Introduction

The greatest challenge in treatment management of metastatic breast cancer (MBC) is to identify early indicators of response to systemic treatment to avoid unnecessary exposure to ineffective but toxic treatment and to enable early prognostication of progression-free (PFS) and overall survival (OS). The presence of ≥5 CTC in 7.5 ml of peripheral blood prior to a new line of treatment is hence an independent factor for poor prognosis in MBC patients [1–6]. The malignant nature and metastasis-initiating potential of CTC is extensively reported in the literature [7, 8]. Previous studies reported that CTC provide more prognostic information than conventional imaging [9], could help identify patients who benefit from a more aggressive treatment regimen, and could be used to monitor treatment response [10–12]. Nevertheless, CTC represent a heterogeneous cell population with phenotypic changes compared to the primary tumor [13, 14] and bear great potential for diagnostic and therapeutic utilization as a liquid biopsy.

Apoptotic CTC (aCTC), which have been reported in 52–79 % of CTC-positive MBC patients as a CTC subtype [15–17], are characterized by altered morphological parameters such as speckled pattern of keratin staining and/or fragmented or disintegrated nuclei. Apoptotic CTC seem to derive from therapy-induced apoptosis and apparently from spontaneous tumor apoptosis as they also appear in patients with progressive disease and no response to systemic therapy [18, 19]. In addition, patients with MBC presented with significantly lower numbers of aCTC compared to patients with early breast cancer [17]. It is suggested that the viability of CTC is related to the stage of disease and aCTC might provide additional prognostic information [20]. Enumeration of disseminated tumor cells (DTC) from bone marrow showed that the appearance of apoptotic DTC (aDTC) is predictive of positive response under neoadjuvant systemic therapy [21, 22]. This underlines the hypothesis that aCTC might serve as a surrogate endpoint for successful systemic therapy [17].

The prognostic impact of both baseline CTC enumeration and kinetics (CTCKIN; change in CTC count from baseline (CTCBL) to first completed cycle of a new line of systemic therapy (CTC1C)) has been demonstrated in recent studies [3]. The present study aimed to prospectively assess the CTC status separately for the apoptotic CTC (aCTC) and intact CTC (iCTC) subtypes in a large group of patients. To this end, we analyzed CTCBL, CTC1C, and CTCKIN for their impact on treatment response, PFS, and OS.

Methods

Patients and study design

This was a prospective, single-center, non-randomized, partially blinded, treatment-based study. Both patients and treating physicians were blinded to CTC status, and hence treatment regimens did not depend on CTC status. All investigators and technical staff who performed or reviewed the CTC data were blinded to patient history and treatment. Independent reviewers confirmed CTC enumeration and characterization. All radiologists performing imaging studies were blinded to the patients’ treatment regimens. The study was conducted at the National Center for Tumor Diseases (NCT), Heidelberg, Germany, and the Department of Obstetrics and Gynecology, University of Heidelberg, Heidelberg, Germany.

Study design

MBC patients were consecutively enrolled between March 2010 and May 2015 at the beginning of new line of systemic therapy. Blood samples were taken prior to treatment to determine the baseline counts (CTCBL) of iCTC and aCTC. Counts ≥5 CTC/7.5 ml peripheral blood were considered CTCBL-positive [23]. After the first cycle of systemic therapy, a second blood sample was taken (CTC1C).

The first 100 patients in this study had no further blood samples taken after a negative CTCBL result. Every 3 months patients were categorized as showing progressive disease (PD), stable disease (SD), complete remission (CR), or partial response (PR) based on the Response Evaluation Criteria in Solid Tumors [24]. Survival status was recorded until death or loss to follow-up.

Enumeration of CTC

For CTC enumeration, 7.5 ml peripheral whole blood was collected in a CellSave tube (J Janssen Diagnostics, LLC, Raritan, NJ, USA). Blood samples were kept at room temperature for ≤96 h until analysis using the CellSearch™ assay (CellSearch™ Epithelial Cell Kit/CellSpotter™ Analyzer, Janssen Diagnostics, LLC, Raritan, NJ, USA). Sample processing and analysis were done strictly according to the manufacturer’s instructions. The assay uses a ferrofluid coated with antibodies to epithelial cell adhesion molecule (EpCAM) to immunomagnetically separate cells of epithelial origin from blood, and fluorescent staining to differentiate between debris, hematopoietic cells, and epithelial cells [13]. It provides high intra-observer, inter-observer, and inter-instrument agreements [2, 5, 15, 25]. Cells enriched by anti-EpCAM antibodies were labeled with the nuclear dye 4′,6-diamidino-2-phenylindole (DAPI) and immunostained with monoclonal antibodies specific for keratins and for the leukocyte common antigen CD45. CTC detection was performed by trained staff using the CellSpotter™ Analyzer, a semi-automated fluorescence-based microscopy system that enables computer-generated reconstruction of cellular images [16, 21, 26]. Morphologically intact, CD45-negative CTC without obvious alterations of nuclei and non-speckled keratin immunofluorescence were defined as iCTC and enumerated by trained operators. Patients with counts ≥5 intact CTC/7.5 ml blood were considered iCTC-positive.

Characterization of apoptotic CTC (aCTC)

Apoptotic CTC were visually characterized by altered morphological parameters such as speckled keratin staining patterns and/or fragmented or disintegrated nuclei. In select cases, classification as aCTC was proven by addition of the M30 antibody for the detection of caspase-cleaved Keratin-18 (VLV bio, 1:100) in the fourth channel of the CellSearch system. The vast majority of CTC with characteristic morphologic changes also was positive for M30 (Fig. 2). Patients with counts ≥5 aCTC/7.5 ml blood were considered aCTC-positive.

Evaluation of CTC kinetic

To analyze CTC kinetics from baseline (CTCBL) until the end of first cycle of systemic therapy (CTC1C), changes were categorized as CTC negative (<5 CTC for both CTCBL and CTC1C), stable (<25 % change), indicating decrease (≥25 % decrease), or indicating increase (≥25 % increase) [12]. Kinetics were separately evaluated for the following cutsoff: ≥5, ≥10, ≥25, and ≥50 CTC/7.5 ml (Supplementary Figs. 1, 2; Supplementary Table 1, 2).

Data collection and analysis

Demographic data and clinical characteristics were described as frequency and percentage, median and range, or mean and standard deviation. Groups were compared using the Wilcoxon rank test or Fisher’s exact test, as appropriate. Kaplan–Meier plots by aCTC and iCTC statuses were generated for PFS and OS (time from enrollment to disease progression and death from any cause, respectively), with data being censored at last follow-up if progression or death had not occurred. PFS and OS times were estimated as medians with 95 % confidence intervals (CIs). Differences in PFS and OS by iCTC and aCTC statuses were assessed by the log-rank test. Statistical analyses were performed using R (version 3.1.2) [27]. A significance level of 5 % was chosen.

Results

CTC status and enumeration at baseline (CTCBL)

442 consecutive patients were enrolled in the study. Figure 1 shows the flow of patients through the study, which also details the reasons for exclusion from or non-availability for further analysis. Patient characteristics at baseline are summarized in Table 1. Of the 442 available patients with CTCBL counts, 155 (35.1 %) were iCTCBL-positive and 128 (29.1 %) patients were aCTCBL-positive (Fig. 2). 108 (24.4 %) patients were both positive for iCTCBL and aCTCBL (pos pos), and 267 (60.4 %) patients were both negative for iCTCBL and aCTCBL (neg neg), as shown in Table 3. Positivity rates for higher cutoff values (≥10, ≥25 and ≥50) are listed in Table 4.

Fig. 1
figure 1

Flow of patients through the study. 466 consecutive patients were assessed for eligibility, 24 (5.2 %) were excluded from the study because no CTCBL data were available, or no follow-up was performed. Of the 442 patients included in the study, 190 were excluded from further analysis for the following reasons. During the initial phase of the study, CTC1C status of the first 100 patients was routinely determined only in CTCBL positive patients, resulting in 64 patients without measured CTC1C counts. Of the remaining 116 patients without CTC1C counts, 18 were excluded because blood samples were not obtained within the predefined study timeframe of 2–14 weeks, 30 did not survive to CTC1C assessment, 10 had no survival data, and 68 patients had not yet proceeded to CTC1C

Full size image
Table 1 Patient characteristics for apoptotic CTC enumeration at baseline (aCTCBL)
Full size table
Table 2 Numbers of patients, OS, and PFS for iCTC and aCTC statuses at baseline (CTCBL)
Full size table
Fig. 2
figure 2

aCTC and iCTC morphologic characteristics. CellSearch image gallery displaying morphologically intact CTC (1–5) and apoptotic CTC (6–10). KER/DAPI represents a composite image of PE-keratin (KER) and DAPI (nuclei). CD45 positivity of leukocytes is demonstrated in the APC channel and applying a FITC-labeled anti-M30 antibody in the fourth channel, M30-FITC-positive CTC present with speckled keratin staining patterns and/or fragmented nuclei

Full size image
Table 3 OS and PFS for iCTC and aCTC at baseline (CTCBL)
Full size table
Table 4 OS for different cutoff values at baseline (CTCBL)
Full size table

While in most cases iCTCBL positivity corresponded with aCTCBL positivity, 47 (10.6 %) patients were iCTCBL-positive and aCTCBL-negative (pos neg) and 20 (4.5 %) were iCTCBL-negative and aCTCBL positive (neg pos) (p < 0.001). Figure 3 illustrates the correlation of iCTC and aCTC counts per 7.5 ml (r = 0.7).

Fig. 3
figure 3

Scatterplot—Numbers of intact to apoptotic CTC. Log scales and points are jittered to avoid overplotting, and cutoffs of five cells are indicated by dotted lines, and the solid line represents identity

Full size image

CTC status and survival at baseline

Follow-up data were available for 442 patients with a median [95 % confidence interval (CI)] follow-up of 34.1 [32.0–36.9] months for OS. Median [95 % CI] OS for iCTCBL-positive versus iCTCBL-negative patients was 13.1 [10.0–15.9] versus 27.0 [24.0–31.7] months (p < 0.001), respectively (Kaplan–Meier plots not shown). Median PFS was 4.9 [4.1–6.1] versus 7.8 [6.9–9.2] months (p < 0.001), respectively (Table 2). Median [95 % CI] OS for aCTCBL-positive versus aCTCBL-negative patients was 10.9 [7.9–15.5] versus 26.9 [22.5–29.9] months (p < 0.001), respectively (Suppl. Material). Median PFS was 4.6 [4.0–6.1] versus 7.7 [6.5–9.2] months (p < 0.001), respectively (Table 2).

Figure 4 shows Kaplan–Meier plots for PFS and OS by iCTC and aCTC status at baseline. Median [95 % CI] OS for iCTCBL-positive/aCTCBL-positive (pos pos), iCTCBL-positive/aCTCBL-negative (pos neg), iCTCBL-negative/aCTCBL-positive (neg pos) and iCTCBL-negative/aCTCBL-negative (neg neg) patients was 10.3 [7.4–15.0], 16.4 [11.1–37.9], 29.9 [10.4–NA], and 27.9 [24.0–32.5] months (p < 0.001) and median PFS was 4.4 [3.5–5.9], 6.1 [4.2–9.8], 6.3 [3.4–19.4] and 7.9 [6.9–9.4] months (p < 0.001) as shown in Table 3.

Fig. 4
figure 4

Kaplan–Meier plots—OS and PFS separated in apoptotic and intact CTC at baseline (aCTCBL and iCTCBL). OS (left) and PFS (right) by CTC status at baseline in 442 patients with MBC; pos pos (iCTCBL-positive and aCTCBL-positive), pos neg (iCTCBL-positive and aCTCBL-negative), neg pos (iCTCBL-negative and aCTCBL-positive), and neg neg (iCTCBL-negative and aCTCBL-negative)

Full size image

This indicates that iCTCBL-positive patients in conjunction with positive aCTCBL (pos pos) had a significantly lower OS compared to iCTCBL-positive/aCTCBL-negative patients (pos neg) (PFS 4.4 vs. 6.1 months, p = 0.166; OS 10.3 vs. 16.4 months, p = 0.012). In contrast, the iCTCBL-negative/aCTCBL-positive (neg pos) group showed no significant impact on PFS and OS compared to iCTCBL-negative and aCTCBL-negative (neg neg) (PFS 6.3 vs. 7.9 months, p = 0.665; OS 29.9 vs. 27.9 months, p = 0.360 univariate).

The group iCTCBL-positive/aCTCBL-positive (pos pos) also showed a significantly higher iCTCBL count when compared to iCTCBL-positive/aCTCBL-negative (pos neg) patients (103.6 (179.0), respectively 28.9 (75.1), mean (STD); Wilcoxon test, p < 0.001).

All proven higher cutoffs for CTC positivity (≥10, ≥25 and ≥50/7.5 ml) were significantly associated with decreasing overall survival (Fig. 8), thereby reducing the prognostic impact of additionally enumerating apoptotic CTC (Table 4).

Table 5 OS and PFS for apoptotic and intact CTC at first cycle (CTC1C)
Full size table

CTC status at first cycle (CTC1C)

The results for CTC1C enumeration are shown in Table 5. The percentage of aCTC-positive patients decreased significantly to 41/252 (16.3 %) from CTCBL to CTC1C (p < 0.001, McNemar test). Figure 5 shows Kaplan–Meier plots for PFS and OS by iCTC and aCTC status after the first cycle of a new line of systemic therapy (CTC1C). Median OS [95 % CI] for iCTC1C-positive and aCTC1C-positive (pos pos), pos neg, neg pos and neg neg patients were 6.7 [5.0–13.9], 13.3 [10.9–22.5], 10.4 [4.1–NA], and 31.8 [26.9–37.9] months (p < 0.001), respectively, 3.5 [3.2–5.5], 5.2 [3.8–7.7], 2.9 [2.7–NA] and 8.0 [6.9–10.1] months for PFS (p < 0.001) as shown in Table 5.

Fig. 5
figure 5

Kaplan–Meier plots—OS and PFS separated in apoptotic and intact CTC at first cycle (aCTC1C and iCTC1C). OS (left) and PFS (right) by CTC status after the first cycle of a new line of systemic therapy 252 patients with MBC; pos pos (iCTC1C-positive and aCTC1C-positive), pos neg (iCTC1C-positive and aCTC1C-negative), neg pos (iCTC1C-negative and aCTC1C-positive), and neg neg (iCTC1C-negative and aCTC1C-negative)

Full size image

The pos pos group showed significantly higher iCTC1C counts than the pos neg group (96.1 (170.1) vs. 24.5 (31.0), mean (STD), p = 0.015).

CTC kinetics from baseline to first cycle (CTCKIN)

Intact CTC and aCTC enumeration for CTCKIN is shown in Table 6. High numbers in the diagonal cells indicate that iCTCKIN and aCTCKIN evolve similarly in most cases (Cramer’s V = 0.50). Most patients were negative for both iCTC and aCTC at CTCBL and CTC1C (50.0 %), followed by decreased iCTC and aCTC levels (17.5 %).

Table 6 CTCKIN from CTCBL and CTC1C for iCTC and aCTC
Full size table

CTCKIN and survival

Figures 5 and 6 show Kaplan–Meier plots for PFS and OS by CTC kinetics. Patients with counts <5 for iCTCKIN, aCTCKIN, and (iCTC + aCTC)KIN combined in both samples had the best outcomes for OS (32.9, 27.9, and 31.8 months, respectively) and PFS (7.9, 7.7, and 7.9 months, respectively). Stable or increased iCTCKIN, aCTCKIN, and (iCTC + aCTC)KIN counts showed the shortest OS (10.7/9.5, 5.1/5.7, and 10.8/7.4 months, respectively) and PFS (4.6/4.0, 4.1/3.2, and 5.5/3.3 months, respectively) as shown in Table 7. Decreased aCTCKIN had most favorable outcomes compared to iCTCKIN and decreased (iCTC + aCTC)KIN for OS (22.2 vs. 16.4 and 21.1 months) and PFS (7.7 vs. 6.1 and 6.1 months). aCTCKIN showed greater discriminatory power for OS and PFS between the favorable groups (<5 CTC at any time or decrease) and the unfavorable groups (stable/increase) than did iCTCKIN. Kaplan–Meier plots are shown in Figs. 5, 6. However, this discriminating power decreases with the increasing CTC cutoff values applied, suggesting that separate counting of aCTC is most meaningful at CTC levels of <10/7.5 ml (Supplementary Figs. 1, 2).

Fig. 6
figure 6

Kaplan–Meier plots—Kinetic of intact CTC between baseline and first cycle (iCTCKIN). OS (left) and PFS (right) by CTC status between baseline and first cycle in 252 patients with MBC: <5 CTC (<5 CTC in cells in both CTCBL and CTC1C), stable (<25 % change), decrease (≥25 % decrease), and increase (≥25 % increase)

Full size image
Table 7 OS and PFS for iCTC and aCTC kinetics
Full size table

CTCKIN and systemic therapy

The regimen of individual treatment for patients depended on the molecular subtype of both primary tumor tissue and metastatic tissue, previous therapeutic response, and clinical situation of the patient, following the German guidelines of the AGO [28]. During the first cycle of systemic therapy, 68 % of patients were treated with chemotherapy (CHT), including mono and poly-CHT after enrollment in the study, 20 % received Bevacizumab in addition to the CHT. 29 % of patients received endocrine therapy (such as Tamoxifen, Fulvestrant, Exemestane, and Letrozole) and 16 % of patients received Anti-HER2 therapy (such as Trastuzumab, Pertuzumab, and Lapatinib).

Table 8 shows the proportion of favorable aCTCKIN (<5 aCTC and decreased aCTC) from baseline to first cycle in relation to systemic therapy during the first cycle after patient enrollment regarding OS and PFS. The patient subset receiving endocrine therapy exhibited the longest OS compared to monochemotherapy, polychemotherapy, and chemotherapy with Bevacizumab (33.4 months vs. 18.1, 22.1, 20.4 months, respectively) and displayed the highest proportion of negative and decreasing aCTC (92 % vs. 89, 86, 87 %, respectively). Anti-HER2 treatment compared to no Anti-HER2 therapy in the first cycle of systemic therapy was associated with longer OS (27.9 vs. 20.4 months) and PFS (8.1 vs. 6.3 months) and a higher proportion of favorable aCTCKIN (97 vs. 87 %).

Table 8 OS, PFS, aCTCKIN sorted by systemic therapy from baseline (CTCBL) to first cycle (CTC1C)
Full size table

Discussion

As reported previously, iCTC at baseline have a high impact on both OS and PFS [3, 10]. Our results from a large cohort of 442 patients confirm that iCTC-positivity has the strongest negative impact on prognosis [3]. However, aCTC might have the potential to tip the scale in undecided situations whether to change, intensify, or stay with the therapy scheme [20].

Prognostic role of aCTC at baseline

The presence of aCTC alone at baseline has a similar favorable outcome compared to the complete absence of CTC (iCTC/aCTC) as evidenced by OS 29.9 versus 27.9 and PFS 6.3 versus 7.9 months. iCTCBL-negative/aCTCBL-positive patients (neg pos) had an even better, though not significantly better OS compared to patients without any CTC (iCTC/aCTC; neg neg). These aCTC in the bloodstream may represent tumor tissue undergoing chemotherapy-induced apoptosis. In line with Fehm et al. [21, 22], aCTCs could be similar to aDTC used as surrogate for systemic therapy response. However, only a small fraction of patients (4.5 %) were iCTCBL-negative/aCTCBL-positive (neg pos) at CTCBL, and the number of events in this subgroup is limited. Therefore, these results should be interpreted with care even though they might support the hypothesis.

Our data suggest that aCTC enumeration in combination with elevated iCTC at baseline compared to negative aCTC with positive iCTC at baseline has a strong impact on OS (10.3 vs. 16.4 months) and PFS (4.4 vs. 6.1 months). Proliferating tumor tissue is associated with necrosis and apoptosis, and therefore aCTC have been considered a sign of tumor proliferation [16, 29]. Following this reasoning, aCTC appear to be the result of therapy-induced apoptosis and tumor cell apoptosis in the context of carcinoma equilibrium and could be regarded as a natural side product of iCTC and tumor tissue [19, 29]. This hypothesis is supported by the fact that iCTCBL-positive/aCTCBL-positive (pos pos) patients had significantly higher amount of iCTC counts and showed the worst prognosis for OS and PFS.

Our results suggest that in the presence of ≥5 iCTC/7.5 ml, aCTC do not predict positive therapy response but outline a heterogeneous picture including tumor cell homeostasis and therapy-induced cell death and thus extend the picture of therapy response status [18, 19]. However, applying increasing cutoff values for iCTC positivity from ≥5 CTC to ≥50 CTC/7.5 ml reduced the additional prognostic impact of aCTC (Fig. 8).

Prognostic role of aCTC at first cycle

To investigate the diagnostic relevance of aCTC, we analyzed the enumeration of aCTC during treatment with the clinical prognosis. The data show a strong relation of iCTC and aCTC kinetics (Cramer’s V = 0.49). Eighty-five percent of patients were either positive for both iCTC and aCTC (pos pos) or negative for both iCTC and aCTC (neg neg). iCTCBL negative patients had the most favorable prognosis for OS (p < 0.001) and PFS (p < 0.001) independent of the aCTC status.

Prognostic role of aCTC kinetics

Differences in OS and PFS were observed for the aCTC kinetics (aCTC decrease (OS 22.2, PFS 7.7) and stable/increased aCTC count (OS 5.1/5.7, PFS 4.1/3.2)). A decrease of ≥25 % in aCTC had a better long-term outcome than stable or increased aCTC count independent of iCTC status.

Apoptotic CTCKIN has more discriminatory power for treatment response than iCTCKIN or the combination of iCTC and aCTC kinetics ((iCTC + aCTC)KIN). The Kaplan–Meier plots for aCTCKIN better reflect the prognostic situation for the single categories than iCTCKIN (as indicated by Figs. 6 and 7). Stronger discriminating results were observed for aCTC compared to iCTC decrease or iCTC + aCTC decrease for OS (22.2 vs. 16.4 and 21.1) and PFS (7.7 vs. 6.1 and 6.1). However, since aCTC levels are significantly associated with iCTC levels, separate enumeration of aCTC is only of prognostic relevance at CTC levels below 10 CTC/7.5 ml Fig. 8. The discriminating power decreases with the increasing CTC cutoff values applied (Supplementary Figs. 1 and 2).

Fig. 7
figure 7

Kaplan–Meier plots—Kinetic of apoptotic CTC between baseline and first cycle (aCTCKIN). OS (left) and PFS (right) by CTC status between baseline and first cycle in 252 patients with MBC: <5 CTC (<5 CTC in cells in both CTCBL and CTC1C), stable (<25 % change), decrease (≥25 % decrease), increase (≥25 % increase)

Full size image
Fig. 8
figure 8

Kaplan–Meier plots—OS with different cutoff values for iCTC and aCTC at baseline. OS by CTC status at baseline in 442 patients with MBC

Full size image

Different types of systemic treatment had no independent influence on aCTC. Treatment groups with high proportions of favorable aCTCKIN (<5 and decreased aCTC) showed longer OS and lower numbers of aCTC at baseline as shown in Tables 1 and 8. Thus, aCTCKIN seems to be mostly influenced by treatment response but by treatment regimen.

Alongside the response of the tumor tissue to treatment, the aCTC levels also appear to subside in the long term. These results underline the hypothesis that aCTC are a side product of metastatic disease [29]. Apoptotic CTC enumeration could thus be used as an extended parameter to monitor systemic therapy in MBC patients.

Limitations

When interpreting CTC kinetics, it must be borne in mind that only patients who survived the first cycle of systemic therapy were included in this statistic, which inherently leads to an overestimation of OS and PFS. This possibly affects CTCBL-positive patients more than CTCBL-negative patients. The first 100 patients in this study had no further blood samples taken after receiving a negative CTCBL result at baseline. This change may have introduced a potential source of bias in the CTC1C results. Also, the sensitivity of CTC detection has limits since certain EpCAM/keratin-negative CTC are not detected via CellSearch©, representing another limitation of the study. Especially systems using EpCAM might miss cells undergoing epithelial–mesenchymal transition (EMT) [2]. This potentially leads to underestimation of CTC levels. Nevertheless, EpCAM-positive CTC measurement is a very reliable and reproducible procedure offering potentially relevant clinical information with little effort and little discomfort to the patient [26]. As an additional prognostic factor, the identification of aCTC as a subgroup of CTC provides further clinically relevant information. Whether morphologic criteria or automated analysis platforms such as M30 epitopes [16] as methods for identification of aCTC are chosen, needs to be evaluated in further trials.

Impact of aCTC in MBC

In summary, our study demonstrates that serial CTC monitoring of iCTC and aCTC is a versatile tool for predicting treatment outcome in MBC and a useful adjunct to standard diagnostic tests for tailoring therapy. The data presented here support the hypothesis that monitoring aCTC is a promising source of biological information toward predicting the course of disease and its responsiveness to targeted agents, thus paving the way for individualized therapy [8, 10, 30].

Conclusions

Separate classification and enumeration of aCTC (≥5 aCTC/7.5 ml) at baseline can tip the scale as a predictive marker for OS. Similarly, increasing cutoff values for unfavorable iCTC counts go along with an increasing discriminating power for OS. Positive aCTCBL in conjunction with positive iCTCBL have an unfavorable prognostic impact on OS. In addition, the decrease in aCTCKIN over time is associated with improved outcomes in terms of both OS and PFS and has more discriminatory power compared to iCTCKIN or (iCTC + aCTCKIN) enumeration.