Patient-Derived Organoids (PDOs) represent a promising technology for therapy prediction in pancreatic cancer, with the potential of enhancing treatment outcomes and allowing more effective, personalized treatment choices. However, classification approaches into sensitive and resistant models remain very variable and are based on single-agent testing only, neglecting synergistic effects of multi-drug combinations. Here, we established 13 PDOs and performed both combinatorial and single-agent drug testing. By comparing different clustering approaches of drug-response metrics and establishing a new classification approach based on pharmacokinetic modelling, we were able to evaluate which score best predicts the clinical response of patients. Our newly developed score considered the Area Under The Curve (AUC) of cell viability curves and reached a prediction accuracy of 85%. Our data supports previous findings for PDOs to constitute an effective platform for translational drug testing. Furthermore, our results suggest that the AUC is a more accurate drug-response metric than the half maximal inhibitory concentration (IC50), and that combinatorial drug testing yields a higher accuracy than single-agent testing. Based on our results, future PDO studies and PDO-based clinical trials should integrate combinatorial drug testing when performing pharmacotyping as it represents a more realistic translational approach towards drug testing and enhances the prediction accuracy of in-vitro drug testing. The methodology and outcomes presented in this study are of critical relevance for future PDO-based translational trials as they allow a new physiology-based approach towards multi-drug testing and classification of organoid response, which improves PDO prediction accuracy.
Research Article
Multi-Drug Pharmacotyping improves Therapy Prediction in Pancreatic Cancer Organoids
https://doi.org/10.21203/rs.3.rs-5304150/v1
This work is licensed under a CC BY 4.0 License
You are reading this latest preprint version
Pancreatic Adenocarcinoma
Patient-derived Organoids
Multi-Drug Response Metrics
Pharmacokinetic Modelling
Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest types of cancer, with a 5-year survival rate of only 11% [1]. Moreover, PDAC is expected to be the second most common cause of cancer-related death in the world by 2030 [1–3]. The poor prognosis of PDAC can be attributed to unspecific symptoms leading to late diagnosis as well as early metastasis and high resistance towards chemotherapy agents. Surgical resection represents the only curative treatment option at a local disease stage, which is possible for 20% of all patients. Once resected, the 5-year survival rate is reported to be 20% [4, 5], while 80% of patients relapse after surgery and adjuvant chemotherapy. The majority of patients are diagnosed at a more advanced disease stage. For patients with borderline resectable or locally advanced PDAC, neoadjuvant chemotherapy can achieve secondary resectability with a chance of 40% [6]. Patients with metastasized or inoperable disease are treated with palliative chemotherapy. Chemotherapy is therefore an integral component of the treatment of all stages of pancreatic cancer as it is used both perioperatively and in a palliative setting. Unfortunately, most patients do not show long-term response to chemotherapy which underlines the high levels of chemotherapeutic resistance and high failure rates in the multi-modal treatment for PDAC [7].
While individualized therapies have emerged in lung- or colorectal cancer [8–10], the driver mutations in PDAC, including KRAS (apart from rare KRASG12C), CDKN2A, TP53 and SMAD4 cannot be therapeutically targeted at the moment [11, 12]. At this point, conventional chemotherapy remains the treatment option in clinical practice. Three different chemotherapeutic regimens are currently used: mFOLFIRINOX (Folic acid, 5-Fluorouracil, Irinotecan and Oxaliplatin) [13], Gemcitabine combined with nab-Paclitaxel [14] and Gemcitabine monotherapy [15]. So far, there a no clinically established biomarkers to predict which patient will respond to which chemotherapy and treatment choices are mostly based on patients’ age and performance status [16]. To predict patients’ response and enable more informed treatment choices, the field of Patient Derived Organoids (PDOs) has gained attention within the last years. PDOs are three-dimensional in-vitro models of epithelial cells grown in an extracellular matrix. They can effectively recapitulate the tumor biology and have become a versatile tool for studying PDAC [17–19]. Importantly, growing evidence suggests a relevant predictive power of PDOs for clinical patients' response [20–24]. However, although individual groups are achieving promising results, there is a lack of standardized methods for the evaluation of drug response in PDOs, resulting in highly variable approaches towards defining drug response. Furthermore, despite patients being administered several drugs at the same time with the drugs giving rise to important and synergistic effects, previous studies only assessed single agent testing on PDO models [25–29]. Since PDO models have great potential to act as dynamic predictive biomarkers, these models are implemented in ongoing and future clinical trials [30]. For future clinical trials to be conducted, however, a reliable way of classifying models into resistant and sensitive should be standardized across the field of PDO-based translational research. The variability of response classification as well as the lack of multi-drug testing in PDAC PDOs represent two major limitations on the way to implementing PDOs in clinical trials.
In the work presented, these limitations were addressed. We evaluate different scores to predict patients’ response and integrate testing chemotherapeutics both as single agents and in combination, in order to more closely replicate the clinical conditions. Moreover, we established statistical clustering and pharmacokinetic models to compare and challenge current definitions of in-vitro PDO response. PDOs were successfully pharmacotyped towards single chemotherapeutic agents as well as the combinatorial therapy of mFOLFIRINOX and gemcitabine/paclitaxel and results of in-vitro drug testing were compared to the clinical response of patients in 13 cases. Our results suggest that multi-agent testing based on physiological pharmacokinetic models is feasible and achieves higher accuracies than single-agent testing, making it the ideal method for future PDO-based, multi-agent clinical trials.
2.1 Patient Selection and Ethics Statement
For this study, tumor samples were obtained from treatment-naïve patients suffering from PDAC and were consequently treated at the Charité-Universitätsmedizin Berlin and at the Waldfriede Hospital Berlin-Zehlendorf between November 2021 and May 2023. The tissue extraction as well as all further steps were performed in accordance with the ethics approval (EA1/157/21) and signed consent was obtained from all patients. Samples were only included if PDAC was histopathologically confirmed and clinical follow-up was known (see Fig. 1A). Patients had to receive at least two cycles of chemotherapy to be included in this study. Maximum follow-up for our patient cohort was 18 months.
2.2 Assessment of Treatment Efficacy
For resected patients, therapy response was clinically evaluated based on radiological imaging after 6 months. Patients were classified as resistant when suffering from a relapse within 6 months after the start of adjuvant chemotherapy. If no relapse occurred, they were classified as sensitive. Treatment response of metastatic patients was evaluated based on RECIST 1.1 criteria [31] of the first CT-staging after a minimum of three months of therapy. Metastatic PDACs were classified as sensitive when partial response or stable disease was detected. When tumors progressed under therapy, they were classified as resistant.
2.3 Isolation and Establishment of Patient-Derived Organoids
Organoids were established from the primary tumor via surgical specimens and from metastatic sides via liver biopsies. Samples were manually cut into small pieces of about 1mm and digested at 37°C in a solution containing 100 µg/ml DNAse I (VWR, Radnor, PA, USA), 100µg/ml Dispase (STEMCELL Technologies, Vancouver, Canada), 125µg/ml Collagenase II (Sigma-Aldrich, Merck, Darmstadt, Germany), 1:2000 Rock-Inhibitor (Abmole Bioscience, Houston, TX, USA) and 1:200 Amphotericin B (Sigma-Aldrich, Merck, Darmstadt, Germany). Biopsy specimens were enzymatically dissociated for 5 to 30 minutes, while tumor specimens were incubated for 2 to 3 hours. After filtering the cells using a sterile 100µm filter and performing red blood cell lysis (Miltenyi Biotec, Bergisch-Gladbach, Germany), cells were plated in Cultrex Reduced Growth Factor BME, Type 2 (R&D Systems, Minneapolis, MN, USA). Domes were incubated for 30 minutes at 37°C to solidify. Afterwards, culture medium as established by L Broutier, A Andersson-Rolf, CJ Hindley, SF Boj, H Clevers, BK Koo and M Huch [32] was added. To prevent contamination, Amphotericin B was added at a dilution of 1:200 for the first week and Mycoplasma tests were regularly performed using the Mycoplasma detection kit (Applied biological Materials, Richmond, Canada). Organoids were split mechanically and enzymatically using TrypLE Express (Thermo Fisher Scientific, Waltham, MA, USA) and plated in Cultrex at a ratio of 1:2.
2.4 Histology and immunohistochemistry
Cultrex was dissolved by adding 500 µl PFA (4% in PBS) to PDOs at 4°C for one hour. PDOs were detached from the plate, washed with PBS and resuspended in histogel (Thermo Fisher Scientific, Waltham, MA, USA). Further steps were performed at the Charité Institute of Pathology. After paraffin-embedding, blocks were cut into 3 µm sections. H&E-staining was carried out with the Tissue Tek Prisma Plus Automated Slide Stainer (SAKURA). Immunohistochemical staining was carried out with the BenchMark XT immunostainer (Ventana Medical Systems, Tucson, AZ). The following antibodies were used: anti-Ki-67 (MIB-1, Dako, 1:50), anti-Vimentin (V9, Dako, 1:5000), anti-CA 19 − 9 (1116-NS-19-9, Dako, 1:500), anti-CK 19 (RCK108, BioGenex, 1:100), anti-SMAD4 (EP618Y, abcam, 1:200), anti-p53 antibody (DO-7, Dako, 1:50), and anti-GATA6 antibody (Q92908, R&D Systems, 1:100). Slides were digitalized and analyzed by a reference pathologist.
2.5 PDO Pharmacotyping
PDO drug response was tested between passages 5 and 15. PDOs were collected and digested to a single-cell suspension by TrypLE (Thermo Fisher Scientific, Waltham, MA, USA). Cells were counted, the proportion of single cells and cell viability was measured using Acridine Orange/Propium Iodide stain (BioCat, Heidelberg, Germany) with the LUNA automated cell counter (BioCat, Heidelberg, Germany). For cell viability assays to be conducted, the proportion of single cells needed to be at least 90%. Having assured to be within the linear range of the CellTiter-Glo® Luminescent Cell Viability Assay (Promega, Walldorf, Germany), 2,000 cells were plated into each well of a 96 Well-plate. 200 µl medium containing 10µM Rock Inhibitor were added after solidification of the domes. PDOs were fully formed by day 3 after plating. After 4 days of incubation, 100 µl of medium were discarded and drugs were added in seven concentrations. For single agents, a 10-fold dilution series was used. The maximum concentrations were 1 mM for 5-FU, 100µM for Oxaliplatin, 50 µM for Gemcitabine and 10 µM for SN-38 and Paclitaxel. Combinatorial therapies were added in a 4-fold dilution series, ranging from 64 x cmax/tissue to 1/64 x cmax/tissue, where cmax/tissue was defined as the pharmacokinetically calculated tissue concentration of each drug (see section 2.6). All cytostatics were purchased from Selleckchem and calcium folinate was purchased from Biozol Diagnostica. Cell viability was measured using the CellTiter-Glo® Luminescent Cell Viability Assay after 3 days of treatment. Outliers in luminescence measurements were identified using the ROUT method in GraphPad Prism (Version 10.2.3). Values were excluded when the Q-value exceeded 0.01. AUC, IC50 and log[IC50] were subsequently calculated in GraphPad Prism based on normalized cell viability values. Dose-response curves for single agents and multi-drug testing and violin plots of AUC- and log[IC50]-values were created using GraphPad Prism.
2.6 Pharmacokinetic Modelling
2.6.1 Identification of Tumor’s Maximum Tissue Concentrations, cmax/tissue
To estimate tissue concentrations of chemotherapeutic agents, mathematical pharmacokinetic models were implemented based on pharmacokinetic values previously reported in the literature [33–53]. In these simulations, the dose and infusion time of each chemotherapeutic agent, the rate of distribution into the tissue as well as the rate of distribution back into the bloodstream were considered. Furthermore, metabolization and elimination of the chemotherapeutic agent were taken into account. Pharmacokinetic modelling was based on one-, two- and three-compartment modelling as described by HA Schiffter-Weinle [54]. Within these models, the body was viewed as a system of different compartments, with a distinction made between the central and the peripheral compartments. When modelling, chemotherapeutic agents were infused into the central compartment (blood) and distributed to the peripheral compartments (tissue) by linear and/or saturable kinetics. Saturation kinetics were used in transport processes that were reported to be dependent on enzymes or proteins. In these cases Michaelis-Menten kinetics were applied. The number of compartments used in each model depended on the drug’s distribution patterns previously reported in the literature [33–53].
One-compartment models were used for drugs that distributed in negligible time within the entire body and were known to have similar kinetics in all body areas (blood and tissue). In contrast, models with two or more compartments were implemented for drugs that were previously reported to distribute heterogeneously within the body. The blood stream always represented the central compartment. Whenever more than one peripheral compartment was employed, the first peripheral compartment was defined to be the one within which the drug was distributed most rapidly (e.g. GI-tract). The second peripheral compartment represented tissues with a lower transfer rate (e.g. fatty tissue). The intercompartmental clearance rates described the speed of drug distribution and the elimination rate described the speed of drug excretion. In the case of Michaelis-Menten kinetics, this depended on the maximum transport velocity and the Michaelis-Menten constant. Tissue concentrations were deduced from the peripheral compartment. When using three-compartment models, the concentration depicting the tumor concentration was calculated in the first peripheral compartment. Extracting tissue concentrations from the first peripheral compartment was based on blood flow rates measured previously in CT-perfusion studies of PDAC patients [55]. To identify diffusion rates and the number of compartments of each chemotherapeutic agent, a systematic literature review was conducted on PubMed. Taking into account that PDAC occurs mainly in elderly patients [56], studies with pediatric patient populations were excluded. All calculations were done in MatLab/Simulink (R2022b).
For one compartment models, the following equation was used to calculate the change in the amount of drug in the body over time, where x1 [mg] represented the amount of drug in the central compartment, D [mg] the dose of the drug, T [h] the infusion time and e [mg] the amount of drug excreted (see Supplementary Table 1).
$$\:\frac{d{x}_{1}}{dt}=\frac{D}{T}-\frac{de}{dt}$$
Excretion of the drug was described by linear kinetics, where Cl [l/h] represented the clearance rate and V1 [l] the volume of distribution of the central compartment.
$$\:\frac{de}{dt}=\frac{Cl}{{V}_{1}}\bullet\:{x}_{1}$$
For two-compartment models, the change in the amount of drug in the central compartment over time was described as
$$\:\frac{d{x}_{1}}{dt}=\frac{D}{T}-\frac{{Q}_{1}}{{V}_{1}}\cdot\:{x}_{1}+\frac{{Q}_{1}}{{V}_{2}}\cdot\:{x}_{2}-\frac{de}{dt}$$
with Q1 [l/h] being the intercompartmental clearance rate between central and peripheral compartment, x2 [mg] the amount of drug in the peripheral compartment and V2 [l] the peripheral volume of distribution.
Elimination was either linear as described above or saturable following Michaelis-Menten kinetics, where vmax [mg/h] equals the maximum speed of excretion in and Km [mg/l] is the Michaelis-Menten constant.
$$\:\frac{de}{dt}=\frac{{v}_{\text{m}\text{a}\text{x}}\cdot\:{c}_{1}}{{K}_{m}+{c}_{1}}$$
The change in the amount of drug in the peripheral compartment, x2 [mg], over time was calculated as follows:
$$\:\frac{d{x}_{2}}{dt}=\frac{{Q}_{1}}{{V}_{1}}\cdot\:{x}_{1}-\frac{{Q}_{1}}{{V}_{2}}\cdot\:{x}_{2}$$
For three-compartment models, the same equations applied for both the elimination of the drug and the amount of drug in the first peripheral compartment. The change in the amount of drug in the central compartment over time was calculated as follows:
$$\:\frac{d{x}_{1}}{dt}=\frac{D}{T}-\frac{{Q}_{1}}{{V}_{1}}\cdot\:{x}_{1}+\frac{{Q}_{1}}{{V}_{2}}\cdot\:{x}_{2}-\frac{{Q}_{2}}{{V}_{1}}\cdot\:{x}_{1}+\frac{{Q}_{2}}{{V}_{3}}\cdot\:{x}_{3}-\frac{de}{dt}$$
where Q2 [l/h] described the intercompartmental clearance rate between central and second peripheral compartment and V3 [l] described the volume of distribution of the second peripheral compartment
2.7. Response classification
Cell viabiliy curves obtained from single agent and multi-drug drug testing as described in section 2.5 were further analyzed with respect to the AUC, IC50 and log[IC50] using statistical clustering methods and pharmacokinetic classifications (see Fig. 1B). These analyses were conducted in Microsoft Excel and R/R-Studio (Version 4.3.2).
2.7.1. Classification of Response using Single Agent Testing
2.7.1.1. Prediction Accuracy of using One Single Agent
First, the prediction accuracy was assessed using single agent testing. This way the clinical response of patients who had received multiple chemotherapeutics in the clinic (mFOLFIRINOX or gemcitabine/paclitaxel), was linked to results of one single agent. AUC- and log[IC50]-values of single agent testing were grouped by Jenks Natural Breaks (JNB) classification into high, intermediate and low responders and were then compared to the clinical response of patients.
2.7.1.2. Summation Score
For the in-vitro results to be matched with the multi-drug treatment of patients in the clinic, in-vitro results of single agent tests had to be combined. This was implemented by adding normalized AUC and normalized log[IC50]-values from single drug testing. AUC-values were normalized with respect to the total area accounting for 100% of untreated live cells at the highest concentration examined (see Fig. 1C). For mFOLFIRINOX, the normalized AUC (AUCnorm) of oxaliplatin, 5-FU and SN-38 were added. For gemcitabine/paclitaxel, the AUCnorm of gemcitabine and paclitaxel were added. log[IC50]-values were normalized with respect to the maximum concentration used for each drug. The same summation score was calculated for normalized log[IC50]-values (log[IC50]norm) for both regimens (see Table 1). JNB classification was used to group PDOs into sensitive, intermediate and resistant based on the sums of AUCnorm- and log[IC50]norm-values described above.
| Summation Score | Calculation |
|---|---|
| AUCnorm(mFOLFIRINOX) | AUCnorm(oxaliplatin) + AUCnorm(5-FU) + AUCnorm(SN-38) |
| AUCnorm(Gem/Pac) | AUCnorm(gemcitabine) + AUCnorm(paclitaxel) |
| Log[IC50]norm(mFOLFIRINOX) | Log[IC50]norm(oxaliplatin) + Log[IC50]norm(5-FU) + Log[IC50]norm(SN-38) |
| Log[IC50]norm(Gem/Pac) | Log[IC50]norm(gemcitabine) + Log[IC50]norm(paclitaxel) |
2.7.2. Direct Clustering using Multi-Drug Testing
In multi-drug tests, PDOs are directly exposed to combinations of chemotherapeutics. Thus, no scoring system as for adding single-agent testing was required. For mFOLFIRINOX, PDOs were treated with folic acid, 5-FU, irinotecan and oxaliplatin at the same time. For gemcitabine/paclitaxel, PDOs were simulatenously treated with gemcitabine and paclitaxel. JNB classification was used to classify PDOs as sensitive, intermediate and resistant based on AUC- and log[IC50]-values for mFOLFIRINOX and gemcitabine/paclitaxel.
2.7.3. Comparison to Established Statistical Score
For a direct comparison to be made to previously published scores, the data collected in this work was directly compared to the score first published by Beutel et al. [15]. In short, JNB classification was used to classify AUC-values of single agents into low, intermediate and high responders. Low-responders were assigned a numerical value of 3, intermediate-responders a value of 2 and high-responders were assigned a value of 1. When patients were treated with multiple chemotherapeutics at the same time (mFOLFIRINOX or gemcitabine/paclitaxel) the mean of the added values was calculated. PDOs were classified as resistant when the mean value was greater than 2.
2.7.4 Pharmacokinetic Classification
In a pharmacokinetic approach, IC50-values of mFOLFIRINOX, gemcitabine/paclitaxel and gemcitabine were compared to the maximum tissue concentrations (cmax/tissue) of all drugs used in the individual regimens calculated in section 2.6.2. PDOs were classified as sensitive if the IC50 was less than cmax/tissue and as resistant if it was higher than cmax/tissue.
2.8 Comparison to Clinical response
All scores were compared to patients’ clinical response and the accuracy for each score was calculated seperately for first- and second-line therapies. When PDOs were defined to be sensitive or intermediate to a chemotherapeutic drug, the in-vitro result was classified as sensitive.
3.1. Successful establishment of 13 therapy-naïve PDO models
Thirteen PDO models were established with an efficacy of 50%. Samples were taken from treatment-naïve PDAC patients. All patients were treated and clinically followed up at between November 2021 and May 2024 at the University Hospital Charité – Universitätsmedizin Berlin and the Hospital Waldfriede Berlin-Zehlendorf.
3.2 Histopathological assessment confirms malignancy in all PDO models
Malignancy was confirmed by histopathological evaluation based on morphology and expression of immunohistochemical markers. Both organoid morphology in H&E staining as well as immunohistochemical stainings (CA19-9, CDX-2, CK19, GATA6, ki-67, p53, SMAD4, Vimentin) showed high similarity of PDO models and corresponding FFPE tissue (see Fig. 2B and 2C). P53 was mutated in 8 of 13 models and SMAD4 was mutated in 3 of 13 models. Ki-67 expression ranged between 25% and 95% with a mean Ki-67-expression of 68%. All cultures were positive for CDX-2 and CK19, 11 of 13 cultures expressed CA19-9. Vimentin was focal positive in 5 cultures (see Fig. 2A).
3.2. Characteristics of patients included
Eight of thirteen patients included in this study were diagnosed at a resectable stage (see Table 2). All eight samples were obtained from surgical resection. R0-resection was achieved in six out of eight patients (6/8), while two patients (2/8) had an R1-resection status. Five samples were obtained at a metastasized stage. The samples collected from metastasized patients were established from liver metastases (4/5) and from the primary tumor (1/5). Samples were collected during palliative surgeries or punch biopsies. The mean age of the patient cohort was 68 years (+/- 10 years), similarly aligned in resected and metastatic groups. On average, 5.5 weeks passed between sample collection and therapy initiation. Therapy started earlier in metastasized patients than in patients who underwent surgery with curative intention (2.8 vs. 7.1 weeks). Seven patients received mFOLFIRINOX as a first-line therapy. Two patients received gemcitabine/nab-paclitaxel and four patients were treated with gemcitabine monotherapy. Six patients showed initial response to mFOLFIRINOX based on the first CT-scan available after the start of therapy. One patient with R0-resection did not respond to mFOLFIRINOX and suffered a relapse five months after surgery. A therapy response was confirmed in four patients treated with gemcitabine as first-line therapy and in two patients who were treated with gemcitabine/nab-paclitaxel. One patient responded to gemcitabine monotherapy as second-line therapy after being treated with mFOLFIRINOX. In another case, the therapy was de-escalated to gemcitabine monotherapy from gemcitabine/nab-paclitaxel due to side effects. The patient had responded to gemcitabine/nab-paclitaxel but showed progression upon treatment with gemcitabine monotherapy. No differences in clinical response were observed between patients according to their TNM-status.
Table 2: Characteristics of patients included for organoid establishment
|
|
Resectable disease (n=8) |
Metastatic disease (n=5) |
|
Age at diagnosis (mean) |
44-84 (65) |
63-79 (70) |
|
Sex Male Female |
3 5 |
2 3 |
|
Tumor specimen Punch biopsy Surgical resection |
0 8 |
2 3 |
|
R-status R0-resection R1-resection |
6 2 |
- - |
|
Sample derived from Primary tumor Metastasis |
8 - |
1 5 |
|
Weeks until therapy initation (mean) |
1-4 (2.8) |
4-12 (7.1) |
|
First-line therapy mFOLFIRINOX Gemcitabine/nab-paclitaxel Gemcitabine monotherapy |
5 0 3 |
2 2 1 |
|
Response to first-line therapy mFOLFIRINOX Gemcitabine/nab-paclitaxel Gemcitabine monotherapy |
4 - 3 |
2 2 1 |
|
Second-line therapy Gemcitabine monotherapy |
1 |
1 |
|
Response to second-line therapy Gemcitabine monotherapy |
1 |
0 |
3.3. Maximum tissue concentrations (cmax/tissue) for chemotherapeutic agents can be determined using pharmacokinetic compartment modelling
Based on previously reported pharmacokinetic compartment models (see Fig. 3A, Fig. 3B and Supplementary Table 2), maximum tissue concentrations (cmax/tissue) of oxaliplatin, SN-38, 5-FU, folic acid, gemcitabine and paclitaxel were simulated (see Table 3). We identified three publications for pharmacokinetic modelling for 5-FU [42–44] all of which reported two-compartment pharmacokinetics with linear distribution [57]. 5-FU is metabolized by the enzyme dihydropyrimidin-dehydrogenase (DPD) and subsequently excreted by renal filtration [58]. Representing the enzymatic metabolization of 5-FU, all publications employed Michaelis-Menten kinetics (see Fig. 3C).
For folic acid, no parameters for multi-compartment models had previously been reported. The drug is known to distribute rapidly throughout the whole body and is therefore best modelled by a one-compartment model. Since folic acid is rapidly excreted renally, linear elimination kinetics were implemented [59]. Folic acid is known to have two isomers which are administered to patients as a racemate. Only the L-isomer has been reported to have anti-cancer effects. Pharmacokinetic modelling of cmax/tissue for folic acid, thus, only took into account the L-isomer concentrations (see Fig. 3D) [45–47, 49, 60].
For oxaliplatin, four studies were included for the pharmacokinetic simulation of tissue concentrations [38–41]. Two-compartment modeling was reported as the most accurate in two publications [38, 39] and two other studies used three-compartment modeling [40, 41]. We applied both two- and three compartment models to take all possible ways of the drug’s distribution into account (see Fig. 3E). Linear elimination was the most accurate for oxaliplatin known to be mainly excreted renally [61].
Irinotecan is known to be metabolized to the active metabolite SN-38 which shows an increased anti-tumoral activity by a 100 to 1000-fold factor [62]. Therefore, SN-38 was used in all experiments and its tissue concentration was calculated. Irinotecan is enzymatically converted to SN-38 by liver carboxylesterases with a conversion rate between 12.0% and 41.2% [37, 63]. In clinically relevant doses, no saturation of this enzymatic system has been observed and the metabolization was described by linear kinetics [64]. We modeled irinotecan distribution using three-and SN-38 two-compartments with linear excretion for both irinotecan and SN-38 (see Fig. 3F) [34–37].
Gemcitabine is reported to show linear distribution. The drug is rapidly deaminated by cytidine deaminase and subsequently excreted, which leads to a short plasma half-life [65]. Gemcitabine excretion occurs in two phases with an initial rapid elimination phase followed by a long terminal elimination [65]. Since these characteristics are typical for drugs with two-compartmental pharmacokinetics, we used a two-compartment model with linear distribution and excretion (see Fig. 3G) [50, 51].
Since paclitaxel is known to be the active metabolite of nab-paclitaxel, paclitaxel has been used in previously reported PDO experiments [20, 21, 66]. Paclitaxel binds to plasma proteins and blood cells in a saturable manner best described by Michaelis-Menten-kinetics. When unbound, paclitaxel can diffuse into the tissue linearly [52, 53]. Three-compartment-models have been reported to be the most accurate for paclitaxel in several pharmacokinetic studies [52, 53, 67–69]. We identified two different publications for calculating paclitaxel pharmacokinetics, both studies employed linear kinetics for paclitaxel distribution and elimination (see Fig. 3H) [52, 53].
|
Drug |
Range of cmax/tissue |
Mean of cmax/tissue |
|---|---|---|
|
5-Fluorouracil [42–44] |
2.23–2.99 µM |
2.64 µM |
|
Folic acid [45–49] |
16.83–31.55 µM |
23.9 µM |
|
Oxaliplatin [38–41] |
0.52–1.1 µM |
0.79 µM |
|
SN-38 [34–37] |
5.13–14.06 nM |
9.42 nM |
|
Gemcitabine [50, 51] |
1.37–2.54 µM |
1.95 µM |
|
Paclitaxel [52, 53] |
48.39–49.35 nM |
48.87 nM |
| cmax/tissue – maximum tissue concentration | ||
3.4. Statistical Clustering approaches reach high prediction accuracies, with Direct Clustering based on multi-drug testing performing best
In a first step, clinical response was linked to the in-vitro results of single-agent testing (see Fig. 4B-F, Fig. 5A and Table 4). Using AUC-based scores, PDO response towards oxaliplatin matched in five of seven cases (71%) and SN-38 in four of seven cases (57%). 5-FU had the highest accuracy for six out of seven patients (86%). When using log[IC50]-based scores, oxaliplatin and SN-38 reached higher prediction accuracy (71%) than 5-FU (57%). For gemcitabine/nab-paclitaxel, gemcitabine pharmacotyping correctly classified response using both AUC and log[IC50] in two of two patients (100%), while paclitaxel pharmacotyping matched with the clinical response in one of two patients (50%) for both AUC- and log[IC50]-values.
When using the summation score for mFOLFIRINOX therapy prediction, clinical response was correctly predicted in five of seven patients (71%) for both AUC- and log[IC50]-values (see Table 4). The summation score for gemcitabine/paclitaxel matched with the clinical response of patients receiving gemcitabine/nab-paclitaxel in one of two patients (50%). This was identical for both AUC- and log[IC50]-based scores. Overall, response was correctly classified in ten of thirteen cases for both AUC based-scores and log[IC50]-based scores, which corresponds to an accuracy of 77% for first-line therapies (see Fig. 4B-H, Fig. 5B, Table 5 and Supplementary Fig. 1).
|
Drugs |
Accuracy AUC |
Accuracy log[IC50] |
|
|---|---|---|---|
|
One agent prediction |
5-FU |
6 of 7 (86%) |
4 of 7 (57%) |
|
Oxaliplatin |
5 of 7 (71%) |
5 of 7 (71%) |
|
|
SN-38 |
4 of 7 (57%) |
5 of 7 (71%) |
|
|
Gemcitabine |
2 of 2 (100%) |
2 of 2 (100%) |
|
|
Paclitaxel |
1 of 2 (50%) |
1 of 2 (50%) |
|
|
Summation Score |
mFOLFIRINOX |
5 of 7 (71%) |
5 of 7 (71%) |
|
Gem/Pac |
1 of 2 (50%) |
1 of 2 (50%) |
Table 5: Overview of predictive accuracy for all statistical scores
When directly clustering AUC-values of multi-drug testing into sensitive, intermediary and resistant, response to first-line therapies was correctly predicted in eleven of thirteen cases, which corresponds to an accuracy of 85%, and in one of two cases for second-line therapies (see Table 5). Based on log[IC50]-values, response to first-line therapies was correctly predicted in eight of thirteen patients which corresponds to an accuracy of 62% (see Supplementary Fig. 1). Thus, the direct clustering approach using AUC values of multi-drug testing showed the highest accuracy (see Fig. 4E, G, H and Fig. 5C).
To compare the results of our newly developed scores to established classifications, we calculated an AUC-based score as published by Beutel et al. [20]. This score correctly predicted response in ten of thirteen patients for first-line therapy, corresponding to an accuracy of 77%, and in one of two patients for second-line therapies. This previously established AUC-based classification proved to have the same prediction accuracy as the summation scores described in section 3.2.3.2 (see Fig. 5D and Table 5).
3.5. The Pharmacokinetic Classification approach demonstrates comparable accuracy to statistical clustering methods
When using the pharmacokinetic classification based on cmax/tissue-values, response towards first-line therapies was correctly classified for ten of thirteen patients. This corresponds to an accuracy of 77% (see Fig. 5E). For second line therapies, the clinical response could not be correctly classified for any patient.
In this study, the potential for therapy prediction of PDOs in pancreatic cancer research was confirmed and the results were used to determine the most ideal way of comparing in-vitro drug response to patients’ clinical response. In contrast to exposing PDO models to single drugs, a more realistic approach is thought to expose PDOs to all drugs of each therapeutic regimen at the same time, i.e. multi-drug testing. Although this takes into account synergistic effects of each drug [70–72], very few studies actually implemented multi-drug testing [73–76]. In previously published studies, results of multi-drug testing were correlated to the patient’s clinical response of one patient only [76]. In this work, a methodology for performing combinatorial drug-tests for both mFOLFIRINOX and gemcitabine/nab-paclitaxel was introduced. In order to assess the potential of multi-drug testing for in-vivo therapy prediction, two different classification methods for response were implemented (direct clustering vs. pharmacokinetic). The AUC-based direct clustering performed best, reaching an accuracy of 85% for first-line therapies, while the pharmacokinetic classification method had a prediction accuracy of 77%. The lower prediction accuracy of the pharmacokinetic classification is taking into account log[IC50] values, which have previously been reported to vary significantly, also being easily affected by curve-fitting. Therefore, the AUC as the integral area of dose response curves is thought to be a more robust value and led to higher prediction accuracies when applied on multi-drug testing [77].
In the clinic, most PDAC patients are treated with combination therapies, either mFOLFIRINOX or gemcitabine/nab-paclitaxel, whereas gemcitabine monotherapy is only recommended for patients with reduced performance status [78]. For both multi-drug regimes, mechanisms of synergy have been identified in-vitro. Inhibitors of topoisomerase-1, including irinotecan, increase the effect of DNA-damaging agents such as oxaliplatin by decelerating the repair of DNA-crosslinks [26]. Folic acid increases the intracellular amount of folates, which leads to an enhanced inhibition of thymidylate synthase by 5-FU [79]. Interestingly, A Di Paolo, P Orlandi, T Di Desidero, R Danesi and G Bocci [80] demonstrated that the enhanced anti-tumor effect of 5-FU combined with folic acid depends on the time of administration and is only present when administering both drugs simultaneously. Paclitaxel has been shown to increase the concentration of gemcitabine in the tumor tissue. This is due to decreased levels of the enzyme cytidine deaminase, which is mainly responsible for eliminating gemcitabine [27]. Beyond in-vitro studies, the synergistic effects of 5-FU and oxaliplatin as well as 5-FU and folic acid were validated in clinical trials [81, 82]. The combination of 5-FU and oxaliplatin showed a significant benefit in patients with advanced PDAC [82]. In our PDO cohort, mFOLFIRINOX was consistently potent across all models, whereas the effectiveness of gemcitabine/paclitaxel was found to be more variable. Considering the evidence of synergistic drug effects both in-vivo and in-vitro as well as the results of this work, multi-drug tests are thought to be critical in future PDO-based translational trials. This is underlined by the high prediction accuracy of multi-drug testing in the study presented, emphasizing synergic effects in organoid-based drug testing. For multi-drug testing, however, the ratio of the drugs needs to be defined. Most researchers previously used IC25-, IC50- or cmax/plasma to determine drug ratios [73, 74, 76], whereas E Hadj Bachir, C Poiraud, S Paget, N Stoup, S El Moghrabi, B Duchene, N Jouy, A Bongiovanni, M Tardivel, LB Weiswald, et al. [75] estimated tumor tissue concentrations of drugs (see Supplementary Table 3). To the best of the authors knowledge, this study is the first, employing pharmacokinetic modeling to deduce optimal concentrations for translational pharmacotyping.
A major limitation of this study remains the small patient cohort. It should be noted, however, that this patient cohort is in the range to previously published PDO cohorts [20, 22, 24, 66, 83, 84]. The majority of previously reported PDO-studies included ten to fifteen patients. AK Beutel, L Schutte, J Scheible, E Roger, M Muller, L Perkhofer, A Kestler, JM Kraus, HA Kestler, TFE Barth, et al. [20] and H Tiriac, P Belleau, DD Engle, D Plenker, A Deschenes, TDD Somerville, FEM Froeling, RA Burkhart, RE Denroche, GH Jang, et al. [66] used a larger cohort of PDOs to develop the in-vitro based statistical classification methods, but were only able to clinically correlate a small fraction of patients. The clinical outcome was known for 16 of 28 patients and 9 of 66 patients respectively. Furthermore, statistical approaches used in previous studies as well as definition of clinical response were highly heterogeneous. It is also important to note that we included patients with both metastatic and primary resectable disease, which represents a heterogeneous patient cohort.
In conclusion, the high accuracy for matching PDO-drug response to the in-vivo response of patients is in accordance to previously reported translational studies [20, 22, 24, 66, 83, 84]. Moreover, this study emphasized the need to uniformly define the way PDAC PDOs are pharmacotyped (single drug testing vs. multi-drug testing) and are further classified into sensitive and resistant models. PDOs have previously been investigated in several studies and have shown promising results for predicting patients’ treatment response. Up until today, however, only single agent testing was used in PDOs to correlate the in-vivo multi-drug regimens, not taking into account essential synergistic effects of drugs. For the first time, multi-drug testing based on pharmacokinetically calculated drug concentrations was employed and results were compared to the clinical response of patients. Based on the findings of this study, multi-drug testing should be conducted in the future.
In this study, the potential of PDOs as a predictive therapy platform was confirmed and different scores for therapy prediction were evaluated using PDAC PDO models. Both statistical clustering and pharmacokinetic classification methods demonstrated good performance and reached prediction accuracies of up to 85%. This is the first study to systematically compare and contrast different classifications of in-vitro response in order to identify the score with the highest prediction accuracy. Our findings demonstrate that multi-drug testing enhances the predictive ability of PDO pharmacotyping to 85%, highlighting the importance of synergistic in-vitro effects. As PDOs represent a valuable platform for therapy response prediction in PDAC, the methods and results described are of high significance for future PDO-based translational trials.
5-FU: 5-Fluorouracil
AUC: area under the curve
AUCnorm: normalized AUC
cmax/tissue, [mmol/l]: maximum concentration of a drug in the tumor tissue, calculated by pharmacokinetic modeling
cmax/plasma, [mmol/l]: maximum plasma concentration of a drug
Gem/Pac: gemcitabine/paclitaxel
IC50, [mmol/l]: half-maximal inhibitory concentration
JNB: Jenks Natural Breaks
Log[IC50], [log[mmol/l]]: logarithmic transformation of the half-maximal inhibitory concentration
Log[IC50]norm, [log[mmol/l]]: normalized log[IC50]
mFOLFIRINOX: Folic acid, 5-Fluorouracil, Irinotecan and Oxaliplatin
PDAC: pancreatic ductal adenocarcinoma
PDO: patient-derived organoid
Author Contributions: All authors approved the work for publication. Conceptualization, C.N., K.W., U.P.,; methodology, C.N., K.W., U.P., M.P.D., J.I.; software, K.W., C.N.; validation, C.N. U.P., K.W.; formal analysis, K.W., C.N. ; investigation, C.N., K.W., U.P. M.P.D., J.I.; resources, C.N., K.W., U.P.; data curation, C.N., K.W., U.P., U.K.; writing—original draft preparation, K.W., C.N.; writing—review and editing, C.N., U.P.,K.W., J.I., G.H., L.V., M.F., J.P., M.L., C.J., A.G., S.S., U.K., M.B.; visualization, K.W., C.N.; supervision, C.N., U.P.; project administration, C.N., U.P.; funding acquisition, U.P., C.N. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the Berliner Krebsgesellschaft, grant number #PEFF202104 and by the „Hans Beger Stiftung zur Erforschung und Bekämpfung von Bauchspeicheldrüsenkrebs“. PD Dr. Pelzer is participant of the BIH-Charité Clinical Fellow, Dr. Neumann of the BIH-Charité Junior Clinician Scientist program and K. Wansch a BIH MD Doctoral scholar funded by the Charité-Universitätsmedizin Berlin and the Berlin Institute of Health. Dr. Ihlow and Dr. Dragomir were supported by the Berlin Institute of Health (Clinician Scientist Program) and Dr. Dragomir by the DKTK Berlin (Young Investigator Grant 2022).
Trial registration: The study was conducted according to the guidelines of the Declaration of Helsinki. Ethical approval was requested and granted by the ethics committee of the Charité-Universitätsmedizin Berlin (EA1/157/21) on 26th May 2021.
Informed Consent Statement: Informed consent was obtained from all enrolled subjects as part of this study.
Availability of Data and Materials: The data included in this study is presented in this published article.
Acknowledgments: The authors are deeply grateful to the support of the jointly affiliated research group of Prof. Ulrich Keilholz, especially S. Liebs, M. Bossen, Anna Kotarac and Anastasia Dielmann. We also want to thank Tobias Janik for his technical assistance. No other acknowledgements have to be declared.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the investigation; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
- Siegel RL, Miller KD, Fuchs HE, Jemal A: Cancer statistics, 2022. CA: A Cancer Journal for Clinicians 2022, 72:1.
- Quante AS, Ming C, Rottmann M, Engel J, Boeck S, Heinemann V, Westphalen CB, Strauch K: Projections of cancer incidence and cancer-related deaths in Germany by 2020 and 2030. Cancer Med 2016, 5:9.
- Rahib L, Wehner MR, Matrisian LM, Nead KT: Estimated Projection of US Cancer Incidence and Death to 2040. JAMA Netw Open 2021, 4:4.
- Nymo LS, Myklebust TÅ, Hamre H, Møller B, Lassen K: Treatment and survival of patients with pancreatic ductal adenocarcinoma: 15-year national cohort. BJS Open 2022, 6:2.
- Strobel O, Lorenz P, Hinz U, Gaida M, König A-K, Hank T, Niesen W, Kaiser Jör, Al-Saeedi M, Bergmann F et al: Actual Five-year Survival After Upfront Resection for Pancreatic Ductal Adenocarcinoma: Who Beats the Odds? Annals of Surgery 2022, 275:5.
- Kunzmann V, Siveke JT, Algul H, Goekkurt E, Siegler G, Martens U, Waldschmidt D, Pelzer U, Fuchs M, Kullmann F et al: Nab-paclitaxel plus gemcitabine versus nab-paclitaxel plus gemcitabine followed by FOLFIRINOX induction chemotherapy in locally advanced pancreatic cancer (NEOLAP-AIO-PAK-0113): a multicentre, randomised, phase 2 trial. Lancet Gastroenterol Hepatol 2021, 6:2.
- Kleeff J, Korc M, Apte M, La Vecchia C, Johnson CD, Biankin AV, Neale RE, Tempero M, Tuveson DA, Hruban RH et al: Pancreatic cancer. Nat Rev Dis Primers 2016, 2.
- Wang M, Herbst RS, Boshoff C: Toward personalized treatment approaches for non-small-cell lung cancer. Nat Med 2021, 27:8.
- Biller LH, Schrag D: Diagnosis and Treatment of Metastatic Colorectal Cancer: A Review. Jama 2021, 325:7.
- Ohishi T, Kaneko MK, Yoshida Y, Takashima A, Kato Y, Kawada M: Current Targeted Therapy for Metastatic Colorectal Cancer. Int J Mol Sci 2023, 24:2.
- Wood LD, Canto MI, Jaffee EM, Simeone DM: Pancreatic Cancer: Pathogenesis, Screening, Diagnosis, and Treatment. Gastroenterology 2022, 163:2.
- Luo J: KRAS mutation in pancreatic cancer. Semin Oncol 2021, 48:1.
- Conroy T, Castan F, Lopez A, Turpin A, Ben Abdelghani M, Wei AC, Mitry E, Biagi JJ, Evesque L, Artru P et al: Five-Year Outcomes of FOLFIRINOX vs Gemcitabine as Adjuvant Therapy for Pancreatic Cancer: A Randomized Clinical Trial. JAMA Oncology 2022, 8:11.
- Otsuka T, Shirakawa T, Shimokawa M, Koga F, Kawaguchi Y, Ueda Y, Nakazawa J, Komori A, Otsu S, Arima S et al: A multicenter propensity score analysis of FOLFIRINOX vs gemcitabine plus nab-paclitaxel administered to patients with metastatic pancreatic cancer: results from the NAPOLEON study. International Journal of Clinical Oncology 2021, 26:5.
- Oettle H, Neuhaus P, Hochhaus A, Hartmann JT, Gellert K, Ridwelski K, Niedergethmann M, Zülke C, Fahlke J, Arning MB et al: Adjuvant chemotherapy with gemcitabine and long-term outcomes among patients with resected pancreatic cancer: the CONKO-001 randomized trial. Jama 2013, 310:14.
- Khomiak A, Brunner M, Kordes M, Lindblad S, Miksch RC, Öhlund D, Regel I: Recent Discoveries of Diagnostic, Prognostic and Predictive Biomarkers for Pancreatic Cancer. Cancers (Basel) 2020, 12:11.
- El-Khoury V, Michel T, Kim H, Kwon Y: Personalized Drug Screening for Functional Tumor Profiling. Elements in Molecular Oncology 2022.
- Low RRJ, Lim WW, Nguyen PM, Lee B, Christie M, Burgess AW, Gibbs P, Grimmond SM, Hollande F, Putoczki TL: The Diverse Applications of Pancreatic Ductal Adenocarcinoma Organoids. Cancers 2021, 13:19.
- Melzer MK, Resheq Y, Navaee F, Kleger A: The application of pancreatic cancer organoids for novel drug discovery. Expert Opinion on Drug Discovery 2023, 18:4.
- Beutel AK, Schutte L, Scheible J, Roger E, Muller M, Perkhofer L, Kestler A, Kraus JM, Kestler HA, Barth TFE et al: A Prospective Feasibility Trial to Challenge Patient-Derived Pancreatic Cancer Organoids in Predicting Treatment Response. Cancers (Basel) 2021, 13:11.
- Driehuis E, van Hoeck A, Moore K, Kolders S, Francies HE, Gulersonmez MC, Stigter ECA, Burgering B, Geurts V, Gracanin A et al: Pancreatic cancer organoids recapitulate disease and allow personalized drug screening. Proc Natl Acad Sci U S A 2019, 116:52.
- Demyan L, Habowski AN, Plenker D, King DA, Standring OJ, Tsang C, St. Surin L, Rishi A, Crawford JM, Boyd J et al: Pancreatic Cancer Patient-derived Organoids Can Predict Response to Neoadjuvant Chemotherapy. Annals of Surgery 2022, 276:3.
- Grossberg AJ, Chu LC, Deig CR, Fishman EK, Hwang WL, Maitra A, Marks DL, Mehta A, Nabavizadeh N, Simeone DM et al: Multidisciplinary standards of care and recent progress in pancreatic ductal adenocarcinoma. CA Cancer J Clin 2020, 70:5.
- Seppälä TT, Zimmerman JW, Suri R, Zlomke H, Ivey GD, Szabolcs A, Shubert CR, Cameron JL, Burns WR, Lafaro KJ et al: Precision Medicine in Pancreatic Cancer: Patient-Derived Organoid Pharmacotyping Is a Predictive Biomarker of Clinical Treatment Response. Clinical Cancer Research 2022, 28:15.
- Inoue Y, Tanaka K, Hiro J, Yoshiyama S, Toiyama Y, Eguchi T, Miki C, Kusunoki M: in vitro synergistic antitumor activity of a combination of 5-fluorouracil and irinotecan in human colon cancer. Int J Oncol 2006, 28:2.
- Zeghari-Squalli N, Raymond E, Cvitkovic E, Goldwasser F: Cellular pharmacology of the combination of the DNA topoisomerase I inhibitor SN-38 and the diaminocyclohexane platinum derivative oxaliplatin. Clin Cancer Res 1999, 5:5.
- Frese KK, Neesse A, Cook N, Bapiro TE, Lolkema MP, Jodrell DI, Tuveson DA: nab-Paclitaxel potentiates gemcitabine activity by reducing cytidine deaminase levels in a mouse model of pancreatic cancer. Cancer Discov 2012, 2:3.
- Fischel JL, Formento P, Ciccolini J, Rostagno P, Etienne MC, Catalin J, Milano G: Impact of the oxaliplatin-5 fluorouracil-folinic acid combination on respective intracellular determinants of drug activity. British Journal of Cancer 2002, 86:7.
- Bertino JR, Mini E, Fernandes DJ: Sequential methotrexate and 5-fluorouracil: mechanisms of synergy. Semin Oncol 1983, 10:2 Suppl 2.
- Piro G, Agostini A, Larghi A, Quero G, Carbone C, Esposito A, Rizzatti G, Attili F, Alfieri S, Costamagna G et al: Pancreatic Cancer Patient-Derived Organoid Platforms: A Clinical Tool to Study Cell- and Non-Cell-Autonomous Mechanisms of Treatment Response. Front Med (Lausanne) 2021, 8.
- Schwartz LH, Litière S, de Vries E, Ford R, Gwyther S, Mandrekar S, Shankar L, Bogaerts J, Chen A, Dancey J et al: RECIST 1.1-Update and clarification: From the RECIST committee. Eur J Cancer 2016, 62.
- Broutier L, Andersson-Rolf A, Hindley CJ, Boj SF, Clevers H, Koo BK, Huch M: Culture and establishment of self-renewing human and mouse adult liver and pancreas 3D organoids and their genetic manipulation. Nat Protoc 2016, 11:9.
- Deyme L, Barbolosi D, Gattacceca F: Population pharmacokinetics of FOLFIRINOX: a review of studies and parameters. Cancer Chemother Pharmacol 2019, 83:1.
- Xie R, Mathijssen RH, Sparreboom A, Verweij J, Karlsson MO: Clinical pharmacokinetics of irinotecan and its metabolites in relation with diarrhea. Clin Pharmacol Ther 2002, 72:3.
- Poujol S, Pinguet F, Ychou M, Abderrahim AG, Duffour J, Bressolle FMM: A limited sampling strategy to estimate the pharmacokinetic parameters of irinotecan and its active metabolite, SN-38, in patients with metastatic digestive cancer receiving the FOLFIRI regimen. Oncol Rep 2007, 18:6.
- Berg AK, Buckner JC, Galanis E, Jaeckle KA, Ames MM, Reid JM: Quantification of the impact of enzyme-inducing antiepileptic drugs on irinotecan pharmacokinetics and SN-38 exposure. The Journal of Clinical Pharmacology 2015, 55:11.
- Klein CE, Gupta E, Reid JM, Atherton PJ, Sloan JA, Pitot HC, Ratain MJ, Kastrissios H: Population pharmacokinetic model for irinotecan and two of its metabolites, SN-38 and SN-38 glucuronide. Clinical Pharmacology & Therapeutics 2002, 72:6.
- Kho Y, Jansman FGA, Prins NH, Neef C, Brouwers JRBJ: Population Pharmacokinetics of Oxaliplatin (85 mg/m2) in Combination With 5-fluorouracil in Patients With Advanced Colorectal Cancer. Therapeutic Drug Monitoring 2006, 28:2.
- Delord J-P, Umlil A, Guimbaud R, Grégoire N, Lafont T, Canal P, Bugat R, Chatelut E: Population pharmacokinetics of oxaliplatin. Cancer Chemotherapy and Pharmacology 2003, 51:2.
- Nikanjam M, Stewart CF, Takimoto CH, Synold TW, Beaty O, Fouladi M, Capparelli EV: Population pharmacokinetic analysis of oxaliplatin in adults and children identifies important covariates for dosing. Cancer Chemotherapy and Pharmacology 2015, 75:3.
- Bastian G, Barrail A, Urien S: Population pharmacokinetics of oxaliplatin in patients with metastatic cancer. Anti-Cancer Drugs 2003, 14:10.
- Terret C, Erdociain E, Guimbaud R, Boisdron-Celle M, McLeod HL, Féty-Deporte R, Lafont T, Gamelin E, Bugat R, Canal P et al: Dose and time dependencies of 5-fluorouracil pharmacokinetics. Clinical Pharmacology & Therapeutics 2000, 68:3.
- Arshad U, Ploylearmsaeng S-a, Karlsson MO, Doroshyenko O, Langer D, Schömig E, Kunze S, Güner SA, Skripnichenko R, Ullah S et al: Prediction of exposure-driven myelotoxicity of continuous infusion 5-fluorouracil by a semi-physiological pharmacokinetic–pharmacodynamic model in gastrointestinal cancer patients. Cancer Chemotherapy and Pharmacology 2020, 85:4.
- van Kuilenburg ABP, Häusler P, Schalhorn A, Tanck MWT, Proost JH, Terborg C, Behnke D, Schwabe W, Jabschinsky K, Maring JG: Evaluation of 5-Fluorouracil Pharmacokinetics in Cancer Patients with a C.1905+1G>A Mutation in DPYD by Means of a Bayesian Limited Sampling Strategy. Clinical Pharmacokinetics 2012, 51:3.
- Greiner P, Zittoun J, Marquet J, Cheron J: Pharmacokinetics of (−)−folinic acid after oral and intravenous administration of the racemate. British Journal of Clinical Pharmacology 1989, 28:3.
- Straw JA, Newman EM: Pharmacokinetic Analysis of (6S)-5-Formyltetrahydrofolate (1-CF), (6R)-5-Formyltetrahydrofolate (d-CF) and 5-Methyltetrahydrofolate (5-CH3-THF) in Patients Receiving Constant i.v. Infusion of High-Dose (6R,S)-5-Formyltetrahydrofolate (Leucovorin). In: The Expanding Role of Folates and Fluoropyrimidines in Cancer Chemotherapy Advances in Experimental Medicine and Biology. vol. 244. New York, NY: Springer; 1988.
- Straw JA, Szapary D, Wynn WT: Pharmacokinetics of the diastereoisomers of leucovorin after intravenous and oral administration to normal subjects. Cancer Res 1984, 44:7.
- Stremetzne S, Streit M, Kreuser ED, Schunack W, Jaehde U: Pharmacokinetic and pharmacodynamic comparison of two doses of calcium folinate combined with continuous fluorouracil infusion in patients with advanced colorectal cancer. Pharm World Sci 1999, 21:4.
- Schilsky RL, Ratain MJ: Clinical Pharmacokinetics of High-Dose Leucovorin Calcium After Intravenous and Oral Administration. JNCI: Journal of the National Cancer Institute 1990, 82:17.
- Boosman RJ, Crombag M-RBS, van Erp NP, Beijnen JH, Steeghs N, Huitema ADR: Is age just a number? A population pharmacokinetic study of gemcitabine. Cancer Chemotherapy and Pharmacology 2022, 89:5.
- Khatri A, Williams BW, Fisher J, Brundage RC, Gurvich VJ, Lis LG, Skubitz KM, Dudek AZ, Greeno EW, Kratzke RA et al: SLC28A3 genotype and gemcitabine rate of infusion affect dFdCTP metabolite disposition in patients with solid tumours. Br J Cancer 2014, 110:2.
- Bulitta JB, Zhao P, Arnold RD, Kessler DR, Daifuku R, Pratt J, Luciano G, Hanauske AR, Gelderblom H, Awada A et al: Mechanistic population pharmacokinetics of total and unbound paclitaxel for a new nanodroplet formulation versus Taxol in cancer patients. Cancer Chemother Pharmacol 2009, 63:6.
- Henningsson A, Karlsson MO, Viganò L, Gianni L, Verweij J, Sparreboom A: Mechanism-Based Pharmacokinetic Model for Paclitaxel. Journal of Clinical Oncology 2001, 19:20.
- Schiffter-Weinle HA: Pharmakokinetik - Modelle und Berechnungen : mit 25 Tabellen und 323 mathematischen Formeln / Heiko A. Schiffter. Stuttgart: Stuttgart : Wiss. Verl.-Ges.; 2009.
- Schneeweiß S, Horger M, Grözinger A, Nikolaou K, Ketelsen D, Syha R, Grözinger G: CT-perfusion measurements in pancreatic carcinoma with different kinetic models: Is there a chance for tumour grading based on functional parameters? Cancer Imaging 2016, 16:1.
- McGuigan A, Kelly P, Turkington RC, Jones C, Coleman HG, McCain RS: Pancreatic cancer: A review of clinical diagnosis, epidemiology, treatment and outcomes. World J Gastroenterol 2018, 24:43.
- Diasio RB, Harris BE: Clinical pharmacology of 5-fluorouracil. Clin Pharmacokinet 1989, 16:4.
- Sharma V, Gupta SK, Verma M: Dihydropyrimidine dehydrogenase in the metabolism of the anticancer drugs. Cancer Chemotherapy and Pharmacology 2019, 84:6.
- Olsen EA: The pharmacology of methotrexate. Journal of the American Academy of Dermatology 1991, 25:2, Part 1.
- Stremetzne S, Schunack W, Jaehde U, Streit M, Kreuser ED: Pharmacokinetic and pharmacodynamic comparison of two doses of calcium folinate combined with continuous fluorouracil infusion in patients with advanced colorectal cancer. Pharmacy World and Science 1999, 21:4.
- Graham MA, Lockwood GF, Greenslade D, Brienza S, Bayssas M, Gamelin E: Clinical Pharmacokinetics of Oxaliplatin: A Critical Review. Clinical Cancer Research 2000, 6:4.
- Kawato Y, Aonuma M, Hirota Y, Kuga H, Sato K: Intracellular Roles of SN-38, a Metabolite of the Camptothecin Derivative CPT-11, in the Antitumor Effect of CPT-11. Cancer Research 1991, 51:16.
- Chabot GG: Clinical Pharmacokinetics of Irinotecan. Clinical Pharmacokinetics 1997, 33:4.
- Chabot GG, Abigerges D, Catimel G, Culine S, de Forni M, Extra JM, Mahjoubi M, Hérait P, Armand JP, Bugat R et al: Population pharmacokinetics and pharmacodynamics of irinotecan (CPT-11) and active metabolite SN-38 during phase I trials *. Annals of Oncology 1995, 6:2.
- Wong A, Soo RA, Yong W-P, Innocenti F: Clinical pharmacology and pharmacogenetics of gemcitabine. Drug Metabolism Reviews 2009, 41:2.
- Tiriac H, Belleau P, Engle DD, Plenker D, Deschenes A, Somerville TDD, Froeling FEM, Burkhart RA, Denroche RE, Jang GH et al: Organoid Profiling Identifies Common Responders to Chemotherapy in Pancreatic Cancer. Cancer Discov 2018, 8:9.
- Gota V, Nookala M, Bonda A, Karanam A, Shriyan B, Kembhavi Y, Gurjar M, Patil A, Singh A, Goyal N et al: Effect of body mass index on pharmacokinetics of paclitaxel in patients with early breast cancer. Cancer Med 2021, 10:9.
- Crombag MBS, de Vries Schultink AHM, Koolen SLW, Wijngaard S, Joerger M, Schellens JHM, Dorlo TPC, van Erp NP, Mathijssen RHJ, Beijnen JH et al: Impact of Older Age on the Exposure of Paclitaxel: a Population Pharmacokinetic Study. Pharm Res 2019, 36:2.
- Joerger M, Huitema AD, Richel DJ, Dittrich C, Pavlidis N, Briasoulis E, Vermorken JB, Strocchi E, Martoni A, Sorio R et al: Population pharmacokinetics and pharmacodynamics of paclitaxel and carboplatin in ovarian cancer patients: a study by the European organization for research and treatment of cancer-pharmacology and molecular mechanisms group and new drug development group. Clin Cancer Res 2007, 13:21.
- Von Hoff DD, Ervin T, Arena FP, Chiorean EG, Infante J, Moore M, Seay T, Tjulandin SA, Ma WW, Saleh MN et al: Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med 2013, 369:18.
- Ducreux M, Mitry E, Ould-Kaci M, Boige V, Seitz JF, Bugat R, Breau JL, Bouché O, Etienne PL, Tigaud JM et al: Randomized phase II study evaluating oxaliplatin alone, oxaliplatin combined with infusional 5-FU, and infusional 5-FU alone in advanced pancreatic carcinoma patients. Ann Oncol 2004, 15:3.
- Bayat Mokhtari R, Homayouni TS, Baluch N, Morgatskaya E, Kumar S, Das B, Yeger H: Combination therapy in combating cancer. Oncotarget 2017, 8:23.
- Hennig A, Baenke F, Klimova A, Drukewitz S, Jahnke B, Brückmann S, Secci R, Winter C, Schmäche T, Seidlitz T et al: Detecting drug resistance in pancreatic cancer organoids guides optimized chemotherapy treatment. The Journal of Pathology 2022, 257:5.
- Farshadi EA, Chang J, Sampadi B, Doukas M, Van 't Land F, van der Sijde F, Vietsch EE, Pothof J, Koerkamp BG, van Eijck CHJ: Organoids Derived from Neoadjuvant FOLFIRINOX Patients Recapitulate Therapy Resistance in Pancreatic Ductal Adenocarcinoma. Clinical Cancer Research 2021, 27:23.
- Hadj Bachir E, Poiraud C, Paget S, Stoup N, El Moghrabi S, Duchene B, Jouy N, Bongiovanni A, Tardivel M, Weiswald LB et al: A new pancreatic adenocarcinoma-derived organoid model of acquired chemoresistance to FOLFIRINOX: First insight of the underlying mechanisms. Biol Cell 2022, 114:1.
- Peschke K, Jakubowsky H, Schafer A, Maurer C, Lange S, Orben F, Bernad R, Harder FN, Eiber M, Ollinger R et al: Identification of treatment-induced vulnerabilities in pancreatic cancer patients using functional model systems. EMBO Mol Med 2022, 14:4.
- Fallahi-Sichani M, Honarnejad S, Heiser LM, Gray JW, Sorger PK: Metrics other than potency reveal systematic variation in responses to cancer drugs. Nature Chemical Biology 2013, 9:11.
- Pankreaskarzinom
- Grem JL: Biochemical modulation of 5-FU in systemic treatment of advanced colorectal cancer. Oncology (Williston Park) 2001, 15:1 Suppl 2.
- Di Paolo A, Orlandi P, Di Desidero T, Danesi R, Bocci G: Simultaneous, But Not Consecutive, Combination With Folinate Salts Potentiates 5-Fluorouracil Antitumor Activity In Vitro and In Vivo. Oncol Res 2017, 25:7.
- O'Connell MJ: A controlled clinical trial including folinic acid at two distinct dose levels in combination with 5-fluorouracil (5FU) for the treatment of advanced colorectal cancer: experience of the Mayo Clinic and North Central Cancer Treatment Group. Adv Exp Med Biol 1988, 244.
- Palmer AC, Sorger PK: Combination Cancer Therapy Can Confer Benefit via Patient-to-Patient Variability without Drug Additivity or Synergy. Cell 2017, 171:7.
- Grossman JE, Muthuswamy L, Huang L, Akshinthala D, Perea S, Gonzalez RS, Tsai LL, Cohen J, Bockorny B, Bullock AJ et al: Organoid Sensitivity Correlates with Therapeutic Response in Patients with Pancreatic Cancer. Clin Cancer Res 2022, 28:4.
- Wu Y-H, Hung Y-P, Chiu N-C, Lee R-C, Li C-P, Chao Y, Shyr Y-M, Wang S-E, Chen S-C, Lin S-H et al: Correlation between drug sensitivity profiles of circulating tumour cell-derived organoids and clinical treatment response in patients with pancreatic ductal adenocarcinoma. European Journal of Cancer 2022, 166.
No competing interests reported.