This is the first study on the quantitative assessment of thymidine phosphorylase (TP) activity in patients with nonsmall lung carcinoma. TP is identical to the platelet-derived endothelial cell growth factor with its angiogenic activity. Thus, it is believed that TP activity in tumor tissues plays an important role in disease progression.
Using a sandwich enzyme immunoassay, the TP activity in lung carcinoma tissues was measured quantitatively in 39 patients with primary lung carcinoma who underwent pulmonary lobectomy between July 1999 and May 2000.
The mean value of TP activity in tumor tissues was significantly higher than the level in normal lung tissues (226 U/mg protein vs. 46 U/mg protein, respectively; P < 0.0001). TP activity in normal lung tissues was high in male patients (male vs. female, respectively: 56.1 U/mg protein vs. 29.3 U/mg protein; P = 0.001) and in heavy smokers (Brinkmann index [BI] ≥ 600 [57.9 U/mg protein] vs. BI < 600 [31.7 U/mg protein]; P = 0.001). Conversely, the TP activity in tumor tissues was correlated with neither gender nor smoking status. Although there was no difference in the TP activity among histologic types, well-differentiated tumors exhibited a significantly lower level of TP activity compared with the activity in both moderately and poorly differentiated tumors. However, the TP activity in tumor tissues was not correlated with disease progression.
Thymidine phosphorylase (TP) is an enzyme that catalyses the reversible phosphorolysis of thymidine and is identical to the platelet-derived endothelial cell growth factor with its angiogenic activity.1, 2 In patients with several different malignant neoplasms, it has been reported that TP activity in tumor tissues is higher compared with the activity in normal tissue counterparts.3–5 Thus, it has been suggested that the high TP activity in tumor tissues plays an important role in disease proliferation and metastasis. Concerning lung carcinoma, a few studies using immunohistologic methods revealed that about 50% of nonsmall cell lung carcinoma cells expressed TP6 and that the level of TP expression was correlated with both disease progression and patient prognosis.7, 8 However, TP also was expressed by nonneoplastic cells in both normal lung tissues and tumor stroma, i.e., bronchiolar epithelial cells, alveolar macrophages, and fibroblasts.6 Stromal cells naturally affect both proliferative and metastatic potential of tumor cells. To clarify the role of TP activity in tumor tissues, therefore, it is considered necessary to measure quantitatively the sum of TP activity in tumor cells and in stromal cells. In the current study, we measured the total TP activity in freshly isolated tumor tissues with an enzyme immunoassay and assessed its correlation with clinicopathologic factors.
MATERIALS AND METHODS
Between July 1999 and May 2000, 39 patients with primary lung carcinoma underwent surgery (pulmonary lobectomy) at National Kyushu Medical Center. Patient characteristics are summarized in Table 1. There were 24 men and 15 women with a mean age of 67 years (range, from 44–79 years). The tumors were classified histologically as adenocarcinoma in 24 patients, squamous cell carcinoma in 11 patients, and adenosquamous cell carcinoma in 4 patients. The pathologic staging of all patients' disease is reported according to the 1997 revised International Staging System for Lung Cancer.9 Prior to surgery, no patients had received any treatment for carcinoma of the lung.
Table 1. Patient Characteristics
M: male; F: female.
Mean, 67 (range, 44–79)
Histologic type (adenocarcinoma/squamous/others)
Grade of differentiation (well/moderate/poor)
Pathologic T factor (T1/T2/T3–T4)
Pathologic N factor (N0/N1/N2)
Pathologic stage (I/II/III/IV)
Preparation of Tumor Tissues
Small pieces (2 mm in greatest dimension) of both tumor tissues and nontumor-bearing lung tissues (referred to hereinafter as normal lung tissues) were obtained immediately from surgically resected lung specimens. The tissues were homogenized in 10 mM Tris-HCl buffer, pH 7.4, containing 15 mM NaCl, 1.5 mM MgCl2, and 50 μM potassium phosphate, then centrifuged at × 10,000 g for 90 minutes. The supernatants were stored at − 80 °C until use.
Enzyme Immunoassay to Determine TP Activity
The TP activity in tumor tissues was measured with a sandwich enzyme immunoassay developed by the Nippon Roche Research Center (Kanagawa, Japan). Detailed systems and procedures for the enzyme immunoassay have been published previously.10 Briefly, test samples and standard solutions of TP (the colon carcinoma cell line HCT 116 homogenate) were dispensed onto a plate coated with the first anti-TP antibody (antibody 104B; immunoglobulin M). The plate was incubated at 37 °C for 1 hour and then washed; incubated with the second anti-TP antibody (antibody 232-2; IgG1) at 37 °C for 1 hour, and washed; incubated with antimouse immunoglobulin G (IgG) conjugated with horseradish peroxidase (Bio-Rad, Hercules, CA) at 37 °C for 30 minutes and washed; and incubated with a substrate solution containing 3,3′,5,5′-tetramethylbenzidine (TMB) and H2O2 (TMB microwell peroxidase substrate system; KPL) for 10–20 minutes at room temperature. The level of TP activity was estimated by measuring its absorbency at 450 nm with a plate reader (model 3550; Bio-Rad) and was calibrated with activity measured in the standard solutions. One unit corresponds to the TP level of the standard enzyme solution, which phosphorylizes 5′-deoxy-5-fluorouridine to 5-fluorouracil (5-FU) at the rate of 1 μg 5-FU per hour.
Values for TP activity were compared between two groups with the Student t test or among three or more than three groups with an analysis of variance (ANOVA). If the ANOVA showed that a comparison was statistically significant, then individual pairs of comparisons were made using the Fisher and Scheff tests. Significance was assigned at P < 0.05.
In all 39 patients examined, the TP activity in tumor tissues was higher compared with the activity in normal lung tissues. The mean value of TP activity in tumor tissues was significantly higher compared with that in normal lung tissues (226 U/mg protein vs. 46 U/mg protein, respectively; P < 0.0001).
Figure 1 shows that there was a significant difference in the TP activity in normal lung tissues between male and female patients and between heavy smokers and the other patients. The mean TP activity in normal lung tissues in male patients was significantly higher compared with the activity in female patients (56.1 U/mg protein vs. 29.3 U/mg protein, respectively; P = 0.001). Conversely, the mean TP activity in normal lung tissues in patients with a Brinkmann index (BI) ≥ 600 was significantly higher compared with the activity in patients with a BI < 600 (57.9 U/mg protein vs. 31.7 U/mg protein, respectively; P = 0.001). The TP activity in normal lung tissues was not correlated with patient age.
Next, the correlations between TP activity in tumor tissues and various clinicopathologic factors were analyzed. Compared with the TP activity in normal lung tissues, the TP activity in tumor tissues was correlated with neither gender nor smoking status (Fig. 1). Although there was no difference in the TP activity among histologic types, the well-differentiated tumors exhibited a significantly lower level of TP activity compared with both moderately and poorly differentiated tumors (P = 0.029; Fig. 2). However, the TP activity in tumor tissues was not correlated with disease progression (tumor [T] factor and lymph node [N] factor).
To our knowledge, this is the first study on the quantitative assessment of TP activity in patients with nonsmall cell lung carcinoma. The TP activity in tumor tissues, as expected, was significantly higher compared with the activity in normal lung tissues. However, the high TP activity in tumor tissues did not reflect the malignant potentials of the disease. There were significant variations in TP activity in normal lung tissues according to gender or smoking status. Those variations in normal (nontumor-bearing) tissues have not been reported in other organs.
Unlike normal lung tissues, the TP activity in tumor tissues exhibited a correlation with neither gender nor smoking status, indicating that the high level of TP activity was not dependent on the nature of individual lungs but was unique to tumor tissues. However, the TP activity in tumor tissues was correlated with neither the histologic tumor type nor the stage of disease. Koukourakis et al.8 reported that abundant stroma was correlated directly with TP overexpression by fibroblasts and that fibroblast TP overexpression accompanied a high degree of macrophage infiltration. We observed that TP activity in patients with granulomatous lesions, such as tuberculosis, was as high as the activity in tumor tissues (398 U; n = 3 patients). Thus, it is possible that the major part of TP activity in tumor tissues may be attributed to the stromal cells, which consist of macrophages, lymphocytes, and fibroblasts. The current finding that TP activity was lower in well-differentiated tumors compared with activity in both moderately and poorly differentiated tumors may be attributable to the amount of stroma, depending on the grade of differentiation.
Conversely, TP activity in normal lung tissues was high in male patients and heavy smokers, patients who are prone to suffer from interstitial inflammation to some extent. The high level of TP activity in normal lung tissues may be attributed to some active inflammatory cells in the interstitium and may reflect the presence of interstitial inflammation. In fact, in 8 of 14 patients with TP levels in normal lung tissues > 50 U/mg protein, interstitial inflammation (infiltration of chronic inflammatory cells) was identified histologically. It has been reported that TP is induced by various inflammatory cytokines, i.e. interferon, interleukin-1, tumor necrosis factor-α, and transforming growth factor-β.11 Therefore, we are now investigating the usefulness of TP activity in lung tissues as a marker for interstitial pneumonia.
The authors thank Nippon Roche Research Center for technical assistance in determining TP activity levels.