Generation of post-surgical minimal residual disease models to investigate metastasis in soft tissue sarcoma patient-derived orthotopic xenografts

Summary

Despite optimal multimodal treatment including surgical resection, 50%–80% of high-grade soft tissue sarcoma (STS) patients metastasize. Here, we present a protocol for the generation and use of post-surgical minimal residual disease models to investigate metastatic relapse in STS patient-derived xenografts. We describe steps for orthotopic engraftment of high-grade STS patient-derived tumor tissue. We then detail procedures for primary tumor resection with broad, negative resection margins and follow-up until metastases using MRI.

For complete details on the use and execution of this protocol, please refer to Fischer et al. (2023).¹

Graphical abstract

Highlights

• •Protocol for modeling minimal residual disease for soft tissue sarcoma in vivo
• •Steps for tissue collection, orthotopic engraftment, and hind limb amputation
• •Protocol for evaluation of primary and metastatic tumor growth with MRI
• •Guidance on processing of generated histologic, genetic, and radiological data

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Despite optimal multimodal treatment including surgical resection, 50%–80% of high-grade soft tissue sarcoma (STS) patients metastasize. Here, we present a protocol for the generation and use of post-surgical minimal residual disease models to investigate metastatic relapse in STS patient-derived xenografts. We describe steps for orthotopic engraftment of high-grade STS patient-derived tumor tissue. We then detail procedures for primary tumor resection with broad, negative resection margins and follow-up until metastases using MRI.

Before you begin

Before you begin, following preparations need to be taken care of.

• 1.Recruitment of high-grade sarcoma patients, eligible for primary tumor biopsy or resection. We advise the selection of treatment-naïve patients, because chemotherapy and radiotherapy could potentially impact engraftment rate or bias therapy testing results.²
• 2.The protocol can be used to set up MRD-PDX from fresh and cryopreserved tumor tissue. In case of the former, preparation of transport medium is necessary to transfer the patient-derived tumor tissue from the operating room (OR) to the laboratory. For the latter, preparation of cryopreservation medium to store fragments (1 mm³) of patient-derived tumor tissue in liquid nitrogen is necessary.
• 3.Preparation of animals: 6-week-old, NOD-SCID-IL2γ−/− (NSG) mice can be purchased from Charles River Laboratories (Wilmington, MA, US) and need to be maintained in pathogen-free facilities and given 2 weeks habituation period before the start of the experiment.

This protocol describes the specific steps for development of an MRD-PDX model for high-grade STS of the trunk and extremities. More specifically, this protocol has been executed for four different STS subtypes: malignant peripheral nerve sheath tumor (MPNST), undifferentiated pleomorphic sarcoma (UPS), rhabdomyosarcoma (RMS) and extra-skeletal osteosarcoma (EOS) but can be used for other STS subtypes of the trunk and extremities as well.

Xenografted tumors need to be resected with broad negative margins (R0) once they reach a size of 250–450 mm³, in order to mimic maximal safe resection as applied in the human clinical setting.³ We recommend resection once tumor size reaches 250–450 mm³, which will allow for tumor take, angiogenesis and dissemination,⁴ while at the same time preventing morbidity associated with the primary tumor.³ To achieve this, we prefer hindlimb amputation because it facilitates a higher likelihood for R0 resection and provides repeatability of the surgical protocol.

Institutional permissions

The collection and processing of human high-grade soft tissue sarcoma tissue was conducted following the Declaration of Helsinki and approved by the Ethics Committee of Ghent University Hospital (EC 2018/0080). Included patients gave their written consent to participate in this study.

NSG mice were handled and used in accordance with the national legislation of Belgium and the European community guidelines for animal studies. All procedures were approved by the Animal Ethical Committee of the Faculty of Medicine and Health Sciences at Ghent University, Belgium (DEC 19–45). All investigators in the study are Category C Animal Experiment Leader certified by the Federation of European Laboratory Animal Science Association. All efforts were made to minimize animal suffering. Animals were housed on a 12–12 light cycle (light on 6 am, off 6 pm), with a maximum of 4 animals per cage (floor area 580 cm²). They were provided with food and water ad libitum, and every cage was provided with nesting material (nestlets) and red-tinted nesting shelters.

When performing these experiments, it is absolutely necessary to first acquire permission from both human and animal ethical committees.

Preparation of cryopreservation vials for long-term tissue storage

Timing: 10 min

Note: All procedures below should be performed in a biological safety cabinet with sterile reagents and sterile equipment to minimize contamination, unless stated otherwise.

Cryopreservation enables the long-term storage of tumor tissue fragments. Tissue fragments are aliquoted in 1.2 mL cryotubes. Cryopreservation media can be prepared and aliquoted fresh, or one month beforehand and stored at ‒20°C up until the moment of use. It should then be thawed and stored at 4°C until the moment of actual usage during the protocol. It can be thawed and stored at 4°C for a maximum of 5 days.

• 4.
  For 10 mL of cryopreservation medium, 9 mL of heat-inactivated fetal bovine serum (FBS) is used, and 1 mL of DMSO.

  • a.
    Add 9 mL of FBS to a 15 mL tube.
  • b.
    Add 1 mL of DMSO.
  • c.
    Vortex for 5 s to mix the two components.
• 5.Aliquot the cryopreservation medium 1 mL/cryovial.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Biological samples
High-grade soft tissue sarcoma tumor fragments	Human	Ethical Committee of Ghent University Hospital approval (EC 2018/0080)
High-grade soft tissue sarcoma tumor fragments	Rodents, this paper	Animal Ethical Committee of the Faculty of Medicine and Health Sciences at Ghent University approved (DEC 19–45).
Chemicals, peptides, and recombinant proteins
DMSO	Chem-Lab	CL00.0422.1000
DMEM	Gibco	41963-039
Fetal bovine serum	PAN-Biotech	P30-3306
Penicillin/Streptomycin	Gibco	15070-063
Ketoprofen	Sanofi	BE132063
UltraDose Enzyme plus ultrasonic cleaning solution	L&R Manufacturing Company	UD038
Ringer solution	B. Braun	3570030
Chlorhexidine 0.2%	N/A	N/A
Experimental models: Organisms/strains
NOD-SCID-IL2γ−/− mice	Charles River Laboratories	614NSG
Software and algorithms
Horos	Horos Project	https://horosproject.org
GraphPad Prism 9	Dotmatics	https://www.graphpad.com
CNpare	Computational Oncology Group, Spanish National Cancer Research Centre (CNIO)	https://github.com/macintyrelab/CNpare5
ParaVision 360	Bruker	N/A
Other
Z880 surgical reconstructive microscope	Carl Zeiss	8801261
7 T small animal MRI	Bruker BioSpin	PharmaScan 70/16
CoolCell LX cell freezing container	Corning	432001
70 μm cell strainer	SPL Life Sciences	93070
Sterile surgical gown	N/A	N/A
Sterile surgical drapes	N/A	N/A
Sterile surgical gloves	N/A	N/A
Carbon steel surgical blades nr 11	Swann-Morton	0203
Carbon steel surgical scalpel handle nr. 3	Swann-Morton	0933
Microsurgery scissors sharp points curved 6″	N/A	N/A
McIndoe scissors	N/A	N/A
Operating scissors (heavy)	N/A	N/A
Debakey forceps	N/A	N/A
Gillies forceps	N/A	N/A
Jeweler’s forceps	N/A	N/A
Jeweler’s forceps swiss style	N/A	N/A
Needle holder	N/A	N/A
Castroviejo needle holder	N/A	N/A
Bone curette	N/A	N/A
Elastic stays	Lone Star	3316-8G
PDS 3/0 suture	Ethicon	Z423E
Vicryl 6/0 suture	Ethicon	V492H
Vicryl 8/0 suture	Ethicon	W9577
Vetbond	3M	1469SB
Surgical clipper	N/A	N/A
30G needle	N/A	N/A
0.3 syringes 8 mm 30G needles	BD Micro-Fine	0382904826031
Silicone pad	N/A	N/A
Small animal heating pad	N/A	N/A
Infrared lamp	N/A	N/A
Absorbent cotton gauze swabs (sterile)	N/A	N/A
Q-tip cotton swabs	N/A	N/A

Materials and equipment

Alternatives:

• •We made use of a Z880 surgical reconstructive microscope for tumor implantation and hindlimb amputation (Z880 surgical reconstructive microscope, serial no. 8801261). However, both procedures can also be performed using surgical binoculars, which is a less expensive alternative. If sufficiently experienced, tissue implantation can also be executed without magnification. Amputation is difficult to be executed without magnification.
• •For PDX analgesics injection, dilutions can be prepared using Ringer solution, Hartmann solution, NaCl 0.9% or PBS.
• •For cell freezing, a Cool Cell LX cell freezing container from Corning was used. This is equal to a Mr. Frosty freezing container from Thermo Scientific.
• •For surgical area disinfection chlorhexidine 0.2% was used. Betadine dermicum can be used as well, but this discolors the tissue which can make suturing more difficult. Ethanol 70% is not recommended as this can cause pain and irritation when in contact with wounds.

Transport medium

Reagent	Final concentration	Amount
DMEM	1x	44.5 mL
Fetal Bovine Serum	1x	5 mL
Penicillin-Streptomycin	1x, 100 U/mL	0.5 mL
Total	N/A	50 mL

[Stored at 4°C for up to six months.]

Cryopreservation medium

• •Add 1 mL of DMSO to 9 mL of Fetal Bovine Serum.

[Stored at ‒20°C for up to one month and at 4°C for up to one week.]

Ketoprofen solution

• •Dilute Ketoprofen 100 mg/2 mL to 1 mg/mL by adding 98 mL of Ringer solution.

Cleaning and sterilization of surgical equipment

For cleaning of surgical equipment, we recommend using an ultrasonic cleaner, with an ultrasonic cleaning solution such as UltraDose Enzyme plus. For sterilization, we recommend using an autoclave.

Step-by-step method details

Sterile collection of patient-derived tumor tissue

Collection of patient-derived tumor tissue can either be carried out in the OR (option 1), or at the pathology department (option 2). Direct harvesting of tissue during tumor resection in the OR before the resection specimen is transferred to the pathology department is preferred, as this provides shorter cold ischemia times, complete sterile execution of the protocol and thus potentially higher engraftment rates.⁶

Note: Patient surgery can take more time than anticipated. To minimize timing issues of sample collection followed by direct animal implantation we recommend cryopreserving tumor tissue followed by later planned (following day) animal xenografting. An uncertainty of the cryopreservation step is that this procedure may lower the tumor engraftment ratio.⁶ If patient surgery and animal xenografting can be well organized by an experienced team, we recommend animal xenografting directly following sample collection. The cryopreservation procedure can be maintained as additional tumor tissue can be used for future xenograft experiments.

Option 1: Rapid sterile collection of patient-derived tumor tissue in the operating theatre

Timing: 20 min

• 1.
  Prepare the sampling table by preparing a sterile field, strictly separated from the surgical field.

  • a.
    Use a mayo table or small instrument table that is kept strictly separate from the patient’s surgery. Set-up a sterile field using surgical drapes.

    • i.
      Prepare the table with a sterile gown, sterile gloves, and sterile surgical equipment. We recommend a pair of scissors, a scalpel with blade nr. 11, Debakey forceps, surgical forceps, needle holder and a suture PDS 3/0. A small curette can be useful in the case of necrotic cores or liquified tumors (see below).
    • ii.
      Use an alcohol-based hand rub for surgical sterile rubbing.
    • iii.
      After surgical rubbing is done, put on the sterile gown and gloves.
  • b.
    Once the tumor specimen is resected, transfer the specimen to the sampling table. The surgeon indicates the side of the tumor which has safe, broad resection margins. Preferably this site is covered by healthy tissue which can be closed again over the sampling area after tissue biopsy (see below).
• 2.
  The tumor is incised at the indicated site over a distance of at least 1 cm (Figure 1A 1^st and 2^nd image).

  • a.
    Depending on the tissue consistency, different surgical equipment will be needed for proper sampling.

    • i.
      If the tissue is more solid, Debakey forceps and surgical scissors or a scalpel can be used for tumor sampling (Figures 1A and 1B).
    • ii.
      If the tissue is more liquified, it will not be possible to hold the tissue with forceps (Figure 1C), and a curette can be used to scrape out tissue from the core of the tumor.
• 3.
  Depending on the purpose of the tissue, following steps have to be followed:

  • a.
    In case of tissue storage in cryopreservation medium: the tissue is diced into multiple pieces of 1–2 mm3 with a scalpel or surgical scissors (Figure 1A, 3^rd and 4^th picture). The pieces are transferred into the vials with forceps. Someone who is not scrubbed-in (e.g., nurse) opens the vials for the person who is doing the sterile sampling, as only the inside of the cryovials are sterile. In every cryovial with 1 mL of cryopreservation medium, 4–6 pieces are stored (Figure 1A, 5^th picture) .

    • i.
      The vials are then transferred into a Cell Freezing container (for a controlled cooling rate of ‒1 °C/min) and immediately transferred to the ‒80°C available in the OR.
  • b.
    In case of direct engraftment: a large tumor tissue piece (±1 cm³, Figure 1A, 3^rd image) is transferred into a 50 mL tube of preheated DMEM 10% FBS and quickly transported to the laboratory. Time from tumor tissue harvesting to tumor implantation should be no longer than 60 min.
• 4.
  Close the tumor at the sampling site with a running suture (PDS 3/0).

  • a.
    Make sure the tumor is then transferred to the pathology department. Document clearly where the sample site is located.

Note: tumor tissue for MRD-PDX engraftment can also be collected during a biopsy. In that case, the same set-up is used, but the surgeon will deliver the tissue biopsy, next to the biopsy that is taken for diagnostic purposes. The protocol then continues from step 3.

CRITICAL: for patient safety and prevention of tumor seeding and contamination, it is absolutely necessary to keep the surgical material used during the tumor sampling, as well as the tumor tissue itself, strictly separated from the patient’s surgery at all times. Surgical instruments used for tumor sampling cannot under any circumstances be used during the continuation of the patient’s surgery!

CRITICAL: The sampling should be supervised strictly by the leading surgeon, in order to not sample at the side where the quality of the resection margin could have an impact on the patient’s prognosis and/or treatment (Figure 1A, 1^st picture).

CRITICAL: tumor tissue pieces suspended in cryopreservation solution should be immediately transferred to a ‒80°C freezer and the time spent at room temperature should be kept as short as possible and should be documented. Ideally, a ‒80°C freezer is available in the OR department. If this is not the case, transfer the tissue in preheated transport medium to a laboratory with access to a ‒80°C freezer, and perform the tissue processing and storage there. Keep the transport time under one hour, as the duration of the cold ischemia time can impact the tumor take ratio.

Figure 1.

Collection of patient-derived tumor tissue

(A) Patient-derived high-grade malignant peripheral nerve sheath tumor metastasis resection specimen, located in the kidney, collected in the operating room (OR). From left to right, first image: dorsal side, region of the resection margin, not recommended for tumor sampling. T = tumor, K = kidney. Second image: ventral side, safe region for tumor sampling (the tumor specimen has been incised). (1) and (2) indicate the layers of healthy tissue that need to be closed with a running suture before sending the sample to the pathology department, (3) indicates the outer border of the tumor, (4) indicates the incised part of the tumor. Third image: sampled tumor tissue in a 6-well plate, size ±2 cm.³ Fourth image: tumor sample sliced into small pieces of 1–2 mm³. Fifth image: 4–6 tumor pieces (indicated by the dotted square) in a 1.2 mL cryovial with 1 mL of cryopreservation medium.

(B) Patient specimen, high-grade malignant peripheral nerve sheath tumor located in the iliopsoas muscle, in close relation to radix L3. T = tumor, surrounded by a thin layer of muscle tissue, M = muscle. First image: tumor specimen received at the Pathology department; no sample was taken in the OR. Second image: tumor tissue, sliced after inking (green). Third panel: macroscopic heterogeneity in the tumor specimen. Blue dotted line: necrosis, yellow dotted line: hemorrhagic necrosis, V = vital tissue.

(C) First image: patient specimen, high-grade chondrosarcoma of the ribs received at the Pathology department. Second image: patient specimen after inking of the tumor (green) and sliced. This sample is illustrative of a sample with little solidity and where tissue sampling is better with a curette.

Option 2: Collection of patient-derived tumor tissue at the pathology department

Timing: 60–75 min

Collection of tumor tissue at the pathology department takes place after inking of the tumor to assess the surgical resection margins. Disadvantages are (1) a longer cold ischemia time and (2) the difficulty to maintain tumor sterility, as the procedure at the pathology department is not aseptic. Advantages are the possibility of taking larger samples and taking samples at different macroscopical regions in the tumor (Figure 1B).

• 5.
  Transport the resection specimen from the operating theatre to the pathology department.

  • a.
    The transportation should be carried out in a sterile container.

CRITICAL: once arrived at the pathology department, do NOT wash the tumor with tap water as is a routine procedure in some pathology departments, this can contaminate the specimen with micro-organisms!

• 6.
  The tumor should be inked by the pathologist and then sliced (exemplified in Figures 1B and 1C).

  • a.
    Sample from the core of the tumor as much as possible. Do not sample at the site of the resection margins!
  • b.
    Document in detail how the tumor was sampled.

    • i.
      We advise to take pictures and indicate the area where the tissue was sampled.
    • ii.
      Describe the macroscopic variations of the tissue, e.g., vital tissue, necrotic tissue, hemorrhagic tissue (Figure 1B), as this is useful to correlate with the microscopical images and could potentially be related to engraftment rate which can be useful to your research question.⁷
  • c.
    From every sampled location, slice the tissue to perform:

    • i.
      formalin fixation, to assess the morphology of the sampled region.
    • ii.
      Cryopreservation and/or xenografting.

Optional: instead of cryopreservation, the tissue can also be transported immediately to the laboratory for direct engraftment. For this, continue the protocol as described above in Option 1, step 3b. We however recommend to additionally cryopreserve tissue. In that case, it is necessary to work with an experienced team, so one person can perform the cryopreservation protocol, and the other person can perform the orthotopic engraftment protocol.

Orthotopic primary tumor implantation into NSG mice

Timing: 15 min/animal

In this step, tumor tissue from high-grade sarcoma patients is engrafted orthotopically into NSG mice. The subtype and localization of the patient’s tumor decides the anatomical localization for xenografting. This protocol has been used to develop MRD-PDX for UPS from diverse localizations, MPNST from diverse localizations and EOS. Examples of patient tumor localization and xenograft tumor localization are illustrated in Table 1; the number indicates the patient pseudonymization.

Table 1.

Examples or orthotopic tumor tissue implantation

Model	Patient tumor	Implantation localization
EOS/045	Extra-skeletal osteosarcoma in the deltoid muscle.	Intra-muscular in the quadriceps femoris muscle.
MPNST/024	Malignant peripheral nerve sheath tumor in the iliopsoas muscle in close relation to radix L3.	Intra-muscular in the quadriceps femoris muscle.
MPNST/058	Malignant peripheral nerve sheath tumor of the femoral nerve.	At the site of the femoral nerve.
MPNST/183	Malignant peripheral nerve sheath tumor of the bone in close relation to the sciatic nerve.	At the site of the sciatic nerve.
UPS/030	Undifferentiated pleomorphic sarcoma, subcutaneously in the upper thigh.	Subcutaneously at the upper tight.
UPS/048	Undifferentiated pleomorphic sarcoma of the quadriceps femoris muscle.	Intra-muscular in the quadriceps femoris muscle.
UPS/059	Undifferentiated pleomorphic sarcoma of the biceps muscle.	Intra-muscular in the quadriceps femoris muscle.

A tumor tissue piece of only ± 1 mm³ is necessary for engraftment in 1 animal. Therefore, numerous PDX can be developed from only a small tissue biopsy. The number of animals xenografted for the patient included in Table 1 ranged between 8‒20 animals/patient. Tissue availability is usually not a limiting factor as sarcomas are very large.

• 7.
  Thawing of the patient-derived tumor samples.

  • a.
    Prepare a 50 mL tube with 37°C preheated DMEM + 10% FBS + Penicillin/Streptomycin 1% and a 6-well plate with 2 wells with 3 mL preheated DMEM + 10% FBS + P/S, place a 70 μm cell strainer in one of the wells filled with medium.
  • b.
    Thaw the tissue from the cryopreservation solution by gently adding preheated DMEM + 10% FBS + P/S 1% to the vial and transferring the thawed and diluted cryopreservation solution to the well with the cell strainer.

    • i.
      Continue until all of the cryopreservation solution is thawed and all tissue pieces are transferred into the cell strainer.
    • ii.
      Then transfer the cell strainer to the second well with DMEM + 10% FBS + P/S, so the tumor tissue is no longer in contact with the diluted cryopreservation solution (containing DMSO) (Figure 2A).
• 8.
  Preparation of the animal.

  • a.
    Install a heating pad to keep the animal at body temperature during the procedure. On top of the heating pad, place a silicone pad.
  • b.
    Anaesthetize an NSG mouse with isoflurane (induction 4%, maintenance 2%–3%, flow 0.8‒1 L/min) and position it on the silicone pad.
  • c.
    The animal is positioned for easy access to the region of interest.

    • i.
      For sciatic nerve implantation, we recommend a prone position with the hind limbs spread and paws taped to the surgical table (Figure 2B, left).
    • ii.
      For femoral nerve implantation, we recommend a supine position with the hind limbs spread and paws taped to the surgical table (Figure 2B, middle).
    • iii.
      For intramuscular implantation, we recommend a side position with the upper leg stretched and the paw taped to the table (Figure 2B, right).
  • d.
    Clip the hair short at the surgical region.
  • e.
    Disinfect the surgical region with chlorhexidine 0.2%.
• 9.
  Implantation of the patient-derived tumor tissue.

  • a.
    For sciatic nerve implantation (Figure 2C):

    • i.
      Make a small proximal incision in the skin (5 mm), longitudinal with and just underneath the femur.
    • ii.
      Spread the underlying muscle fibers (biceps femoris) until the sciatic nerve can be clearly visualized. Use curved forceps to spread the muscle fibers.
    • iii.
      In this deep pocket, position a tumor tissue fragment of ± 1 mm³.
    • iv.
      Enclose the muscle fibers over the nerve and the tissue fragment to close the pocket with 1 stich of Vicryl 8/0.
    • v.
      Close the skin as instructed above.
  • b.
    For femoral nerve implantation (Figure 2D):

    • i.
      Make a small inguinal incision in the skin (5 mm) with surgical scissors and forceps.
    • ii.
      Dissect the underlying fascia until the femoral nerve can be clearly visualized.
    • iii.
      Loosen the nerve from the surrounding tissue and create a pocket deep from the femoral nerve to position the tissue fragment. Make sure to do this below the level of the lateral circumflex femoral artery and the proximal caudal femoral artery, otherwise later amputation will be compromised.
    • iv.
      Enclose the surrounding muscle over the nerve and the tissue fragment (± 1 mm³) to close the pocket with 1 stich of Vicryl 8/0.
    • v.
      Close the skin as instructed above.
  • c.
    For intra-muscular implantation (Figure 2E):

    • i.
      The same method as instructed in (a) can be used, although here it is not necessary to go as deep, as contact with the sciatic nerve is not necessary. Make a small incision of max. 5 mm with surgical scissors and forceps in the skin on the lateral region of the hind limb.
    • ii.
      Split the underlying muscle fibers (m. quadriceps femoris) by spreading the scissors or curved forceps. Create a small pocket within the muscle.
    • iii.
      Insert one tumor fragment (± 1 mm³) into the pocket.
    • iv.
      Close the pocket with 1 stich of coated Vicryl 8/0.
    • v.
      Close the skin with a running suture of Vicryl 6/0 and 1 drop of Vetbond (tissue glue).

Optional: If necessary, the wound can easily be held open by using elastic stays with a blunt tip. The elastic ends can be fixed into the silicone pad with 30G needles. By the use of this, there is no additional assistance needed to retract the wound.

Note: Due to the MRI imaging and image artefacts, insertion of a metal chip to mark the animals is not possible. We recommend purchasing animals with a tattooed tail or marking of the ears during the implantation procedure.

Note: for rodent lower limb muscular and vascular anatomy, we recommend following referenced papers.^8,9

CRITICAL: A drop of Vetbond on top of the skin suture is essential to prevent the animals from opening the wound with their nails or teeth.

Figure 2.

Orthotopic implantation of high-grade soft tissue sarcoma

(A) High-grade soft tissue sarcoma tissue pieces thawed in DMEM supplemented with 10% FBS, ready for implantation.

(B) First image: positioning for sciatic nerve implantation, second image: positioning for femoral nerve implantation, third image: positioning for intramuscular implantation.

(C) Sciatic nerve implantation. First row, left to right: skin incision with scissors; opening of the wound with elastic stays; opening of the fascia and spreading of the muscle fibers; visualization of the sciatic nerve (indicated by the scissors). Second row, left to right: placement of a tumor tissue piece in the intramuscular pocket; closing of the surrounding muscle fibers over the tumor tissue piece with Vicryl 8/0; result of muscle approximation; closure of the skin with 2 stiches Vicryl 6/0 and surgical glue.

(D) Femoral nerve implantation. First image: skin incision longitudinal with the inguinal ligament, wound retraction with 3 elastic stays. The blue arrow indicates the inguinal ligament. Second image: visualization of the neurovascular bundle (femoral artery, vein and nerve, black arrow). The purple arrow indicates the level of the proximal caudal femoral artery.

(E) Intra-muscular implantation. First image: a small incision is made at the same location for sciatic nerve implantation. Second image: creation of an intra-muscular pocket with a tumor tissue piece already inserted (black arrow).

Imaging and tumor volume rendering with MRI of the orthotopic PDX

Timing: 20 min/animal (+ setup of the instrument: 20 min)

A 7-tesla magnetic resonance imaging (MRI) scanner for small animal imaging research is used to visualize the tumor growth (Figure 3A).

Note: We recommend the first scan 3 weeks after tumor implantation and then monthly. Usually, macroscopical tumor growth will not be visible yet, but it can potentially already be detectable on MRI. If the MRI is negative, we recommend repeating the investigation in one month and further clinically monitoring the animal. If the MRI scan remains negative, we recommend only scanning again when macroscopical tumor growth is visible. If macroscopically visible tumor growth occurs prior to the first 3 weeks, we recommend scanning more regularly as the tumor will grow exponentially. Close clinical monitoring is crucial in this case.

• 10.
  Start and install the MRI equipment.

  • a.
    Use a mouse body volume coil (¹H resonator, 75/40 mm) (Figure 3A).
  • b.
    Prepare a water-based heating blanket to maintain the animal’s body temperature (Figure 3B).
  • c.
    Connect the anesthesia tubing and the pressure sensor to monitor the animal’s respiratory rate (Figures 3B and 3C).
• 11.
  Preparation of the animal.

  • a.
    Anaesthetize an NSG mouse with isoflurane (induction 5%, maintenance 2%, flow 0.5‒0.8 L/min).
  • b.
    Transfer the animal to the heating pad of the MRI and place it in a prone position, headfirst.
• 12.
  Start the MRI protocol.

  • a.
    First, two localizer scans are acquired to identify the global position of the animal and the precise location of the pelvic and lower limb region.
  • b.
    Then, a T2-weighted TurboRARE scan is acquired. The main parameter settings of this accelerated spin echo sequence are listed below:

Repetition time TR	3660.9 ms
echo time TE	37.1 ms
field-of-view	30 mm × 25 mm
matrix	250 × 208
in-plane resolution	120 μm × 120 μm
slice thickness	600 μm
number of slices	30
number of averages NA	18 mm
RARE factor	8
acquisition time TA	6′20″

CRITICAL: closely watch the animal during imaging, keep track of the respiratory rate, and don’t overdose the amount of isoflurane. If the respiratory rate goes down, isoflurane dosages should be lowered. Animal death due to isoflurane overdose can be expected up to 5% of animals during MRI imaging.

• 13.Download the images for volume rendering.
• 14.
  Calculate the tumor volume with Horos (Figure 3D).

  • a.
    Open the application.
  • b.
    Open the axial TurboRARE-T2 image set of the animal of interest.
  • c.
    Change the mouse button function to “closed polygon”.
  • d.
    Indicate the tumor borders on every image with this function.
  • e.
    When finished, go to “ROI”, “ROI volume”, “compute volume”.

Figure 3.

Follow-up with MRI

(A) Set-up of the MRI. The black arrow indicates the mouse body volume coil (¹H resonator, 75/40 mm). The coil needs to be inserted into the MRI for visualization.

(B) Set-up for rodent placement in the mouse body volume coil. 1 = Nose cone for isoflurane and oxygen insufflation, 2 = pressure sensor to monitor the animal’s respiratory rate, 3 = water-based heating blanket to maintain the animal’s body temperature.

(C) Pressure sensor unit, which is located outside the MRI cabinet, showing the respiratory rate of the rodent on the screen.

(D) Volume rendering in Horos using the Region of Interest (ROI) application. Left image: indication of ROI (= purple circle = tumor) in every slice. F^R = right femur, F^L = left femur, S = spine, A = abdomen. Right image: volume calculation.

Amputation of the tumor-bearing hind limb

Timing: 30 min/animal

Time until this stage is reached: 9–38 weeks from tumor implantation

Amputation of the tumor-bearing hind limb (Figure 4) mimics tumor resection with broad negative surgical resection margins (R0). Using limb amputation, an identical method is used for every animal in the study. This is important as the completeness of resection will impact metastases and animal survival.¹⁰ As such, using an identical methodology for all animals should minimize this confounding factor. Negative surgical resection margins are evaluated through H&E microscopy by an experienced pathologist.

• 15.
  Prepare the experimental set-up, similar to the tumor tissue implantation.

  • a.
    Install a heating pad and place the silicon pad on top of the heating pad.

    CRITICAL: Make sure there is an infrared lamp present in the room, this can be used if the animal cools down too much during the procedure and has difficulties waking up. Using the infrared lamp during the procedure itself is not recommended because it complicates identification of anatomical structures due to the red color.

  • b.
    Switch on the surgical microscope and install the height of the chair, the operating table and the oculus in such a way that it is comfortable for you. Make sure to sit up with a straight back and relaxed shoulders.
  • c.
    Prepare the surgical set-up. We recommend the following tools (Figure 4A).

    • i.
      Surgical drape.
    • ii.
      Small cotton gauze swabs (5 × 5 cm, or smaller).
    • iii.
      Sterile q-tip cotton swabs.
    • iv.
      Ringer solution or any of the recommended alternatives.
    • v.
      0,3 mL syringes with 8 mm 30G needles.
    • vi.
      Ketoprofen 0.1 mg/100 μL Ringer solution.
    • vii.
      Elastic stays with blunt tip.
    • viii.
      Surgical scissors (McIndoe scissors and micro scissors).
    • ix.
      Surgical forceps.
    • x.
      Debakey forceps.
    • xi.
      Curved micro forceps (2).
    • xii.
      Fine needle holders, type Castroviejo (2).
    • xiii.
      Regular needle holder.

      Note: Use sterile equipment during these procedures!

• 16.
  Prepare the animal.

  • a.
    Anaesthetize a NSG mouse with isoflurane (induction 4%, maintenance 2%–3%, flow 0.8‒1 L/min) and position it in a supine position on the heated silicone pad.
  • b.
    Clip the hair short of the affected limb. Make sure to clip the hair short ventral and dorsal side of the animal.
  • c.
    Disinfected the surgical region with chlorhexidine 0.2%.
  • d.
    Before you begin, administer following substrates.

    • i.
      200 μL intraperitoneal Ringer solution to supplement potential blood loss.
    • ii.
      100 μL subcutaneous 1 mg/mL ketoprofen solution (Ketoprofen dose in rodents: 1–5 mg/kg).
• 17.
  Amputate the affected limb (Figures 4B and 4C).

  • a.
    Make an 8 mm inguinal incision using McIndoe surgical scissors and surgical forceps.
  • b.
    Dissect the skin loose from the underlying muscles of the limb and the abdomen using the curved forceps and curved micro scissors.

    • i.
      For blunt dissection, use two curved micro forceps.
    • ii.
      For sharp dissection, use micro scissors and curved micro forceps.
  • c.
    Use the elastic stays to retract the wound.

    Note: during dissection, but also during the whole procedure, make sure the tissues don’t desiccate. You can keep them lubricated by moisturizing a q-tip cotton swab with Ringer solution, and gently tapping the surgical area.

  • d.
    Identify the inguinal fat tissue overlaying the inguinal ligament (Figure 4B, lower panel). This can be absent, definitely in younger animals. If necessary, dissect the fat pad loose from the inguinal ligament and the surrounding muscle. If necessary for visual exposure, remove the adipose tissue completely. Make sure you do not cut the inguinal ligament.
  • e.
    Identify the neurovascular bundle (femoral nerve, artery and vein). The neurovascular bundle exits the abdomen under the inguinal ligament.
  • f.
    Dissect the neurovascular bundle loose from the surrounding tissue with both blunt and sharp dissection, above the level of proximal caudal femoral artery.

    • i.
      Be careful not to injure the lateral circumflex femoral artery which rises from the dorsal site of the femoral artery above the level of the proximal femoral artery during this maneuver, as it can cause bleeding that will be difficult to control.
  • g.
    Ligate the neurovascular bundle by preference proximal from the lateral circumflex femoral artery.

    • i.
      Using the curved forceps, go under the neurovascular bundle.
    • ii.
      Place a ligature (Vicryl 8/0, no needle) in the tip of these curved forceps and pull the ligature under the neurovascular bundle. Place the ligature as close as possible to the inguinal ligament. Tie the knot by performing a singular sliding knot 3 times.
    • iii.
      Repeat the previous maneuver but place the second ligature distally from the previous ligature. This is the second proximal ligature.
    • iv.
      Repeat the previous maneuver but leave some space in between the previous ligature the cut the vessel. This is the distal ligature.
    • v.
      Cut the neurovascular bundle.

      Note: due to the anatomical variation, it can be that it is not possible to place all ligatures proximal from the lateral circumflex femoral artery. If that is the case, place the 2 proximal ligatures proximal from this artery, and the distal ligature distal, but still proximal of the proximal caudal femoral artery. It can be useful to add an extra ligature just for the lateral circumflex femoral artery itself, as it can cause a great amount of backflow when cutting the neurovascular bundle at this level.

  • h.
    Transect the muscles at the level of the transection of the neurovascular bundle until the femur is completely free.

    • i.
      Dissect the muscle of the upper hindlimb loose from the skin using blunt and sharp dissection.
    • ii.
      Start with the transection under the level of the external obturator muscle and above the medial hamstring muscle. By doing this, you avoid cutting the deep femoral artery at a proximal level.

      CRITICAL: it could be that due to tumor size it is necessary to transect the muscle at a level where the volume of the deep femoral artery is too large to cut it without avoiding significant bleeding. If that is the case, it is important to ligate the vessel in a similar way as is described for the femoral artery.

  • i.
    Transect the femur with a bone cutter or heavy scissors.

    CRITICAL: bleeding can be present coming from the bone itself. You can stop this by applying pressure with a q-tip cotton swab for several minutes to the bleeding area. If some bleeding still remains, enclose the bone with a muscle flap using a Vicryl 8/0 cross-stich.

  • j.
    Close the skin over the stump using a running suture (Vicryl 6/0) and multiple drops of Vetbond.

    CRITICAL: A drop of Vetbond on top of the skin suture is essential to prevent the animals from opening the wound with their nails or teeth.

  • k.
    End anesthesia and observe the recovery. Place the animal under an infrared lamp to keep it warm if necessary and put it back in its cage before complete awakening. Recovery from amputation should be fast and within 30 min the animals should be behaving as usual, without need for segregation. Food and water intake should be the same as before the surgical procedure.

Figure 4.

Amputation of the tumor-bearing hind limb

(A) instrument set-up. Upper row, from left to right: Debakey forceps, surgical forceps (2) Castroviejo Needle Holder (2), curved micro scissors, Cottle-Masing Plastic Surgery Scissors (heavy scissors), needle driver. Lower row: elastic stays, curved forceps (2).

(B) Upper image: incision of the skin right below the inguinal ligament, lower image: opening of the underlying fascia. The inguinal ligament is indicated by the blue arrow.

(C) Amputation sequence. First row, left to right: (1) black arrow indicates the neurovascular bundle, the femoral nerve being the most lateral structure. The dotted arrow indicates the proximal caudal femoral artery. It is important to perform the ligation proximal of this structure. (2) freeing up the neurovascular bundle from the surrounding tissue and proximal ligation, (3) distal ligation, (4) placement of a second proximal ligature. Second row, left to right: (1) neurovascular bundle has been transected, (2) further incision of the surrounding skin, (3) transection of the muscles and transection of the bone, (4) amputation complete, arrow indicates the left-over skin flap. Third row, from left to right: (1) resection of the resection margin, (2) closure of the skin flap with a Vicryl 6/0 running suture, (3) application of skin glue, (4) healed amputation wound.

Follow-up of the MRD-PDX model, including imaging and harvesting of metastases

Timing: time until metastasis 12–75 weeks since implantation for first-generation (G1) animals

MRI: 20 min/animal. Necropsy and harvesting of metastasis: 15 min/animal.

After primary tumor resection, animals need to be followed-up closely to monitor disease progression. This can be done by regular clinical evaluation, including measuring weight weekly, or MRI. As MRI is costly and time-consuming, a strategy of watch and wait for clinical signs followed by MRI when disease progression is expected can be adopted.

• 18.
  Clinical evaluation.

  • a.
    Check the amputation site for swelling as a sign of local tumor recurrence.
  • b.
    Check the locations of the salivary glands, and axillary and inguinal lymph nodes for swelling as these can be locations for metastasis and are easily recognized.
  • c.
    Check behavior and body conditions such as weight, coat and skin, ocular discharge, eye/ear positioning, posture as indicators of disease progression.
• 19.
  MRI imaging.

  • a.
    The same protocol can be used as described above.
  • b.
    Metastasis in the lungs cannot be visualized with MRI.

Note: although the lungs are one of the main localizations for metastasis, follow-up in this region is difficult. MRI is not possible due to chest movement during breathing. Our consortium has explored the option of CT, but good visualization of the nodules is only possible when they’ve reached a size in which animals show systemic symptoms. This is why we recommend clinical follow-up, although the option of micro-CT should be further explored.

• 20.
  Necropsy and harvesting of distant metastasis (Figure 5).

  • a.
    Animals can be euthanized by cervical dislocation.
  • b.
    Lungs, liver, spleen, salivary glands, axillary and inguinal lymph nodes, macroscopic metastases and local recurrence if present, should be harvested.
  • c.
    Even when no macroscopic local recurrence is present, it is important to sample the amputation border to exclude microscopic local recurrence.
  • d.
    If clinical progression is present in the patient from whom the xenograft was derived, it is recommend harvesting PDX tissue from the matching anatomical areas where metastasis are present in the patient, even if PDX metastasis were not macroscopical or radiographical observed in those areas, as microscopical metastasis can be present.

Figure 5.

Examples of disease progression in MRD-PDX

(A) Lung metastasis of an undifferentiated pleomorphic sarcoma. L = lung, M = metastasis.

(B) multiple distant metastasis localizations, indicated by black arrows. Upper row, from left to right: (1) metastasis at the site of the axillary lymph nodes, (2) abdominal metastasis, (3) subcutaneous metastasis, (4) kidney metastasis. Lower row, from left to right: (1) kidney metastasis, (2) salivary gland metastasis, (3) multiple liver metastases, (4) multiple retroperitoneal metastases.

(C) Local recurrence of an extra-skeletal osteosarcoma at the amputation site. The black arrow indicates the local recurrence. Right image: transection of the local recurrence, showing a heterogeneous tumor.

Expected outcomes

This protocol holds model-specific advantages. First, because orthotopic implantation to develop MRD-PDX of STS of the trunk and extremities is at the hindlimb, the models have the unique characteristic that tumors are easily fully resected by hindlimb amputation. This reduces the heterogeneity between animals and treatment cohorts that normally would be created by the surgery itself, such as completeness of resection and wound response. Although it is difficult to rule out shedding and dissemination of cancer cells during surgery, the procedure of hindlimb amputation reduces surgery-induced cancer cell shedding to an absolute minimum, as amputation does not require direct manipulation of the tumor and direct surrounding area. This is ascertained by negative surgical margins, evaluated by hematoxylin and eosin histology. In addition, negative surgical margins are a prerequisite to avoid local relapse, a confounding aspect to study MRD and distant metastasis.

Second, metastasis arises spontaneously and do not require in vivo enrichment, selection and reinjection of cancer cells. Moreover, the distribution of organs affected by metastatic disease and their histological architecture is highly reminiscent to the metastatic spectrum observed in human STS, as has demonstrated by our group¹ and is currently being confirmed in ongoing experiments (unpublished data).

MRD-PDX primary tumor growth

Based on the MRI images and volume rendering in combination with time, different parameters can be calculated, such as primary tumor growth rate, tumor take ratio and probability of tumor engraftment.

Primary tumor growth rate is calculated as number of weeks since implantation until the tumor reaches a volume of 250 mm³. Tumor volumes are visualized with MRI (Figure 6A) and calculated with Horos (Figure 3D). Usually, tumor growth curves are visualized showing volume changes over time. An exponential growth equation can be fitted (exemplified for one animal, Figure 6B). For^. G1 MRD-PDX, first signs of tumor development on MRI can be expected 3–32 weeks post-implantation (n = 12). Time until tumor size reaches >250 mm³ can be expected 9–38 weeks post-implantation (n = 12). Time between first detection on MRI and hind limb amputation ranged between 3‒14 weeks. For G2 MRD-PDX (n = 28), time until first detection on MRI was 3–9 weeks. Primary tumor growth ratio is defined as the percentage of PDX that show primary tumor growth. For G1 MRD-PDX developed by patient primary tumor implantation this ranged between 75%‒100% (n = 8).¹ When implanting patient-derived metastasis, take ratios of 100% can be expected (n = 4).¹ For G2 MRD-PDX, take ratios were 100% for all models (n = 28). Probability of tumor growth can be visualized for groups and can be used to compare groups. It is visualized using Kaplan-Meier curves (Figure 6C). Usually, the time until a tumor reaches the size of 250 mm³ or time until amputation is used as the event.

Figure 6.

Examples of expected outcomes

Figure reprinted and adapted with permission from Fischer et al., 2023.¹

(A) Example of tumor growth after malignant peripheral nerve implantation attached to the femoral nerve, visualized by MRI. In the first image (week 8), no tumor growth is present. Orange arrows in the follow-up images indicate the primary tumor.

(B) Example of exponential growth curves of the primary tumor (black) and metastasis (pink) for one patient-derived xenograft. The minimal residual disease period is the time between amputation and detection of metastases, which is 18 weeks in this example.

(C) Example of comparison of probability of tumor growth between two groups using the Kaplan-Meier method. A p-value can be calculated.

(D) Examples of histochemical analysis of the tumor resection specimen (upper row, first image), negative resection margin (upper row, second image) and metastasis (lower row). Mu = muscle tissue, T = tumor, arrow indicates infiltrating tumor into the surrounding muscle tissue, B = bone, M = metastasis, F = fat tissue.

(E) (Immuno)histochemical comparison between patient and PDX samples for a patient with a high-grade undifferentiated pleomorphic sarcoma (UPS). Upper row are patient samples (primary tumor left and metastasis right), lower row are matched PDX samples. H&E = hematoxylin and eosin staining, ⍺SMA is alpha smooth muscle actin, a marker than can be positive in UPS. Both cell morphology and staining patterns show a high similarity between the patient and PDX tumors.

MRD-PDX disease progression

For disease progression, similar parameters can be calculated. However, metastasis or local recurrence growth is usually not calculated, as animals are quickly euthanized once disease progression is detected. We calculated a MRD period, metastasis ratio and probability of metastasis.

The MRD period is defined as the time between hind limb amputation with negative resection margins, and disease progression. For G1 MRD-PDX (n = 12) this was 14.61 ± 6.09 weeks, and ranged between 3 days and 40 weeks. The MRD period is illustrated in Figure 6B. Metastasis ratio is calculated as the percentage of PDX that show metastasis. Expected metastasis ratios are between 25%‒100% in G1 MRD-PDX (n = 12). Metastasis ratios after mouse metastasis implantation (G2 mO-PDX) is expected up to 100% (n = 4).¹ Localization of PDX metastasis is expected to be similar to patient metastasis. PDX metastasis have been observed in salivary glands, axillar and inguinal lymph nodes, abdominal lymph nodes, lungs, liver, kidney, bone and brain (Figures 5A and 5B) and are usually growing rapidly. Local recurrence was only detected in G2 MRD-PDX models which had positive resection margins (7%, n = 28). An example of a local recurrence is illustrated in figure FC. Probability of metastasis can be visualized for groups and can be used to compare groups. It is visualized using Kaplan-Meier curves, similar to probability of tumor growth (Figure 6C). Time until metastasis detection is used as the event.

Immunohistochemical analysis

Orthotopic tumors show infiltration of the surrounding tissue, as can be illustrated by nerve and vascular encasement, infiltration of fat, muscle and bone tissue (Figure 6D). A high morphological similarity is to be expected between patient and PDX primary tumors, as well as patient and PDX metastasis. Immunohistochemical makers specific to the tumor subtype are suspected to have a stable expression for both patients and corresponding xenografts, as well as their primary tumors and metastasis¹ (Figure 6E).

Genomic stability

Copy number variation (CNV) analysis allows to evaluate genomic similarity between patients and their corresponding xenografts. First and secondary generations can be compared, as well as primary tumor and metastasis.¹¹ We used CNpare (R-package)⁵ to calculate genome difference.¹

Resection margins

Resection margins should always be harvested and checked for tumor positivity (illustrated in Figures 4C and 7D). In our study, none of the resection margins in first generation PDX with tumor size < 450 mm³ were positive. For our G2, 2/28 animals had positive resection margins, but both these animals had a primary tumor size > 450 mm³.

Figure 7.

Troubleshooting for amputation

(A) Example of amputation in a rodent where the tumor tissue was implanted under the level of the proximal caudal femoral artery. There is a large distance between the tumor (T) (margin indicated with the full curved line) and the inguinal ligament (dotted line), which makes ligation of the neurovascular bundle more convenient. The right image shows the amputated hindlimb.

(B) Example of amputation in a rodent where the tumor tissue (T) was implanted too proximal, e.g., proximal to the level of the proximal caudal femoral artery. The black full arrow indicates the neurovascular bundle which is completely attached to the tumor. The dotted line shows attachment to the abdominal wall. Second image: placement of elastic stays to stretch away the abdomen, resulting in better visualization of the neurovascular bundle (arrow) and the tumor (T). Due to the positioning of the tumor, chances for negative resection margins are lower.

Applicability

MRD-PDX models can be used to study the biology of metastasis and as a preclinical tool to test metastasis-preventive agents.¹² These interventional studies can either be conducted in neo-adjuvant setting (before-surgery, primary tumor setting) or in adjuvant setting (post-surgery, MRD setting) in which metastasis-free survival is investigated as endpoint or metastatic burden is quantified at a predefined time-point. Metastasis-preventive agents could block a number of steps in the metastatic cascade: invasion of cells from the primary tumor into blood- and lymph vessels, survival in the circulation, arrest and egress from the circulation, dormancy and successful colonization of a distant organ. Potential other applications include investigating liquid biopsy biomarkers such as circulating tumor DNA¹³ and tumor-derived extracellular vesicles¹⁴ to determine the detection limit of MRD and to monitor metastatic relapse.

Quantification and statistical analysis

Horos was used for tumor volume rendering (exemplified in Figure 3A).

GraphPad Prism was used to develop tumor growth curves using exponential growth equations (Figure 6B). Comparison of tumor growth between groups (considering both take ratio and growth rate) can be performed using the Kaplan-Meier method (Figure 6C).

Limitations

Limitations of the protocol include both technical as biological aspects. First, orthotopic implantations and limb amputation require surgical skills. This step-by-step protocol provides ample documentation, ensuring that non-experts acquire the mastery of surgical orthotopic implantation/amputation. Basic surgical kills that should be learned before starting the protocol include suturing and rodent anatomy. Second, due to the absence of an intact immune system, evaluation of immunotherapy, which looks promising for some STS subtypes,¹⁵ is not possible. Last, only soft-tissue sarcomas of the trunk and extremities are eligible for this protocol. The protocol has not been tested for retroperitoneal sarcomas, such as well-differentiated and dedifferentiated liposarcoma, although these are only rarely resected with R0 margins in a clinical setting. We believe that due to the heterogeneity of STS, no preclinical model will exist that is a good surrogate for all STS subtypes and all possible treatment modalities, but this MRD-PDX is an important complementary addition to further study and target metastasis in sarcoma patients.

Troubleshooting

Problem 1: Lack of tissue consistency

When harvesting tumor tissue, it can be difficult to cut the tumor, depending on the tissue consistency. When a tumor is too soft, it can only be pulled apart or scraped instead of cutting into individual pieces.

Potential solution

• •
  Related to sterile collection of patient-derived tumor tissue:

  • ○
    Use a curette to obtain tumor tissue from the core of the tumor after tumor resection. Pull the tumor tissue apart using surgical forceps. Store a maximum of 6–10 mm³ tumor tissue in 1 cryovial containing 1 mL freezing medium.
• •
  Related to tumor implantation:

  • ○
    As the tissue consistency of the tumor is not eligible for suturing, it is important to completely close the muscle over the implanted tumor tissue, so it cannot slide out of the pocket. Preferentially, close the muscle with a running suture in this case.

Problem 2: Bleeding during the surgical procedure

• •Related to tumor implantation.

Implantation should always be bloodless. Don’t cut muscle with scissors for the creation of a pocket, as this will cause bleeding. Apply blunt dissection by using curved forceps and opening/spreading the pair of scissors for creation of the intra-muscular pocket. If bleeding occurs anyway, use a q-tip cotton swab to put pressure on the blood vessel, and hold this for at least 2 min for a venous bleeding and at least 5 min for an arterial bleeding. This will stop the bleeding in almost all cases.

If this doesn’t help, place an 8.0 stich over the bleeding, using the surrounding muscle tissue for compression.

• •Related to hind limb amputation:

There are multiple critical steps during this procedure when bleeding can occur.

Transection of the neurovascular bundle. This is an essential, but potential life-threatening step of the procedure. First, preferentially never grasp the neurovascular bundle with forceps. The femoral vein is the most vulnerable structure in the bundle; it can easily be damaged and can cause significant bleeding. As such, if you do need to grasp the neurovascular bundle, use Debakey forceps and grasp it at the site of the femoral nerve. Second, free-up the neurovascular bundle from the surrounding tissue for as much distance in the upper part of the limb as you possible can. This makes the ligation easier and chances of damaging the bundle during the continuation of the procedure smaller. Third, use two proximal ligation sutures instead of one, as extra safety measurement. If the proximal suture would be accidentally transected during the transection of the bundle, there is a spare ligation suture preventing bleeding. Fourth, when ligating, don’t pull the sutures with force, as the suture can transect the vessels. To prevent this, it is useful to start with a singular knot instead of the regular double knot when tying the sutures. If a bleeding still occurs, apply pressure more proximal than the site you want to transect with a q-tip cotton swab. Use a suture and go through the underlying muscle with a deep stitch to stop the bleeding by using the muscle tissue to compress the bleeding.

Bleeding after transection of the bone. Bone marrow is well vascularized and can bleed heavily when transected. This can be stopped by applying pressure with a q-tip cotton swab during several minutes. As an extra safety, the surrounding muscle tissue can be sutured together over the transected bone to apply extra pressure.

Problem 3: Tumor growth is too close to the amputation plane

When implanting tumor tissue attached to the femoral nerve, it is essential to implant the tumor tissue below the level of the proximal caudal femoral artery, to leave enough distance at the femoral artery to perform the ligation and obtain negative resection margins. This is illustrated in Figure 7, panel A showing an example where the tissue was implanted distally and panel B illustrating where the tissue was implanted more proximally. In the example of 7B, obtaining negative resection margins is difficult, but can be aided by pulling the abdomen away with an elastic stay, and if necessary, transecting the inguinal ligament to lengthen the femoral artery as much as possible.

Problem 4: The metastatic tissue is completely necrotic

It is possible that harvested PDX metastasis appear to be completely necrotic. This is due to the rapid tumor growth. There are however still plenty of viable, aggressive cells available in this tissue, and they can be visualized on H&E staining. This seemingly necrotic tissue can still be xenografted, in our study their engraftment ratio was 100% (n = 4).

Problem 5: Opening of the implantation/amputation wounds

To prevent wound dehiscence, it is really important to use both stiches and wound glue, otherwise the animals will scratch, or bite open the stiches. Check the wounds of the animals again <8 h after implantation. If the wound opened up, do not try to close it again with stiches, but cover the wound in skin glue to protect the wound from getting dirty. They are expected to heal properly.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Olivier De Wever (Olivier.dewever@ugent.be).

Technical contact

Further information and questions adressing the protocol should be directed to and will be fulfilled by the technical contact, Suzanne Fischer (suzanne.fischer@ugent.be).

Materials availability

Tissue samples are available at the Laboratory of Experimental Cancer Research, Ghent University, Ghent, Belgium, and will be happily provided for collaboration.

Data and code availability

Data is available upon reasonable request.