Potential predictive value of plasma heat shock protein 90α in lung cancer.

OBJECTIVE: Heat shock protein 90α (HSP90α) is associated with cancer development, progression, and metastasis. This study assessed the relationships of plasma HSP90α levels with treatment efficacy and prognosis in lung cancer.
METHODS: In this retrospective cross-sectional study, 231 patients with lung cancer were enrolled from 1 September 2016 to 31 December 2019. HSP90α levels were measured before and after treatment, and their relationships with outcomes were assessed.
RESULTS: Patients with elevated HSP90α levels before treatment had a better overall response rate (ORR, 44.1% vs. 30.6%), whereas the disease control rate did not differ between patients with elevated and normal HSP90α levels (81% vs. 78.5%). Median progression-free survival (PFS) was 6.9 months in patients with elevated baseline HSP90α levels, versus 9 months in patients with normal HSP90α levels, whereas the median overall survival (OS) times in these groups were 12 and 14.1 months, respectively. Concerning HSP90α levels after treatment, ORR (20% vs. 47.1%) and DCR (67.3% vs. 90.9%) were lower in patients with increased HSP90α levels, and PFS and OS were also significantly different between the groups.
CONCLUSIONS: HSP90α levels before and after treatment were associated with treatment response and patient prognosis in lung cancer.

Introduction

Lung cancer is the most common cancer worldwide and the leading cause of malignant tumor death.[1,2] The incidence of lung cancer in China has significantly increased, making this disease a common health hazard.[3,4] Because many patients are diagnosed after disease progression, early detection and diagnosis are the most important measures for preventing lung cancer death. There has been substantial progress in the treatment of lung cancer recently, such as the development of targeted therapy and immunotherapy, but the identification of biomarkers has been slow.[6,7] In fact, there is no good biomarker to predict and monitor therapeutic efficacy in lung cancer at present. In the clinic, many techniques are used to diagnose and monitor lung cancer, such as the use of specific tumor markers, X-ray, computed tomography, and positron emission tomography–computed tomography, but they all have certain limitations. Therefore, there is an urgent need to identify new strategies for efficiently monitoring or forecasting treatment efficacy in lung cancer.

Heat shock proteins (HSPs) are named according to their molecular weight. HSP90 has four isoforms, namely HSP90α, HSP90β, tumor necrosis factor receptor-associated protein 1, and glucose related protein 94. HSPs are associated with several cancer types, such as breast, gastric, and colorectal cancers.[9-11] At present, it is known that HSP90α is associated with cancer development, progression, and metastasis, mainly because of the presence of HSP90α on the cell surface.[12-14] However, to the best of our knowledge, few studies with large sample sizes have reported the relationship of plasma HSP90α levels with therapeutic efficacy or prognosis in lung cancer.

In this study, we measured HSP90α levels in patients with locally advanced or metastatic lung cancer before and after treatment to assess the relationships of HSP90α levels and their variations with treatment response and prognosis. Subgroup analyses were also conducted to further determine whether HSP90α is a good biomarker for evaluating treatment efficacy and prognosis.

Materials and Methods

Patients

Patients treated for lung cancer between 1 September 2016 and 31 December 2019 at Sichuan Cancer Hospital were eligible for enrollment. The inclusion criteria were as follows: newly diagnosed lung cancer confirmed by pathology; locally advanced or metastasis disease; no receipt of radical surgery for primary or metastasis lesions; presence of measurable lesions according to the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST 1.1); completion of least two cycles of treatment and an efficacy evaluation; and available HSP90α test results before and after treatment. The exclusion criteria were as follows: presence of other malignancies; therapy was delayed for more than 28 days; and baseline or post-treatment HSP90α data were unavailable. This study was approved by the Institutional Review Board of Sichuan Cancer Hospital (approval number: SCCHEC-01A-2021-011). As a retrospective study, the requirement for informed consent was waived, and all patient details have been de-identified.

Methods

Baseline patient characteristics, treatment responses, and plasma HSP90α protein levels were recorded before and after two cycles of treatment. Plasma HSP90α levels protein were detected using an ELISA kit (Yantai Protgen Biotechnology Development Co., Ltd, Yantai, China). Treatment responses were evaluated using radiological images and the electronic medical record. According to RECIST 1.1, treatment responses were evaluated as complete response (CR), partial response (PR), stable disease (SD), or progression disease (PD). The disease control rate (DCR) was calculated as follows: DCR = (CR + PR + SD)/n (n is the total number of patients). The overall response rate (ORR) was calculated as follows: ORR = (CR + PR)/n. Progression-free survival (PFS) was defined as the period from date of histological diagnosis to that of confirmed progression disease. Overall survival (OS) was defined as the time from diagnosis to death from any cause with censoring at the time of last follow-up. The reporting of this study conforms to the STROBE guidelines.

Statistical analysis

The cutoff baseline HSP90α level was obtained using receiver operator characteristic curve analysis. ORR and DCR were compared between the groups using the χ2 test (Fisher’s exact test if necessary). Survival analysis in the different cohorts was performed using Kaplan–Meier curves and the log-rank test. All statistical analyses were performed using GraphPad Prism 7 (GraphPad Software, Inc., La Jolla, CA, USA), and P < 0.05 indicated statistical significance.

Results

Patient characteristics

Data for more than 900 consecutive patients lung cancer treated between 1 September 2016 and 31 December 2018 at Sichuan Cancer Hospital were reviewed, but most patients were excluded because they underwent surgical excision, lacked evaluable lesions, or failed to undergo HSP90α measurements before or after treatment. Finally, 231 patients with lung cancer were included in the analysis. The cohort included 158 (68.4%) men and 73 (31.6%) women. In total, 175 (75.7%) patients were younger than 65 years old, whereas 56 patients were at least 65 years old (24.3%). Regarding the lung cancer type, non-small cell lung cancer (NSCLC) was present in 160 (69.3%) patients, whereas 71 (30.7%) patients were diagnosed with small cell lung cancer (SCLC). Most patients (191, 82.7%) had stage IV NSCLC or extensive SCLC, whereas only 40 (17.3%) patients had stage III NSCLC or limited SCLC. More patient details are presented in Table 1.

Table 1.

Baseline characteristics of patients with lung cancer.

Group	Number (%)	Median baseline HSP90α, ng/mL (range)
Entire cohort	231 (100)	78.2 (20.1–606.9)
Sex
Male	158 (68.4)	78.6 (20.1–606.9)
Female	73 (31.6)	76.6 (31.1–261.7)
Age (years)
<65	175 (75.7)	82.2 (20.1–606.9)
≥65	56 (24.3)	73.1 (26.4–233.7)
Histologic type
Squamous cell carcinoma	47 (20.3)	78.4 (26.4–512.5)
Adenocarcinoma	105 (45.5)	74.3 (20.1–253.0)
SCLC	71 (30.7)	83.5 (27.5–606.9)
Others	8 (3.5)	111.3 (34–268.6)
Tumor stage
III (NSCLC)	23 (10.0)	97.2 (35.6–512.5)
IV (NSCLC)	138 (59.7)	74.7 (20.1–467.5)
Limited (SCLC)	17 (7.3)	63.4 (27.5–174.4)
Extensive (SCLC)	53 (23.0)	92.6 (31.1–606.9)

NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; HSP90α, heat shock protein 90α.

HSP90α is associated with treatment response in lung cancer

Among patients with elevated baseline HSP90α levels (≥82.06 ng/mL), 49 patients achieved CR or PR, and 40 patients were deemed to have SD. Meanwhile, among patients with normal baseline HSP90α levels (<82.06 ng/mL), 37 patients achieved CR or PR, and 58 had SD. ORR significantly differed between the elevated and normal HSP90α level groups (44.1% vs. 30.6%, P < 0.05). Among patients with increased HSP90α levels after treatment, 22 achieved CR or PR, and 52 had SD. Among patients with decreased HSP90α levels after treatment, 57 achieved CR or PR, and 53 patients had SD. ORR (20% and 47.1%, P < 0.001) and DCR (67.3% vs. 90.9%, P < 0.001) were both significantly lower among patients with increased HSP90α levels after treatment (Table 2).

Table 2.

Associations of baseline and post-treatment HSP90α levels with the treatment response.

	Baseline (%)		P	Post-treatment (%)		P
	≥82.06 ng/mL	<82.06 ng/mL		Increased	Decreased
ORR	44.1	30.6	0.033*	20.0	47.1	<0.001*
DOR	81.0	78.5	0.651	67.3	90.9	<0.001*

*Significant at P < 0.05.

HSP90α, heat shock protein 90α; ORR, objective response rate; DCR, disease control rate.

PFS and OS according to the HSP90α level

Median PFS was 6.9 months in patients with elevated baseline HSP90α levels, versus 9 months among those with normal baseline HSP90α levels (P < 0.05), whereas the median OS times among patients with elevated and normal baseline HSP90α levels were 12 and 14.1 months, respectively (P < 0.05, Figure 1a–b). Further subgroup analysis was also conducted. Among patients with SCLC, elevated baseline HSP90α levels indicated a worse prognosis, albeit without significance (Figure 1c–d). For patients with adenocarcinoma, normal baseline HSP90α levels represented a good prognostic factor (P < 0.05, Figure 1e–f). For squamous cell carcinoma patients, there was no significant difference in survival between the two groups (Figure 1g–h).

Figure 1.

PFS and OS according to baseline HSP90α levels in patients with lung cancer. (a) PFS in the entire study cohort. (b) OS in the entire study cohort. (c) PFS among patients with SCLC. (d) OS among patients with SCLC. (e) PFS among patients with adenocarcinoma. (f) OS among patients with adenocarcinoma. (g) PFS among patients with squamous cell carcinoma. (h) OS among patients with squamous cell carcinoma.

PFS, progression-free survival; OS, overall survival; HSP90α, heat shock protein 90α; SCLC, small cell lung cancer.

Outcomes also differed between patients with increased and decreased HSP90α levels after treatment. Median PFS was 6.9 months among patients with increased levels, versus 9.1 months among those with decreased levels (Figure 2a–b). Among patients with SCLC, elevated HSP90α levels indicated a worse prognosis, albeit without significance (Figure 2c–d). Among patients with adenocarcinoma, outcomes did not differ according to the HSP90α level (Figure 2e–f). Regarding patients with squamous cell carcinoma, normal HSP90α levels portended a good prognosis (P < 0.05, Figure 2g–h).

Figure 2.

PFS and OS according to the changes of HSP90α levels after treatment. (a) PFS in the entire study cohort. (b) OS in the entire study cohort. (c) PFS among patients with SCLC. (d) OS among patients with SCLC. (e) PFS among patients with adenocarcinoma. (f) OS among patients with adenocarcinoma. (g) PFS among patients with squamous cell carcinoma. (h) OS among patients with squamous cell carcinoma.

PFS, progression-free survival; OS, overall survival; HSP90α, heat shock protein 90α; SCLC, small cell lung cancer.

Discussion

Lung cancer accounts for more than 20% of cancer-related deaths globally, making it the deadliest cancer. A good tumor biomarker is extremely important for improving diagnosis and survival. In this study, we analyzed the relationships of baseline plasma HSP90α levels and their changes after treatment with treatment efficacy and prognosis. We found that baseline HSP90α levels and their changes after treatment were associated with treatment responses and prognosis in patients with locally advanced or metastasis lung cancer. Subgroup analysis illustrated that the prognosis of patients with adenocarcinoma differed between those with elevated and normal baseline HSP90α levels. After grouping patients based on the changes of HSP90α levels after treatment, the results illustrated lower HSP90α levels after treatment were associated with better therapeutic efficacy and longer survival. Similarly, the prognosis of patients with squamous cell carcinoma also differed according to the change of HSP90α levels after treatment.

A few studies demonstrated that altered HSP90α levels can serve as a biomarker for different cancers. Wang and colleagues demonstrated that HSP90α levels were higher in most patients with lung cancer and associated with cancer development, suggesting the utility of HSP90α as a potential diagnostic and prognostic marker in clinical practice. Some previous studies illustrated that HSP90 inhibitors had potent antitumor activity, and they could overcome resistance to chemotherapy and radiotherapy, also suggesting that HSP90α may be a potential biomarker for cancers, including lung cancer.[18-23] Shi and colleagues reported that HSP90α levels significantly differed between patients with PD and SD. Our study further demonstrated that changes of plasma HSP90α protein levels could be a useful biomarker for monitoring treatment response and predicting survival in patients with lung cancer.

In a recent study, Zhong’s group explored the association of HSP90α with chemotherapy in advanced NSCLC. They found that plasma HSP90α levels can be considered a valuable predictor of early chemotherapy effectiveness in patients with advanced NSCLC, and its levels were positively correlated with tumor remission after chemotherapy. This was similar to our findings. Based on previous research, cancer cells are more dependent on HSP90α than normal cells, and high HSP90α expression thus encourages the growth and survival of tumors.[25,26] Rong’s group indicated that HSP90 upregulation was associated with poor post-surgical survival time and lymphatic metastasis in patients with lung cancer. Our results demonstrated that patients with elevated baseline HSP90α levels had shorter PFS and OS. The higher ORR of patients with elevated baseline HSP90α levels may also indicated that HSP90α levels are associated with sensitivity to treatment. Nevertheless, further studies are needed to confirm this finding.

In clinical practice, many plasma proteins have been used as biomarkers for cancer diagnosis, such as CEA and CA125. However, few measurable and useful tumor markers are available, especially markers that can predict treatment efficacy and survival benefit. In our study, we found that both baseline and post-treatment HSP90α levels were associated with ORR, DCR, PFS, and OS. The results are interesting, as they indicated that baseline HSP90α elevation is a good marker for ORR and that decreased HSP90α levels after treatment represented a good biomarker for lung cancer. As a retrospective study, this study had several limitations. First, it was a single-center study. Second, the sample size was not large. Thus, more prospective clinical trials are needed to confirm these results.

Conclusions

This study may be the first to reveal the role of dynamic changes of HSP90α for monitoring the response to systemic treatment and predicting survival in patients with lung cancer. The study demonstrated that baseline HSP90α levels and their changes after treatment are both associated with treatment response and prognosis.