Radiogenic Lymphangiogenesis in the Skin

【摘要】  The time course of microvascular changes in the environment of irradiated tumors was studied in a standardized human protocol. Eighty skin biopsies from 40 patients with previously treated primary breast cancer were taken from irradiated skin and corresponding contralateral unirradiated control areas 2 to 8 weeks, 11 to 14 months, or 17+ months after radiotherapy (skin equivalent dose 30 to 40 Gy). Twenty-two biopsies of 11 melanoma patients who had undergone lymph node dissection were used for unirradiated control. We found an increase of total podoplanin+ lymphatic microvessel density resulting mainly from a duplication of the density of smallest lymphatic vessels (diameter <10 µm) in the samples taken 1 year after radiation. Our findings implicate radiogenic lymphangiogenesis during the 1st year after therapy. The numbers of CD68+ and vascular endothelial growth factor-C+ cells were highly elevated in irradiated skin in the samples taken 2 to 8 weeks after radiotherapy. Thus, our results indicate that vascular endothelial growth factor-C expression by invading macrophages could be a pathogenetic route of induction of radiogenic lymphangiogenesis.
--------------------------------------------------------------------------------
Most knowledge on structural reorganization of microvessels after radiotherapy is derived from studies of acute radioreaction or on late changes that occur several years after radiation. In comparison with highly proliferative tissues, acute effects of radiation appear relatively late in the microvasculature of normal tissue. The nadir of microvessel density is to be found here within 50 to 70 days after a radiation dose equivalent to that usually used in solid neoplasms (50 to 60 Gy).1 The microvasculature is almost completely destroyed by very high doses (100 Gy) and must be rebuilt by the tissue following the principles of angiogenesis.2 In contrast to posttraumatic wound healing, neovascularization is ineffective in irradiated tissue, despite high vascular endothelial growth factor (VEGF) levels and endothelial cell activation.3 Atrophy and fibrosis are typical late effects of radiotherapy. They frequently first become evident by ischemia and radiogenic ulcer 10 or more years after the radiogenic damage.4 In this, rarefaction of the blood microvascular densities of the terminal stream bed and formation of bypass vessels that are clinically visible as telangiectases play a central role.5 Interestingly, the time between acute and late effects, which can be regarded as the time window for development of local recurrences or metastases in most solid neoplasms, is not well studied. We partially close this gap in this study.
The effect of an intermediate radio dose is another important aspect that has not yet been sufficiently investigated. Although the studies cited above mainly focused on tumoral target doses, investigation of the effect of a "tumor-environmental dose" (about two-thirds of the target volume dose) seems to be most promising, because the tumor environment is the major site of development of local recurrences or skin metastases.6
The detection of molecules that are relatively specifically expressed by lymphatic endothelial cells, like podoplanin, lymphatic vessel endothelial hyaluronate receptor 1 (LYVE-1), vascular endothelial growth factor receptor (VEGFR)-3, prospero-related homeobox gene PROX-1, desmoplakin-1 (2.17), and ß-chemokine receptor D6, has facilitated new insights into the molecular mechanisms that control lymphatic vessel development.7-12 Lymphangiogenesis is stimulated during embryogenesis, after trauma, in ischemic tissue, chronic wounds, acute lymphedema, or by malignant tumors.13-16 Two members of the VEGF family, VEGF-C and VEGF-D, are important for the induction of lymphangiogenesis.7,17-21 These factors are ligands of the lymphatic endothelial VEGF receptor-3 that is expressed on the surface of lymphatic endothelial cells. VEGF-C is produced by macrophages, dendritic cells, endothelial cells, platelets, basophilic granulocytes, lymphocytes, and tumor cells.15,22 Like other VEGFs, VEGF-C controls not only angiogenesis of blood and lymphatic vessels but also microvessel permeability and migration of endothelial cells.7,11,17,23
In this study, we investigated the changes of the lymphatic and blood microvasculature during the first years after postoperative radiotherapy for breast cancer in a standardized group of patients by comparing intraindividual control samples of unirradiated skin taken from exactly symmetrically contralateral areas. In addition, we used for comparison a group of melanoma patients who received no radiotherapy but were treated by operation (lymph node dissection).

【关键词】  radiogenic lymphangiogenesis

Materials and Methods

The study received approval by the local Ethics Committee of the Medical Faculty at the University of Halle. Informed consent was obtained individually from the patients.

Patients

We studied 80 samples of 40 consecutive female patients of the Breast Cancer Center of Martin Luther University Halle-Wittenberg after completed standard radiation therapy because of breast cancer (Table 1) . Patients were treated by operation, chemotherapy, and/or hormone therapy according to the standards of care (St. Gallen protocol).24 At the time of biopsy, patients were free of local recurrence or metastatic disease. Local exclusion criteria regarding the sampling regions were any clinical signs of radiation dermatitis or other inflammatory disease, palpable or sonographically detectable axillary lymph node swelling, and lymphedema of the biopsy area and its direct neighborhood (axilla and pectoral region). In contrast, clinically detectable lymphedema restricted to the arm of the irradiated site was allowed and regarded in the analysis (10 of 40 patients). General exclusion criteria were diabetes mellitus, autoimmune disease, human immunodeficiency virus infection, and reduced general condition (Karnofsky index <80%). Co-morbidity included hypertension (n = 8), arrhythmia (n = 3), bronchial asthma (n = 2), previously treated inactive pulmonary tuberculosis (n = 2), rheumatic polyarthritis (n = 1), angina pectoris (n = 1), psoriasis (n = 1), or chronic hepatitis C (n = 1). Besides specific anticancer therapy (Table 1) , patients were treated with L-thyroxin (n = 4), acetylsalicylic acid (n = 3), ß-blockers (n = 2), angiotensin-converting enzyme inhibitors (n = 2), or calcium antagonists (n = 2).

Table 1. Characteristics of the Patients and Tumors

Radiotherapy

Before sampling, patients had received standard radiotherapy with 6-MV photons via tangential fields (Table 1) . Some patients with large breasts and larger dose inhomogeneities received part of the radiation dose with 10-MV photons, mainly about 40% of the total dose (20 of 50 Gy). Radiotherapy was delivered in conventional fractionation with five fractions per week in all patients. No bolus was used. Three of 40 patients with four or more positive nodes received additional radiotherapy to the supraclavicular fossa; treatment was delivered via a single anterior portal with 6-MV photons, and the radiation field did not cover the skin biopsy side (see below). These patients with additional supraclavicular radiotherapy belonged to the patient groups "2 to 8 weeks" (n = 1), "11 to 14 months" (n = 1), or "17 to 25 months after radiotherapy" (n = 1). In 38 of 40 patients, the total dose calculated at the International Commission on Radiation Units and Measurements reference point within the target volume was 50.4 Gy (in 28 single doses of 1.80 Gy). One patient of 40 received 54 Gy (30 x 1.8 Gy), and one patient of 40 received 44.8 Gy (28 x 1.6 Gy). Both of these dose-deviant patients belonged to the same patient subgroup "2 to 8 weeks after radiotherapy" (see below). The estimated skin dose was 60 to 70% of the reference dose (see Discussion).

Skin Samples

Punch biopsies (6 mm) were taken 2 to 8 weeks (median, 4 weeks; n = 13 patients), 11 to 14 months (median, 12 months; n = 11), 17 to 25 months (median, 24; n = 8), or 37 to 157 months (median, 43.5; n = 8) after radiotherapy from the skin of the superior anterior axillary line that was directly included in the target region of all fractions of the radiotherapy (2 to 3 cm proximal from lymph node dissection scar) and from the corresponding unirradiated skin area of the opposite anterior axillary line of each patient for control. Biopsies were divided into two parts. One was snap-frozen in liquid nitrogen and stored at C70??C, and the other was fixed overnight in 4% formalin and paraffin-embedded.

Skin Samples from Melanoma Control Patients

Twenty-two samples from 11 melanoma patients who had not undergone radiation but underwent similar procedure of operation (lymph node dissection) were used for comparison. Eleven biopsies were taken 2 to 3 cm proximal to a lymph node dissection scar either from the axilla (9 of 11 cases) or from inguinal region (2 of 11), and 11 additional biopsies were taken from the exactly contralateral skin area of the same 11 patients: age, 61.8 ?? 12.8 years; sex, 10 female, one male; time between operation and biopsy median, 14.4 ?? 2.8 months (minimum, 9; maximum, 19 months); stage of disease (American Joint Committee on Cancer 200225 ), 1b/2a/3b/3c, n = 1/1/8/1; type of operation, radical lymph node dissection (9 of 11) or sentinel node biopsy (2 of 11); and number of dissected lymph nodes, 13.9 ?? 7.2. At the time of biopsy, patients were free of local recurrence or metastatic disease. Biopsies were fixed overnight in 4% formalin and paraffin-embedded.

Antibodies

Polyclonal rabbit anti-human podoplanin IgG was raised against the recombinant human homolog of the rat 43-kd glycoprotein podoplanin. Rabbit sera were affinity-purified using nitrocellulose strips containing recombinant protein, as described previously.26 Other primary antibodies were as follows: monoclonal mouse anti-human CD34, class II clone QBEnd10, isotype IgG1, (Dako, Glostrup, Denmark); monoclonal mouse anti-human CD68, clone KP1, isotype IgG1, (Dako); monoclonal mouse anti-human CD68, clone PG-M1, fluorescein isothiocyanate conjugate (Dako); rabbit anti-human LYVE-1 (AngioBio, Del Mar, CA); rabbit anti-human VEGF-C (Zymed Inc., South San Francisco, CA); and goat anti-human VEGF-D (R&D Systems, Minneapolis, MN). Fluorochrome-labeled secondary monoclonal antibodies were as follows: goat anti-rabbit IgG, F(ab')2-tetramethylrhodamine isothiocyanate (Santa Cruz Biotechnology Inc., Santa Cruz, CA) and donkey anti-goat IgG, F(ab')2-tetramethylrhodamine B isothiocyanate (Santa Cruz).

Histological Preparation of Paraffin Sections

Paraffin sections (5 µm) of each sample were stained with hematoxylin and eosin (H&E). For immunohistology, consecutive sections were deparaffinized in a graded ethanol series. Endogenous peroxidase was blocked (0.3% H2O2 in methanol). Then, the sections were microwaved (10 mmol/L citrate buffer, pH 6, four times for 5 minutes, and 600 W) for antigen demasking. The sections were incubated with anti-podoplanin (1:250), anti-CD34 (1:40), anti-CD68/KP1 (1:50), or anti-VEGF-C (1:100) for 30 minutes at room temperature. In VEGF-C or CD68 staining, undiluted rabbit or mouse primary antibody isotype controls (08-6199 or 08-6599; Zymed) were used for negative control. Further steps were performed with the Elite-ABC kit (Vector Laboratories, Burlingame, CA) according to the manufacturer??s instructions. Finally, the sections were counterstained with 25% hemalum (3 minutes; Merck, Darmstadt, Germany), rinsed with tap water, and mounted (gelatin mounting medium; Dako).

Immunofluorescence Investigations

Cryostat sections (5 µm) were triple-stained with fluorescein isothiocyanate conjugate of anti-CD68/PG-M1 (1:50; 30 minutes; room temperature), anti-VEGF-C (1:100; 30 minutes; room temperature), or anti-VEGF-D (1:80; 30 minutes; room temperature) and DNA fluorochrome (Hoechst 33258; Sigma). Tetramethylrhodamine isothiocyanate-fluorochrome-labeled secondary monoclonal antibodies were both applied at 1:100 (30 minutes). For demonstration of LYVE-1+ lymphatic microvessels, 8-µm cryostat sections were double-stained with anti-LYVE-1 (1:100; 30 minutes; room temperature) and DNA fluorochrome. Tetramethylrhodamine isothiocyanate-fluorochrome-labeled goat anti-rabbit IgG, F(ab')2 was applied 1:100 (30 minutes).

Negative Controls

In VEGF-C, VEGF-D, or LYVE-1 staining, 0.5 µg/ml primary antibody isotype controls of rabbit (Zymed) or goat (Abcam, Cambridge, UK) were used for negative control. In VEGF-C, negative controls were additionally preblocked with human recombinant VEGF-C (5 µg/ml; BioVision, Mountain View, CA).

Analysis

Sections were studied for general description of changes in comparison with the controls. Immunohistologically treated sections were evaluated in a "blinded" setting. At least three different sections of each sample were evaluated in the upper horizontal dermal plexus. Densities were calculated relative to skin surface (n/mm). The lymphatic vessels were grouped by diameter: <10 µm, 10 to 17 µm, 18 to 24 µm, and 25 µm (only in podoplanin investigations), as were the blood vessels: <12 µm, 12 to 15 µm, and 16 µm. For this, the shortest transversal axis of each vessel section was assessed under a measuring field ocular. The numbers of the visible microvessel sections were counted using the diameter groups. Only clearly identifiable microvessels were taken into account (Figure 1, j and k) . The interobserver coefficient of variation of this method was calculated in sixfold counting of the podoplanin-stained sections of 10 different cases with a total of 57 sections at 0.076 ?? 0.045.

Figure 1. Microphotographs of the upper dermal plexus of irradiated and naive control skin. aCf: H&E staining. a: Four weeks after radiation, sparse perivascular mononuclear infiltrate (arrow) and shrunken sebaceous glands (asterisk) were evident. b and c: Details of a (compare with asterisk or arrow). d: Greater enlargement shows microvessel degeneration (arrows). e: Twelve months after radiation, lymphatic spaces (arrow, inset), and disfigured microvessels were visible. f: Similar microvessel changes and, in this case, sweat duct changes (inset) could be found 23 months after radiotherapy. gCk: Podoplanin immunostaining (hematoxylin counterstained). g: In periaxillar normal control skin, small and large lymphatic vessels (arrows) can be found regularly in horizontal orientation. h: Four weeks after radiation, lymphatic vessels (asterisk) showed vertical enlargement of the lumina, but the number of small lymphatics (arrow) was comparable with the controls. i: One year after radiotherapy, the number of small horizontally oriented lymphatics (>10 µm diameter, arrows) was increased dramatically. j and k: Higher enlargement of podoplanin staining 11 months (j) or 17 months (k) after radiation: Only clearly identifiable podoplanin+ longitudinal structures were evaluated as lymphatics (arrows; E, epidermis). lCn: CD68 immunostaining (red). l: Low density of CD68+ cells (macrophages, arrows) in the normal control. m and n: High densities of CD68+ cells were characteristic 4 weeks after radiation. o and p: VEGF-C immunostaining (brown). Normal control (o) and irradiated skin (4 weeks after radiation) (p) showed significant difference in the densities of VEGF-C+ cells (arrows). Scale bars: 100 µm (a, eCi, l, and m); 50 µm (bCd, j, nCp, and insets in e and f); 25 µm (k).

Statistics were performed by SPSS 12.0.1 (Chicago, IL). The following statistical tests were used: Pearson correlation coefficient (r); Spearman correlation coefficient (); t-test for paired samples (t-test); Wilcoxon test, exact, two-sided significance (Wilcoxon test); Mann-Whitney U-test, exact, two-sided significance (U-test); and logistic regression. Categorical, ordinal covariates were median-dichotomized in logistic regression. In LYVE-1 immunofluorescence investigations, the number of positive cells or microvessel segments was calculated from five sections of each analyzed sample.

Results

Micromorphological Changes in Irradiated Skin

In the H&E-stained sections of irradiated skin of the breast cancer patients, a sparse perivascular-oriented mononuclear infiltrate and qualitative degeneration of microvessels, hair follicles, or sebaceous glands were evident in the samples harvested 2 to 8 weeks after the end of radiation (Figure 1, aCd) . Venules presented diameter enlargement. The basal epidermis showed single-cell necroses. Twelve months after radiation, a proportion of 10 to 30% disfigured microvessels was visible (Figure 1e) . These changes were persistently found in other samples from later stages that included a time up to 157 months after radiotherapy (Figure 1f) . In addition, basophilic degeneration of the subepidermal connective tissue could be found in some cases >3 years after radiation.

Lymphatic and Blood Microvessel Densities

Absolute Density Values

The mean density of total podoplanin+ lymphatic vessels in the samples of the treated skin of the breast cancer patients proved 18% higher than that of the individual control skin. This difference was significant (Table 2) . Detailed analysis showed that this was mainly a consequence of an increase of the smallest lymphatics (with a diameter <10 µm) in the irradiated skin. In all patient subgroups together, the mean density of small lymphatics <10 µm was 44% higher in the treated side versus control side (Table 2) . The source of this effect could be localized by time-subgroup analysis: We found more than duplication (107%) of the mean density of small lymphatics <10 µm in the subgroup of patients who had undergone radiotherapy 11 to 14 months earlier (Table 2 ; Figure 1, gCk ). This effect was not yet visible in samples gathered in the first weeks after therapy and tended to persist beyond 14 months after therapy (Table 2) . Neither the larger lymphatic microvessel calibers (10 µm) nor the total CD34+ blood microvessels showed significant radiogenic effect regarding the vessel densities (Table 2) .

Table 2. Densities of Microvessels, CD68+, VEGF-C+, or VEGF-D+ Cells in Skin of the Therapeutic and Contralateral Control Sides of 40 Cases of Breast Cancer

There was neither a significant difference nor a recognizable trend in all investigated microvessel density parameters between the two "late" subgroups (17 to 25 months and 37 to 157 months after radiotherapy). In addition, the size of each of these subgroups was small. Thus, both subgroups were evaluated and represented together (17+ months after radiotherapy, Table 2 ).

Intraindividual Correlation to the Controls

We found intraindividual correlation between the microvessel densities of the treated skin of the breast cancer patients and those of unirradiated individual control skin (total lymphatics, r = 0.415, P = 0.008; small lymphatics <10 µm diameter, r = 0.357, P = 0.024; small blood microvessels <12 µm, r = 0.534, P < 0.001; medium blood microvessels 12 to 15 µm, r = 0.668, P < 0.001; n = 40). A similar tendency of intraindividual correlations between the therapeutic and control sides of the lymphatic microvessel densities was found even in the melanoma control patients (total lymphatics, r = 0.549, P = 0.081; small lymphatics <10 µm diameter, r = 0.653, P = 0.029; lymphatics 10 to 17 µm, r = 0.559, P = 0.074; lymphatics >25 µm, r = 0.522, P = 0.100; n = 11). There was no significant intraindividual correlation in all other vessel diameters (Supplemental Table 1 at http://ajp.amjpathol.org). Interindividual differences between the normal control skin of different patients were considerable in all investigated vessel diameter groups (Table 2) . Moreover, there were stochastic differences in the normal values of the control skin between the "time subgroups." Thus, individual microvessel density in irradiated skin seemed to be a result of the normal individual state and the therapeutic effects. To overcome this problem, we used the individual ratio of treated skin to untreated control in addition to the absolute density.

T/C Ratios of Microvessel Densities

The T/C ratios of the densities of total podoplanin+ lymphatics or CD34+ blood microvessels of the different time groups are presented in Figure 2, a and b . Separate analysis of each vessel diameter subgroup revealed a clear effect of the time between radiotherapy and biopsy on the T/C ratios. Although the densities of the smallest podoplanin+ lymphatics (<10 µm) in the biopsies taken shortly after radiation (2 to 8 weeks) showed a T/C ratio of 0.91 ?? 0.74 (mean ?? SD), it was 2.04 ?? 1.09 in the biopsies taken 11 to 14 months after radiation.

Figure 2. Ratios of "treated/control" densities of different microvessel calibers during different time stages of the postradiation period. a: Lymphatics with four groups of microvessel diameter and three time groups after radiotherapy, respectively. b: Analogous figure of blood microvessel densities with three groups of microvessel diameter ranges. Intertime group significance analysis by U-test. Boxes include all values from 25th to 75th percentile; C, median; , largest to smallest observed values that are not outliers; , outlier more than 1.5 box lengths from the 25 or 75 percentile; and asterisk, outlier value >2 x SD from mean.

In the CD34+ blood microvessels, there was significant difference regarding the number of bigger vessels detectable. In the biopsies taken shortly after radiotherapy (2 to 8 weeks), the bigger blood microvessels >16 µm diameter showed a comparably high density with a mean T/C ratio of 2.09 ?? 1.67, but in the biopsies taken 11 to 14 months after radiation, this density was equal to values of the controls with a mean T/C ratio 0.93 ?? 0.65 (Figure 2b) .

Density of LYVE-1+ Lymphatic Microvessels

To audit the results of podoplanin staining, the density of LYVE-1+ lymphatic microvessels was investigated in the subgroup of patients who had undergone radiotherapy 11 to 14 months before biopsy (Supplemental Figure 1, A and B, at http://ajp.amjpathol.org). In total lymphatics, we found 3.24 ?? 0.94/mm in the irradiated skin and 1.98 ?? 0.59/mm in the unirradiated control skin (or skin of unirradiated controls). This difference was significant (P = 0.017, n = 10, Wilcoxon test).

These values showed individual correlation to the total lymphatic density results of podoplanin investigations (r = 0.698, P = 0.003, n = 20, Pearson correlation). Moreover, the mean values of the podoplanin results (given in Table 2 , first line) did not differ significantly from that of the results of LYVE-1 investigations given above (P = 0.14, n = 10 or P = 0.11, n = 10 for the irradiated or unirradiated biopsies, respectively, all Wilcoxon test).

CD68 and VEGF-C Expression

Increased densities of CD68+ and VEGF-C+ cells in the samples from the first weeks after radiotherapy (2 to 8 weeks) were some of the most striking effects found in this work (Table 2 ; Figures 1, lCp, and 3, a and b ; Supplemental Figure 1C at http://ajp.amjpathol.org). In the samples obtained 2 to 8 weeks after radiotherapy, the densities of CD68+ cells were elevated in the irradiated skin (94.7 ?? 31.4/mm) in comparison with the controls (62.8 ?? 22.3/mm). Likewise, the densities of the VEGF-C+ cells were elevated in this time group in the irradiated sites (21.1 ?? 7.7/mm) in comparison with the controls (12.4 ?? 6.3/mm). Although the number of CD68+ cells remained elevated in the samples taken thereafter, the frequency of VEGF-C+ cells showed "normalization" to the control values in the biopsies taken later (Figure 3, a and b) .

Figure 3. Densities of CD68+ cells (a), VEGF-C+ (b), or VEGF-D+ (c) cells. Dependence on the time after radiation and comparison with the data of unirradiated control skin. Wilcoxon test, comparison of controls with irradiated samples; U-test, between irradiated samples. Boxes include all values from 25th to 75th percentile; C, median; , largest to smallest observed values that are not outliers; , outlier more than 1.5 box lengths from the 25 or 75 percentile.

Coexpression of CD68 and VEGF-C

We found high intraindividual correlation between CD68 and VEGF-C expression in all samples (r = 0.510; P < 0.001, n = 44, paraffin-embedded material). We could demonstrate by triple immunofluorescence that VEGF-C was highly coexpressed with CD68 (cryosections). Evaluation of 100 randomly selected VEGF-C+ cells (defined as VEGF-C+ cells with DNA fluorochrome staining nucleus) of eight sections revealed 82% coexpression of VEGF-C with CD68 (Supplemental Figure 1C at http://ajp.amjpathol.org).

Expression of VEGF-D

In the cryosections, we could demonstrate that VEGF-D expression was low (Figure 3c ; Supplemental Figure 1D at http://ajp.amjpathol.org). Furthermore, we found no significant difference between the irradiated skin and the individual control samples (densities, 3.9 ?? 2.9 and 3.6 ?? 2.5/mm VEGF-D+ cells, respectively; P = 0.351, Wilcoxon test, n = 20).

Correlation with Other Patient Features

In the irradiated skin, we found a negative correlation of chronological age with the T/C ratio and the absolute densities of small lymphatics <10 µm (r = C0.384, P = 0.017, n = 38; or r = C0.343, P = 0.03, n = 40, respectively).

Clinically detectable lymphedema occurred in 10 of 40 cases (four cases in the patient group 2 to 8 weeks after radiotherapy, and six cases of persistent lymphedema in the patients 11 months after radiotherapy). In the six samples from previously irradiated skin of patients with persistent lymphedema, the mean density of small lymphatics (<10 µm) was 0.43 ?? 0.40/mm (T/C ratio, 1.05 ?? 0.63), whereas in the 21 cases without lymphedema, this value was 1.23 ?? 1.02/mm (T/C ratio, 2.16 ?? 1.76) (P = 0.033 for absolute values and P = 0.209 for T/C ratios, U-test). There was no significant difference in the other diameter groups of the lymphatic microvessels (Supplemental Table 2 at http://ajp.amjpathol.org). No correlation of histological tumor features (T or N status, number of positive nodes, primary grading, estrogen, or progesterone receptor status), drugs, or co-morbidity was detectable with microvessel densities, macrophage, or VEGF-C+ macrophage densities.

Influence of Therapy

In the 26 samples harvested 2 to 8 weeks after radiotherapy, we found correlations between previous radiotherapy and total CD68+ macrophage density ( = 0.49, P = 0.012) as well as between radiotherapy and the density of VEGF-C+ macrophages ( = 0.55, P = 0.006). Axillary dissection, too, correlated with total macrophage density in these samples ( = 0.45, P = 0.020) but not with the number of VEGF-C+ macrophages. Nevertheless, logistic regression revealed radiotherapy as the only significant factor that predicted total macrophage density (odds ratio, 7.50; 95% confidence interval, 1.31 to 43.03; P = 0.024). The number of dissected lymph nodes, mastectomy (versus breast-conserving therapy), or systemic therapies (chemotherapy and hormone therapy) showed correlation with neither the total macrophage density nor the density of VEGF-C+ macrophages.

In the 54 biopsies that were taken 11 or more months after radiotherapy, the density of small lymphatics (<10 µm) correlated with previous radiotherapy ( = 0.29, P = 0.031). All other therapeutic factors (axillary dissection, number of dissected lymph nodes, mastectomy versus breast-conserving therapy, or systemic therapies) showed no correlation with the density of small lymphatics. Of all therapeutic procedures, logistic regression demonstrated that radiation was the only significant predictor of the density of small lymphatics (<10 µm) in these samples (odds ratio, 4.04; 95% confidence interval, 1.30 to 12.58; P = 0.016).

To delimit the influence of surgery from radio effects, a group of unirradiated patients was investigated for the presence of podoplanin+ lymphatic vessels in the upper dermis using the same conditions. Results were compared with those of the irradiated patients. In this comparison, 22 skin samples from 11 melanoma patients who had undergone lymph node dissection (but no radiation) 9 to 19 months previously were studied. We found no significant difference between the absolute lymphatic vessel counts of samples from operated and contralateral control areas (Table 3) . T/C ratios showed no significant distinction from normal (Table 3) . Furthermore, both the operated and the control skin of the melanoma patient control group showed high agreement with the normal control values of the breast cancer patients (Tables 2 and 3) .

Table 3. Densities of Lymphatic Microvessels in the Skin of the Treated and Contralateral Control Sides of 11 Melanoma Patients Who Had Undergone Lymph Node Dissection but No Radiotherapy

Discussion

In the present work, we examined for the first time, to our knowledge, the influence of radiotherapy on lymph and blood microvessel quantities in normal skin in a standardized protocol in humans. The model fulfilled the following criteria: realistic dose schedule, tumor environment radiation dose, focus on the first years after radiotherapy, and true comparability of the samples with compensation of known intraindividual differences in the microvasculature of different body regions27 as well as of interindividual differences. The latter demonstrated to be important, because we could show that lymphatic and blood microvessel densities of irradiated skin are highly influenced by the individual normal state. Thus, in addition to absolute values, we introduced the ratio of microvessel densities "of the treated skin to the control skin" ("T/C ratios") as an instrument to investigate the therapy-induced changes in our patients. The here-established strategy of strict intraindividual controls taken from exactly the same contralateral position is recommendable for similar investigations in the future.

A major result of our study is that, in contrast to studies with higher doses,2,28-30 neither blood vessels nor lymphatics are reduced in the early after-radiation phase in the tumor environment that received an intermediate radiation dose. This, together with our finding of VEGF-C increase and the known regulatory effects of VEGFs on tumor cells,31-34 might be of importance for the potency of local recurrence and metastasis development in case of remaining tumor cells in the tumor environment.

Summarizing our results on lymphatic vessels after radiotherapy, we found significantly higher densities of lymphatic microvessels in biopsies taken 1 or more years after radiotherapy than in the control skin and in the skin shortly after radiotherapy. The increase was mainly produced by a duplication of the density of the small lymphatic capillaries within the 1st year after radiation. This constellation could be regarded as a strong sign of interim lymphangiogenesis. It should be pointed out in this context that the analyzed subgroups were relatively small.

One of the most interesting aims of this study was to provide evidence that lymphogenic effects do not result from surgery. In all patients, the biopsy sides were clearly spatially separated from the sides of previous surgical activities (>2 cm) while they were covered completely by radiation. Thus, this study was primarily designed to investigate the influence of radiation but not that of wound healing. In accordance, logistic regression demonstrated that the increase in lymphatic densities in the breast cancer patients was linked to previous radiotherapy but not to operative therapy or other covariates. Furthermore, we compared our results with a group of unirradiated patients who had undergone lymph node dissection for melanoma. In these unirradiated controls, we found no similar increase of lymphatic densities in the 1st year after therapy (neither significance nor tendency). This allows the conclusion that the increase in the density of lymphatics during the 1st therapeutic year is primarily radiogenic.

It should be noted that this effect was observed after a total dose of less than 40 Gy administered with single doses of about 1.5 Gy or less because the radiation dose in the superficial skin in the treatment of breast cancer (with the radiation technique mentioned) is about 30% lower than the reference dose in the target volume. According to standard radiobiological models such as the linear-quadratic model, this dose (assumed skin dose of 35 Gy in single doses of 1.4 Gy) would be considered equivalent to a total dose of 30 Gy in single doses of 2 Gy (assuming an -to-ß ratio of 3 Gy for late effects). This dose is far lower than the tolerance dose currently assumed for blood vessels. Thus, significant and persistent changes with regard to lymphangiogenesis might be provoked by radiation doses that are much lower than previously assumed. This hypothesis is supported by recent data suggesting that microenvironmental changes and especially angiogenic cytokines may play a critical role in the radiation response of normal and tumor tissues.35

In our study, the chronological age of the patients seemed to suppress the individual lymphangiogenic potential. Furthermore, cases with coexisting chronic lymphedema showed a reduction of radiogenic effect in the small lymphatics. Lymphedema occurs in up to 40% of breast cancer patients after axillary lymph node dissection in combination with radiotherapy.36 Two peaks of incidence of lymphedema can be distinguished: reversible acute lymphedema (in the first 3 months after radiotherapy) and chronic lymphedema (starting frequently after 1 or more years). Our data give first indication of a possible role of reduced individual ability to develop small lymphatics in the development of chronic lymphedema.

Looking for the potential mechanism of induction of lymphangiogenesis, we found increased numbers of CD68+/VEGF-C+ cells in the first weeks after radiation, suggesting a regulation of lymphangiogenesis by VEGF-C-expressing macrophages. Interestingly, a slightly elevated number of CD68+ macrophages persisted during the first years after radiotherapy, whereas VEGF-C expression was reduced to the control values. Approximately 20% of the CD68+ cells were VEGF-C+ in the early samples (taken 2 to 8 weeks after radiation) in our study. A similar frequency of VEGF-C expression was observed previously in tumor-associated macrophages of the peritumoral tissue.15 Moreover, we have previously shown that the density of macrophages correlates with lymphatic capillary proliferation.15 Macrophages seem to support lymphangiogenesis in two different ways, either by trans-differentiating and direct incorporation into the endothelial layer or by stimulating division of pre-existent local lymphatic endothelial cells.37 VEGF-C itself stimulates recruitment of macrophages and increases inflammatory response in malignant tumors.13 In this, VEGF-C seems to be chemotactic for macrophages via VEGFR-3.13 The role of VEGF-C stimulation on lymphangiogenesis could be demonstrated recently by blockade of VEGFR-3.38 Other studies have shown that lymphangiogenesis is driven by co-stimulation with VEGF-C.39,40 In contrast to VEGF-C, we found no major stimulation of VEGF-D expression. VEGF-D (beside VEGF-C) is another ligand of the tyrosine kinase VEGFR-3 receptor. It has previously been shown that VEGF-D is important in neonatal and tumor lymphangiogenesis but seems to play a subordinate role in comparison with VEGF-C.15,18,41-43

Metastasis of breast cancer occurs primarily through the lymphatic system. It has previously been shown that induction of tumor lymphangiogenesis by tumor cell-derived VEGF-C promotes breast cancer metastasis.10 Thus, it can be hypothesized that both lymphangiogenesis and the direct tumor cell-stimulating potential of VEGF-C might contribute to the development of local recurrences in case of surviving tumor cells in the after-irradiation tumor environment. It should be pointed out that our results do not contradict the indisputable curative antitumoral effect of the radiation in the target fields. Radiogenic lymphangiogenesis can be regarded as a factor for local recurrences only in case of tumor cells outside of surviving tumor cells in the environment of the irradiated tumor. Furthermore, radiation therapy clearly increases the risk of lymphedema. Thus, the increase in small lymphatic vessels with radiation does not seem to increase the functional lymphatic drainage in and around an irradiated tumor bed.

It should be further noted that some aspects cannot be completely clarified with our material and the setting of our study (sampling followed routine follow-up schedule). The first aspect is that the number of VEGF-C-expressing macrophages was highest at 0.5 to 2 months after therapy, whereas high lymphatic vessel numbers were observed in samples of approximately 1 year after therapy. Thus, the time window of the increase of lymphatic density can be identified between >2 and <11 months after therapy by our material. From the early increase of VEGF-C expression, it is probable that the increase of lymphatic density might take place in the earlier part of the identified time window??a period that needs further investigation. Second, it is reasonable to assume that some factors other than VEGF-C may be contributing as well, possibly at the time point later than 2 months after therapy. Third, although observation on the role of macrophages in lymphogenesis has been made in peritumoral tissue or renal transplants,15,44 our findings indicate, but cannot completely prove, a similar process in the skin.

We used CD34 for investigation of blood microvessels. It has been shown previously that CD34 is expressed in blood endothelial cells and lymphatic endothelial cells of human malignancies.45 Nevertheless, in our setting, it can be regarded as a marker for blood vessels in the environment of non-neoplastic skin.45,46 According to the results of our study, blood microvessels seem to be less influenced by intermediate-dose radiotherapy. We found qualitative microscopic radiation effects on microvessels (disfiguring). The only quantitative alteration of blood microvessels was an early and transient increase of the number of bigger vessels (>16 µm) as had already been shown previously in porcine skin.1 It can be concluded from the vessel diameter and microstructure that this effect was generated by an acute postexpositional increase in the number of enlarged venules.27 Most probably, this is an effect of higher tissue perfusion in the postinflammatory phase of the early postradiation period.

Summarizing, we demonstrated a considerable increase in the density of small lymphatic microvessels during the 1st year after an intermediate (tumor-environmental) dose of radiotherapy in the skin. We could demonstrate that radiogenic lymphangiogenesis might be mediated by macrophage-expressed VEGF-C. Age and occurrence of lymphedema were found to be linked with impaired capacity to develop lymphatic capillaries. Blood microvasculature revealed a transient expansion of postcapillary venules during the first weeks, but neither signs of angiogenesis nor substantial depletion of microvessel density during the first years after intermediate-dose radiotherapy was observed.

【参考文献】
  Archambeau JO, Ines A, Fajardo LF: Correlation of the dermal microvasculature morphology with the epidermal and the endothelial population changes produced by single X ray fractions of 1649, 2231 and 2619 rad in swine. Int J Radiat Oncol Biol Phys 1985, 11:1639-1646

Takahashi S, Sugimoto M, Kotoura Y, Sasai K, Oka M, Yamamuro T: Long-term changes in the haversian systems following high-dose irradiation: an ultrastructural and quantitative histomorphological study. J Bone Joint Surg Am 1994, 76:722-738

Tsai JH, Makonnen S, Feldman M, Sehgal CM, Maity A, Lee WM: Ionizing radiation inhibits tumor neovascularization by inducing ineffective angiogenesis. Cancer Biol Ther 2005, 4:1395-1400

Fajardo LF: The pathology of ionizing radiation as defined by morphologic patterns. Acta Oncol 2005, 44:13-22

Reinhold HS: The influence of radiation on blood vessels and circulation. Chapter IV. Structural changes in blood vessels. Curr Top Radiat Res Q 1974, 10:58-74

Rofstad EK, Mathiesen B, Henriksen K, Kindem K, Galappathi K: The tumor bed effect: increased metastatic dissemination from hypoxia-induced up-regulation of metastasis-promoting gene products. Cancer Res 2005, 65:2387-2396

Jussila L, Alitalo K: Vascular growth factors and lymphangiogenesis. Physiol Rev 2002, 82:673-700

Ord?ñez NG: Podoplanin: a novel diagnostic immunohistochemical marker. Adv Anat Pathol 2006, 13:83-88

Breiteneder-Geleff S, Matsui K, Soleiman A, Meraner P, Poczewski H, Kalt R, Schaffner G, Kerjaschki D: Podoplanin, novel 43-kd membrane protein of glomerular epithelial cells, is down-regulated in puromycin nephrosis. Am J Pathol 1997, 151:1141-1152

Skobe M, Hawighorst T, Jackson DG, Prevo R, Janes L, Velasco P, Riccardi L, Alitalo K, Claffey K, Detmar M: Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nat Med 2001, 7:192-198

Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, Saksela O, Kalkkinen N, Alitalo K: A novel vascular endothelial growth factor, VEGF-C, is a ligand for the Flt4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. EMBO J 1996, 15:290-298

Hong YK, Harvey N, Noh YH, Schacht V, Hirakawa S, Detmar M, Oliver G: Prox1 is a master control gene in the program specifying lymphatic endothelial cell fate. Dev Dyn 2002, 225:351-357

Skobe M, Hamberg LM, Hawighorst T, Schirner M, Wolf GL, Alitalo K, Detmar M: Concurrent induction of lymphangiogenesis, angiogenesis, and macrophage recruitment by vascular endothelial growth factor-C in melanoma. Am J Pathol 2001, 159:893-903

Karpanen T, Alitalo K: Lymphatic vessels as targets of tumor therapy? J Exp Med 2001, 194:F37-F42

Schoppmann SF, Birner P, Stockl J, Kalt R, Ullrich R, Caucig C, Kriehuber E, Nagy K, Alitalo K, Kerjaschki D: Tumor-associated macrophages express lymphatic endothelial growth factors and are related to peritumoral lymphangiogenesis. Am J Pathol 2002, 161:947-956

Karpanen T, Egeblad M, Karkkainen MJ, Kubo H, Yla-Herttuala S, Jaattela M, Alitalo K: Vascular endothelial growth factor C promotes tumor lymphangiogenesis and intralymphatic tumor growth. Cancer Res 2001, 61:1786-1790

Ferrara N: VEGF: an update on biological and therapeutic aspects. Curr Opin Biotechnol 2000, 11:617-624

Achen MG, Jeltsch M, Kukk E, Makinen T, Vitali A, Wilks AF, Alitalo K, Stacker SA: Vascular endothelial growth factor D (VEGF-D) is a ligand for the tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Proc Natl Acad Sci USA 1998, 95:548-553

Veikkola T, Jussila L, Makinen T, Karpanen T, Jeltsch M, Petrova TV, Kubo H, Thurston G, McDonald DM, Achen MG, Stacker SA, Alitalo K: Signalling via vascular endothelial growth factor receptor-3 is sufficient for lymphangiogenesis in transgenic mice. EMBO J 2001, 20:1223-1231

Jeltsch M, Kaipainen A, Joukov V, Meng X, Lakso M, Rauvala H, Swartz M, Fukumura D, Jain RK, Alitalo K: Hyperplasia of lymphatic vessels in VEGF-C transgenic mice. Science 1997, 276:1423-1425

Karkkainen MJ, Haiko P, Sainio K, Partanen J, Taipale J, Petrova TV, Jeltsch M, Jackson DG, Talikka M, Rauvala H, Betsholtz C, Alitalo K: Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat Immunol 2004, 5:74-80

Wartiovaara U, Salven P, Mikkola H, Lassila R, Kaukonen J, Joukov V, Orpana A, Ristimaki A, Heikinheimo M, Joensuu H, Alitalo K, Palotie A: Peripheral blood platelets express VEGF-C and VEGF which are released during platelet activation. Thromb Haemost 1998, 80:171-175

Kaipainen A, Korhonen J, Mustonen T, van Hinsbergh VW, Fang GH, Dumont D, Breitman M, Alitalo K: Expression of the fms-like tyrosine kinase 4 gene becomes restricted to lymphatic endothelium during development. Proc Natl Acad Sci USA 1995, 92:3566-3570

Goldhirsch A, Glick JH, Gelber RD, Senn HJ: International Consensus Panel on the treatment of primary breast cancer. V: update 1998. Recent Results Cancer Res 1998, 152:481-497

Balch CM, Buzaid AC, Soong SJ, Atkins MB, Cascinelli N, Coit DG, Fleming ID, Gershenwald JE, Houghton A, Jr, Kirkwood JM, McMasters KM, Mihm MF, Morton DL, Reintgen DS, Ross MI, Sober A, Thompson JA, Thompson JF: Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol 2001, 19:3635-3648

Breiteneder-Geleff S, Soleiman A, Kowalski H, Horvat R, Amann G, Kriehuber E, Diem K, Weninger W, Tschachler E, Alitalo K, Kerjaschki D: Angiosarcomas express mixed endothelial phenotypes of blood and lymphatic capillaries: podoplanin as a specific marker for lymphatic endothelium. Am J Pathol 1999, 154:385-394

Braverman IM, Keh-Yen A: Ultrastructure of the human dermal microcirculation. IV. Valve-containing collecting veins at the dermal-subcutaneous junction. J Invest Dermatol 1983, 81:438-442

Brown WR, Thore CR, Moody DM, Robbins ME, Wheeler KT: Vascular damage after fractionated whole-brain irradiation in rats. Radiat Res 2005, 164:662-668

Li YQ, Chen P, Jain V, Reilly RM, Wong CS: Early radiation-induced endothelial cell loss and blood-spinal cord barrier breakdown in the rat spinal cord. Radiat Res 2004, 161:143-152

Archambeau JO, Ines A, Fajardo LF: Response of swine skin microvasculature to acute single exposures of X rays: quantification of endothelial changes. Radiat Res 1984, 98:37-51

Fukumoto S, Morifuji M, Katakura Y, Ohishi M, Nakamura S: Endostatin inhibits lymph node metastasis by a down-regulation of the vascular endothelial growth factor C expression in tumor cells. Clin Exp Metastasis 2005, 22:31-38

Ueda M, Hung YC, Terai Y, Kanda K, Kanemura M, Futakuchi H, Yamaguchi H, Akise D, Yasuda M, Ueki M: Vascular endothelial growth factor-C expression and invasive phenotype in ovarian carcinomas. Clin Cancer Res 2005, 11:3225-3232

Watanabe O, Kinoshita J, Shimizu T, Imamura H, Hirano A, Okabe T, Aiba M, Ogawa K: Expression of a CD44 variant and VEGF-C and the implications for lymphatic metastasis and long-term prognosis of human breast cancer. J Exp Clin Cancer Res 2005, 24:75-82

Tsai PW, Shiah SG, Lin MT, Wu CW, Kuo ML: Up-regulation of vascular endothelial growth factor C in breast cancer cells by heregulin-ß 1: a critical role of p38/nuclear factor-B signaling pathway. J Biol Chem 2003, 278:5750-5759

Anscher MS, Chen L, Rabbani Z, Kang S, Larrier N, Huang H, Samulski TV, Dewhirst MW, Brizel DM, Folz RJ, Vujaskovic Z: Recent progress in defining mechanisms and potential targets for prevention of normal tissue injury after radiation therapy. Int J Radiat Oncol Biol Phys 2005, 62:255-259

Morrell RM, Halyard MY, Schild SE, Ali MS, Gunderson LL, Pockaj BA: Breast cancer-related lymphedema. Mayo Clin Proc 2005, 80:1480-1484

Kerjaschki D: The crucial role of macrophages in lymphangiogenesis. J Clin Invest 2005, 115:2316-2319

Pytowski B, Goldman J, Persaud K, Wu Y, Witte L, Hicklin DJ, Skobe M, Boardman KC, Swartz MA: Complete and specific inhibition of adult lymphatic regeneration by a novel VEGFR-3 neutralizing antibody. J Natl Cancer Inst 2005, 97:14-21

Cursiefen C, Chen L, Borges LP, Jackson D, Cao J, Radziejewski C, D??Amore PA, Dana MR, Wiegand SJ, Streilein JW: VEGF-A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J Clin Invest 2004, 113:1040-1050

Goldman J, Le TX, Skobe M, Swartz MA: Overexpression of VEGF-C causes transient lymphatic hyperplasia but not increased lymphangiogenesis in regenerating skin. Circ Res 2005, 96:1193-1199

Niki T, Iba S, Tokunou M, Yamada T, Matsuno Y, Hirohashi S: Expression of vascular endothelial growth factors A, B, C, and D and their relationships to lymph node status in lung adenocarcinoma. Clin Cancer Res 2000, 6:2431-2439

Karpanen T, Wirzenius M, Makinen T, Veikkola T, Haisma HJ, Achen MG, Stacker SA, Pytowski B, Yla-Herttuala S, Alitalo K: Lymphangiogenic growth factor responsiveness is modulated by postnatal lymphatic vessel maturation. Am J Pathol 2006, 169:708-718

Baldwin ME, Halford MM, Roufail S, Williams RA, Hibbs ML, Grail D, Kubo H, Stacker SA, Achen MG: Vascular endothelial growth factor D is dispensable for development of the lymphatic system. Mol Cell Biol 2005, 25:2441-2449

Kerjaschki D, Huttary N, Raab I, Regele H, Bojarski-Nagy K, Bartel G, Krober SM, Greinix H, Rosenmaier A, Karlhofer F, Wick N, Mazal PR: Lymphatic endothelial progenitor cells contribute to de novo lymphangiogenesis in human renal transplants. Nat Med 2006, 12:230-234

Fiedler U, Christian S, Koidl S, Kerjaschki D, Emmett MS, Bates DO, Christofori G, Augustin HG: The sialomucin CD34 is a marker of lymphatic endothelial cells in human tumors. Am J Pathol 2006, 168:1045-1053

Hirakawa S, Hong YK, Harvey N, Schacht V, Matsuda K, Libermann T, Detmar M: Identification of vascular lineage-specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol 2003, 162:575-586

Singletary SE, Allred C, Ashley P, Bassett LW, Berry D, Bland KI, Borgen PI, Clark GM, Edge SB, Hayes DF, Hughes LL, Hutter RV, Morrow M, Page DL, Recht A, Theriault RL, Thor A, Weaver DL, Wieand HS, Greene FL: Staging system for breast cancer: revisions for the sixth edition of the AJCC Cancer Staging Manual. Surg Clin North Am 2003, 83:803-819

作者单位：From the Departments of Dermatology,* Gynaecology, and Radiotherapy, Martin Luther University Halle-Wittenberg, Halle, Germany; the Department of Radiotherapy, University of L?beck, L?beck, Germany; the Institute of Pathology,¶ Medical University of Vienna, Vienna, Austria; and the Department o