Cancer Chemotherapy and Pharmacology

, Volume 73, Issue 4, pp 827–837

Simvastatin suppresses head and neck squamous cell carcinoma ex vivo and enhances the cytostatic effects of chemotherapeutics

Authors

    • Department of Otolaryngology, Head and Neck SurgeryUniversity of Leipzig
  • Christian Mozet
    • Department of Otolaryngology, Head and Neck SurgeryUniversity of Leipzig
  • Andreas Boehm
    • Department of Otolaryngology, Head and Neck SurgeryUniversity of Leipzig
  • Achim Aigner
    • Rudolf-Boehm-Institute for Pharmacology and Toxicology, Clinical PharmacologyUniversity of Leipzig
  • Andreas Dietz
    • Department of Otolaryngology, Head and Neck SurgeryUniversity of Leipzig
  • Gunnar Wichmann
    • Department of Otolaryngology, Head and Neck SurgeryUniversity of Leipzig
Original Article

DOI: 10.1007/s00280-014-2412-1

Cite this article as:
Stoehr, M., Mozet, C., Boehm, A. et al. Cancer Chemother Pharmacol (2014) 73: 827. doi:10.1007/s00280-014-2412-1

Abstract

Purpose

Simvastatin (Sim) is approved as lipid-controlling drug in patients with cardiovascular risk to reduce hypercholesterolemia. Recent publications indicate possible inhibitory effects of Sim on tumor cell lines, and epidemiological data suggest activity in cancer patients. Still, its therapeutic efficacy, particularly in head and neck squamous cell carcinoma (HNSCC), remains to be elucidated. This study analyzes the effects of Sim on HNSCC cell lines (KB, HN5, FaDu) and on a larger set of primary HNSCC cells by employing a short-time ex vivo colony formation test (FLAVINO assay). Possible additive or synergistic effects of Sim combinations with established chemotherapeutics are determined as well.

Methods

Biopsies of 49 HNSCC were tested in the FLAVINO assay with Sim alone or in combination with cisplatin (Cis) or docetaxel (DTX). Cell lines were studied for reference. Epithelial HNSCC cells were stained by Cy2-labeled anti-cytokeratin antibodies facilitating the detection of colony formation (CF) by immunofluorescence. Drug combinations were analyzed regarding their interaction.

Results

Sim alone acted suppressive on tested cell lines and increased the cytostatic efficacy of Cis and DTX. 18/49 HNSCC qualified for FLAVINO-based dose-response analyses, and Sim significantly suppressed CF in 18/18 primary HNSCC. Moreover, Sim increased cytotoxic effects of Cis and DTX, primarily in an additive mode of action.

Conclusions

The ex vivo tumor cell inhibition of Sim and its additive effects upon combination with established cytostatics provide the basis for epidemiological and clinical studies on statins, potentially directed toward co-medication in future treatment regimens.

Keywords

Head and neck squamous cell carcinoma (HNSCC)StatinsCisplatinDocetaxelFLAVINO assay

Introduction

Statins are widely used to reduce the risk of cardiovascular events in cases of hypercholesterolemia, as demonstrated in several randomized trials and meta-analyses [1, 2]. Numerous studies have postulated anticancer properties of statins attributable to their pleiotropic molecular effects. These include anti-proliferative, pro-apoptotic, cell differentiating, anti-invasive and radio-sensitizing properties [35], but the published clinical and experimental data have remained controversial so far [69]. Simvastatin (Sim) is a widely used lipophilic member of the statin family that reduces endogenous cholesterol biosynthesis and elevates low-density lipoprotein reuptake into hepatocytes by blocking 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA-R) and its subsequent product mevalonate [10]. The latter is a precursor molecule of several compounds, thus exerting multiple and pleiotropic biological effects also on signal transduction pathways.

Recently, Nielsen et al. [11] published data on a significant statin-intake-related mortality reduction, especially among head and neck squamous cell carcinoma (HNSCC) patients. This encourages further investigations regarding the potential benefits of statin use in HNSCC, since HNSCC patients still have a poor prognosis. Previous in vitro studies revealed anticancer effects of statins in cell lines from gliomas [12, 13], gastrointestinal carcinomas [14, 15], gynecological cancers [16, 17] and also HNSCC [18]. However, these data are based on stable, immortalized standard cell lines rather than primary tumor cells, thus suggesting that a direct transfer of statin results into novel clinical treatment strategies of HNSCC may be premature. Furthermore, these studies provide only limited information regarding concentrations needed to achieve therapeutic efficacy in HNSCC, and so far, the effects of statins in combination with well-established cytostatic drugs to the best of our knowledge have not been analyzed for HNSCC treatment ex vivo at all. To determine the efficacy of Sim alone or in combination with well-established cytostatics cisplatin (Cis) or docetaxel (DTX) under conditions that reflect the in vivo situation more closely, we employed primary cells from tumor samples and performed a short-time ex vivo colony formation assay [19, 20]. This FLAVINO assay has been shown previously to allow for the determination of cytostatic effects of Cis within in vivo tolerable concentration ranges and also the assessment of the effects of test drug combinations, as shown for Cis and the EGFR tyrosine kinase inhibitor lapatinib [21]. Furthermore, lipid metabolism and in particular cholesterol proved to be essential for the hedgehog signaling pathway (HhP) [22, 23]. Since embryonic malformations such as holoprosencephalopathy [24] are among potential side effects of statin exposure during pregnancy, and also the natural HhP inhibitor cyclopamine suppressed colony formation of HNSCC [25], we hypothesized that Sim might potentially affect HNSCC in a comparable way. We hence employed the FLAVINO assay with primary HNSCC cells to evaluate the effects of Sim alone or in combination with Cis or DTX.

Materials and methods

Chemical compounds

Sim (Sigma-Aldrich, St. Louis, MO, USA) was obtained as solid powder and dissolved in pure ethanol. Cis was purchased from Sigma (Steinheim, Germany) in solid form and dissolved in pure ethanol, while DTX (Taxotere®; Sanofi-Aventis, Berlin, Germany) was obtained as liquid pharmaceutical preparation.

Cell lines and cell culture

The HNSCC cell lines KB [26], HN5 and FaDu were grown under standard conditions in flavin-free cell culture medium (TM) without antibiotics. The FLAVINO assay was performed as described previously [19, 20], thus excluding the risk of common artifacts in chemoresponse testing due to photoactivation of riboflavin (vitamin B2) and other flavins (FAD, FMN). More specifically, photoactivation was avoided by exclusively employing yellow light illumination with an Osram L36W 62 light source (λ ≥ 550 nm) and by using flavin-free cell culture medium (TM) for cell culture and reagent preparation. TM is a modified medium on the basis of RPMI1640 without phenol red or riboflavin (Biochrom, Berlin, Germany), buffered with 1.134 g/l NaHCO3 and 14.5 mM HEPES and supplemented with 10 % fetal calf serum (FCS; Invitrogen, Darmstadt, Germany), amikacin, nystatin, penicillin and streptomycin (all from Sigma). Thus, upon addition of FCS, TM contained only 20 nM riboflavin. Cells from HNSCC were cultured in TM and incubated under standard conditions for epithelial cells (36.5 °C, 3.5 % CO2, 95 % relative humidity).

MTT and BrdU assays

Ten thousand cells/well were seeded into the wells of collagen-coated 96-well microtiter plates containing drugs at the concentrations indicated in the figures and tables. After 3 or 6 days of incubation, the number of viable cells was assessed utilizing MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide]. The MTT turnover was photometrically measured at λ = 570 nm using a Synergy2 system (BioTek Instruments, Bad Friedrichshall, Germany).

Furthermore, a colorimetric immunoassay to quantify cell proliferation by measuring the 5-bromo-2′-desoxyuridine (BrdU) incorporation during DNA synthesis was performed according to the manufacturer’s protocol (Roche, Mannheim, Germany). Incorporated BrdU was detected after incubation with anti-BrdU-POD and tetramethylbenzidine (TMB) by measurement of the absorbance at λ = 450 and 690 nm.

Patients and HNSCC specimens

The ex vivo study was approved by the ethics committee of the Medical Faculty of the University Leipzig (No. 201-10-12072010 and No. 202-10-1207210). After declaring their informed consent, biopsies of primary or recurrent HNSCC were obtained from 49 patients under general anesthesia during panendoscopy or definitive tumor resection and stored in tubes containing TM at room temperature. For chemoresponse testing in the FLAVINO assay, the material was immediately transferred to the laboratory. Inclusion criteria were samples derived from histopathologically confirmed HNSCC (ICD-O type 8070) and sufficient colony formation (≥4 vital colonies consisting of ≥3 pan-cytokeratin-positive cells with morphological features of viability at time of ethanol fixation in all control wells; see “FLAVINO assay of colony formation” section). Accordingly, 18 of 49 tested tumor samples were included in the further analyses (dropout rate 63 %). Characteristics of these 18 HNSCC according to the criteria of the Union internationale contre le cancer (UICC) [27] are shown in Table 1.
Table 1

Localization, stage (TNM, UICC) [27] and grading (G) of the investigated head and neck squamous cell carcinomas (HNSCC) according to the inclusion criteria

No.

Sex

Localization

TNM

UICC

Grading

1

Male

Paranasal sinus

T4a N0 M0

IVA

G2

2

Male

Larynx

T3 N3 M0

IVB

G2

3

Male

Larynx

T2 N2b M0

IVA

G3

4

Male

Tonsil

T3 N2c M0

IVA

G3

5

Male

Oropharynx

T4 N2c MX

IVA

G2

6

Male

Larynx

T2 N2b M0

IVA

G3

7

Female

Metastasis

TX N3 M0

IVB

G3

8

Male

Tonsil

T3 N2b M0

IVA

Not defined

9

Male

Metastasis

T0 NX M1

IVC

G2

10

Male

Metastasis

T3 NX M0

III

G2

11

Male

Nasopharynx

T4a N2b M0

IVA

G1

12

Male

Oral cavity

T4a N0 M0

IVA

G2

13

Male

Larynx

T4a N1 M0

IVA

G2

14

Female

Larynx

T3 N2c M0

IVA

G2

15

Male

Oropharynx

T4b N3 M0

IVB

G2

16

Male

Larynx

T4 N0 M0

IVA

G2

17

Male

Oral cavity

T2 N0 M0

II

G3

18

Male

Metastasis

TX N2b M0

IVA

G3

FLAVINO assay of colony formation

The specimens were mechanically cut into pieces <1 mm3 and enzymatically digested with collagenase type IV (230 mU/ml; Sigma) for 16 h. In medium, serial dilutions of Sim and of the chemotherapeutics were prepared at appropriate concentrations, based on data specifying the tolerable plasma level (TPL) in vivo (Cis: 6,667 nM [28], DTX: 550 nM [29]), and 200 μl was transferred into 48-well microtiter plates (Greiner Bio-One, Nürtingen, Germany) coated with extracellular matrix (ECM) proteins (collagen I, fibronectin and laminin; Roche, Mannheim, Germany). Thereafter, HNSCC digest was added to a final median concentration of 2.0 mg HNSCC per well [interquartile range (IQR) 1.8–2.3 mg] in a total of 600 μl. Control wells contained 0.1 % ethanol (v/v, for dilution of Sim). After incubation for 3 days (±1 day) under standard conditions, supernatants were removed, and cells were carefully washed with PBS and fixed with ethanol, followed by blocking with FCS. Colonies were then incubated with murine anti-pan-cytokeratin antibodies (Santa Cruz Technologies, Santa Cruz, CA, USA) and, upon washing, with Cy2-labeled goat-anti-mouse IgG (Jackson Immune Research, Suffolk, UK). Colony formation was assessed utilizing an inverted fluorescence microscope (Axiovert 200M; Carl Zeiss, Jena, Germany). According to the FLAVINO protocol, a quality control was performed by parallel testing of KB cells with the serial test drug dilutions from above and MTT assays.

Analysis of drug interaction

The modified probability sum test according to Jin [30] was applied to assess the mode of interaction, according to the formula: q = P[A + B]/(P[A] + P[B] − P[A] × P[B]). Jin defined “observed value/expected value (sum of probability of independent events)” ratios q < 0.85 as antagonism, while q = 1 ± 0.15 indicates additivity and q > 1.15 synergism.

Statistical analysis

Cutoff analyses were conducted in experiments with sufficient colony formation (see “Patients and HNSCC specimens” section). Calculation of the median and IQR was conducted to summarize the results of colony formation assays. Due to large heterogeneity in basic colony formation observed in controls of the eighteen HNSCC, summarizing statistics were calculated based on the relative values of colonies obtained after normalization to mean colony formation in ethanol-treated controls of the individual HNSCC. Mean, standard deviation and confidence intervals of MTT turnover and BrdU incorporation in cell lines were calculated, and normalized values of epithelial cell colony formation of the individual treatments were analyzed by a two-tailed t test for paired samples, using SPSS Statistics 18.0 for Windows, version 18.0.1 (SPSS Inc., Chicago, IL, USA). Differences were regarded as significant at p < 0.05.

Results

Effect of Sim on HNSCC cell lines in MTT, BrdU and in FLAVINO quality control assays

Initial testing of Sim on KB cells in the MTT assay after 3 or 6 days revealed inhibitory effects (Fig. 1a). A mean reduction in cell viability to 70 % at 5 μM Sim (95 % CI 57–83 %) was calculated for 3-day exposure. At this time point, the comparison of the numbers of viable cells at 0.15 and 0.5 μM Sim with ethanol-treated controls revealed a significant decrease (p < 0.05). Compared to control, all p values for Sim concentrations equal or above 1.5 μM were below 0.001. A significantly increased efficacy was detected by Sim dose escalation from 0.5 to 1.5 μM, and from 1.5 to 5 μM (both p < 0.05) as well as from 5 to 15 μM (p < 0.001). After 6 days, a statistically significant inhibition of proliferation as compared to the ethanol-negative control was already observed at 0.05 μM Sim. At 5 μM, Sim cell viability was decreased to 62 % (95 % CI 55–67 %). Concentrations of Sim ≥0.15 μM reached p < 0.001 compared to the control (Fig. 1a).
https://static-content.springer.com/image/art%3A10.1007%2Fs00280-014-2412-1/MediaObjects/280_2014_2412_Fig1_HTML.gif
Fig. 1

Effects of a simvastatin on KB cells after 3 and 6 days of incubation (MTT assay; n = 7 tests) and b simvastatin, cisplatin or docetaxel on KB cells (FLAVINO assay quality control; n = 32 tests, 3 days of incubation; mean and 95 % CI, *p < 0.05, **p < 0.01; ***p < 0.001). MTT (c) and BrdU assays (d) for monitoring the effects of simvastatin, cisplatin or various combinations of both, on FaDu cells after 3 days of incubation (mean, SE of the mean)

KB cell experiments according to the FLAVINO protocol were used for quality control and confirmed the finding of suppressive effects of Sim (Fig. 1b). Furthermore, multiple repetitions (n = 32) demonstrated the reproducibility and robustness of the FLAVINO assay. Sim at concentrations of 5 μM reduced the number of viable cells in the MTT assay after 3 days of incubation to 54 % (95 % CI 49–59 %, Fig. 1b). Due to the large number of tests, effects of all Sim concentrations reached significance (p < 10−10) in comparison with ethanol-treated controls. Moreover, 1.5 and 5 μM Sim showed significantly increased efficacy (p < 0.01) over lower doses.

To address the question of Sim effects on HNSCC cell lines, FaDu and HN5 cells were analyzed in MTT and BrdU assays. The MTT-based quantization revealed a Sim dose-dependent decrease in viable cell numbers, reaching ~25 % reduction upon incubation with the highest Sim concentration (50 μM; Fig. 1c, dark gray bars). Similar results were obtained in HN5 cells (data not shown). These findings were confirmed in BrdU assays which, when compared to the MTT assays, revealed similar Sim inhibitory effects on BrdU incorporation (Fig. 1d, dark gray bars).

Inhibitory effects of Sim on HNSCC primary cells

Next, we analyzed Sim effects on primary HNSCC samples in the FLAVINO assay with regard to the formation of epithelial (cytokeratin positive) colonies. This was done in all evaluable HNSCC (18/18; Fig. 2a, b, dark gray bars). Compared to the cell lines KB, FaDu and HN5, already lower concentrations of Sim led to suppressive effects on HNSCC primary cells. The median colony formation of these 18 HNSCC, normalized to their individual ethanol-treated controls, was significantly reduced at all concentrations of Sim (Fig. 2a, b; see Table 2 for p values). Increasing Sim doses exhibited significantly higher efficacy (Fig. 2a, b). At a concentration of 0.5 μM Sim, colony formation was reduced to 45 % (34–60 %, median, IQR), which was approximately 10 times lower than the concentration of a comparable reduction in proliferation of KB cells as determined in the MTT assay.
https://static-content.springer.com/image/art%3A10.1007%2Fs00280-014-2412-1/MediaObjects/280_2014_2412_Fig2_HTML.gif
Fig. 2

Combined effects of simvastatin with cisplatin (Cis) and docetaxel (DTX) on colony formation of n = 18 primary HNSCC in the FLAVINO assay. a Mean and 95 % CI. b Median and interquartile range

Table 2

Suppressive effects on colony formation of HNSCC

Docetaxel (nM)

Cisplatin (nM)

Simvastatin (μM)

0.15

0.5

1.5

5

p

0

0

1.630E−08

3.782E−18

6.275E−23

3.221E−24

137.5

0

3.571E−12

7.946E−16

2.122E−17

9.274E−27

275

0

1.281E−18

2.625E−23

2.577E−27

1.989E−18

550

0

1.717E−24

2.373E−25

1.658E−25

3.137E−25

0

1,667

2.128E−17

1.758E−12

2.391E−19

2.711E−20

0

3,333

1.555E−18

1.466E−20

3.064E−23

7.332E−23

0

6,667

7.649E−28

1.202E−33

5.979E−31

1.494E−33

p values obtained from t tests for paired samples (comparison of ethanol-treated negative controls with cells exposed to various concentrations of docetaxel or cisplatin, alone or combination with different concentrations of simvastatin)

Cis and DTX reduce HNSCC viability and colony formation in a dose-dependent manner

The number of viable KB cells (n = 32) was dose dependently reduced in MTT assays according to the FLAVINO protocol upon treatment with the guideline-conforming chemotherapeutic agents Cis and DTX for 3 days (Fig. 1b, center and right panels). More specifically, a decrease to 77 % was observed at 400 nM Cis (71–83 %; mean, 95 % CI) and to 66 % at 0.62 nM DTX (59–73 %; mean, 95 % CI). The MTT-based quantization revealed a slight ~20 % decrease in viable cell numbers to 81 % (75–86 %; mean, 95 % CI) upon incubation with the highest Cis concentration (6.7 μM; Fig. 1c, leftmost panel). Despite some differences in cell sensitivity toward treatment, similar results were obtained in HN5 cells (data not shown). These findings were confirmed in BrdU assays which, when compared to the MTT assays, revealed substantially more profound Cis effects on BrdU incorporation (Fig. 1d; 42 %; 29–55 %).

Cis treatment of primary HNSCC cells inhibited colony formation in 18/18 cases to 51 % at a Cis concentration of 1,667 nM (35–64 %; median, IQR; Fig. 2b, left panel) with the significance level already at p < 0.001, compared to ethanol-treated control. Likewise, DTX suppressed the colony formation in 100 % of analyzed HNSCC (44 % at CDTX = 137.5 nM, 30–54 %; median, IQR; Fig. 2b, left panel) at all DTX concentrations (p < 0.001).

Sim enhances inhibitory effects of Cis and DTX

As shown above, the MTT-based quantization of viable FaDu cells revealed a slight ~20 % decrease in viable cell numbers upon incubation with the highest Cis concentration (6.7 μM; Fig. 1c, leftmost panel) or dose dependently with increasing Sim concentrations. Notably, combinations of Cis and Sim led to more profound tumor cell inhibitory effects (Fig. 1c). Likewise, Sim enhanced inhibitory effects of Cis on BrdU incorporation in FaDu cells in an additive mode of action, e.g., addition of 1.5 μM Sim reduced BrdU incorporation to 15 % (6–24 %).

To identify possible additive or synergistic effects of Sim in combination with Cis or DTX on primary cells, HNSCC samples were treated ex vivo with binary mixtures of Sim, Cis and DTX at various concentrations. Upon treatment with all Sim plus Cis combinations, colony formation of HNSCC was significantly suppressed compared to the ethanol-treated control (Fig. 2a, b), with p < 0.001 using a two-tailed t test for paired samples (Table 2). When combined with 1.5 μM Sim, Cis profoundly decreased colony formation to 14 % (10–23 %; median, IQR) at half of the maximum tolerable dose (1/2 MTD: CCis = 3,333 nM) compared to ethanol control (Fig. 2b). Likewise, DTX in combination with Sim inhibited HNSCC colony formation significantly (p < 0.001) compared to ethanol-treated control (Table 2). Again, at half the MTD (CDTX = 275 nM) in a binary mixture with 1.5 μM Sim, colony formation was decreased to 14 % (10–20 %; median, IQR; Fig. 2b). More importantly, the comparison of effects exerted by Cis or DTX alone (leftmost panels in Fig. 2a, b) with Cis or DTX in binary combinations with Sim revealed increased cytotoxicities especially at high Sim concentrations. For example, at a given sub-therapeutic concentration of 1,667 nM Cis, the combination with 5 μM Sim significantly reduced colony formation from 51 % (35–64 %) to 10 % (7–27 %, median, IQR, Fig. 2b), which is a relative decrease of further 80 %. Likewise, at a baseline concentration of 137.5 nM DTX, the binary mixture with 5 μM Sim yielded a relative reduction in colony formation by 70 % (from 44 %, 30–54 % to 13 %, 7–17 %, median, IQR, Fig. 2b). Notably, this Sim-mediated enhancement of cytostatic effects was true for low as well as for high Cis or DTX concentrations and prompted us to analyze these interactions in more detail.

Analysis of drug interactions

The assessment of q values according to Jin (see “Analysis of drug interaction” section) revealed additivity as predominant mode of action regarding suppressive effects of either Cis or DTX in combination with Sim (Fig. 3a–d; Table 3). In binary mixtures of Sim and 1/2 MTD of either Cis (3,333 nM) or DTX (275 nM), the effects on HNSCC colony formation were additive in most cases. However, in some HNSCC, q values below 0.85 were determined, suggesting a mode of action that is best described as functional antagonism (combined effect below additivity). This definition is applicable since both substances individually exerted significant suppressive effects on the colony formation of epithelial cells, but in these dose combinations and in certain HNSCC, the gain in efficacy was below the lower level of the interval defining additivity. Compared to combinations with 6,667 nM, lower Cis concentrations tended to interact even more below additivity (lower q values) with decreased concentrations of Sim (Fig. 3a).
https://static-content.springer.com/image/art%3A10.1007%2Fs00280-014-2412-1/MediaObjects/280_2014_2412_Fig3_HTML.gif
Fig. 3

Interaction of simvastatin with either cisplatin (Cis; a, b) or docetaxel (DTX; c, d) on colony formation of n = 18 primary HNSCC. The q values obtained according to Jin [30] for n = 18 HNSCC (cf. Table 3) are summarized and presented. Antagonism, additivity or synergism are defined by q < 0.85, q = 1 ± 0.15 or q > 1.15, respectively. Median and interquartile range (a, c). Mean and 95 % CI (b, d)

Table 3

Analysis of drug interactions in binary combinations of either cisplatin or docetaxel with simvastatin at the indicated concentrations

Cisplatin (nM)

Docetaxel (nM)

Simvastatin (μM)

Antagonism

Additivity

Synergism

q < 0.85

0.85 ≤ q ≤ 1.15

q > 1.15

n

%

n

%

n

%

1,667

0

0.15

6

33.33

9

50.00

3

16.67

1,667

0

0.5

5

27.78

10

55.56

3

16.67

1,667

0

1.5

5

27.78

11

61.11

2

11.11

1,667

0

5

5

27.78

13

72.22

0

0

3,333

0

0.15

4

22.22

12

66.67

2

11.11

3,333

0

0.5

3

16.67

15

83.33

0

0

3,333

0

1.5

4

22.22

12

66.67

2

11.11

3,333

0

5

2

11.11

16

88.89

0

0

6,667

0

0.15

3

16.67

15

83.33

0

0

6,667

0

0.5

1

5.56

17

94.44

0

0

6,667

0

1.5

2

11.11

16

88.89

0

0

6,667

0

5

1

5.56

17

94.44

0

0

0

137.5

0.15

8

44.44

8

44.44

2

11.11

0

137.5

0.5

4

22.22

14

77.78

0

0

0

137.5

1.5

6

33.33

11

61.11

1

5.56

0

137.5

5

2

11.11

16

88.89

0

0

0

275

0.15

6

33.33

11

61.11

1

5.56

0

275

0.5

1

5.56

17

94.44

0

0

0

275

1.5

3

16.67

14

77.78

1

5.56

0

275

5

3

16.67

15

83.33

0

0

0

550

0.15

2

11.11

16

88.89

0

0

0

550

0.5

3

16.67

15

83.33

0

0

0

550

1.5

3

16.67

15

83.33

0

0

0

550

5

2

11.11

16

88.89

0

0

Results of the drug interaction analysis are given as number (n) and percentage (%) of 18 evaluable HNSCC and indicate antagonism, additivity or synergism

Classified according to the modified probability sum test proposed in Jin [30]

Combinations of Sim and DTX yielded smaller median q values than binary mixtures of Sim and Cis (Fig. 3a, c). More specifically, in the case of lower DTX and Sim concentrations, additive effects were observed in an only smaller number of HNSCC. Thus, compared to higher doses of DTX and Sim, low-level concentrations fostered rather antagonistic effects (Fig. 3c). This is exemplified by the combination of 0.15 μM Sim and 137.5 nM DTX, which resulted in 8/18 HNSCC responding antagonistically, while only 8/18 or 2/18 tumors, respectively, showed additive or synergistic effects (Table 3).

Discussion

Our experiments utilizing the FLAVINO assay-based ex vivo chemoresponse testing of primary HNSCC demonstrate a clear anticancer effect of Sim. While this is in full agreement with effects found here and previously in stable cell lines from different tumor entities, prompting large-scale epidemiological trials [11], we have now extended our studies toward biologically more relevant primary HNSCC cells. Doubtlessly, the FLAVINO assay cannot totally reflect the in vivo environment, including all pharmacologic and dynamic effects and feedback mechanisms. Also, despite the avoidance of trypsin digestion by only using collagenase IV besides mechanical disintegration, cell–cell contacts and tumor integrity are still impaired. Nevertheless, the experimental setup of the FLAVINO assay overcomes some serious limitations in chemoresponse assessment ex vivo by avoiding the somewhat artificial situation of cells growing two-dimensionally in a monolayer and photoactivation-induced chaotropic effects, and by allowing for the detection of chemoresponse at in vivo tolerable concentrations [1921, 25]. More importantly, the FLAVINO assay is an excellent model for the assessment of effects on primary cells, and our study indeed demonstrates differences between effects on the stable reference cell line KB, HNSCC cell lines and primary HNSCC. However, our rigorous quality management led to exclusion of all primary HNSCC of which colony formation could not be reliably assessed. Consequently, even colony-forming HNSCC were excluded when a large heterogeneity within the quadruplicate control wells was observed, i.e., when the criterion of at least four colonies per well was not fulfilled. The majority of HNSCC exhibiting ex vivo growth failed to attach to the surface of the plate while rather forming spheroids (spherical colonies growing in suspension and failing to tightly attach to the ECM-coated bottom of the well, as detected by transmitted light microscopy prior to ethanol fixation), and thus did not qualify for analysis according to the FLAVINO protocol. Such a behavior was observed predominantly in samples derived from recurrent or metastatic HNSCC. Since many HNSCC analyzed in this study were recurrent or metastatic (16/49 HNSCC, 33 %) disease, this contributed to a poor adhesion in several cases and a comparably low number of samples that allowed for their inclusion in our analyses. Despite these effects leading to a lowered number of HNSCC which could be included into our analyses, a number of n = 18 HNSCC should allow for a reliable assessment of Sim effects on primary HNSCC ex vivo.

The statin-intake-related effects in the clinical context both in terms of single agent use and in combination with cytostatic drugs are partially controversial, and results are poorly comparable so far. In a retrospective controlled trial, a statin-intake-associated risk reduction of 20 % in cancer incidence was observed for a multitude of organ sites and histological types, especially when statins were administered for more than 4 years (OR 0.64), or at higher doses [31]. A tendency toward a reduction in deaths from cancer was described under rosuvastatin treatment (p = 0.02) [1]. A significant reduction in breast cancer [32, 33] or prostate cancer [34] recurrence under statin treatment has been shown as well. In contrast, an increased incidence of breast cancer after pravastatin use was observed in one trial [9], and no benefit was documented in patients with gastric adenocarcinoma or glioblastoma receiving lovastatin [7, 8]. These somewhat controversial results might be due to heterogeneity in statin effects related to different tumor entities and histological type, but also to the fact that most of the studies were not originally designed to investigate anticancer effects of statins. Recently, Nielsen et al. [11] published data on statin-related mortality in n = 295,925 cancer patients. Multivariable-adjusted hazard ratios (HR) for death from cancer were 0.85 (95 % CI 0.82–0.87) for statin users compared to those without statin intake. Remarkably, patients with primary tumors of the larynx (n = 3,958; HR 0.66, 95 % CI 0.55–0.80) and pharynx (n = 7,932; HR 0.67, 95 % CI 0.51–0.87) presented with strongest benefits from statin treatment. Item, statin intake was associated with a 16 % risk reduction in gastric cancer in a conservative estimation (adjusted OR 0.84; 95 % CI 0.78–0.90) [35]. In a clinical phase I study with prolonged administration of lovastatin, 23 % of the patients (n = 26) suffering from advanced HNSCC or cervical carcinoma achieved stable disease for more than 3 months, but this failed to be significant [36].

Various studies have provided clear evidence that statins exert anti-proliferative effects, as shown here in KB and FaDu cells (Fig. 1), and other effects on various tumor cell lines from different tumor entities in vitro. Significant antineoplastic concentrations of Sim are described in ovarian cancer cell lines A2780 and UCI101 at 1 μM [17] or at 5 μM in esophageal adenocarcinoma cell line BIC-1 [14] and HNSCC cell line Tu167 [18]. Also, a significant anti-proliferative concentration of 10 μM Sim is depicted in glioma cell line C6 [13], esophageal adenocarcinoma cell line OE33 [14], colon cancer cell lines COLO205 and HCT116 [15], as well as HNSCC cell line JMAR [18]. We now extend these finding toward primary cells in a very prominent tumor entity and in the biologically more relevant FLAVINO assay. Our data reveal a significant anti-proliferative and colony formation suppressing effect of Sim at concentrations of 0.15 μM and above.

In general, statins are regarded as safe and well-tolerated drugs with serum LDL-lowering, triglyceride-reducing and HDL-increasing properties. Severe side effects like muscle toxicity or rhabdomyolysis are rare. This is also true for the fungal derivative Sim, which, despite extensive hepatic first-pass excretion that results in a low maximum bioavailability of 5 %, is found at peak plasma concentrations of 20–81 nM after a single 40 mg dose of Sim [10, 37, 38]. The standard daily dose of Sim administered is 40 mg/day; however, dosages up to 80 mg are approved and mostly well tolerated. Thus, it is likely that the dosages used in our assays may indeed reflect feasible local concentrations at the tumor site in a therapeutic setting in vivo.

The exact mechanism of the tumor suppressive effect of statins remains controversial, and even clinical data are conflicting. It is likely that statin-mediated effects on tumor growth and development accrue from a variety of different factors. Beyond studies indicating that anticancer effects induced by statins, such as inhibition of cell proliferation, induction of apoptosis or inhibition of angiogenesis [36], are due to the suppression of the mevalonate pathway, there is also evidence for more direct effects of statins based on oncogene repression. While this is beyond the scope of this study, our FLAVINO assay provides an excellent platform to address this point in further studies. Clendening and Penn [39] provided a brilliant overview on the metabolic interaction of statins in tumor cells. Furthermore, cholesterol synthesis blockade by statins is cross-linked with the HhP, which is essential in a variety of human cancers, including HNSCC [40], and partially may be inhibited by statins [2224, 41]. Recent studies and our own data have demonstrated the significance of the HhP in HNSCC [30, 40, 42, 43].

It remains an open question whether in vivo tolerable doses of Sim in combination with chemotherapeutics like Cis or DTX can significantly improve the patients’ outcome with regard to reduction of tumor burden, increased survival and/or less side effects due to lower dosages required. Our results demonstrate a reduction in FaDu cell proliferation and in the growth of primary HNSCC ex vivo, as determined by inhibited colony formation, and highlight the predominantly additive interactions of Sim in combination with Cis or DTX. Hence, our data provide the basis for the further evaluation of Sim in the context of established Cis- or DTX-based treatment regimens for HNSCC in vivo. Obviously, statins have gained interest beyond their lipid-lowering effects in terms of therapeutic value for anticancer treatment and may well contribute to future multimodal therapy regimens.

Acknowledgments

The authors would like to thank the colleagues from the Institute of Pathology (C. Wittekind, J. Bertolini and U. Bauer) for providing histopathological data. The laboratory staff (A. Reiche, G. Müller and K. Herrmann) is acknowledged for technical assistance and in particular for material transfer from surgery and for handling of specimens. M. Stoehr and C. Mozet contributed equally to this study under the supervision of G. Wichmann.

Conflict of interest

None.

Copyright information

© Springer-Verlag Berlin Heidelberg 2014