Increased heat shock protein 70 (Hsp70) serum levels and low NK cell counts after radiotherapy – potential markers for predicting breast cancer recurrence?

With 2,088,849 newly diagnosed cases in 2018, reflecting 25.4% of all tumor entities, breast cancer is the most common invasive cancer in women worldwide (World Cancer Research Fund). In early tumor stages, a breast-conserving surgery with adjuvant RT is the standard of care. A comparison of accelerated partial breast irradiation using multicatheter brachytherapy and whole breast irradiation after breast conserving surgery revealed no significant differences in the 5 year local control, disease-free and overall survival [1]. Postoperative RT of the tumor bed results in lower rates of local recurrence [2] because remaining viable tumor (stem) cells might be depleted [3]. Irradiation- as well as chemotherapy-induced stress can result in the release of damage-associated molecular patterns (DAMPs) by tumor cells which induce immunogenic cell death [4–7]. These DAMPs including calreticulin, extracellular ATP, high mobility group box 1 (HMGB1), HSPs derived from programmed, caspase-independent cell death termed necroptosis [8] are known to trigger acute inflammation and thereby might elicit protective antitumor immunity [9]. In contrast to acute, chronic inflammation has been shown to promote tumor growth [10]. Therefore, the impact of DAMPs in the context of acute and chronic inflammation need further investigations. The highly conserved member of the 70 kDa heat shock protein family Hsp70 is considered as a DAMP when released as a free molecule by dying cancer cells [11] whereas exosomal Hsp70 (exHsp70) which is actively released by viable tumor cells serves as a marker for viable tumor mass [12, 13]. Elevated exHsp70 serum levels as detected by the lipHsp70 ELISA [14] that detects both, free and exosomal Hsp70 have been found in breast cancer patients [15], patients with hepatocellular carcinoma, lung cancer [16], colorectal carcinoma [17, 18] and other cancer entities [19]. Although high exHsp70 levels are generally associated with an adverse prognosis also in breast cancer patients [20, 21], less is known about the role of free Hsp70 as a predictive marker for therapy response. Therefore, we investigated serum levels of free Hsp70 as well as the immunophenotype in breast cancer patients before, during, after RT and in the follow-up period up to 1 year to study potential immunomodulatory effects of circulating, free Hsp70.

The response assessment included clinical examination, mammography and sonography. Three and 5 years after the end of RT, 32 patients (14 and 18 patients, respectively) were disease-free. One patient developed an endometrial carcinoma and two patients showed a contralateral recurrence or developed distant metastases 2 years after RT (Fig. 1; Table 1). The clinical outcome of five patients remained unknown (Fig. 1), and therefore these patients were not considered in the analysis.

However, a comparison of the free Hsp70 serum levels with the outcome revealed significant differences. Although the number of patients who showed recurrence was low both breast cancer patients who developed distant metastases or contralateral recurrence within the first 2 years after RT exhibited significantly higher median serum Hsp70 levels directly after RT (t3; p = 0.03) which remained significantly upregulated up to 6 weeks after RT (t4; p = 0.007) compared to the group of patients who remained disease-free for 3 and 5 years, respectively (Fig. 2a). The Hsp70 serum levels of the patient who was diagnosed with a secondary endometrium carcinoma 2 years after RT ranged within that of healthy individuals (2.74 + 1.07 ng/ml, n = 131 healthy donors) throughout the course of disease (Table 6). Blood samples of patients with other tumor entities such as squamous cell carcinoma of the head and neck (SCCHN), lung cancer, or glioblastoma at diagnosis revealed higher Hsp70 serum levels, as determined with the R&D ELISA compared to those of healthy controls (Table 6).

Healthy controls vs tumor patients	Number	Hsp70 serum levels (ng/ml)
Healthy controls	131	2.74 ± 1.07
SSCHN	22	3.51 ± 3.58
Lung cancer	168	5.40 ± 4.56
Glioblastoma	141	14.1 ± 10.63

A measurement of the serum samples of the breast cancer patients at the different time-points after surgery using the lipHsp70 ELISA revealed similar results to those assessed by the R&D ELISA. However, due to the higher sensitivity of the lipHsp70 ELISA, the values were higher than that measured with the R&D ELISA (Fig. 2b). Since all patients were tumor-free at the time-point of blood collection, we assume that the lipHsp70 ELISA only detects free but not exosomal Hsp70. Due to the limited amount of serum samples of the patients the lipHsp70 ELISA could be performed only once in triplicates.

The analysis of lymphocyte subpopulations in the peripheral blood revealed a drop in the percentage of NK cells 6 and 12 months after RT in recurrence/metastasize patients compared to disease-free patients, whereas the percentage of T and B cells remained nearly unaltered in the follow-up period (Fig. 3). The absolute numbers of NK cells in patients with contralateral recurrence/metastases remained always below that of disease-free patients throughout the whole course of disease, however, due to the low number of patients the data failed to show statistical significance (Table 7).

	N	Before RT	After 30 Gy	End of RT	6 weeks after RT	6 months after RT	1 year after RT
		NK cell counts
Disease-free	32	0.15 ± 0.08	0.12 ± 0.06	0.09 ± 0.05	0.14 ± 0.06	0.16 ± 0.09	0.18 ± 0.1
Recurrence/metastases	2	0.14 ± 0.03	0.07 ± 0.01	0.06 ± 0	0.11 ± 0	0.14 ± 0	0.11 ± 0

Two breast cancer patients out of 35 with elevated Hsp70 serum levels at the end of RT that appear to persist for 6 weeks after adjuvant RT developed contralateral recurrence or distant metastases 2 years after RT. Since the patients of this study were diagnosed as tumor-free after breast conserving surgery (R0 resection) and RT, it is assumed that the circulating, free Hsp70 which was measured in the serum by the R&D and lipHsp70 ELISA rather originates from chronically inflamed normal tissue than from dying tumor cells, which was shown in a lung cancer study, previously [13]. One might hypothesize that tumor cell regrowth might be promoted by chronic inflammation which is reflected by a long-lasting release of DAMPs, such as free Hsp70 for up to 6 weeks after RT. Chronic inflammation is known to promote tumor growth [23] via several mechanisms. Inflammatory mediators such as tumor necrosis factor-alpha (TNF-α), Interleukin-6 (IL-6), Interleukin-10 (IL-10) and transforming growth factor-beta (TGF-β) play important roles in tumorigenesis and the development of cancer in different entities [24]. In addition, reactive oxygen and nitrogen species are released in chronically inflamed tissues and thereby affect DNA integrity, which in turn might also increase the risk of cancer formation [25]. In an inflamed microenvironment, different DAMPs, which are normally localised in the cytosol, are released into the extracellular milieu where they can activate the immune cells, stimulate tissue repair mechanisms and thereby might mediate protective antitumor immunity [6, 10]. However, a chronic, long-term exposure of immune cells to DAMPs can induce immune-tolerance which promotes tumor growth and thereby enhances the risk of relapse or distant metastases [26]. Free Hsp70 as a representative for DAMPs was found to play an important role in tumor formation of different cancer entities, such as bladder cancer [27], colorectal carcinoma [28, 29], osteosarcoma [28], fibrosarcoma [30], melanoma [31] and T cell leukemia [32]. In the present study, free Hsp70 levels were elevated for at least 6 weeks after RT in two patients who showed recurrence or developed distant metastases. In contrast, a patient with secondary endometrial carcinoma had Hsp70 levels in the range of healthy individuals throughout the whole course of disease. This finding might be indicative for lower Hsp70 levels in this specific cancer type.

The function of intracellular Hsp70 as a molecular chaperone plays a major role in the correct folding of nascent proteins, and its antiapoptotic activity protects cancer cells against environmental stress [19]. Elevated extracellular Hsp70 levels fulfil dual roles, on the one hand they can stimulate antitumor immune responses, on the other hand, in the case of chronic exposure they might induce immune-tolerance. We hypothesize, that apart from Hsp70 also other DAMPs could be released into the circulation that further propagate chronic inflammation which in turn suppress antitumor immune responses and promote tumor growth. The patients with a long-term disease-free survival exhibited significantly lower free Hsp70 values in the serum that might be indicative for a faster clearance of the inflammatory response after RT.

Apart from free also exosomal Hsp70 which is actively released from viable tumor cells [14, 33] has been found at elevated levels in tumor patients. Since blood can be taken repeatedly by minimal invasive methods, the course of Hsp70 (free and exosomal) levels during therapy can be easily determined. Therefore, we speculate that measuring free as well as exHsp70 concentrations in the serum of tumor patients before, during and after therapy might provide predictive value with respect to determine chronic inflammation and viable tumor mass. These data might allow a more rapid and earlier therapy adaptation in patients with high-risk tumors. Since the percentage of T and B cells remained nearly unaltered, but that of NK cells dropped significantly in the follow-up period of patients with unfavourable prognosis, we speculate that NK cells might be important players in tumor control. In line with these findings we could show previously that a low NK cell infiltration in squamous cell carcinoma of the head and neck (SCCHN, N = 145) and a high Hsp70 expression are predictive for an adverse clinical outcome [34]. With respect to the absolute numbers, NK cell counts remained below that of disease-free patients throughout the whole course of disease in recurrence/metastases patients which might be due to the radiation therapy as shown by others [35–37]. The presence of extracellular Hsp70 derived from dying tumor cells in a pro-inflammatory environment provides a stimulatory signal for NK cell activation, as shown by in vitro studies [38, 39].

However, due to the low number of breast cancer patients who developed recurrence or distant metastases in our clinical trial further studies with larger patient cohorts are necessary to validate the value of Hsp70 serum levels and NK cell counts in the peripheral blood as predictors of clinical responses.

Two breast cancer patients who developed contralateral recurrence or distant metastases showed longitudinal higher serum levels of free Hsp70 after RT than patients who remained disease-free. These findings are in line with data showing an unfavourable clinical outcome in SCCHN patients with a low NK cell infiltration and a high Hsp70 expression. Furthermore, NK cell counts in patients with adverse prognosis were always below that of disease-free patients. This finding might be explained by an irradiation-induced NK cell killing. In case that these findings could be confirmed in a larger patient cohort, longitudinal elevated levels of free Hsp70 levels after RT together with a drop in NK cell counts might be considered as a predictive marker for an unfavourable prognosis.