Establishment of Patient-Derived Organoids from Hepatocellular Carcinoma: Preliminary Data on Yield, Histopathological Concordance, and Methodological Challenges

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

 

Thanks for the opportunity to review this interesting article. 

The study presents a methodologically robust, dual-center, prospective interventional approach which assesses patient-derived organoids (PDOs) from hepatocellular carcinoma (HCC) patients and focuses on yield, histopathological concordance, and methodological challenges.

The study included 56 patients with diverse HCC backgrounds, documenting clinical and pathological features systematically. The establishment rate of PDOs is discussed with coverage of the G2 and G3 PDOs primarily. I do note that there is no discussion of G1 or G4 PDOs here, and a short comment on these in view of transparency would be appropriate if the authors feel they are able to.

It would also be useful to know how the sample was clinically processed for the purposes of histology and RNA processing. Were the samples histopathologically assessed at the time or placed in formalin for later? Removal of macroscopically normal tissue is also not commented on, which can affect downstream organelle development.

I also note that there is a comment about the PDO expression of CK7 and CK19 which are differentially expressed in these compared to tissue. Do these organelles appear therefore to have a cholangiocyte phenotype rather than a hepatocyte one? Would this affect your downstream analyses? A comment on this within the limitations would also seem appropriate at this stage.

Otherwise, Funding sources are transparently declared, and conflicts of interest are reported as absent. The author group is multidisciplinary, with affiliations documenting both clinical and translational expertise.

I am happy to recommend and look forward to future results.

 

Author Response

1. The study included 56 patients with diverse HCC backgrounds, documenting clinical and pathological features systematically. The establishment rate of PDOs is discussed with coverage of the G2 and G3 PDOs primarily. I do note that there is no discussion of G1 or G4 PDOs here, and a short comment on these in view of transparency would be appropriate if the authors feel they are able to.

A: We thank the reviewer for this constructive comment. In our cohort, no PDOs could be established from G1 and G4 HCC tumors. Although these cases were included in the clinical and pathological description of the patient population, only G2 and G3 tumors yielded viable long-term PDO cultures, which is why the detailed analyses presented in the manuscript focus on these grades. We have now clarified this point in the Results section for transparency. Page 7, paragraph 3.1

 

2. It would also be useful to know how the sample was clinically processed for the purposes of histology and RNA processing. Were the samples histopathologically assessed at the time or placed in formalin for later? Removal of macroscopically normal tissue is also not commented on, which can affect downstream organelle development.

A: We thank the reviewer for this important observation. We have now clarified the clinical processing of the samples to improve transparency and reproducibility. After surgical resection, the specimen is immediately transferred from the operating room to the pathology unit. A dedicated pathologist assesses the biopsy upon arrival, evaluates tumor quality (as detailed in Section 3.3), and selects the most appropriate tumor fragment for the establishment of HCC organoids. The specimen is placed in a sterile dish on ice and carefully inspected. Using a sterile scalpel, the pathologist precisely trims the tissue to remove macroscopically normal liver parenchyma and any regions not required for organoid culture. Only the visually confirmed tumor lesion is retained for downstream applications. The remaining portion of the biopsy is fixed in formalin at the time of collection for immunohistochemical and histopathological analyses of tumor markers. In parallel, another portion of the tumor tissue is snap-frozen at −80 °C for potential future DNA and RNA extraction (out of the purposes of the current manuscript). Once the histopathological nature of the tumor is defined by the pathology unit, the corresponding organoid is classified accordingly. We want to specify that RNA-based analyses are not performed in our current workflow, and organoid classification therefore relies exclusively on the histopathological features of the matched tumor. These details are now included in the revised manuscript to clearly outline the timing of tissue handling, histological assessment, and removal of non-tumoral areas, all of which are relevant for downstream organoid development. Material and Method section, paragraph 2.1, page 3.

 

3. I also note that there is a comment about the PDO expression of CK7 and CK19 which are differentially expressed in these compared to tissue. Do these organelles appear therefore to have a cholangiocyte phenotype rather than a hepatocyte one? Would this affect your downstream analyses? A comment on this within the limitations would also seem appropriate at this stage.

A: We thank the reviewer for the helpful observation. In our cohort, CK7/CK19 expression patterns differ substantially by tumour grade. In G2 tumors, CK7 and CK19 positivity is almost exclusively restricted to entrapped non-neoplastic biliary epithelium within the tumor mass (Figure 3). In contrast, G3 tumors show CK7 expression in malignant hepatocytes, whereas CK19 is present only in very small foci, largely corresponding again to entrapped biliary ducts. As we also discussed in the manuscript, CK19 expression in G3-derived PDOs is therefore more likely to reflect culture-driven progenitor reactivation/dedifferentiation than a high CK19 burden in the primary lesions. For G2-derived PDOs instead, we assessed MUC1, a well-established marker of biliary epithelium and cholangiocarcinoma. MUC1 staining was negative in both the original tumor tissue and the PDOs (Figure 1 of the rebuttal). The lack of MUC1 expression in the tumor strongly argues against the PDOs originating from non-malignant residual ducts, which are typically MUC1-positive. Instead, these findings support the interpretation that the CK7/CK19-positive phenotype observed in the PDOs reflects intrinsic features of the tumor cells. We acknowledge that additional cholangiocyte markers (e.g., SOX9) could further strengthen this analysis; however, due to limited remaining material we were unable to perform further stainings. Nonetheless, we believe that the concordant CK7/CK19 profile and the matching MUC1-negative pattern in both the tumor and PDOs provide sufficient evidence to support our interpretation. We have now commented this aspect in the discussion section of the revised manuscript. Discussion section, page 11.

See PDF here attached

Figure 1. MUC1 IHC in HCC tissue and Tumor-derived PDO.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1. Please provide additional details on the time interval between tumor resection and organoid culture initiation, as this may influence organoid
2. The manuscript occasionally alternates between “HCC organoids” and “liver tumor organoids.” For clarity, please use consistent terminology throughout.
3. In Figures 2 and 3, the resolution appears low, making it difficult to visualize morphological features. Higher-quality images would improve interpretability.
4. The composition of the extracellular matrix and medium supplements could be described more precisely, including supplier information and catalog numbers to enhance reproducibility.
5. Please include exact p-values (rather than only stating significance) in the results section to provide more transparent statistical interpretation.
6. Several sentences, particularly in the Introduction and Methods sections, would benefit from minor grammatical revisions for clarity and readability.
7. The ethical approval number is mentioned, but please also specify whether informed consent was obtained from all participants or their legal guardians.

Author Response

1. Please provide additional details on the time interval between tumor resection and organoid culture initiation, as this may influence organoid.

A: We appreciate reviewer’s comment and we agree that the timing is crucial for organoid development. The tumor biopsy is processed as quickly as possible, typically within 1 hour from surgical resection. To preserve tissue viability, the sample is never allowed to dry out; instead, it is immediately transferred in AdDF+++ medium (Advanced DMEM/F12 supplemented with 1× GlutaMAX, 10 mM HEPES, penicillin, and streptomycin) during transport from the pathology unit to the laboratory. The tissue processing begins immediately upon arrival. The interval between tissue processing and organoid culture initiation is also minimized. As soon as the enzymatic digestion and mechanical dissociation steps described in the Methods section are completed, the resulting single-cell suspension is promptly embedded in pre-cooled BME2 (Basement Membrane Extract, Type 2, Bio-Techne). The cell aggregates are seeded in 40-µL drops onto 24-well plates pre-warmed to 37 °C. After BME gelation at 37 °C for 20–30 minutes, the HCC culture medium (AdDF+++) is added. Therefore, organoid culture is initiated immediately after digestion, ensuring minimal delay between tumor resection, tissue processing, and the establishment of organoids. We have now clarified the timing of sample handling and organoid initiation in the revised manuscript. Material and Method section, page 4.

3. In Figures 2 and 3, the resolution appears low, making it difficult to visualize morphological features. Higher-quality images would improve interpretability.

A: We thank the reviewer for this helpful comment. We apologize for the low resolution of the images originally provided in Figures 2 and 3, which may have hindered the visualization of key morphological features. We have now replaced these panels with higher-resolution images to improve clarity and interpretability in the revised manuscript.

 

4. The composition of the extracellular matrix and medium supplements could be described more precisely, including supplier information and catalog numbers to enhance reproducibility.

A: We agree with the reviewer and we apologize for the missing information. We are now providing them in the current rebuttal and in the revised version of the manuscript. For organoid embedding, we used Cultrex Basement Membrane Extract (BME), Type 2, Path Clear (Bio-Techne/R&D Systems; Cat. No. 3532-010-02). This preparation is derived from Engelbreth–Holm–Swarm (EHS) mouse sarcoma and contains the major basement-membrane components laminin, collagen IV, entactin/nidogen, and heparan sulfate proteoglycans. The BME is supplied at a total protein concentration of 8–12 mg/mL in DMEM without phenol red supplemented with 10 µg/mL gentamicin. Manufacturer testing reports endotoxin ≤ 8 EU/mL and confirms sterility. The BME remains liquid on ice and polymerizes at 37 °C within 20–30 minutes, enabling stable dome formation for organoid culture.

Culture Medium Supplements: After BME polymerization, organoids are cultured in AdDF+++ medium, consisting of Advanced DMEM/F12 (Thermo Fisher Scientific; Cat. No. 12634010), supplemented with 1× GlutaMAX (Thermo Fisher; Cat. No. 35050061), 10 mM HEPES (Thermo Fisher; Cat. No. 15630080), and penicillin–streptomycin (100 U/mL–100 µg/mL) (Thermo Fisher; Cat. No. 15140122). This formulation provides enhanced buffering capacity, stable glutamine, and antimicrobial protection suitable for primary hepatocellular tumor cultures. The medium was freshly prepared and replaced every 2–3 days.

These details have been incorporated into the revised methods section to ensure complete transparency and allow accurate replication of our organoid cultures protocol (page 4)

 

5. Please include exact p-values (rather than only stating significance) in the results section to provide more transparent statistical interpretation.

A: We thank the reviewer for this suggestion. In the revised manuscript, we have performed quantitative analyses of marker expression between primary tumors and PDOs by using multiple unpaired t tests analysis. Exact p-values are now reported for all comparisons, providing a more transparent and rigorous statistical interpretation of the results. These values are included in the Results section and in the updated Figures.

 

6. Several sentences, particularly in the Introduction and Methods sections, would benefit from minor grammatical revisions for clarity and readability.

A: Grammatical revision of the Introduction and Methods sections was performed.

 

7. The ethical approval number is mentioned, but please also specify whether informed consent was obtained from all participants or their legal guardians.

A: We have already specified that informed consent was obtained from all participants (page 3 line 118).

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript by Lo Rei et al, described the preparation of HCC patients-derived organoids. The aim of their study, as stated on the introduction, is "...to establish a standardized protocol for generating HCC PDOs from resected specimens and to assess their morphological and immunophenotypic fidelity to the primary tumor."  

Major comments:

1. Results need major improvement. Section 3.1 describe how the PDOs maintain histological features of the tumor, but the results presented only show H&E staining and ki67 and no other histological features. Also, the images need to be better described, indicating in the figure what are the similarities between organoid and tissue of origin.
2. In the same way, the ki67 staining showed in figure should indicate is the staining is blue or brown (I assumed is the blue, but I am not totally sure). Also quantification of the positive area would be helpful to compare between PDO and tumor.
3. In figures 3 and 4 they show IHC results for a series of stage markers. Including quantification of the markers would be helpful to compare between PDO and tumor.
4. It is interesting to see the high expression of CK7 and CK19 in the PDO. It would be helpful if the authors could measure other cholangiocyte markers in the PDO to discard to possibility of the organoids being derived from the positive ducts observed in the tumor only. 
5. The authors indicate that 56 samples were used in the study, but only representative from one tumor were shown. Including quantification results for the used markers in several tumor-pdo pairs will increase the importance of the results.
6. The methods indicate that the organoids are passaged every 1-4 weeks, what passage were the organoids used for histology comparisons? Also, during how many passages the organoids maintain the described features?

Minor comments

Figure legends needs to describe better what it is shown, like inset in figure 4A, where there is no description.

Author Response

Major comments:

1. Results need major improvement. Section 3.1 describe how the PDOs maintain histological features of the tumor, but the results presented only show H&E staining and ki67 and no other histological features. Also, the images need to be better described, indicating in the figure what are the similarities between organoid and tissue of origin.

A: We thank the reviewer for this helpful suggestion. We have now expanded our morphological description of PDOs and explicitly highlighted the similarities with the corresponding primary tumors, including epithelial cytology, nuclear atypia, architectural disorganization and maintenance of a high proliferative fraction. This information has been incorporated into the Results section (paragraph 3.1, page 7).

 

2. In the same way, the Ki67 staining showed in figure should indicate is the staining is blue or brown (I assumed is the blue, but I am not totally sure). Also quantification of the positive area would be helpful to compare between PDO and tumor.

A: We thank the reviewer for this observation. We clarify that Ki67 staining in the presented figure is brown. As requested, we have now performed and included the quantification of the Ki67-positive area, which enables a direct comparison between PDOs and tumor samples. These results have been added to the revised version of the manuscript (Results section, paragraph 3.1, page 7).

3. In figures 3 and 4 they show IHC results for a series of stage markers. Including quantification of the markers would be helpful to compare between PDO and tumor.

         A: We acknowledge the reviewer’s comment and have now quantified all markers. As shown in the revised Figures 3 and 4, CK20, HepPar1 and Glypican-3 display comparable levels between PDOs and primary tumors, confirming their phenotypic similarity. CK19 (and CK7 to a lesser extent) is more highly expressed in PDOs, which we attribute to an in vitro dedifferentiation process during culture expansion. We have clarified this point in the Discussion section (page 11).

1. It is interesting to see the high expression of CK7 and CK19 in the PDO. It would be helpful if the authors could measure other cholangiocyte markers in the PDO to discard to possibility of the organoids being derived from the positive ducts observed in the tumor only.

A: We thank the reviewer for this important comment. To further exclude the possibility that the PDOs originated from the CK7/CK19-positive biliary ducts present in the tumor, we assessed the expression of additional cholangiocyte markers in the original HCC tissue and in the derived PDOs. We evaluated MUC1, a well-established marker of biliary epithelium and cholangiocarcinoma. MUC1 staining resulted negative in both samples (Figure 1 of the rebuttal). The absence of MUC1 in the original tumor argues against the PDOs being derived from non-malignant residual ducts, which are typically MUC1-positive, and instead supports the interpretation that the CK7/CK19-positive phenotype observed in the PDOs reflects the characteristics of the tumor cells themselves. We acknowledge that additional cholangiocyte markers (e.g., SOX9) could further strengthen this analysis; however, due to limited remaining material we were unable to perform further stainings. Overall, the combined marker profile (high CK7/CK19 with absent or minimal MUC1) supports a hepatocellular, rather than biliary, identity of the PDOs. These MUC1 data have been included in the rebuttal and can be provided as supplementary material if required.

The figure is provided in the attached PDF file.

Figure 1. MUC1 IHC in HCC tissue and Tumor-derived PDO.

2. The authors indicate that 56 samples were used in the study, but only representative from one tumor were shown. Including quantification results for the used markers in several tumor-pdo pairs will increase the importance of the results.

A: We thank the reviewer for this comment. As shown in Table 2, while 56 tumor samples were used in the study, 37 were HCC, and organoids were successfully established from 21 of these. For quantitative analyses of IHC markers, we focused on a representative subset of 8 organoids, covering both G2 and G3 tumors. This subset was selected to provide a balanced and informative representation of tumor differentiation grades. As shown in point 3 of the rebuttal, the markers quantification in these organoids consistently reflects the patterns observed in the primary tumors, supporting the conclusions of the study. We have clarified in each figure legend the number of samples quantified and the statistical method applied for the comparison, which we consider sufficient to illustrate the reproducibility and biological relevance of the findings across multiple tumor–PDO pairs.

3. The methods indicate that the organoids are passaged every 1-4 weeks, what passage were the organoids used for histology comparisons? Also, during how many passages the organoids maintain the described features?

A: Thank you for the comment. The organoids used for histological comparisons are collected at passages 4–5, a stage at which they show stable morphology and reproducibly display the histopathological features reported in the manuscript. From passage 4–5 onward, the organoids consistently maintain these morphological and histological characteristics across multiple subsequent passages under our culture conditions. This observation is in line with published experience showing that patient-derived liver-cancer organoids preserve the histological architecture, gene-expression patterns and genomic alterations of the parental tumors after extended in-vitro expansion (including both early and late passages)  (Broutier et al, Nature Medicine 2017, doi: 10.1038/nm.4438).

Literature evidence further supports that tumor-derived organoids are generally stable over time in culture: several groups report long-term expansion of liver and other tumor organoids while retaining parental tumor features for months and across numerous passages, although the exact duration and passage number of stable maintenance can vary with tumour type, initial viability, and the specific culture conditions used (Nuciforo et al, Cell Reports 2018 DOI: 10.1016/j.celrep.2018.07.001; Vlachogiannis et al, Science 2018, doi:10.1126/science.aao2774)

We therefore report (and have added to the Methods) that:

histology comparisons was performed on organoids at P4–P5; and under our conditions these organoids retain the described morphology and marker expression for multiple subsequent passages (routinely observed over the next several passages in our hands), consistent with prior reports of long-term stability of liver cancer organoids.

Caveats and additional details: the exact number of passages during which all features remain identical can be influenced by starting tissue quality, tumor subtype, seeding density, and medium/ECM composition; for these reasons, we monitor morphology at regular intervals. Published studies use different endpoints to define “long-term” (e.g., >2 months, >4 months or many passages), so reported passage-stability ranges vary between cohorts and labs; we therefore avoid overgeneralizing and report our histology time point (P4–P5) and the empirical observation of continued stability thereafter. (Artegiani et al, Human Molecular Genetics 2018, https://doi.org/10.1093/hmg/ddy187; Christopher et al, Cancer Res. 2016, doi:10.1158/0008-5472.CAN-15-2402.)

4. Minor comments Figure legends needs to describe better what it is shown, like inset in figure 4A, where there is no description.

A: We thank the reviewer for this helpful comment. We have revised the legend of Figure 4 to provide a clearer and more detailed description of the content presented. Specifically, we now explicitly describe the inset shown in Figure 4A, noting that it highlights a focal area of Glypican-3–positive staining and HepPar-1 within the tumor. The updated legend reads as follows:

Figure 4. Immunohistochemical characterization of patient-derived organoids (PDOs) compared with the original tumor tissue.

1. Upper panel: Hematoxylin and eosin (H&E) staining and immunohistochemistry for CK7, CK20, CK19, HepPar-1, and Glypican-3 in the primary HCC tumor tissue (scale bar: 100 µm). For Glypican-3 and HepPar-1, an inset highlights a focal area of positive staining within the tumor, corresponding to the region magnified in the main panel.
2. Lower panel: PDOs derived from the same tumor, showing comparable morphological features and marker-expression patterns across the same panel of antibodies (scale bar: 80 µm).

Author Response File: Author Response.pdf