Clinical data was collected from 14 patients (mean age, 55.5 years) with advanced and heavily pre-treated epithelial tumours who received an anti-cancer treatment in the UNIFONTIS clinic (Tübingen, Germany) and who were followed up by PET-CT, with additional monitoring of CTC numbers in the blood (Table I). The majority of the patients (57%) underwent primary tumour removal. The remaining 43% of patients did not undergo primary tumor removal due to unresectable tumors or patient refusal. A total of 93% had histologically confirmed distant metastases at the time of treatment.

A total of 10 patients underwent treatment with salicylate therapy. The treatment with diflunisal or other salicylates (ASA and PAS) was applied following the recommendations of Kreutz (5). Usually, the drugs were administered intravenously for 4 days a week, 2 weeks in a row. The initial dose of salicylates on the first day was 35 mg/kg, administered intravenously. On the second to fourth days, patients received a dose of 30 mg/kg, administered intravenously. Intravenous application was selected to avoid gastrointestinal side effects, which may occur in up to 20% of patients treated with diflunisal. Salicylates were provided by OncoAdvance GmbH (Staufen, Germany). The patients who qualified for such a treatment were in the extreme advanced stage of the disease and had previously undergone multiple treatments. In such cases, the off-label use is justified in Germany. The patients were treated with salicylates only after an extensive informative conversation with the physician and all were required to provide written informed consent.

One patient received metronomic therapy in combination with the administration of capecitabine and trofosfamide combined with bevacizumab for 61 days. Metronomic low-dose chemotherapy was administered as a long-running daily oral application of cytostatics (capecitabine or trofosfamide) in extremely low doses (1 g/m² or 1 g absolute oral). In 2008, Reichle and Vogt (16) introduced this treatment schedule in combination with anti-inflammatory therapies and angiostatic therapies.

Tumour response was assessed by PET-CT. The optical evaluation of the response to a therapy followed the Response Evaluation Criteria In Solid Tumors (RECIST) criteria (17). PET was used to visualize the metabolic activity of the tissue. The metabolic assessment of tumour activity was documented, but metabolic activity data were not included in this analysis. The PET-CT was performed by an independent facility, the Radiological Clinic of the University Tübingen (Tübingen, Germany).

For the RECIST criteria, the number of target lesions is equal to ≤2 per organ, with up to 5 in total. A complete response (CR) is classified as the disappearance of all target lesions. The short axis of any pathological lymph nodes must be reduced to <10 mm. A partial response (PR) is classified as a ≥30% decrease in the sum of the diameters of the target lesions (SLD), using the baseline sum of diameters as a reference. Progressive disease (PD) is defined as a ≥20% increase in the SLD, using the smallest sum on the study as a reference. The SLD must also show an absolute increase of ≥5 mm. Stable disease (SD) is classified as insufficient shrinkage to qualify for PR or an insufficient increase in SLD to qualify for PD.

The quantification of CTCs from the whole blood was performed as previously described (14). In short, red blood cells from 1 ml blood were lysed and the remaining white blood pellet was analysed. Epithelial cells were detected by laser scanning cytometry using the Olympus ScanR screening station (Olympus Corporation, Tokyo, Japan). For staining, the cells were incubated at 4°C overnight with a monoclonal fluorescein-isothiocyanate (FITC)-conjugated EpCAM antibody (mouse α-human; catalog no. 130-080-301; 1:100; Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). On the second day, dead cells were detected with propidium iodide (PI), which is excluded from viable cells but incorporated in dead cells. The baseline CTC assessment for each treatment was performed prior to (maximum, 4 months previously) or during the treatment, and the follow-up assessment was performed at the earliest 2 weeks after the beginning of treatment and between 1 and 6 months after the baseline assessment. A >5-fold decrease in CTCs was considered a ‘significant decrease’, while a >5-fold increase was considered a ‘significant increase’ in CTCs. A <5-fold increase or decrease was considered a stable CTC course. The chemosensitivity testing of the patients' CTCs was performed as described previously (18). CTCs among the white blood cells from the whole blood were labelled with the FITC-conjugated EpCAM antibody, as aforementioned. The cytotoxicity of the drug in question was determined by exposing the cells to three different concentrations: The physiological concentration of the drug in the blood of a patient under treatment, plus a 10-fold higher and a 10-fold lower concentration. Dead cells could be distinguished from living cells by PI and EpCAM antibody staining, and subsequent quantification with the laser scanning cytometry. The chemosensitivity rate was calculated as the ratio of dead cells to the total cell number in the sample. The quantification of CTCs from the whole blood and the chemosensitivity testing was performed in the diagnostic laboratory of Dr Ulrich Pachmann (Transfusion Medical Centre Bayreuth, Bayreuth, Germany).

All statistical analysis was performed using SigmaPlot software (version 13.0; Systat Software GmbH, Erkrath, Germany). Data was analyzed using Pearson's χ² test with 4 degrees of freedom. P<0.05 was considered to indicate a statistically significant difference.

For the present study, 25 patients were screened using the inclusion criteria. For inclusion in the study, the patients had to have at least one chemosensitivity testing, at least one medical treatment followed by PET-CT response assessment and at least two CTC counts, the first one (baseline) shortly prior to or shortly after the beginning of treatment and the second one (follow-up) between 1 and 6 months later. In total, 14 patients met the inclusion criteria. The in vitro sensitivity of the CTCs to salicylates, selected cytostatics and tyrosine kinase inhibitors (TKIs) was tested for each patient individually. Diflunisal was found to be an effective cytotoxic agent in 11 patients (mean chemosensitivity-index, 86%) and PAS was found to be an effective cytotoxic agent in 8 patients (mean chemosensitivity-index, 73%), as shown in the results of the in vitro chemosensitivity test (Table II). The cytostatic drugs used in these patients (cisplatin, capecitabine, trofosfamide, paclitaxel, docetaxel and carboplatin) showed high cytotoxic activity ranging from 40 to 95%. The TKIs pazopanib and imatinib also showed cytotoxic activity of 95% in 2 patients.

The treatment response data of all patients assessed by PET-CT was collected. A total of 21 treatments were assessed (Table III). Additionally, the patient's condition was assessed without previous treatment 6 times (Table III). The median age of the patients was 55.5 years. PET-CTs that were performed no earlier than 2 weeks after the beginning of treatment and no later than 4 months after the treatment were included in the evaluation. CR to a treatment was observed in 1 case after diflunisal treatment. PR to a treatment was observed in 5 cases, three times after treatment with cytostatic drugs in combination with bevacizumab, once after treatment with diflunisal and once after a combined treatment with ASA, tamoxifen and bevacizumab. SD after treatment with diflunisal or other salicylates was observed in 7 cases. SD after treatment with cytostatic drugs was observed in 1 case and SD without previous treatment was observed in 1 case. The objective response rate, which is the sum of the CR and PR rates, was 22%. PD following a treatment was observed in 7 cases and a PD without previous treatment was documented 5 times.

An increase in cell numbers is assumed to correspond to high tumour activity and poor prognosis (13). The results from the present study showed that the behaviour of CTC numbers correlates well with the imaging response. The positive imaging response (CR or PR) correlated with a CTC reduction in 5/6 (83%; P=0.030) of cases. The assessment of a SD correlated with a stable CTC course in 7/9 cases (78%; P<0.001). The assessment of a PD revealed no correlation with an increase in cell numbers in 6/12 (50%; P=0.368) of cases (Table IV). The χ² test confirmed that PET-CT outcome and CTC course were significantly correlated (P=0.0018). Fig. 1 shows three examples of individual treatment courses. CTC quantity was measured during the treatment periods and for follow-up. A patient who was suffering from an endocervical carcinoma started with stable CTC numbers during the initial diflunisal treatment (Fig. 1A). The following decrease from 600 to 0 cells was accompanied by a CR, and the following rise was accompanied by PD. Thereafter, a tamoxifen/bevacizumab treatment resulted in a PR, which was accompanied by stable CTC numbers (Fig. 1A). The patient in Fig. 1B exhibited rising CTC numbers during the paclitaxel/bevacizumab treatment, but then falling tumour cell numbers indicated successful treatment. Rising cell numbers during paclitaxel treatment have been previously observed and are assumed to originate from enhanced cell dissemination from the primary tumour and most probably also from metastases (19). The treatment success was also reflected by a PR. After the chemotherapy, cell numbers were rising and could be again reduced by paclitaxel therapy. However, 3 months later, PD was observed despite previously falling numbers. Whether a repeated increase in cell numbers, as had been observed prior to the previous progression, had occurred could not be evaluated, since the next analysis was performed only after an additional treatment. Thereafter a stable CTC course resulting from treatments with doxorubicin/bevacizumab and cyclophosphamide/bevacizumab treatment was reflected by findings of SD in PET-CT. Furthermore, the CTC course of a breast cancer patient (Fig. 1C) showed a decrease in the number of CTCs after a 4-week diflunisal treatment. Subsequent rising CTC numbers were accompanied by PD. Another diflunisal treatment (for 4 weeks) and the following low-dose chemotherapy resulted in a decrease in CTC numbers, while the disease continued to progress, as indicated in Fig. 1C. Again, it was assumed that this was an incomplete response characterized by a positive response on the CTC level, but PD observed by the imaging technique.

The principles outlined in the Declaration of Helsinki and the German Data Protection Act were followed during data collection. Informed consent for the publication of clinical data and CTC count data was obtained from the participants involved.