Cutaneous tissues

The appropriate handling of cutaneous tissues procured for research purposes can be facilitated by having a pathologist or dermatopathologist working together with the surgeon in the actual procurement of the tissue. This is particularly important when dealing with small tumor samples like primary melanoma and other skin cancers to prevent compromise of diagnostic specimens. Specimen collection containers should be prelabeled with barcode/donor ID for high-volume collections to improve workflow and ensure accurate labeling and specimen tracking.

If not processed immediately, specimens should be placed in a clean or sterile container on wet ice (2°C–8°C) for transport from surgery to pathology or to the repository. If samples are to be used for establishing cell lines or tumor-derived xenografts, it is ideal for the specimen to be transported immediately to the pathology laboratory. Most institutions will have a biorepository technician available to do this when notified ahead of time.

To avoid desiccation and compromise of subsequent analyses, cutaneous specimens should not be resected on a dry towel or other absorbent material. Unless previously specified, tissue provided to the repository may be placed directly in appropriately labeled clean containers of cold biopreservation media (2°C–8°C) for transport to the repository for processing. If the tissue is to be frozen immediately, it is not necessary to place it in preservation media. Such media may cause ice crystals to form on the outside of the specimen when freezing.

Data should be maintained and tracked on the time that elapses between relevant time points (collection, processing, preservation, and storage). A date and time stamp can be used for maintaining these records efficiently. Information from the pathologist on the characteristics of the biopsy or surgical material (eg, percentage normal, percentage tumor, percentage necrosis, and/or percentage fibrosis) as determined by microscopic evaluation should be obtained on a per-specimen basis and recorded.

An important consideration for immunotherapy trials is the fact that tumors in the skin and dermis provide ready access to multiple and sequential biopsies. This can be the case for those on neoadjuvant treatment protocols as well as those on treatment trials for metastatic disease who still have their primary (or local recurrence) intact. In these cases, small (2 or 3 mm) punch biopsies should be done to sample the tumor while still allowing for response assessment. Additionally, small punch biopsies can simultaneously sample normal adjacent skin.

Lymph nodes (LNs) and nodal basin tissue

Regional LNs are the primary site of immune responses to vaccines and intratumoral immune therapies; thus, procurement of LN tissue provides critical information on the immune response to regional cancer immunotherapy. Palpable tumor-involved nodes can be biopsied at the patient bedside with a spring-loaded core needle biopsy instrument. Successful sampling depends on obtaining sufficient tissue to be representative of the tumor specimen. The volume of tissue obtained with a core needle biopsy depends on the inner diameter of the core needle and the ‘throw distance’, which defines the length of the core obtained (see table 2). Our routine practice is to obtain at least four core biopsies on each biopsy date, using a 14 G needle with a 2 cm throw distance. These can be obtained under local anesthesia through one small skin incision, with each sample placed immediately into preservation conditions. This is ideally done with staff at the bedside in the procedure room, with liquid nitrogen, RNAlater, formalin and tissue culture media. Table 3 shows the preferred preservation conditions based on the type of biospecimen analyses being planned.

Biopsy of putatively normal LNs commonly requires localization using the techniques of sentinel LN biopsy, such as ^99mTc sulfur colloid localization. This can be done with minimal morbidity under local anesthesia in a clinic setting.44 LN biopsy can rarely cause lymphedema, but we find that if the injection is in the lateral–anterior proximal thigh, removing an inguinal LN should be associated with a very low risk of lymphedema. Axillary nodes can also be biopsied, but some are deep and more challenging to remove. The node can be mechanically disaggregated, providing approximately 100 million immune cells for functional immune analyses, to be assessed fresh or viably cryopreserved in aliquots for later analyses.44 Example text describing these methods is provided in the online supplemental appendix.

Muscle and other soft tissues

The sampling of either primary or metastatic tumors in soft tissues and muscle, in the context of immunotherapy, is dependent on several variables, including anatomical location, size of the target lesion, intent and extent of the resection and invasiveness to the patient. The prioritization of tissue is often necessary, whether it is a full excisional biopsy or several core needle biopsies. The average tissue collection from a single 18 G core needle biopsy with a 2 cm throw distance is a little over 10 mm³ (~10 mg), but can be highly variable, depending on the consistency of the tissue and adequacy of biopsy.45 Thus, a larger core can be considered or more samples can be collected. When possible, larger core needle biopsies including 12, 14 and 16 G yield a higher amount of tissue as previously noted (see table 2), although the weight and amount of actual tumor tissue obtained are related to the consistency and composition of the tumor and its stromal elements.

The sample size is crucial as it may directly influence the type of testing and the number of analyses performed. Specifically, IHC performed for tumor-associated immune markers such as PD-L1 has been shown to be directly dependent on the amount of tissue analyzed46; therefore, the greater the tissue quantity obtained, the greater the potential accuracy. Typically, samples are immediately weighed, segmented, and frozen with a single slide for hematoxylin and eosin (H&E) staining created to confirm the adequacy of the specimen including bona fide tumor tissue. The percent of neoplastic nuclei, necrosis, and overall cellularity are recorded for each specimen as it directly influences the yield of DNA or RNA that can be extracted. In general, samples with 30%–40% neoplastic nuclei and less than 20% necrosis are acceptable for most uses. Tissue should be collected fresh to allow for flow cytometry, RNAseq, and mass spectrometry as formalin fixation will preclude several types of analyses. Immediate snap freezing in liquid nitrogen can be performed at the site of tissue collection. The remaining tissue portions can then undergo formalin fixation and IHC or ISH can be used to generate a picture of immune cell subsets. The peripheral versus central area of the tumor can be defined by simple IHC and comparison to H&E slides, which may reflect an evolving immune response if samples are taken across time. The actual time limit for tissue collection and extracting adequate RNA is variable with several factors, including devascularization of tissue in surgical specimens, processing time for standard pathology assessments, and time to quality-check the material. RNA may still be of good quality if the total time to freezing or processing is less than 2 hours.47

Breast tissue

Management recommendations for patients with breast cancer have centered on immunohistochemical expression of hormone receptors and human epidermal growth factor receptor 2 (ER/PR/HER2). mRNA-based gene expression assays render molecular signatures that provide additional guidance for treatment and relapse risk assessment. Tissue removed for clinical and research studies includes core needle biopsies, lumpectomy, and total mastectomy. Core needle biopsies should be placed immediately into 10% neutral buffered formalin.

Up to 60%–70% of laboratory-associated errors are due to preanalytical factors.48 A major problem in tissue procurement from resection specimens is the lack of quality standardization for preanalytical variables,48 49 most importantly, cold ischemia time and total time in formalin. Best practice guidelines for ER/PR/HER2 assays were recently updated.49 These guidelines recommend a cold ischemia time of <60 min and total time in formalin of 6–72 hours based on assessment of ER/PR status. Further studies are needed to define the impact of cold ischemia time on tissue collected from other anatomical sites. For breast tissue, cold ischemia time, fixative type, and time the sample was placed in neutral buffered formalin must be recorded. Further delays in tissue processing may be incurred when surgery centers are located far from academic/research centers. Compliance with these guidelines is not mandatory, leaving breast surgeons and pathologists to voluntarily collaborate to develop and implement standards of practice. In addition to rapid collection, a clear chain of custody, including time-stamped documentation of the preanalytical variables, should be required. Figure 1 describes a detailed time flowchart from tissue procurement through prioritization into samples for clinical and research labs.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Biopsy tissue acquisition in immuno-oncology clinical trials is a team approach. A standard process from one academic center that employs a tight chain of custody and real-time pathologist triage of patient samples in order to maximize the utility of biopsied material. A teams-based approach to review protocols helps develop specimen acquisition laboratory workflows that keep priorities for tissue allocations clear to each member of the team involved (yellow). Cold ischemia time begins at tissue removal and involves team coordination: nursing, clinical and research coordinators provide specific instructions for collection to the surgeon and the pathology retrieval team (green). The research technologist picks up the specimen at the collection site, establishes a chain of custody and time stamp on the tissues, and transports to clinical trials pathology lab (blue), where the pathologist examines tissues for viable tumor, necrosis, crush and hemorrhage artifact, grosses in and splits the tissue (orange), and sorts according to priority list (purple). In our experience, formalin fixation (purple, priorities 1 and 2) still represents the highest priority preservation because of its wide usage and relative standardization. When possible, grossing the tissues into two blocks, one allocated for clinical diagnosis and one allocated for research, is best as preservation of the clinical specimens is a regulatory requirement. The next most useful sample is a single-cell preparation in RPMI (purple, priority 3), where single-cell RNA sequencing (eg, 10× genomics) and highly multiparametric mass cytometry (Cytometry by time of flight, CyTOF), particularly in conjunction with multiplex immunohistochemistry, can yield deep insights into mechanisms of response and non-response to immuno-oncology agents. Finally, as in purple, priority 4, any remaining tissue after formalin fixation and tumor disassociation is flash frozen, either in liquid nitrogen or cryopreserved in OCT compound. FFPE, formalin-fixed paraffin-embedded; OCT, optimal cutting temperature; RPMI: Roswell Park Memorial Institute.

Head and neck tissues

For head and neck squamous cell carcinomas (HNSCCs), the threshold in tumor size for tissue collection is typically ≥1 cm at the primary site and/or in the metastatic tumor-involved cervical LNs. These are general guidelines that help minimize potential compromise to clinical care and may vary, depending on the institution and the involvement of the pathologists at the time of collection. In HNSCCs, the presence of pathological extranodal extension in the metastatic cervical LN can influence the staging of HNSCCs, in particular, human papillomavirus (HPV) -negative HNSCCs, and subsequent treatment guidelines. Thus, involvement of a pathologist to procure the cancer tissue is highly encouraged, if not required.

For immunological-based assays, non-cancerous involved tissue may often be needed as a comparative ‘control’, and this can include either adjacent normal tissue or non-tumor involved LNs. As part of a quality assurance measure in assessing the adequacy of a neck dissection and for accurate staging purposes, the pathologist reports on both the number of involved nodes and the total number of excised LNs. Thus, coordination with a pathologist in the sampling of LNs is also necessary to provide optimal clinical care and to ensure accurate reporting of the number of LNs removed at the time of the neck dissection.

Since the oral cavity is a major gateway of entry into the body, many head and neck cancers have background bacterial and/or fungal infections that have implications when trying to establish short-term and/or long-term cancer cell cultures. Thus, antifungals and/or antibiotics often need to be placed in the transport and/or culture medium to facilitate the preferential growth of the tumor cells in vitro. Oral rinse and/or saliva samples may also be included in the collection of biological specimens for patients with head and neck cancer. However, oral rinse and/or saliva samples may be most useful for genomic-based work since there is controversy in the field on the utility of using these samples to study the oral microbiome based on background contamination of the oral cavity. Unique to HNSCCs is also the field’s distinction between HPV-positive and HPV-negative HNSCCs. When collecting HNSCC tissue, it is recommended to record these two distinct categories whenever possible, because this will facilitate the analysis and/or subanalysis of the tissue samples in the future. Lastly, given the accessibility of head and neck cancers based on their location, there is an opportunity to obtain serial tumor biopsy samples during therapy, which can inform biomarker development, as well as provide insight into mechanisms of investigational therapies and associated changes induced within the tumor microenvironment (TME).

Lung and other thoracic tissue

Multiple prospective trials have demonstrated improved survival with administration of immunotherapeutic agents (PD-1/PD-L1 blockade) either alone or in combination with cytotoxic chemotherapy in advanced NSCLC, and more recently for small cell lung cancer.50–56 Treatment selection considerations in this population include tumor histology, PD-L1 levels and the presence of epidermal growth factor receptor, anaplastic lymphoma kinase, or ROS proto-oncogene 1 driver mutations. Thus, in advanced disease, pretreatment biopsy is critical for selection of first-line therapy. In most patients presenting with advanced disease, the primary tumor or metastatic sites are readily available for percutaneous core needle biopsy. Central tumors may also be amenable to bronchoscopic approaches, although with generally lesser quantities of tissue. Collection of bronchoalveolar lavage specimens at the time of bronchoscopy facilitates microbiome studies.

The efficacy of checkpoint inhibition in patients with metastatic NSCLC has prompted much interest in the application of immunotherapeutic agents in resectable disease. Recent phase I/II studies have demonstrated impressive pathological responses to neoadjuvant checkpoint inhibition57–61 and multiple phase III studies powered for survival endpoints are in various stages of accrual. Studies to date employ major pathological response and complete pathological response (CPR) as surrogate endpoints for efficacy. Thus, standardization of the processing and interpretation of pathological materials is paramount for comparison across histologies, neoadjuvant treatments and trials. The International Association for the Study of Lung Cancer has recently published recommendations for the pathological assessment of lung cancer resection specimens following neoadjuvant therapy.62

Central mediastinal LNs (ie, paratracheal and subcarinal) are most commonly biopsied via endobronchial ultrasound-guided needle aspiration. However, when more substantial tissue quantities are required, mediastinoscopy may be necessary. Anterior mediastinotomy (Chamberlain procedure) can provide tissue from anterior mediastinal or aortopulmonary window LNs not accessible by percutaneous techniques. Similar approaches can be used to obtain tissue from the thymus or other anterior mediastinal tumors. Processing of nodal specimens should be performed as delineated elsewhere in this review (see table 3).

Colorectal tissue

Immunotherapy is useful for patients with colorectal cancer (CRC) with microsatellite instability (high), and high mutational burden,63 64 but is typically more effective if checkpoint targets are present, along with increased CD8+ T-cell infiltration.65 In patients with deficient mismatch repair, as high as 90% CPR has been reported.66 Chalabi et al reported a 95% major response and a 60% complete response in early-stage, mismatch repair deficient (dMMR) colon cancers treated with ipilimumab+nivolumab.67 For patients with intact mismatch repair, 27% showed response to therapy, with CD8+/PD-1+ T-cell infiltration predicting response.67 These data support the role of tumor histology, MMR status, PD-1/PD-L1 levels and CD8+ T-cell infiltration, along with the mutational status of BRAF and KRAS, as important biomarkers in CRC that may be useful in patient selection for immunotherapy.

Assessment of tumor immune infiltrates in CRC is complicated by the high level of immune cells in the gut, including gut-associated lymphoid tissue rich in lymphocytes that protect the epithelial barrier.68 The lamina propria contains antigen-presenting cells, innate lymphoid cells, CD4+ and CD8+ T cells, and B cells that affect immune responses throughout the body. Subverted organization of colonic crypts, chronic inflammation, and oxidative stress expose cancer stem cells in CRC to non-canonical signaling pathways, affecting their self-renewal and adaptability to therapy.69–71 The microbiome also plays a critical role in determining therapeutic responses and in considering tissue collection.72 Differences in geography, lifestyle, and diet, along with microbial diversity between individuals and between mucosal versus luminal sites within an individual add another layer of complexity.73–75 Existing microbiota can be characterized using metagenomic and metatranscriptomic sequencing74 75; however, bacterial contamination and the dynamic relationship between the gut microbiome and immune system complicates this analysis. Obtaining high-quality tissue necessitates tumor cell purification for accurate molecular analysis of the heterogeneous CRC TME and is essential for the success of immuno-oncology clinical trials. Similar to head and neck tissues, addition of antibiotics may be helpful when viable cells need to be derived from gastrointestinal tract tumors.

Pancreatic tissue

Continued research into the complex biology of pancreatic ductal adenocarcinoma (PDAC) is key to personalizing treatment and improving outcomes. Guidelines exist for standardized procurement of pancreatic tissues.76 Clinical trials mainly enroll patients with locally advanced unresectable or metastatic disease. The anatomical location and heterogeneity makes obtaining tissue challenging.77 Endoscopic ultrasound is the norm for obtaining fine needle aspirate (FNA) specimens, which are frequently inadequate for meaningful analysis of the TME. The feasibility of getting appropriate DNA quantity is improved in newer studies.78 One hundred percent driver gene concordance was noted when comparing FNA to matched tumor tissue identifying different mutations for locally advanced PDAC. Fibrotic reactions around pancreatic tumors are a problem, and laser microdissection is often used to extract tumor cells out of surrounding stromal non-malignant cells.

Histologically, PDACs have thicker desmoplastic structure than most other cancers with an erratic distribution of lymphoid components. There is minimal to moderate CD3+, CD4+, and CD8+ T-cell infiltration, predominantly stromally, and metastatic PDACs have lower immune infiltrates compared with resectable primaries.79 The TME is immunosuppressive, but functional TILs have been successfully expanded from pancreatic tumors and are able to respond to tumor-associated antigens.80 Major issues with pancreatic tissue include the presence of proteases, RNAses, and DNAses, which require expert care and immediate handling in order to maintain quality of planned analytes.

For genomic analysis, high quality, well-annotated tissue repositories are needed. Specimens should be placed in preservative immediately, and transferred rapidly to pathology, ideally within 15 min, to reduce RNA degradation. Minimum sample weight of 200 µg is preferred for DNA isolation. Preservation techniques are based on the study intended (eg, flash freezing in liquid nitrogen for RNA and DNA isolation, while FFPE shows better protein structure preservation). Collecting matched peripheral blood samples to isolate germline DNA, as well as serum and pancreatic juice, for future correlation with soluble biomarkers is recommended.81 Creation of coordinated clinically annotated regional and national cancer tissue banks is encouraged.82 Recently, interest has been generated in research autopsy (a postmortem medical procedure with primary goal of collecting tissue to support basic and translational research) especially to obtain advanced-stage tumor tissue and to facilitate creation of cell lines and PDX models.83

Liver tissue

The liver is an immune organ, with unique components such as Kupffer cells and abundant anticancer effector cells like natural killer (NK) and NK T cells, and a complex relationship exists with chronic inflammation and anticancer immune responses.84 Few patients qualify for surgical resection and research specimens are hard to come by, mainly obtained from diagnostic biopsies. Many biorepositories and biobanks contain primarily FFPE College of American Pathologists-graduated older samples with corresponding clinical data. Hepatocellular carcinoma is unique due to various mutagenic agents contributing to genomic alterations.85 Exome sequencing shows imprints of mutagenic exposures and has identified new genes contributing to tumorigenesis. Protocols to characterize the landscape of pharmacogenomic interactions have discovered gene–drug associations and predictive biomarker candidates.86

Core needle biopsies of liver lesions have a high failure rate (>50%), mainly due to insufficient tumor content or low RNA yield despite a median of 3 x passes per biopsy and a median of three cores with an average maximum length of 0.7 cm. Liver tissues have high endogenous biotin activity and may simulate positive staining when using a detection method based on biotin labeling, so positive and negative tissue controls have to be used for each antibody. Cold ischemia impacts alterations in RNA transcript levels and protein expression, and significant RNA degradation has been reported after 12 hours.76 RNA from FFPE tissue suffers from strand breakage and cross-linking during tissue handing. More than 2000 ng of mRNA is necessary for NGS.87

The hepatic TME is enriched with Tregs, tissue resident memory CD8+ T cells, resident NK cells and tumor-associated macrophages. PD-1⁺ tissue-resident T cells have been identified as the predominant T-cell subset responsive to anti-PD-1 treatment and were significantly reduced with tumor progression. Using high-dimensional proteomic and transcriptomic analyses, studies showed that an immunosuppressive gradient was identified across the TME, non-TME and peripheral blood in primary HCC that manipulated the activation status of TILs and rendered them immunocompromised against tumor cells.88

The rich blood supply of the liver makes it the most common site of visceral metastasis, especially from CRC. The liver TME also provides autocrine and paracrine signals that creates a metastatic niche for tumor growth. Four different phases of liver metastatic development have been described based on both location of cancer cells in the liver parenchyma and phase-dependent interactions between cancer cells and the liver microenvironment89; the biological mediators of these phases are potential targets for clinical investigation. Histopathological growth patterns of colorectal metastasis (the specific interface between the tumor and the surrounding normal liver), especially the desmoplastic type associated with the inflamed immune phenotype, could predict efficacy of antiangiogenic therapies in metastatic CRC and may provide a foundation for biomarker analyses of tumor immunotherapy agents in HCC.90

Urological tissue

Despite advances in the treatment of genitourinary malignancies, as a group, they still result in significant cancer-related mortality worldwide.91 Biospecimen collection of genitourinary tissue may be invasive and associated with complications (eg, renal hemorrhage and depletion of tissue). Using alternatives to patient-derived tissue, such as immortalized cell lines, to develop anticancer drugs is discouraging, as they often fail to represent fundamental features of human tumors.92 Similarly, the use of two-dimensional monolayer cultures for drug screening and resistance studies has often led to inaccurate results and thus has failed to predict chemoresistance of various tumor types.93 There is a growing need to develop patient-derived cell models as these retain patient-specific features and are amenable to a variety of experimental applications, including assessment of genitourinary immunotherapy.94 95

Urological tissue should be collected from either biopsy or surgical specimens. Body fluids like urine or blood are collected during routine clinical care activity and can be collected for biospecimen banking as well. Blood is often collected as serum, plasma and buffy coats and provides germline data. Although urine can be valuable in order to assess the microbiome and perhaps response to therapy, it has multiple caveats that prevent it being used to adequately sample urological tumors. For one, it may not accurately reflect one particular urological organ (as it could represent the inner lining of the kidney, ureters, bladder, prostatic urethra, and the penile urethra). Second, cellular yield is affected by the time of day and method of collection. Therefore, for organ-specific samples, biopsy and surgical specimens are still preferred, but urine may represent a non-invasive future method of sampling.

Special care should be taken in biopsy of primary and metastatic clear cell renal cancers as they are highly vascular and hemorrhage is more common than in other epithelial malignancies.96 Urological tissue may be obtained using image guidance, and the reported overall mortality of abdominal fine needle biopsies is only 0.031%.97 Biopsies of some urological tissue (eg, prostate and select bladder specimens) may be performed in the outpatient clinic under a local anesthetic. Prostate biopsies are performed under transrectal/perineal ultrasound guidance, and small bladder biopsies can be performed in the office only if they require minimal to no coagulation. However, most substantial specimens are obtained from the operating room. After collection, the biospecimens must be processed, aliquoted, and frozen in containers suitable for long-term cryopreservation.98 Storage at −80°C or in liquid nitrogen freezers is preferable in order to preserve tissue integrity and to avoid repetitive freezing and thawing.99 100 The involvement of a pathologist is crucial for gross examination and dissection of the specimen in order to allocate what portion of tissue will be used for routine pathological analysis and the portion that will be used for biobanking.98 The majority of tumor tissue will be collected and processed as FFPE tissue blocks. If feasible, tumors can also be collected at the time of procurement and be snap (fresh) frozen for future analysis or placed into OCT media and then frozen (see table 3). Such samples allow molecular analysis and sampling of tumor tissue-infiltrating immune cells (eg, lymphocytes) using a variety of methodologies. Other methods in which viable human tissue may be propagated include the generation of patient-derived cell lines, organoids and xenografts. Similar to other tissues, urological biospecimens are subject to several challenges in biobanking and specimen collection, including cost, patient consent, reporting back of genetic aberrations, ongoing storage and limited specimens for multiple analyses.

Gynecological tissues

The role of immunotherapy has been shown to be promising in several cancers given its durable response and lower toxicities. However, in gynecological cancer, these results have been underwhelming as monotherapy, except in endometrial cancer with deficient mismatch repair. The response rates in recurrent platinum-resistant ovarian cancer and recurrent endometrial cancer with proficient MMR are low, ranging from 11% to 15% compared with up to 53% in recurrent endometrial cancer with dMMR.101–104 Several combination approaches are currently under investigation. Ovarian cancer and the majority of endometrial cancers (except dMMR subtype) are characterized by an immunosuppressive TME and therefore, targeting the immunosuppressive factors within the TME is a potential attractive approach.

To optimize biomarker data, it is important to have rational translational objectives and a standardized approach to collection of gynecological samples (including blood, tissue and ascitic fluid). One approach is to collect these samples at the time of surgery, which will be our focus here. Another approach is to use interventional radiology with image-guided approaches, especially in recurrent settings where surgery is not indicated. At time of surgery, tumor tissue tends to be bulky, especially in patients with ovarian cancer, but it is critical to collect viable tumor with surrounding tissue and avoid necrotic tumor tissue as much as possible. Transferring tissue into culture media might be better than normal saline especially if cell dissociation and single-cell analyses are planned.

Obtaining high-quality tissue is important especially with recent technological advances, such as single-cell RNA sequencing. Tissue obtained at the time of surgery can be immediately processed by dissociating cells, including both tumor cells and tumor-infiltrating immune cells for future analyses using RNA sequencing, single-cell sequencing and/or flow cytometry. Further, ascitic fluid, which is common in ovarian cancer, is an important resource to analyze the TME as it tends to be rich with immune and tumor cells, as well as cytokines. Collection of ascitic fluid can be done not only at the time of surgery but also from patients with recurrent disease with an abdominal catheter placed to drain ascites for palliative symptomatic indications.

Brain and central nervous system (CNS) tissue

Despite improvements in non-invasive techniques, such as advanced imaging and peripheral blood testing, surgical biopsy remains the gold standard to enable definitive diagnosis and management of CNS lesions. Tumor tissue within the brain can be acquired via a traditional open craniotomy or through the less-invasive stereotactic needle biopsy. Stereotactic needle biopsy is employed most often when lesion diagnosis is the sole goal, and resection is either not indicated or not feasible. This can be true, for instance, in cases of multifocal lesions, suspected small cell lung or lymphoma diagnoses, or in highly eloquent (the primary motor, sensory or visual cortices and their associated fiber tracts) or deep (medulla, pons, midbrain, thalamus and basal ganglia) locations within the brain where an open procedure may put the patient at high risk of postoperative neurological deficits or complications. Needle biopsy offers the advantages of small incision size, no craniotomy, theoretical reduced anesthesia time, and ability to target lesions within critical or difficult-to-access locations in the brain. However, it is limited by the size of the needle aspiration port (~1.0×0.1×0.1 cm, generally requiring aspiration of multiple cores), the lack of direct visualization of the needle tip, and slight inaccuracies inherent to frameless neuronavigation systems, with the resultant potential for sampling error.

For patients presenting with neurological symptoms attributable to mass effect, edema, or hydrocephalus (eg, speech/cognition deficits, motor weakness or neglect, and/or seizures), biopsy alone will not confer rapid symptom resolution or reduction/discontinuation of adjunct therapies, such as corticosteroids. In this scenario, open biopsy via craniotomy plus resection has the advantages of direct visualization of the tissue being obtained (especially useful for small or superficial biopsies), ability to make use of anatomical functional mapping of speech and motor cortices, as well as perform adjunctive tumor debulking or resection for local disease control and symptom relief (with resultant larger tissue yield for histopathologic or research purposes). Downsides to this approach are related to the more invasive nature of the procedure and risks therein (ie, larger incision, removal/replacement of bone flap, creation of larger corridor to subcortical lesions, and theoretical increased anesthesia time). In rare cases, open biopsy without resection can be used if technical limitations preclude needle biopsy.

Regardless of surgical technique, high resolution MR or CT images should be obtained in the preoperative setting for image guidance. These scans are imported into the neuronavigation software, where trajectory planning and biopsy sampling regions are selected. Major considerations for trajectory planning include incision location (facial avoidance for esthetic reasons), skull angle of entry (as perpendicular of a trajectory as possible to prevent drill skiving), and shortest possible path through the parenchyma that avoids transiting blood vessels, sulci and eloquent or deep brain regions, among others. For lesions that are contrast-enhancing (eg, high-grade gliomas or metastases), sampling is typically targeted to a region of confluent and/or nodular enhancement. For lesions that do not enhance, an intralesional region of T2-weighted-fluid-attenuated inversion recovery (T2/FLAIR) hyperintensity should be targeted. Necrotic or cystic regions of the lesion should be avoided. This is to maximize the amount of viable, diagnostic tumor tissue, given the expansion in molecular testing performed at most major centers.

Bone tissue

Tissue banking of primary bone sarcomas presents several unique challenges, and may also apply to metastatic carcinoma to bone for which immunotherapy trials may be valuable. Less than 3500 cases of primary bone sarcoma are diagnosed annually in the USA. Therefore, the rarity of primary bone tumors can lead to reluctance from the pathologist to release tissue for research and banking purposes as tissue supply may be exhausted for diagnostic purposes. Often, once the diagnosis is finalized, small amounts of tissue, if any, remain for banking purposes. This conflict can be further amplified when biopsy specimens are obtained by image-guided needle techniques, which are often performed in hard-to-reach anatomical locations such as the spine and pelvis. The literature supports an average number of five to six core samples (16–18 G) in solid, non-sclerotic soft tissue or bone lesions for diagnosis alone. In sclerotic medullary or cortical bone lesions, a larger gage (commonly 11 G) is more likely to provide usable tissue. Given the lack of stratifiable biomarkers or actionable targets in bone sarcomas, specimen acquisition exclusively for banking purposes remains controversial.

Most bone sarcomas are treated with preoperative chemotherapy or radiation leading to a paucity of treatment-naïve specimens. The standard of care for the treatment of the most common primary bone sarcomas most often consists of preoperative chemotherapy followed by local control.105 106 While the post-treatment surgical specimens provide a considerable volume of tissue, therapeutic manipulation likely changes the innate biology. Efforts need to be focused on acquisition of the pretreated tumor and other related correlative specimens (ie, blood and bone marrow).

Mineralized bone often requires decalcification for embedding and sectioning. Demineralization of specimens destroys some of the cellular and extracellular components, thereby rendering the tissue unsuitable for certain techniques, such as ISH or PCR genetic testing. Nucleic acid, in particular, is degraded by the acid demineralization process. In general, routine demineralization of bone specimens should be discouraged. When feasible, tissue should be divided into sections to allow processing of softer elements without decalcification.107 One proposed tissue allocation scheme is outlined in table 4.

Local therapy and assessment of response

The ability of systemic therapy to mediate important antitumor effects is now well established, adding immunotherapy to chemotherapy and targeted therapy as established therapeutic options. Well within the surgeon and radiation oncologist’s bailiwick is the notion of local therapies, allowing development of systemic responses.111 112 This has been best developed with the local injection of oncolytic viruses or alternatively immunostimulatory moieties, including TLR9 and STING agonists, developing a local interferon or interleukin-2 signal to promote immune reactivity. These ‘pre-tumor-infiltrating lymphocyte’ strategies will allow broader application of TIL to other tumor types by recruiting,113 even if transiently, T cells that can be sampled and administered to patients, similar to what has been successfully done in the setting of melanoma.114 115

Repertoire analysis

Diversity was originally thought to be due to stochastic processes but now is recognized to be due to combinatorial diversity of encoded germline elements with ‘trimming’ and N-region diversification (additional nucleotides).116 The B-cell receptor is expressed on the surface of B cells with the possibility of generating as many as 10²⁶ different antibody molecules.117 Adult B-cell numbers in any individual are actually <10¹² cells, with similar numbers predicted of T cells, bearing as many as 10¹⁵–10²⁵ different receptors.118–121 Effective immunity to tumors requires an intact adaptive immune system. We have made great progress in allowing assessment of the complexity of the T-cell receptor (TCR) repertoire in patients undergoing immunotherapy.122 A deep understanding of the breadth and depth of the immune repertoire has only come about as a consequence of modern technologies capable of assessing each of the now apparent rearranged receptors in totality. We should plan to carefully examine starting cell populations (from the tumor) and paired peripheral blood prior to, during and sequentially following therapy to examine the evolution of an immune response. Analysis of maturation markers, exhaustion markers, and expression of T-cell, NK and B-cell markers, as well as baseline diversity measures may provide valuable insight. The ability to sequentially harvest spatially distinct regions of the tissue/tumor will be increasingly important in the personalized evaluation of an individual’s own tumor.

Integration of tumor tissue biopsy and blood ‘liquid biopsy’ biospecimens

Tumor-reactive and neoantigen-reactive TCRs can be identified based on their frequency in the tumor by TCR sequencing without knowledge of cognate epitopes, and can be exploited for developing personalized neoantigen-based vaccines or TCR–gene therapy.123 TCR sequencing of both tumor tissue and peripheral blood allows us to identify TIL clonotypes (TIL–TCRs, both alpha beta and gamma delta), as well as B cells (IgH chain and kappa and lambda light chains) within the tumor and peripheral blood. The frequency of circulating TIL–TCRs can be monitored over the course of treatment to evaluate response to immunotherapy.124 Furthermore, high-dimensional single-cell profiling platforms (eg, 10× chromium) have been transformative for understanding complex cell populations. Combining TCR and scRNAseq should allow us to profile paired TCR αβ sequences, cellular proteins, gene function, and transcriptional signatures, and to identify novel transcriptional, genomic, and cellular markers of clonally expanded TIL–TCRs in peripheral blood, unveiling mechanisms of response and resistance to immunotherapy.