Introduction
Despite the development of targeted anticancer therapies such as Tyrosine Kinase Inhibitors (TKI) and Immune Checkpoint Inhibitors (ICI), Cytotoxic Chemotherapy Agents (CCA) remain essential agents in the neo-adjuvant, adjuvant and metastatic settings in the management of most solid tumors. Choice of monotherapy or combination chemotherapy regimens is largely based on clinical guidelines with minimal or no guidance from molecular or functional indications. This inability to inform optimal therapy in individual patients and subsequently low response rates reflect the limitations of such non-personalized therapy selection. For example, in metastatic breast cancer, first-line therapy with Standard of Care (SoC) weekly Taxol typically produces an overall response rate of ~ 30% with a further ~ 30% of patients achieving stable disease [
1], implying that 40% of patients will derive no benefit at all but will incur toxicity.
The failure of chemotherapy can be attributed to resistance of tumors (innate and acquired) towards CCA and is a significant impediment to successful management of solid tumors [
2,
3]. Resistance to CCA is random, and hence unpredictable, and becomes apparent only at response evaluation imaging or clinical assessment. This inability to detect emerging sub-clinical drug resistance in real time is the undeniable Achilles heel of purposive strategic vigilance against treatment failure.
Understanding the resistance/sensitivity profile of each patient’s case prior to initiation of treatment offers the ability to optimize treatments and time-dependent clinical outcomes not only at first but at all subsequent lines of therapy. Prior attempts at in vitro chemoresistance profiling (CRP) of tumor tissue-derived cells (TDCs) showed inadequate clinical correlation and hence is not widely adopted in clinical practice [
4,
5]. There have been prior attempts [
6‐
10] to develop real-time non-invasive means to monitor cancer sensitivity to CCA based on circulating tumor cells (CTCs), but these generally have suffered from low yields of CTC which hinders any meaningful clinical application of the concept. The scope of CTC investigations has been largely limited to enumeration for the purposes of prognostication [
11].
We have recently described a method that permits detection, enrichment and harvest of viable circulating tumor-associated cells (C-TACs: EpCAM + , Pan-CK + , CD45 ±) from the peripheral blood of patients with various solid organ cancers [
12]. We employed this approach to enrich and harvest C-TACs from 5,090 patients with prior diagnosis of either of 17 types of solid organ cancers, irrespective of treatment status and extent of disease. In a subset of 230 patients, viable TDCs were harvested from concurrently biopsied tumor tissue. In vitro response profiling of C-TACs against a panel of CCA was performed to determine concordance in response with TDCs, concordance with radiological treatment response, and to identify innate and acquired resistance in therapy naïve and pretreated cases. We present findings that establish CRP of C-TACs as an accurate and patient friendly means to non-invasively monitor resistance to CCA and guide selection of optimal treatments.
Discussion
Though it is agreed that timely identification of drug resistance is critical for optimal therapy management, there are presently no technologies or biomarkers for real time surveillance or prospectively determining drug resistance. Upfront knowledge of innate drug resistance and early detection of emerging acquired resistance thus have major clinical and financial implications, especially if such knowledge can be obtained non-invasively and in real-time. There are a few commercial assays which examine the CRP of TDC, which, however, require a substantial amount of tissue from an invasive biopsy, have extended turn-around times, and have low or no correlation with clinical outcomes [
16,
17]. CRP of TDCs is clinically unviable for two further reasons, (a) tumor evolution and heterogeneity render CRP from diagnostic biopsy rapidly obsolete with time and disease progression, and (b) repetitive invasive biopsies to obtain cells from tumor tissue are most often clinically unadvisable. Together, these factors have greatly restricted the adoption of these platforms into routine clinical practice.
It is well accepted that blood is a viable option for real-time sampling of tumor analytes. We hence describe the use of C-TACs for functional chemo-response profiling of cancers. C-TACs include CTCs (EpCAM + , PanCK + , CD45-) as well as CD45 + cells such as TAM and TAF that can be profiled to obtain clinically informative data. Prior reports have acknowledged the therapeutic potential of targeting tumor associated cells (such as TAM) owing to their role in suppressing antitumor immunity and promoting tumor progression [
18]. The negative enrichment approach we developed [
12] for harvesting of C-TACs using epigenetically activated media yielded consistently high numbers of viable C-TACs across all cancer types and permits meaningful evaluation of chemotherapy sensitivity/resistance using a broad panel of cytotoxic anticancer agents. Direct functional interrogation of viable C-TACs can provide actionable information, which is clearly more useful clinically than simple numerical or molecular correlation of such circulating malignant cells with disease status [
5,
19]. Since the presence of viable tumor cells in peripheral blood has been causatively linked to metastasis, understanding their drug sensitivities may aid detection of emergent chemotherapy-resistant clonal sub-populations [
20,
21].
The study findings demonstrate a robust correlation between CRP of C-TACs and TDCs implying that C-TACs accurately represent and report the chemotherapy sensitivity characteristics of the tumor from which they derive in the vast majority of cases.
In the arm of therapy naïve patients, C-TACs displayed widely variable innate resistance consistent with the clinical setting where lack of response to first line CCA is commonly encountered in multiple cancer types. For example, in metastatic breast cancer, response rates to first-line Taxol or Capecitabine are typically around 30% with stable disease achieved in a further 30%. With sub-optimal Pathological Complete Response (pCR) rates in the neo-adjuvant setting [
22,
23]. Similar treatment failures have been reported following resistance to 5-Fluorouracil combination regimens (FOLFOX, FOLFIRI) have been reported in Colorectal cancers [
24]. Likewise, resistance to first line platinum regimens are encountered in cancers of the Head and Neck, Oesophagus, Stomach, Colorectum, Ovary, Breast, Lung and Gallbladder [
25]. Detection of chemo-resistant C-TACs in therapy naïve patient samples can be predictive of sub-optimal response as well as eventual disease progression, which is clearly advantageous prior to initiation of treatment. Similarly, CRP of C-TACs from previously treated patients detected higher chemoresistance in a majority of samples indicating acquired resistance following failure of/exposure to prior therapies. The ability to detect emergent (acquired) resistance indicates high accuracy for longitudinal monitoring where it would be possible to identify such ‘resistance-educated’ C-TACs. Additionally, it is also possible to identify agents from prior regimens that may be used to re-challenge the tumor in a subsequent line of therapy.
The clinical utility of CRP of CTACs was investigated in the real-world scenario by assessing concordance between in vitro findings and objective (radiological) measurement of treatment response. Patients who were therapy naïve at initial CRP were followed-up to determine response to first line treatments. Within this subgroup, we observed ~ 97% concordance between CR or PR and in vitro sensitivity of C-TACs to CCA. On the other hand, a lower in vitro sensitivity was associated with lower chance of radiological response to treatment. In the first line setting, in vitro sensitivity in CRP was thus more predictive of response to therapy. In the second subset-arm, we evaluated CRP in patients who were already receiving CCA prior to a follow-up radiological scan. Among the patients with radiologically evident PD, we observed ~ 87% concordance between treatment response/resistance and in vitro sensitivity/resistance of C-TACs to CCA. CRP of C-TACs can non-invasively determine failed treatments with high accuracy and can be used for longitudinal monitoring of patients. In this pretreated population, chemoresistant C-TACs were observed even in patients with radiologically evident partial response (PR) to treatment. Since PR, by definition, indicates slower or no response to treatment in a proportion of the tumor, it is likely that the resistant C-TACs emerged from the non- or weakly responding tumors, indicating the presence of a surviving resistant tumor cell population which could pose a risk of treatment failure and resurgence.
Significant inter-patient variability within all cancer types was observed which indicated the need and potential value of this approach prior to any line of therapy including neo-adjuvant. CRP can avoid several pitfalls of present treatment structures, especially in pretreated patients, following failure of multiple lines of multi-drug regimens by identifying and eliminating potentially sub-optimal drugs and reduce the risk of unnecessary toxicity arising from sub-optimal agents. In vitro chemotherapy sensitivity/resistance profiling of C-TACs is a non-invasive, uncomplicated, cost-effective process to determine cancer cell sensitivity to CCA in real time. CRP can be performed not only at diagnosis (prior to first line therapy selection), but also routinely during ongoing cancer treatment to achieve a previously unattainable level of synchronicity, precision and personalization. Therapy selection based on CRP of C-TACs may not only be able to reduce the risk of progression or recurrence due to treatment failures, but also the expenses of sub-optimal treatments as well as the accumulated drug toxicities associated with failed treatments. The ability to obtain treatment insight in real time and non-invasively has profound clinical significance. This approach is not only mature for adoption in clinical practice but also for improving efficiency of clinical trials aimed at expanding the scope of approved CCAs for use in additional cancers apart from those that are included in the labelled indication.
Acknowledgements
The authors are grateful towards all study participants and their caregivers. The contributions of Nueclear Healthcare Limited, HCG Cancer Centres, Avinash Cancer Centre, Apex Wellness, Chandak Hospital, all staff and Scientists of study sponsor (DCG) towards managing various clinical, operational and laboratory aspects of the study are also acknowledged with gratitude.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.