TNF receptor p55 and IL-8₇₂ and IL-8₇₇ isoforms: blood and urine levels in breast cancer patients

Abstract

In the preliminary study reported here, 37 patients with breast cancer and 10 healthy volunteers were analyzed for soluble TNF-R p55 and two variants of IL-8 consisting of 72 and 77 amino acid residues (IL-8₇₂ and IL-8₇₇, respectively) in their blood and urine with novel ELISA test systems. The clinical/prognostic values of determining these inflammatory cytokines at different stages of the cancer process appeared to depend on the treatment course being evaluated. In contrast to expectations, it was noted that there was a stabile tendency for decreased TNF-R p55 and IL-8₇₂ levels in the plasma and urine of breast cancer patients as compared with levels observed with healthy controls. Moreover, patients that underwent polychemotherapy treatments were notable for significant decreases in IL-8₇₂ and TNF-R p55 levels in their blood plasma; these findings contrasted with significant increases in these parameters in these patients’ urine. Interestingly, the IL-8₇₇ isoform that now appeared both in the urine and plasma of patients was not detectable before initiation of the polychemotherapy. In spite of all these findings, individual fluctuations among these parameters still do not allow us to establish, at this time, any strong correlations between these values with any particular breast cancer stage or a type of treatment. Nonetheless, while the results here are preliminary, they demonstrate that testing for TNF-R, along with IL-8 isoforms, in the blood plasma and urine could potentially present a valid means for monitoring of the overall immune and disease progress/remission status in breast cancer patients. Ongoing studies with larger patient sample sizes, as well as collecting and analyzing samples at multiple timepoints—to minimize the potential influence of any inherent variability in cytokine levels in humans—will hopefully allow us to specify what these preliminary results reported here suggest, i.e., the potential utility of this experimental approach for determining disease progression or efficacy of treatment in cancer patients.

Keywords::

• Tumor necrosis factor
• interleukin-8
• breast cancer

Introduction

According to our current understanding, both tumor necrosis factor (TNF) and interleukin-8 (IL-8) occupy important places within the family of pro-inflammatory cytokines. The biological effects of each of these cytokines are believed to play an important role in the pathogenesis of the growth of malignancies (Aggarwal et al., 2006).

Tumor necrosis factor can be produced by tumor cells and other cell populations, including macrophages, monocytes, and neutrophils, that help to create in the microenvironment of tumors. In general, TNF plays a large role in the development of anorexia and cachexia in various cancer diseases. It is shown that tissue macrophages produce TNF on the tumor perimeter and express mRNA for TNF (Ahmed et al., 2001). However, the presence of TNF in blood, secretions, or tissues has no significant diagnostic value, since the protein is quickly absorbed by cells and tissues. For example, in the process of septic shock, TNF levels in the blood reach a peak concentration over 1.5 hr; however, after 3.5–4.0 hr, levels are not detected because of binding to target tissues (Redl et al., 1995; Casey, 2000; Lin et al., 2007). However, the role of TNF in pathology can be readily identified. This is possible because the process of TNF binding to a specific receptor protein (p55) leads to the conversion of p55 into a soluble form. The soluble TNF receptor p55 receptor (TNF-R p55) then circulates in the blood (for at least one day) and it is a stable marker of systemic and local inflammatory reactions induced by TNF (Redl et al., 1995).

In studies attempting to relate p55 levels with cancers, highly interesting data was obtained when patients with different forms of cancer, including malignant lymphoma, acute or chronic myeloma, were examined (Or et al., 1996). An average level of p55 in patients with lymphoma and myeloma was 3.5 ng/mL (range of 1.8–18.8 ng/mL). An increased p55 level (i.e., 3.0–10.0 ng/mL) was also observed in all patients suffering from a variety of oncogynecologic cancers (e.g., ovarian cancer, cervical cancer, etc.) (Grosen et al., 1993; Strieter et al., 1993; Warzocha et al., 1997). It is also established that a high level of soluble TNF-R p55 in the blood is a sensitive marker of actively progressing and recurrent and metastazing tumors. This means that the changes in TNF-R p55 levels in the blood are important characteristics of the malignant process, and that the definition of TNF-R p55 levels in cancer patients may have a prognostic value.

IL-8 is a non-glycated protein with molecular weight of ≈ 8 kDa. IL-8 is a very important factor for angiogenesis and is one of the main chemokines for neutrophils. The role of IL-8 in neutrophil chemotaxis is well illustrated by the fact that injection of IL-8 induces an inflammatory reaction accompanied by massive leukocyte infiltration (Or et al., 1996). In contrast, neutralization of IL-8 in vivo via injection of antibodies against IL-8 almost completely inhibits the inflammation and infiltration in response to bacterial lipopolysaccharide (LPS) endotoxin, with a >90% reduction in leukocyte levels in tissues in response to the LPS by >90% (Dumont et al., 2000). The special role that IL-8 plays in the pathogenesis of malignant tumor growth is due to its great ability to promote angiogenesis and matrix production (Kuhlmann et al., 2009), processes that are necessary for the growth of solid tumors and the avoidance against immune attack(s). Such influence by IL-8 is due, in part, to its effect on endothelial cells—through expression bacterial lipopolysaccharide (LPS) endotoxin, with a >90% reduction in leukocyte levels in tissues in response to the LPS—that leads to their active proliferation (Belperio et al., 2000; Wang et al., 2005; Pohl and Lenz, 2008). In addition, an important feature of the regulation of the synthesis of IL-8 in cells is activation of gene expression for IL-8 under the influence of either/both hypoxia and/or acidosis (Shi et al., 1999; Xu et al., 1999).

It is known that tumors of a size <1 mm require vascularization for their growth (Folkman, 1971). In rapidly-growing tumors, hypoxic conditions and acidosis form due to insufficient output of metabolites (Shubik, 1982; Xu and Fidler, 2000). On one hand, this inhibits the growth and proliferative activity of the tumor cells, making them less susceptible to cytostatic agents and irradiation (Tannock, 1982). One the other hand, at the same time, the hypoxia and acidosis create conditions supporting the activation of IL-8 synthesis by tumor cells, a process that has the ability to result in strong angiogenic activity. Many authors have established a correlation between the intensity of malignant tumor growth and an ability to IL-8 production (Xu and Fidler, 2000; Hantschel et al., 2008; Uchikawa et al., 2009). However, it remains unknown which isoform of IL-8 (i.e., IL-8₇₂ or IL-8₇₇) is the primary form that tumor cells produce in vivo, despite the fact that the answer to this question is of great interest. This interest is reinforced by the fact that IL-8₇₇ has specific pro-apoptotic and chemotactic activity for malignant cells (Terui et al., 1998, 1999; Ramjeesingh et al., 2003) and is predominantly produced in the fetus (Maheshwari et al., 2009).

Data in the literature indicate that IL-8 production by tumor cells in different parts of the tumor is different, and that IL-8 levels around the perimeter of solid tumors could be significantly high. Along with this, the relevant data concerning IL-8 levels in the blood of cancer patients does not have a clear character, i.e., during inflammatory processes, high levels of IL-8 appear in the urine (Mohkam et al., 2008) and at sites of local inflammation – but not in the blood (Uchikawa et al., 2009). This might be due to the previously-described phenomenon of the formation of IL-8 complexes with IgG antibodies, a reaction that then interferes with the identification of IL-8 in blood using the conventional ELISA for an IL-8₇₂ isomer (Sylvester et al., 1992). Thus, the commercial ELISA kits available today for the determination of IL-8 in biological fluids is not fully suitable for use in scientific and diagnostic research.

The sensitivity of commercial ELISA test systems for determining TNF-R p55 levels in human blood is at 20–50 ng/mL. Voitenok and colleagues have developed several laboratory ELISA test systems, in particular one for TNF-R p55 that is comparable to the best commercial analogues, and ELISAs for the two isoforms of IL-8, i.e., IL-8₇₂ and IL-8₇₇ (Ko et al., 1992). The ELISA for IL-8₇₇ allows any Investigator to identify the levels of IL-8 in the blood of cancer patients and, until recently, had no analogues. It is assumed that the N-pentapeptide of IL-8₇₇ molecules that is recognized with the ELISA for IL-8₇₇ is located on the outside of the complexes that form between IL-8 and any blocking IgG.

Based on the above facts, we hypothesized that the identification/measurements of the TNF-R p55 and IL-8 isoforms in the blood and urine of cancer patients, taking into account the clinical features of the malignant process, could be useful for prediction and evaluation the effectiveness of anti-cancer treatments. The specific selection of breast cancer as a focus of our research here was due, in great part, to the increase in the frequency of this pathology in both the Ukraine and other countries.

Materials and methods

Patients

A total of 37 patients with histologically-verified diagnoses of breast cancer (age range 25–77 years; median = 52 years) who were undergoing examination and/or treatment at the Breast Cancer Branch of the National Cancer Institute (Kiev, Ukraine), were involved in this study. Among them, 5 were verified to have Stage I disease, 16 with Stage IIA, 9 with Stage IIB, 5 with Stage IIIA, and 2 with Stage IIIB. All patients were examined before or between rounds of (and after) treatment (specifically, polychemotherapy anti-relapsing courses) at least at 1–2 timepoints after their cancer was diagnosed. Integrated treatment for breast cancer consisted of neo-adjuvant bloc (polychemotherapy, radiation therapy of regional zone metastasis) and/or radical surgery operation, and/or adjuvant bloc of treatment (polychemotherapy and radiation therapy) alone. The control group in these studies was composed of 10 healthy (i.e., non-cancer-bearing) women, age range 21 to 47 years (median = 35 years).

Preparation of urine

Morning urine (collected in a special tank provided to each patient/control subject) was used for analysis. From the tank, 10 mL of urine was removed and placed in centrifuge tubes, and then centrifuged at 300 xg (1500 rpm) for 15 min to remove any cells/debris. The interval between sample collection and centrifugation did not exceed 30 min, and all samples were held at 18–20°C during the post-collection period. The cleared samples of urine (1.0–1.2 mL) were then aliquoted into small Eppendorf tubes (at least two tubes/each patient) and then placed in a −20°C freezer (Rosenlew, Pori, Finland) until analysis.

Preparation of blood plasma

Fasting blood (10 mL) from the cubital vein was collected (in the morning) into polypropylene tubes containing heparin (Deltalab, Barcelona, Spain). The blood was mixed with the heparin by gentle rocking (at least five times) and avoiding foaming. The cellular elements were then removed by centrifugation at 200 xg (1000 rpm) for 10 min. The interval between the collection and centrifugation of the blood did not exceed 30 min and all samples were held at 18–20°C during the post-collection period. The resultant plasma (1.0–1.2 mL) was then carefully recovered and aliquoted into small Eppendorf tubes (at least two tubes/each patient). The tubes were marked/logged and then placed in a −20°C freezer (Rosenlew) until analysis.

ELISA for quantification of soluble receptor TNF p55 and isoforms of IL-8

In general, ELISA was performed as described previously (Nashkevich et al., 2002). Mouse monoclonal antibodies (B5 and A11) specific to natural human TNF-R p55 antigen, as well as mouse monoclonal antibodies N11 (specific to IL-877) and 4C (for use in the capture steps of the assay), and the clone products 3A and H6 (used for the latter detection steps) were provided by Nikolai N. Voitenok (Fund for Molecular Hematology & Immunology, Moscow, Russia). As IL-872 is a constituent of IL-877 (i.e., actual antigen used to generate the three ‘anti-IL-872’ antibodies is the 72-mer protein + an AVLPR [Arg-Val-Leu-Pro-Arg] pentapeptide), each of these latter three antibodies by their nature must recognize both IL-877 and IL-872 (i.e., ‘total’ IL-8). Thus, in these studies, the specific amount of IL-872 in any given sample was deduced from the mathematical differences between the respective ELISA levels of “total IL-8” and that specifically of IL-877.

For the actual assays, capture monoclonal mouse anti-human anti-p55 antibody B5 or monoclonal anti-IL-8 antibodies 4C or N11 were absorbed (at 5 μg/ml in 0.1 M carbonate buffer [pH 9.6] in volumes of 100 μl/well) overnight at 4°C onto 96-well flat-bottom plates (Costar, Corning Inc., Corning, NY) (at 5 μg/mL in 0.1 M carbonate buffer (pH 9.6) in volumes of 100 μL/well). The plates were then washed twice with 170 μL/well of 10 mM sodium phosphate buffer containing 0.3 M NaCl and 0.05% Tween-20. The wells were then filled with incubation buffer (10 mM sodium phosphate buffer containing 0.3 M NaCl, 0.05% Tween-20, and 0.5% bovine serum albumin [BSA, Sigma-Aldrich, St. Louis, MO]) in volumes of 100 μL/well, and incubated for 40 min at room temperature on a shaker platform (Titertek, Flow Laboratories, Meckenheim, Germany). The plates were then washed again twice.

After the washings, to the first 14 wells (i.e., 7 doublets) were added the dilutions of standard TNF-R p55, i.e., natural purified human TNF-R p55 (Engelmann et al., 1990) provided by Dr. David Wallach (Weizmann Institute, Rehovot, Israel) or recombinant eukaryote expressed human p55 (Centocor Inc., Malvern, PA), recombinant IL-8₇₂ (Institute of Ultrapure Biopreparates, St. Petersburg, Russia), or IL-8₇₇ (Peprotech, Rocky Hill, NJ), in volumes of 100 μL/well; each standard had been prepared in incubation buffer at two-fold dilutions (i.e., final concentration/well ranging from 1.0 ng/mL to 0.016 ng/mL). The wells of one set of doublets were filled with an equivalent volume of incubation buffer as a negative control. To the other wells on the plate were added samples of urine or blood plasma (100 μL/well) that were diluted in incubation buffer at least two or more times (i.e., 1:1 dilutions and beyond).

The plates were then incubated for 1 hr at room temperature on a shaker platform. After the incubation, all wells were washed three times as described above, and then the wells received biotin-labeled upper monoclonal mouse anti-human antibodies anti-p55 A11 or anti-IL-8 3A (100 μl/well) prepared in the incubation buffer (at dilutions of 1:2000 or 1:5000, respectively). The plates were then incubated a further 40 min at room temperature. Following this incubation, all wells were washed three times as described above, and then the wells received peroxidase-conjugated streptavidin (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA; 100 μL/well; 1:3000 in incubation buffer). The plates were then incubated 30 min at room temperature, then washed four times as above. To detect levels of antibody bound, TMB solution (prepared according to manufacturer’s [Serva, Heidelberg, Germany] instructions) was added to each well (100 μL/well). The reaction was stopped after 10 min (by addition of 1 N H₂SO₄ solution; 100 μL/well), and the optical density in each well then determined at 450/650 nm using an ELISA processor plate reader (Behring ELISA Processor II, Frankfurt, Germany). The sensitivities of ELISAs in these analyses were 0.005 ± 0.002 ng/mL for IL-8₇₂, 0.030 ± 0.015 ng/mL for IL-8₇₇, and 0.015 ng/mL for TNF-R p55.

Statistical analysis

Statistical analysis was performed using the software package Statistica 6.0 (StatSoft, Inc., Tulsa, OK). Because there were not normal distributions in the groups and the groups were small, non-parametric statistics for continuous variables were applied. Continuous variables were expressed as medians and ranges. The comparative analysis between any two groups was carried out using the Mann–Whitney U-test. To compare the values of measured parameters within a group, the Wilcoxon rank test was applied. To establish the correlation links between measured parameters within a group, the Spearman’s rank correlation coefficient (r) was calculated. Statistical significance was defined as P < 0.05 (two-sided).

Results

Content of soluble receptor TNF p55 and IL-8 isoforms in plasma and urine of patients with breast cancer

According to the data presented in Table 1, the content of soluble TNF receptor p55 and IL-8₇₂ in the urine of the patients with breast cancer, though trending to be on average lower, was not significantly different when compared with that in healthy donors (i.e., P > 0.05). These parameters had no differences in blood plasma (P > 0.05) without taking into account the specific of treatment and stage of malignant process. In general, there was a wide value range of TNF-R p55 content and IL-8₇₂ in the plasma and urine of the cancer patients and healthy donors. The level of the IL-8₇₇ isoform in the peripheral blood and urine of donors and patients with breast cancer in general was <0.030 (± 0.015) ng/mL, a value that was below the sensitivity of this ELISA test system. A significant positive correlation between the measured parameters in the plasma and urine was found only for IL-8₇₂ in the healthy donors group.

Table 1.  Comparative characteristic of content soluble TNF receptor p55 and isoforms of IL-8 in blood plasma and urine of patients with breast cancer and healthy donors.

Investigated groups	Blood plasma^a (all as ng/mL)			Urine^a (all as ng/mL)			Plasma vs. urine (^bP and ^cr values)
	p55	IL-872	IL-877	p55	IL-872	IL-877	p55	IL-872
Patients with breast cancer (n=37)	4.300 (1.100–13.300)	0.020 (0.003–0.160)	<0.030^†	3.500 (0.015–23.400)	0.020 (0.003–0.650)	<0.030^†	0.439 (0.075)	0.529 (0.163)
Healthy donors (n=10)	4.300 (1.800–7.200)	0.025 (0.007–0.180)	<0.030^†	4.900 (2.000–10.700)	0.045 (0.007–0.160)	<0.030^†	0.533 (0.399)	0.315 (0.889)^*
Patients vs. donors (^bP values)	0.404	0.481	—	0.184	0.310	—

^aContinuous variables are expressed as medians and ranges (in parentheses). ^†Below ELISA sensitivity.

^bMann–Whitney U-test (continuous variables). All differences are statistically insignificant (P>0.05).

^cSpearman’s rank correlation coefficient (r values are in parentheses). ^*Positive correlation is statistically significant (P < 0.05).

Content of soluble TNF receptor p55 and IL-8 isoforms in plasma and urine of patients with breast cancer depending on a stage of disease

When the patients were sub-categorized into groups depending on disease stage, the assays revealed that there were, in general, no significant differences for all the measured parameters either between any stage of breast cancer and healthy donors or between Stage I and any other stages of disease (P > 0.05) (Table 2). A statistically significant difference was found only with respect to IL-8₇₂ plasma levels at Stage IIA of the disease (median = 0.010 ng/mL) as compared to those levels in the healthy donors’ samples (median = 0.025 ng/mL). At this Stage, the statistically-significant positive correlation between TNF-R p55 plasma levels (median = 3.550 ng/mL) and TNF-R p55 urine levels (median = 1.700 ng/mL) was also noted (r = 0.542; P < 0.05). However, due to the large dispersion of individual levels within the group, these TNF-R p55 median values did not reach the level of statistically-significant differences (P > 0.05) required by our protocol. With regard to the patients at all stages of the disease, both the TNF-R p55 and IL-8₇₂ levels in the plasma and urine were either at or below/above those in the control volunteer samples without any stabile regularity. The only significantly lower value was seen in the samples from patients with Stage IIIA disease.

Table 2.  Comparative characteristic of content soluble TNF receptor p55 and IL-8₇₂ in plasma and urine of patients with breast cancer in the context of cancer stage.

Investigated groups	^aBlood plasma (all ng/mL)		^aUrine (all ng/mL)		Plasma vs. urine (^bP and ^cr values)
	p55	IL-872	p55	IL-872	p55	IL-872
Patients with breast cancer (n=37)
Stage I (n = 5)	4.800 (2.000–13.300)	0.030 (0.005–0.050)	5.500 (0.015–23.400)	0.030 (0.007–0.319)	1.000 (0.400)	0.690 (0.205)
Stage IIA (n = 16)	3.550 (1.500–5.800)	0.010* (0.003–0.040)	1.700 (0.015–12.900)	0.012 (0.003–0.514)	0.224 (0.542)*	0.491 (0.330)
Stage IIB (n = 9)	4.700 (1.100–7.000)	0.040 (0.005–0.160)	5.200 (0.400–16.000)	0.040 (0.007–0.650)	0.605 (0.067)	0.796 (0.066)
Stage IIIA (n = 5)	3.600 (1.700–5.000)	0.030 (0.003–0.048)	3.300 (0.200–4.300)	0.007 (0.003–0.100)	0.421 (0.800)	1.000 (0.100)
Stage IIIB (n = 2)	7.400 (5.100–9.700)	0.050 (0.040– 0.060)	5.500 (4.200–6.800)	0.055 (0.030–0.080)	0.667 (NA)	1.000 (NA)
Healthy donors (n = 10)	4.300 (1.800–7.200)	0.025 (0.007–0.180)	4.900 (2.000–10.700)	0.045 (0.007–0.160)	0.533 (0.399)	0.315 (0.889)*

NA: not applicable.

^aContinuous variables are expressed as medians and ranges (in parentheses). *Difference from healthy donors is statistically significant (P < 0.05). All other differences, including Stage I vs. Stages IIA, IIB, IIIA, and IIIB, are statistically insignificant (P>0.05). Mann–Whitney U-test (continuous variables).

^bMann–Whitney U-test (continuous variables). All differences are statistically insignificant (P < 0.05) (without parentheses).

^cSpearman’s rank correlation coefficient (r values are in parentheses). ^*Positive correlation is statistically significant (P < 0.05).

Among breast cancer patients (at all Stages) as well as with the healthy donors, the levels of the IL-8₇₇ isoform—both in the plasma and urine—were below the level of detectability by the ELISA system employed here (data not shown).

Content of soluble TNF receptor p55 and IL-8 isoforms in plasma and urine of patients with breast cancer after a treatment

Analysis of these cytokines’ concentrations in the plasma and urine of patients with breast cancer—who were distributed in groups depending on the type of treatment course (Table 3)—also revealed an absence any stable regularity in the levels of the measured parameters. With respect to the plasma or urine p55 and IL-8₇₂ levels in the patients in any of the treatment groups, there were neither statistically-significant differences/trends (compared to the values seen with the healthy controls) that were detected nor any significant correlation(s) between its levels in plasma and urine within a group.

Table 3.  Comparative characteristic of content soluble TNF receptor p55 and IL-8₇₂ in plasma and urine of patients with breast cancer before and after treatment.

Investigated groups	^aBlood plasma (all ng/mL)		^aUrine (all ng/mL)		Plasma vs. urine (^bP and ^cr values)
	p55	IL-872	p55	IL-872	p55	IL-872
Before treatment (n = 5)	1.700* (1.500–5.200)	0.005 (0.003–0.048)	3.500 (0.200–18.600)	0.007 (0.003–0.319)	0.421 (0.462)	0.690 (0.300)
Neo-adjuvant bloc (n = 6)	2.300 (1.400–9.700)	0.045 (0.003–0.100)	2.850 (0.400–6.800)	0.014 (0.003–0.080)	0.818 (0.609)	0.180 (0.147)
Surgery (n = 11)	4.300 (1.500–13.300)	0.017 (0.003–0.050)	2.100 (0.015–23.400)	0.030 (0.003–0.514)	0.519 (0.100)	0.193 (0.021)
Neo-adjuvant bloc, surgery (n = 9)	4.600 (1.100–5.900)	0.020 (0.005–0.040)	4.300 (0.600–16.000)	0.007 (0.003–0.100)	0.796 (0.126)	0.436 (0.144)
Surgery, adjuvant bloc (n = 10)	3.100 (1.900–7.000)	0.009 (0.004–0.160)	3.500 (0.900–6.700)	0.034 (0.007–0.794)	0.739 (0.109)	0.052 (0.056)
Neo-adjuvant bloc, surgery, adjuvant bloc (n = 2)	3.900 (2.000–5.800)	0.018 (0.006–0.030)	2.508 (0.015–5.000)	0.122 (0.007–0.236)	0.667 (NA)	0.667 (NA)
Healthy donors (n = 10)	4.300 (1.800–7.200)	0.025 (0.007–0.180)	4.900 (2.000–10.700)	0.045 (0.007–0.160)	0.533 (0.399)	0.315 (0.889)*

NA; not applicable.

^aContinuous variables are expressed as medians and ranges (in parentheses). *Difference from healthy donors is statistically significant (P < 0.05). All other differences, including the group before treatment vs. any group after treatment, are statistically insignificant (P>0.05); Mann–Whitney U-test (continuous variables).

^bMann–Whitney U-test (continuous variables). All differences are statistically insignificant (P>0.05) (without parentheses).

^cSpearman’s rank correlation coefficient (r values are in parentheses). ^*Positive correlation is statistically significant (P < 0.05).

Only in the case of patients examined prior to any treatment initiation were the values of plasma TNF-R p55 found to be significantly different (i.e., lower) than those in samples from the healthy control donors (P< 0.05). As was the case cited above, the levels of IL-8₇₇ isoform were below the ELISA assay sensitivity, for all measurements (data not shown).

Dynamics of TNF receptor p55 and IL-8 isoforms in plasma and urine of patients with breast cancer before and after adjuvant polychemotherapy courses

The data presented in Table 4 demonstrate that the patients with breast cancer who have received any courses of anticancer therapy prior to a course of polychemotherapy presented with a stable tendency to have lower levels of TNF-R p55 and IL-8₇₂ in their blood plasma and urine as compared to values associated with healthy control donors. However, immediately after polychemotherapy, a significant decrease in plasma IL-8₇₂ and TNF-R p55 levels was revealed. Interestingly, in conjunction with these drops in plasma levels, the levels of these parameters were strongly increased in the patients’ urine (P < 0.05). Furthermore, IL-8₇₇ isoform also was now detected in the urine and plasma; before the polychemotherapy treatment, this isoform had routinely been undetectable.

Table 4.  Dynamics of TNF receptor p55 and IL-8 isoforms in plasma and urine of patients with breast cancer before and after adjuvant polychemotherapy courses.

Investigated groups		^aBlood plasma (all ng/mL)			^aUrine (all ng/mL)			Plasma vs. urine (^bP and ^cr values)
		p55	IL-872	IL-877	p55	IL-872	IL-877	p55	IL-872
Patients with breast cancer (n = 10)	BAP	3.000 (1.100–4.800)	0.014 (0.003–0.100)	<0.030^†	2.150 (0.400–15.100)	0.007 (0.003–0.514)	<0.030^†	0.796 (0.689)*	0.739 (0.240)
	AAP	2.300* (1.900–5.200)	0.006*, ** (0.003–0.015)	0.058 (0.045–0.105)	4.200 (2.400–6.500)	0.031 (0.010–0.794)	0.048 (0.045–0.076)	0.002* (0.556)	0.0002* (0.031)
Healthy donors (n = 10)		4.300 (1.800–7.200)	0.025 (0.007–0.180)	<0.030^†	4.900 (2.000–10.700)	0.045 (0.007–0.160)	<0.030^†	0.533 (0.399)	0.315 (0.889)*

^aContinuous variables are expressed as medians and ranges (in parentheses). *Difference from healthy donors is statistically significant (P < 0.05); Mann–Whitney U-test (continuous variables). **Difference from BAP is statistically significant (P < 0.05); Wilcoxon’s test (continuous variables).

^†Below ELISA sensitivity.

^bMann–Whitney U-test (continuous variables). *Difference is statistically significant (P < 0.05) (without parentheses).

^cSpearman’s rank correlation coefficient (r values are in parentheses). ^*Positive correlation is statistically significant (P < 0.05).

AAP; after adjuvant polychemotherapy; BAP; before adjuvant polychemotherapy.

Individual levels of TNF receptor p55 and IL-8 isoforms in plasma and urine of patients with breast cancer before and after treatments

Table 5 presents the set of individual data for blood plasma and urine levels of TNF-R p55 and IL-8 isoforms in patients with breast cancer before (Patients #1–5) and after treatment (Patients #5-10). Patients #6-10 were tested at least twice during the course of their respective treatments. The first tests analyzed these patients immediately after surgery, while the second took place after a course of polychemotherapy; this represented a 2-3-wk interval between the takings of blood and urine samples. All patients included in this group were diagnosed with primary Stage IIA breast cancer, and they had normal biochemical and clinical parameters in their blood and urine. These all relate also to patient #5 (Stage IIA), who was tested before treatment and after a course of polychemotherapy, and also to Patient #4 (Stage IIA), who was tested only before treatment. Patients #1–3, who were at different stages of breast cancer (i.e., Stages I, IIB, or IIIA) than these other patients, were assessed only once before treatment; they also had normal biochemical and clinical parameters in their blood and urine.

Table 5.  Individual levels of TNF receptor p55 and IL-8 isoforms in plasma and urine of patients with breast cancer.

Patient number	Patient age	Cancer stage	Treatment	Blood plasma (all ng/mL)			Urine (all ng/mL)
				p55	IL-872	^†IL-877	p55	IL-872	^†IL-877
1	43	I	Before treatment	2.000	0.005	<0.045	18.600	0.319	<0.045
2	46	IIB	Before treatment	5.200	0.040	<0.015	3.500	0.300	<0.015
3	47	IIIA	Before treatment	1.700	0.048	<0.045	0.200	0.004	<0.045
4	66	IIA	Before treatment	1.700	0.004	<0.045	3.700	0.007	<0.045
5	53	IIA	Before treatment	1.500	<0.003^†	<0.045	3.200	<0.003^†	<0.045
			Polychemotherapy	1.900	0.015	<0.045	2.400	0.048	<0.045
6	67	IIA	Surgery	3.500	0.017	<0.045	7.200	0.043	<0.045
			Polychemotherapy	2.000	0.005	<0.045	4.100	0.011	<0.045
7	47	IIA	Surgery	2.500	0.008	<0.045	12.900	0.514	<0.045
			Polychemotherapy	2.300	0.004	<0.045	3.400	0.794	<0.045
8	40	IIA	Surgery	1.500	0.004	<0.045	10.000	0.166	<0.045
			Polychemotherapy	2.800	0.004	<0.045	2.900	0.034	<0.045
9	47	IIA	Surgery	4.300	0.030	<0.015	1.100	<0.007^†	<0.015
			Polychemotherapy	1.900	0.007	<0.045	3.500	0.010	<0.045
10	47	IIA	Surgery	4.500	0.010	<0.015	0.600	0.017	<0.015
			Polychemotherapy	3.500	0.004	<0.045	37.600	0.036	<0.045
Healthy donors (n = 10) (medians)				4.440	0.041	<0.015	5.710	0.051	<0.015

^†Below ELISA sensitivity.

In compliance with the common data (described above), the analysis of individual TNF-R p55 and IL-8 isoforms levels in the plasma and urine of patients with breast cancer—before and after treatments—does not show any stable correlations between these markers with any stage of cancer or type of cancer treatment. However, presenting the data as medians (Table 6) clearly demonstrates a tendency that correlates with the data presented above in Table 4. In particular, as compared to healthy control donors and within of Group 2 (Surgery vs. Polychemotherapy), the lower TNF-R p55 and IL-8₇₂ levels in the plasma and urine of patients with breast cancer before treatment displayed a tendency to increase after surgery and again decrease immediately after polychemotherapy.

Table 6.  Comparative characteristic of content soluble TNF receptor p55 and IL-8₇₂ in plasma and urine of patients with breast cancer before treatment (Group 1) and after treatment (Group 2) with surgery (Point 1) and polychemotherapy (Point 2).^a

Patients	^bBlood plasma (all ng/mL)		^bUrine (all ng/mL)		Plasma vs. urine (^cP and ^dr values)
	p55	IL-872	p55	IL-872	p55	IL-872
Before treatment (Group 1, n = 5, Patients 1–5)	1.700* (1.500–5.200)	0.005 (0.003–0.048)	3.500 (0.200–18.600)	0.007 (0.003–0.319)	0.421 (0.462)	0.690 (0.300)
Surgery (Group 2, n = 5, Patients 6–10)	3.500 (1.500–4.500)	0.010 (0.004–0.030)	7.200 (0.600–12.900)	0.043 (0.007–0.514)	0.690 (0.900)*	0.151 (0.800)
Polychemotherapy (Group 2, n = 5, Patients 6–10)	2.000* (1.900–2.800)	0.005* (0.004–0.015)	3.400 (2.400–4.100)	0.034 (0.010–0.794)	0.016* (0.051)	0.032* (0.308)
Healthy donors (n = 10)	4.300 (1.800–7.200)	0.025 (0.007–0.180)	4.900 (2.000–10.700)	0.045 (0.007–0.160)	0.533 (0.399)	0.315 (0.889)*
^eP values (Group 2, surgery vs. Group 2, polychemotherapy)	0.138	0.068	0.225	0.686	–	–

^aArises from information provided in Table 5.

^bContinuous variables are expressed as medians and ranges (in parentheses). *Difference from healthy donors is statistically significant (P < 0.05). Mann–Whitney U-test (continuous variables).

^cMann–Whitney U-test (continuous variables). *Difference is statistically significant (P < 0.05).

^dSpearman’s rank correlation coefficient (r values are in parentheses). Positive correlation is statistically significant (P < 0.05).

^eWilcoxon’s test (continuous variables). Differences are statistically insignificant (P >0.05).

Discussion

The overall analysis of the presented data shows a wide range of individual fluctuations of TNF-R p55 and IL-8₇₂ in the plasma and urine of patients with breast cancer. The individual fluctuations in the control group are virtually identical and therefore medians of data, when comparing the results of the survey of patients and healthy donors, are not significantly different. There is only a tendency toward a decrease in the levels of TNF-R p55 and IL-8(72) in the urine of patients with breast cancer. These indicators are not demonstrative in blood plasma; these may reflect the current understanding of the blocking of IL-8(72) by serum IgG (Sylvester et al., 1992) and of the short-time presence of these soluble factors in blood (Redl et al., 1995; Lin et al., 2007). Contrary to our expectations, in general, there were no detectable levels of IL-8₇₇ isoform in either fluid assayed.

When the patients were distributed in groups depending on the cancer process or type of treatments, there was also not found to be any stable significant differences in the levels of TNF-R p55 and IL-8(72) isoform when they were compared with corresponding values in healthy donors OR between any stages of breast cancer OR any type of cancer treatment(s).However, there was a stable tendency toward a decrease in these parameters (compared to levels in the healthy controls)- not only in the urine but also in the plasma- among the patients before any treatment appearances. Moreover, courses of polychemo-therapy were clearly accompanied by a significant decrease in IL-8₇₂ and TNF-R p55 levels in the blood plasma; interestingly, these findings contrasted with significant increases in these parameters within the breast cancer patients’ urine. Furthermore, the IL-8₇₇ isoform that was not detectable prior to initiation of polychemotherapy now appeared both in the urine and plasma of the patients. In general, the dynamics of reducing the inflammatory cytokines in the blood, and their concurrent appearance/increase in the urine, likely reflected both the damaging effects of the given treatment on the cancer cells and an increased clearance of these cells and cytokines from the body of the treated patient. This explanation is supported, in part, by data from the studies of other Investigators (i.e., see Redl et al., 1995; Lin et al., 2007; Mohkam et al., 2008).

Still, at this moment, the analyses of these breast cancer patients does not allow us to establish strong stable correlations between stage of cancer/type(s) of treatments and the levels of TNF-R p55 and/or isoforms of IL-8₇₂ and IL-8₇₇ in the patients’ plasma and urine. Individual fluctuations in these markers, rather, are apparently situational moreso than directly related to breast cancer. Furthermore, in contrast to any other type of reported cancer cases showing apparent increased levels of inflammatory cytokines (Strieter et al., 1993; Grosen et al., 1993; Or et al., 1996; Warzocha et al., 1997; Xu and Fidler., 2000; Hantschel et al., 2008; Uchikawa et al., 2009), in the studies we report here there was only a stable tendency toward decreases in TNF-R p55 and IL-8₇₂ levels (as compared to corresponding values associated with healthy donors).

In conclusion, the study reported here is preliminary, yet it illustrates the potential feasibility of the simultaneous detection of TNF-R p55 and two isoforms of IL-8 in the blood plasma and urine of patients with breast cancer. This novel ability could be useful in several ways before and during cancer patient treatment, i.e., for evaluation of immune status and treatment efficacy and the monitoring of cancer progression. However, before this assay can be readily included in any arsenal of tests routinely employed by immunologists, oncologists, radiologists, pharmacologists, etc., the results here require: deeper analysis; more data on groups of patients with breast cancer, including multipoint analysis of individual patients; and, accumulated inclusion/analyses of patients with several other types of cancers, to be undertaken.

Acknowledgments

The study was supported by the Government Found of Fundamental Investigations of Ukraine, Grant #ф14/296-2007 and Belarusian Republican Foundation for Fundamental Research, Grant B-07-K-052.

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