESULTS: Low CAIX-fraction was prognostic for worse regional control and overall survival in patients treated with AR. Patients with a low CAIX-fraction treated with ARCON had better regional control and metastasis-free survival compared to AR (RC 97% vs 71%, p < 0.01 and MFS 92% vs 69%, p = 0.06). Patients with a perinecrotic CAIX staining pattern had a significantly worse local control, metastasis-free and overall survival compared to patients with a diffuse pattern (65% vs 84%, p = 0.01, 70% vs 96%, p < 0.01 and 42% vs 71%, p < 0.01 respectively), and this could not be improved with ARCON. After multivariate analysis CAIX pattern and N-stage emerged as significant predictors for metastasis-free survival and overall survival. CONCLUSIONS: ARCON improves regional control and metastasis-free survival only in patients with low CAIX expression. The different patterns of CAIX expression suggest different mechanisms of upregulation and have important prognostic value.

Combretastatin A-4 phosphate affects tumor vessel volume and size distribution as assessed using MRI-based vessel size imaging.

PURPOSE: Combretastatin A-4 disodium phosphate (CA4P) is a promising vascular disrupting agent (VDA) in clinical trials. As CA4P acts on dividing endothelial cells, we hypothesize that CA4P affects vessels of certain sizes. The aim of this study was to evaluate the effect of CA4P by the MRI-based vessel size imaging (VSI). EXPERIMENTAL DESIGN: C3H mammary carcinomas were grown to 200 mm(3) in the right rear foot of female CDF(1) mice. A control group of mice received no treatment, and a treatment group had CA4P administered intraperitoneally at a dose of 250 mg/kg. VSI was conducted on a 3 Tesla MR scanner to estimate the tumor blood volume (ζ(0)) and mean vessel radius (R). Vascularization was also estimated histologically by endothelial and Hoechst 33342 staining. RESULTS: ζ(0) and R showed different spatial heterogeneity. Tumor median and quartile values of ζ(0) were all significantly reduced by about 35% in the CA4P-treated group as compared with the control group, and the median and upper quartile of R were significantly increased. Histograms of ζ(0) and R showed a general decrease in ζ(0) following treatment, and values of R in a certain range (≈20-30 µm) were decreased in the treatment group. The drug-induced change in ζ(0) was in agreement with histology and our previous dynamic contrast enhanced MRI (DCE-MRI) data. CONCLUSIONS: Tumor blood volume and mean vessel radius showed a clear response following treatment with CA4P. VSI may prove valuable in estimation of tumor angiogenesis and prediction of response to VDAs.

Expression of EGFR under tumor hypoxia: identification of a subpopulation of tumor cells responsible for aggressiveness and treatment resistance.

PURPOSE: Overexpression of epidermal growth factor receptor (EGFR) and tumor hypoxia have been shown to correlate with worse outcome in several types of cancer including head-and-neck squamous cell carcinoma. Little is known about the combination and possible interactions between the two phenomena. METHODS AND MATERIALS: In this study, 45 cases of histologically confirmed squamous cell carcinomas of the head and neck were analyzed. All patients received intravenous infusions of the exogenous hypoxia marker pimonidazole prior to biopsy. Presence of EGFR, pimonidazole binding, and colocalization between EGFR and tumor hypoxia were examined using immunohistochemistry. RESULTS: Of all biopsies examined, respectively, 91% and 60% demonstrated EGFR- and pimonidazole-positive areas. A weak but significant association was found between the hypoxic fractions of pimonidazole (HFpimo) and EGFR fractions (F-EGFR) and between F-EGFR and relative vascular area. Various degrees of colocalization between hypoxia and EGFR were found, increasing with distance from the vasculature. A high fraction of EGFR was correlated with better disease-free and metastasis-free survival, whereas a high degree of colocalization correlated with poor outcome. CONCLUSIONS: Colocalization of hypoxia and EGFR was demonstrated in head-and-neck squamous cell carcinomas, predominantly at longer distances from vessels. A large amount of colocalization was associated with poor outcome, which points to a survival advantage of hypoxic cells that are also able to express EGFR. This subpopulation of tumor cells might be indicative of tumor aggressiveness and be partly responsible for treatment resistance.

Spatial relationship of phosphorylated epidermal growth factor receptor and activated AKT in head and neck squamous cell carcinoma.

BACKGROUND: Overexpression of EGFR correlates with decreased survival after radiotherapy in head and neck squamous cell carcinoma (HNSCC). However, the contribution of the activated form, pEGFR, and its downstream signaling (PI3-K/AKT) pathway is not clear yet. METHODS: Fifty-eight patients with HNSCC were included in the study. pEGFR, pAKT, hypoxia, and vessels were visualized using immunohistochemistry. Fractions (defined as the tumor area positive for the respective markers relative to the total tumor area) were calculated by automated image analysis and related to clinical outcome. RESULTS: Both pEGFR (median 0.6%, range 0-34%) and pAKT (median 1.8%, range 0-16%) expression differed between tumors. Also, a large variation in hypoxia was found (median pimonidazole fraction 3.9% 0-20%). A significant correlation between pEGFR and pAKT (r(s) 0.44, p=0.004) was seen, however, analysis revealed that this was not always based on spatial coexpression. Low pAKT expression was associated with increased risk of regional recurrence (p<0.05, log-rank) and distant metastasis (p=0.04). CONCLUSION: The correlation between expression of pEGFR and pAKT is indicative of activation of the PI3-K/AKT pathway through phosphorylation of EGFR. Since not all tumors show coexpression to the same extent, other factors must be involved in the activation of this pathway as well.

Parametric mapping of immunohistochemically stained tissue sections; a method to quantify the colocalization of tumor markers.

BACKGROUND: Automated analysis of immunohistochemically stained tissue sections is of great importance in cancer research to detect tumor-specific prognostic markers and make therapy decisions. Here, an automated quantitative analysis is presented to assess the colocalization of CAIX, a membrane-bound hypoxic marker and Ki-67, a nuclear proliferation marker. METHODS: Tissue sections of 104 biopsies from 89 patients were stained for CAIX and Ki-67 with diaminobenzidine and haematoxylin counterstain. Image scans of whole tumor sections were recorded and image maps were created with parametric mapping to quantify the markers and assess the colocalization. RESULTS: The fraction of CAIX showed a range of 0-93%. The interobserver correlation and the correlation between manual scores and automated analysis were both very strong (rs=0.96, p<0.0001, and rs=0.97, p<0.0001). The labelling index of Ki-67 exhibited a range of 0-42% with less strong interobserver and manual to automated analysis correlations (rs=0.90, p<0.0001, and rs=0.71, p<0.0008). The relative tumor area positive for both markers varied from 0-76%. CONCLUSION: Parametric mapping of immunohistochemically stained tumor sections is a reliable method to quantitatively analyze membrane-bound proteins and assess the colocalization of various tumor markers in different subcellular compartments.

PET of hypoxia with 89Zr-labeled cG250-F(ab')2 in head and neck tumors.

UNLABELLED: Hypoxic tumor cells are resistant to radiotherapy and various chemotherapeutic agents. The pretherapeutic assessment of intratumoral hypoxia may allow selection of patients for intensified treatment regimens. Carbonic anhydrase IX (CAIX) is an endogenous hypoxia-related protein involved in pH regulation and is upregulated in many tumor types. Radionuclide imaging using a monoclonal antibody against CAIX, such as cG250, may allow noninvasive PET of hypoxia in these tumor types. The aims of this study were to investigate whether (89)Zr-labeled cG250-F(ab')(2) allowed visualization of tumor hypoxia using small-animal PET and whether the tracer showed spatial correlation to the microscopic distribution of CAIX-expressing cells in a human head and neck xenograft tumor model. METHODS: Athymic mice with subcutaneous human head and neck carcinoma xenografts (SCCNij3) were imaged with small-animal PET after injection of (89)Zr-cG250-F(ab')(2). PET images were parameterized in terms of standardized uptake values (SUVs). After injection with the nitroimidazole hypoxia marker pimonidazole and the perfusion marker Hoechst 33342, the animals were sacrificed, tumors excised, and CAIX- and pimonidazole-marked hypoxia and blood perfusion were analyzed immunohistochemically. (89)Zr-cG250-F(ab')(2) tumor uptake was analyzed by ex vivo activity counting and by autoradiography of tumor sections. RESULTS: As early as 4 h after administration, accumulation of (89)Zr-cG250-F(ab')(2) in the tumor had occurred and tumors were clearly visualized by PET, with reduced uptake by 24 h after injection. Pixel-by-pixel analysis showed a significant positive spatial correlation between CAIX expression and (89)Zr-cG250-F(ab')(2) localization (r = 0.57-0.74; P < 0.0001). Also, significant correlations were found between pimonidazole staining intensity and (89)Zr-cG250-F(ab')(2) activity concentration, although less strong (r = 0.46-0.68; P < 0.0001). Tumor maximum SUV correlated significantly with tumor uptake determined ex vivo (r = 0.93; P = 0.0067), as did fractions of CAIX and pimonidazole in tumor sections (r = 0.75; P = 0.03 and r = 0.78; P = 0.02, respectively). CONCLUSION: (89)Zr-labeled cG250-F(ab')(2) small-animal PET showed rapid accumulation in a head and neck xenograft tumor model with good correlation to CAIX expression on a microscopic level.

Hypoxia in larynx carcinomas assessed by pimonidazole binding and the value of CA-IX and vascularity as surrogate markers of hypoxia.

Tumour hypoxia as driving force in tumour progression and treatment resistance has been well established. Assessment of oxygenation status of tumours may provide important prognostic information and improve selection of patients for treatment. In this study, a large homogenous group of 103 laryngeal carcinomas has been investigated in the presence of hypoxia by pimonidazole binding and the usefulness of Carbonic anhydrase IX (CA-IX) and vascular parameters as surrogate markers of hypoxia. These parameters are further related to clinical and biological characteristics. One hundred and three patients with T2-T4 larynx carcinoma were included. They were given the hypoxia marker pimonidazole intravenously (i.v.) 2h prior to taking a biopsy. Expression of all the parameters was examined by immunohistochemistry, excluding large necrotic areas. Among tumours a large variation in pimonidazole positivity (hypoxic fraction based on pimonidazole, HFpimo) (range 0-19%) and CA-IX expression (hypoxic fraction based on CA-IX staining, HFCA-IX) (range 0-34%) was observed. In 67% of the tumours, hypoxia involved 1% of the viable tumour area. HFpimo and HFCA-IX correlated significantly albeit weak (p=0.04). Both parameters showed weak inverse correlations with the relative vascular area (RVA) (p=0.01). HFpimo was further associated with histopathological grade, with poorly differentiated tumours being more hypoxic. The fraction of the tumour area positive for both pimonidazole and CA-IX correlated significantly with N stage. From these results, it was concluded that CA-IX and RVA have only limited value for measuring hypoxia and are not as robust as pimonidazole, probably due to the influence of other factors in the microenvironment. A combination of staining patterns of exogenous and endogenous markers might give important additive information about tumour biology and behaviour.

Hypoxia in relation to vasculature and proliferation in liver metastases in patients with colorectal cancer.

PURPOSE: To investigate hypoxia measured by pimonidazole binding, glucose transporter 1 (GLUT1) and carbonic anhydrase IX (CA-IX) expression, proliferation, and vascularity in liver metastases of colorectal cancer and to compare GLUT1 and CA-IX expression in corresponding primary tumors. METHODS AND MATERIALS: Twenty-five patients with liver metastases of colorectal cancer, planned for metastasectomy, were included. The hypoxia marker pimonidazole and proliferation marker iododeoxyuridine were administered before surgery. After immunofluorescent staining of the frozen metastases, pimonidazole binding, vascularity, and proliferation were analyzed quantitatively. Thirteen paraffin-embedded primary tumors were stained immunohistochemically for GLUT1 and CA-IX expression, which was analyzed semiquantitatively in primary tumors and corresponding liver metastases. RESULTS: In liver metastases, pimonidazole binding showed a pattern consistent with diffusion-limited hypoxia. The mean pimonidazole-positive fraction was 0.146; the mean distance from vessels to pimonidazole-positive areas was 80 microm. When expressed, often co-localization was observed between pimonidazole binding and GLUT1 or CA-IX expression, but microregional areas of mismatch were also observed. No correlation between the level of pimonidazole binding and GLUT1 or CA-IX expression was observed. In some patients, a large fraction (up to 30%) of proliferating cells was present in pimonidazole-stained areas. Expression of CA-IX in primary tumors and metastases showed a significant correlation, which was absent for GLUT1 expression. CONCLUSIONS: Compared with other tumor types, liver metastases of colorectal cancer contain large amounts of hypoxic cells. The lack of correlation with pimonidazole binding brings into question the value of GLUT1 and CA-IX as endogenous markers of hypoxia.

Erythropoietin receptor is not a surrogate marker for tumor hypoxia and does not correlate with survival in head and neck squamous cell carcinomas.

BACKGROUND AND PURPOSE: To evaluate erythropoietin receptor (EPOR) expression in human head and neck squamous cell carcinomas and correlate this to the presence of tumor hypoxia and treatment outcome. PATIENTS AND METHODS: Eighty-five patients with locally advanced tumors of the head and neck were included. Of these, 34 were given the hypoxia marker pimonidazole i.v. 2 h prior to biopsy taking. Contiguous paraffin embedded biopsies were stained for EPOR expression and, if administered, for pimonidazole binding. Immunohistochemical staining for EPOR was interpreted semiquantitatively according to a composite scale, ranging from 0 to 200. Pimonidazole positivity was quantitatively analyzed in a semiautomatic way. RESULTS: Diffuse weak-to-moderate cytoplasmic and membrane EPOR immunostaining was observed in 80 of 85 biopsies (94%) and staining scores ranged from 0 to 198 (median 100). No correlations were found between EPOR expression, and the primary tumor site, T-stage or N-stage. Also, There was no association between EPOR expression and treatment outcome. The degree of tumor hypoxia represented by the relative area of pimonidazole binding varied between 0 and 26% (median 7%). Contiguous biopsy sections showed a lack of colocalization between EPOR and pimonidazole binding. CONCLUSION: EPOR expression was demonstrated in the majority of the head and neck tumors. No colocalization was found between EPOR expression and pimonidazole binding indicating that the presence or absence of hypoxia did not necessarily indicate a distinct pattern of EPOR expression. The level of EPOR expression was not of prognostic significance in patients with head and neck cancer, although small effects of EPOR cannot be excluded because of the sample size of this study.

Thymidine analogues to assess microperfusion in human tumors.

PURPOSE: To validate the use of the thymidine analogues as local perfusion markers in human tumors (no labeling indicates no perfusion) by comparison with the well-characterized perfusion marker Hoechst 33342. METHODS AND MATERIALS: Human tumor xenografts from gliomas and head-and-neck cancers were injected with iododeoxyuridine (IdUrd) or bromodeoxyuridine (BrdUrd) and the fluorescent dye Hoechst 33342. In frozen sections, each blood vessel was scored for the presence of IdUrd/BrdUrd labeling and Hoechst in surrounding cells. The percentage of analogue-negative vessels was compared with the fraction of Hoechst-negative vessels. Collocalization of the two markers was also scored. RESULTS: We found considerable intertumor variation in the fraction of perfused vessels, measured by analogue labeling, both in the human tumor xenografts and in a series of tumor biopsies from head-and-neck cancer patients. There was a significant correlation between the Hoechst-negative and IdUrd/BrdUrd-negative vessels in the xenografts (r = 85, p = 0.0004), despite some mismatches on a per-vessel basis. CONCLUSIONS: Thymidine analogues can be successfully used to rank tumors according to their fraction of perfused vessels. Whether this fraction correlates with the extent of acute hypoxia needs further confirmation.

Colocalization of carbonic anhydrase 9 expression and cell proliferation in human head and neck squamous cell carcinoma.

PURPOSE: Tumor cells undergo a variety of biological changes under sustained hypoxic conditions, allowing cells to survive and retain their clonogenic potential. The purpose of this study is to relate the expression of the hypoxia marker carbonic anhydrase 9 (CA9) to the uptake of iododeoxyuridine (IdUrd), a marker of proliferation, in head and neck squamous cell carcinomas. Colocalization of IdUrd and CA9 may identify an important subpopulation of tumor cells that might be responsible for repopulation and disease progression. EXPERIMENTAL DESIGN: Expression of CA9, IdUrd labeling, and colocalization between IdUrd and CA9 was examined by immunohistochemistry in biopsies of head and neck squamous cell carcinomas. Biopsies were taken from 51 patients recruited between 1998 and 2001 after administration of the proliferation marker IdUrd. RESULTS: A large variation was observed between the tumors in CA9 expression (range 0-39%), IdUrd labeling (range 0-81%), and colocalization between IdUrd and CA9 [FId(CA9); range 0-53%]. FId(CA9), the fraction of IdUrd-labeled cells positive for CA9, was highest at an intermediate distance from the blood vessels (100-150 microm). IdUrd labeling was higher in T4 carcinomas relative to lower stage tumors (P = 0.04). High FId(CA9) correlated with the worst disease-free survival rates (P = 0.04). CONCLUSIONS: Colocalization between IdUrd labeling and CA9 expression was observed in head and neck squamous cell carcinomas, suggesting the presence of a population of tumor cells under intermediate hypoxic conditions which still has proliferative capacity. The size of this subpopulation may be indicative of tumor aggressiveness and is associated with the worst disease-free survival rates.

Tumor vascular architecture and function evaluated by non-invasive susceptibility MRI methods and immunohistochemistry.

PURPOSE: To investigate the physiological origins responsible for the varying blood oxygenation level dependent (BOLD) magnetic resonance imaging (MRI) responses to carbogen (95% O(2)/5% CO(2)) breathing observed with different tumor types. MATERIALS AND METHODS: Susceptibility contrast-enhanced MRI using the exogenous blood pool contrast agent NC100150 to determine blood volume and vessel size, and immunohistochemical-derived morphometric parameters, were determined in GH3 prolactinomas and RIF-1 fibrosarcomas, both grown in mice, which exhibited very different BOLD responses to carbogen. RESULTS: Administration of NC100150 increased the R(2)* and R(2) rates of both tumor types, and indicated a significant four-fold larger blood volume in the GH3 tumor. The ratio deltaR(2)*/deltaR(2) showed that the capillaries in the GH3 were two-fold larger than those in the RIF-1, in agreement with morphometric analysis. Carbogen breathing induced a significant 25% decrease in R(2)* in the GH3 prolactinoma, whereas the response in the RIF-1 fibrosarcoma was negligible. CONCLUSION: Low blood volume and small vessel size (and hence reduced hematocrit) are two reasons for the lack of R(2)* change in the RIF-1 with carbogen breathing. BOLD MRI is sensitive to erythrocyte-perfused vessels, whereas exogenous contrast agents interrogate the total perfused vascular volume. BOLD MRI, coupled with a carbogen challenge, provides information on functional, hemodynamic tumor vasculature.

Assessment of the neovascular permeability in glioma xenografts by dynamic T(1) MRI with Gadomer-17.

The uptake of Gadomer-17, as probed by fast dynamic T(1) measurements, was used to assess the vascular permeability surface-area product per leakage volume of tissue (k(Tofts)) of human glioma xenografts implanted in mice. With this approach we could discriminate between two types of glioma xenograft lines with a known difference in the perfused vascular architecture and degree of hypoxia. The T(1) data were analyzed according to the Tofts-Kermode compartment model. The fast-growing E102 tumor demonstrated a homogeneous distribution of the vascular permeability surface area across the tumor (mean k(Tofts) value = 0.18 +/- 0.05 min(-1)). The slowly growing E106 tumor showed a more heterogeneous pattern. Three perfused tumor areas with differences in vascular permeability surface area could be distinguished: a well-perfused periphery with high k(Tofts) values (0.24 +/- 0.04 min(-1)), perfused capillaries inside the tumor with low k(Tofts) values (0.108 +/- 0.026 min(-1)), and perfused capillaries adjacent to necrotic regions with high k(Tofts) values (0.29 +/- 0.10 min(-1)). On a different series of tumors, the hypoxic fractions were measured, and these were significantly higher in E106 tumors (0.14 +/- 0.05) compared to tumors of the E102 line (0.03 +/- 0.02).