Concomitant Active Tuberculosis Prolongs Survival in Non-Small Cell Lung Cancer: A Study in a Tuberculosis-Endemic Country

Background Adjuvant tumor cell vaccine with chemotherapy against non-small cell lung cancer (NSCLC) shows limited clinical response. Whether it provokes effective cellular immunity in tumor microenvironment is questionable. Concomitant active tuberculosis in NSCLC (TBLC) resembles locoregional immunotherapy of tumor cell vaccine; thus, maximally enriches effective anti-tumor immunity. This study compares the survival and immunological cell profile in TBLC over NSCLC alone. Methods Retrospective review of NSCLC patients within 1-year-period of 2007 and follow-up till 2010. Results A total 276 NSCLC patients were included. The median survival of TBLC is longer than those of NSCLC alone (11.6 vs. 8.8 month, p<0.01). Active tuberculosis is an independent predictor of better survival with HR of 0.68 (95% CI, 0.48∼0.97). Squamous cell carcinoma (SCC) (55.8 vs. 31.7%, p<0.01) is a significant risk factor for NSCLC with active TB. The median survival of SCC with active tuberculosis is significantly longer than adenocarcinoma or undetermined NSCLC with TB (14.2 vs. 6.6 and 2.8 months, p<0.05). Active tuberculosis in SCC increases the expression of CD3 (46.4±24.8 vs. 24.0±16.0, p<0.05), CXCR3 (35.1±16.4 vs. 19.2±13.3, p<0.01) and IP-10 (63.5±21.9 vs. 35.5±21.0, p<0.01), while expression of FOXP3 is decreased (3.5±0.5 vs. 13.3±3.7 p<0.05, p<0.05). Survival of SCC with high expression of CD3 (12.1 vs. 3.6 month, p<0.05) and CXCR3 (12.1 vs. 4.4 month, p<0.05) is longer than that with low expression. Conclusions Active tuberculosis in NSCLC shows better survival outcome. The effective T lymphocyte infiltration in tumor possibly underlies the mechanism. Locoregional immunotherapy of tumor cell vaccine may deserve further researches.


Introduction
Non-small cell lung cancer (NSCLC) remains the leading cause of cancer death worldwide [1].Chemotherapy alone has been the standard care in advanced disease, albeit the effect is suboptimal [2,3].Adjuvant immunotherapy in combination with chemotherapy has been reported to be an alternative, which improves quality of life without the increase of side effect.Hence, understanding the host immune responses to NSCLC is essential to tailor the use of immunotherapeutic strategies.
The ''immune surveillance theory'' holds that all tumor cells express antigenic markers capable of eliciting immune responses which prevent the outgrowth of malignant cells [4].However, cellular immunity is usually ineffective because neoplasms have mechanism to evade the host immune response [5,6].Firstly, tumor antigens are usually poorly immunogenic because they are perceived by the immune system as ''self'' and ''alter self'' antigens.
Secondly, tumor cells may induce functional suppression of T cell, and conferring resistance to T cells-induced apoptosis.In contrast, increased T cell infiltration in tumor site has been reported to be positively associated with survival outcome in NSCLC [5,6].
Therefore, correction of cancer-related host immune dysregulation accounts for an appealing strategy of anti-cancer treatment.Purified protein derivative of M. tuberculosis (PPD) enhances the production of inducible protein-10 (IP-10), the ligand of CXCR3, in human lungs [7,8] to recruit activated T cells.Increased CXCR3 expression in tumor nests is associated with prolonged survival and more inflammatory cell infiltration in patients with non-small cell lung cancer (NSCLC) [9].Tumor cell vaccine, such as tuberculosis (TB) Bacillus Calmette-Gue ´rin (BCG) and heatkilled Mycobacterium vaccae suspension (SRL172), have been used as non-specific immunostimulator against several types of human cancers with variable clinical response [10,11].For advanced NSCLC, combination of SRL172 with chemotherapy showed limited improvement of survival but improvement in quality of life [12].
The wide variation of efficacy of tumor cell vaccine raised a question of whether and to what extent is the T cell immunity provoked by systemic injection of tumor cell vaccine enriches in the tumor microenvironment to potentiate effective immunity against tumor [13,14] In a tuberculosis endemic country, it is possible to approach it [15] by looking at the immunological cell profile in tumor microenvironment when NSCLC has superimposed active TB happening in close vicinity [15].Furthermore, this clinical scenario resembles the locoregional immunotherapy of tumor cell vaccine for NSCLC, and suggests the development of a constant and maximally enriched T cell immunity nearby tumors.
This study primarily aims to compare the outcome in NSCLC patients with concomitant active TB over NSCLC alone patients, in terms of overall survival.Furthermore, the expression of immunological cell profile adjacent to tumors in either group is also studied.

Study population
We performed a retrospective review of active tuberculosis in advanced NSCLC patients in Chang Gung Memorial Hospital (a tertiary referral medical center) during the 1-year period of 2007 and followed up their clinical outcomes till 2010.Informed consents were obtained for all stored samples used for this study.Determination of active TB infection started from the review of positive acid-fast smear from sputum, bronchial wash or broncho-alveolar lavage fluid in advanced NSCLC patients, followed by the specific focus on those with positive TB growth to assure the existence of viable mycobacteria.Hence, active TB infection qualified both positive acid-fast smear and positive culture for TB.Subsequently, the chest X-ray or CT scan were reviewed to include patients presenting same lobar distribution of TB and cancerous lesions.

Staging and treatment of advanced NSCLC
The scheme of 2009 AJCC 7 th edition for TNM staging was used for study patients.Platinum-based chemotherapy was the standard care for stage IV and stage IIIB patients.Stage IIIA patients received concurrent chemoradiotherapy (CCRT) or platinum-based neoadjuvant chemotherapy, and surgical resection was considered for whom have been downstaged successfully.Performance status is measured by ECOG scoring system runs from 0 to 5 [16].Patients with an ECOG performance status more than two were treated with single-agent chemotherapy or best supportive care alone decided by in charge physician.

Treatment for tuberculosis
Patients with concomitant active tuberculosis received standard treatment with two months isoniazide, rifampicin, pyrazinamide and ethambutol combination, following by four months combination with isoniazide, rifampicin and ethambutol.For patients whose treatment of NSCLC overlapped with that of TB; the former was postponed at least one month after the start of anti-tuberculosis therapy.The appropriate timing to resume the NSCLC treatment was decided by in charge physician based on patient's performance status.

Immunohistochemical (IHC) staining
Paraffin-embedded specimens were retrieved.Rabbit anti-human antibodies to CD3 (Dako A/S, Glostrup, Denmark), CD4 (NeoMarkers, Thermo Fisher Scientific, Fremont, CA, U.S.A.), CD8 (BioCare Medical, Concord, CA, U.S.A.), CD56 (Dako A/S, Glostrup, Denmark), CD 68 (Dako A/S, Glostrup, Denmark), iNOS (Santa Cruz Biotechnology, CA, U.S.A.), CXCR3 (BD Biosciences Pharmingen, NJ, USA), IP-10 (Santa Cruz Biotechnology, CA, U.S.A.), CD25 (Novocastra, Newcastle upon Tyne, UK) and FOXP3 (ab20034; Abcam, Cambridge, UK) were obtained.The staining intensity in tumor cells was assessed in comparison to normal bronchial epithelium as an internal positive control.Staining specificity was evaluated by negative controls in which the primary antibody had taken the place of normal mouse (for CD3, CD4, CD56, CD68, CD25 and FOXP3) and normal rabbit (for CD8, iNOS, CXCR3 and IP-10) non-immune IgG.The IHC slides were assessed under a light-microscope at X400 magnification.All nucleated cells in tumor islet cells and stroma cells (including tumoral, structural and infiltrative inflammatory cells) were counted by reader blinded for survival outcome information.The percentage of cell marker-positive nucleated cells of total nucleated cells in selective tumor islets and stromal areas was calculated and expressed as cells/100 nucleated cells.

Statistical analysis
All quantitative data were expressed as mean values and standard deviations.Non-parametric approaches were used since most data were not normally distributed.The Mann-Whitney test was used to compare numeric variables between two groups, and the x 2 test was used for categorical variables.Univariate analyses of survival used the log-rank test to examine effects of baseline clinical factors.Variables with a value of p,0.1 was made to enter a multivariate analysis with Cox's regression model for identifying independent predictor of survival.All analyses were conducted using the

Active Tuberculosis as an independent predictor of survival outcome
Predictors of survival outcome are shown in Table 2.In univariate analysis, better performance status, TB infection, stage III of NSCLC and receiving treatment are significant predictors for survival outcome with HR of 0.43 (95% CI, 0.33,0.52),0.61 (95% CI, 0.43,0.86),0.62 (95% CI, 0.48,0.79)and 0.54 (95% CI, 0.41,0.72),respectively.In multivariate analysis, active TB remains one of the independent predictors of survival outcome with HR of 0.68 (95% CI, 0.48,0.97).Figure 1 shows the survival curve of NSCLC patients with or without active TB.The median survival of concomitant active TB is longer than those without TB (11.6 vs. 8.8 month, p,0.01).

Squamous cell carcinoma is a risk factor for concomitant active tuberculosis
Table 3 shows the clinical factors of NSCLC patients with or without active tuberculosis.The risk of squamous cell carcinoma (55.8 vs. 31.7%),male gender (82.7 vs. 66.5%) and smoker (78.8 vs. 58.5%)are significantly higher for NSCLC with active TB.Multivariate analysis shows that squamous cell carcinoma (ORs 2.09; 95% CI, 1.06,4.14,p,0.05) is the single independent risk factor.

Concomitant active tuberculosis shows better survival outcome in squamous cell carcinoma patients
Comparison of survival outcome stratified by tuberculosis infection and histology are shown in Figure 2A (with TB) and Figure 2B (without TB).In NSCLC with concomitant active TB, the median survival is significantly longer in squamous cell carcinoma than adenocarcinoma and undetermined NSCLC (14.2 vs. 6.6 and 2.8 months, p,0.05).For patient without TB, survival is not significantly different between three histology groups (8.2, 9.1 and 8.9 month, p = 0.27).For adenocarcinoma with concomitant active TB, significantly higher expression of CD3 and CXCR3 are also noted.(Table S1)

Increases in CD3, CXCR3 expression is associated with better survival outcome in squamous cell carcinoma patients
The survival outcome analysis of high or low immunohistochemical marker indices divided by median in squamous cell carcinoma is shown in Table 5. Tumor with high expression of CD3 (12.1 vs.  3.6 month, p,0.05) and CXCR3 (12.1 vs. 4.4 month, p,0.05) shows better median survival than those with low expression, while other markers are not associated with survival outcome.Figure 3 representatively demonstrates the images showing high and low expression of CD3, CXCR3 and FOXP3.For adenocarcinoma, level of expression for each inflammatory marker is not associated with survival outcome.(Table S2)

Discussion
The present study has demonstrated that NSCLC with active TB is more frequently found in squamous cell carcinoma.With concomitant active TB, NSCLC patients show a better survival outcome; especially in squamous cell carcinoma, which is associated with increased CD3-and CXCR3-expressing cells within tumor.[17,18,19] A clinicopathological analysis reported by Tamura et al. have shown that mycobacteriosis is often adjacent to preexisting lung cancer.[20] The walls of encapsulated caseous nodules may be invaded by NSCLC, especially squamous cell carcinoma featuring local invasiveness; thereby, leading to tuberculosis reactivation.Our result reveals that active TB infection in NSCLC is an independent predictor of better survival.There is abundant evidence in the earlier literature that remission, regression and even total disappearance of tumors may occur when mycobacterium infection co-exists.[21] However, treatment with tumor cell vaccine preparation from heatkilled Mycobacterium vaccae suspension (SRL172) for advanced NSCLC has only shown limited survival benefit.[12] One of the aims of tumor cell vaccine therapy is correcting the imbalance of Th1 and Th2 cell immunity established by cancer cells.[22,23] These include: induction of T lymphocytes accumulation [24], upregulation of Th1 cytokines, such as interferon-c [25], and suppression of Th2 immunity [26].However, whether these immune reactions provoked by intradermal injection of SRL172 are effectively enriched nearby cancer cells has never been confirmed.
By studying NSCLC in close vicinity to active tuberculosis, it resembles locoregional immunotherapy of tumor cell vaccine for lung cancer.The inflammatory profile nearby tumor in this condition is higher than those without TB, in terms of CD3, CXCR3 and IP-10 for squamous cell carcinoma, and CD3 and CXCR3 for adenocarcinoma.Moreover, when analyzing inflammatory burden to survival, it reveals that higher infiltration of CD3-and CXCR3-expressing cells are associated with better survival in squamous cell carcinoma.This result suggests that locoregional immunotherapy possibly enables enrichment of effective anti-tumor immunity nearby tumor, but appears to be important for the clinical outcomes in patients with squamous cell carcinoma.
On the other hand, our results show that CD3 and CXCR3 rather than IP-10 are the markers associated with survival benefit.There are two possible explanations to it.Firstly, IP-10 is not the ligand exclusively required by CXCR3 in order to activate T lymphocyte [8], whereas CXCL9 and CXCL11 could play the similar function.Secondly, the activation of T lymphocyte not only  depends on the ligand stimulation, but also on the immunologically suppressive cells recruited and tamed by tumor.The emerging understanding of regulatory T cells (Treg) significantly improves our knowledge of immune system evasion and how it enables tumor progression.[22,27,28] Our study also provides evidence that expression of FOXP3; as a marker for Treg, is decreased in squamous cell carcinoma with active TB.Dumitriu et al. has shown that NSCLC promotes the differentiation of Treg through dendritic cell via a TGF-beta 1 dependent mechanism [29].In light of this, report from MacKinnon et al. revealed that blockade of galectin-3 could possibly inhibit the IL-4/IL-13 driven alternative macrophage activation which is critical in building a tumor friendly extracellular matrix [30].Therefore, gaining better insight of the immunosuppressive property of tumor microenvironment underlies the responsiveness of tumor cell vaccine treatment.The major limitation of current study is the retrospective property per se; thereby, the samples qualified for immunohistochemical stain are limited due to lack of prospective standardization.However, the trend of survival outcome is similar with the main clinical results.Additionally, the possibility of synergistic effect or interaction between the anti-tuberculosis medication and the chemotherapeutic agents could not be excluded.
In conclusion, this study shows that NSCLC with active TB has better survival outcome.The effective T cell immunity nearby tumor possibly underlies the mechanism.Locoregional treatment of tumor cell vaccine for NSCLC may deserve further researches.

Table 2 .
Predicting factors of survival outcome by univariate and multivariate analysis.

Table 3 .
Analysis of clinical factors associated with active tuberculosis in NSCLC patients.

Table 4 .
Immunohistochemical marker indices of squamous cell carcinoma with or without active tuberculosis.

Table 5 .
Survival outcome analysis of high or low immunohistochemical marker indices divided by median in squamous cell carcinoma. doi:10.1371/journal.pone.0033226.t005