Feasibility of Primary Tumor Culture Models and Preclinical Prediction Assays for Head and Neck Cancer: A Narrative Review

Abstract

Primary human tumor culture models allow for individualized drug sensitivity testing and are therefore a promising technique to achieve personalized treatment for cancer patients. This would especially be of interest for patients with advanced stage head and neck cancer. They are extensively treated with surgery, usually in combination with high-dose cisplatin chemoradiation. However, adding cisplatin to radiotherapy is associated with an increase in severe acute toxicity, while conferring only a minor overall survival benefit. Hence, there is a strong need for a preclinical model to identify patients that will respond to the intended treatment regimen and to test novel drugs. One of such models is the technique of culturing primary human tumor tissue. This review discusses the feasibility and success rate of existing primary head and neck tumor culturing techniques and their corresponding chemo- and radiosensitivity assays. A comprehensive literature search was performed and success factors for culturing in vitro are debated, together with the actual value of these models as preclinical prediction assay for individual patients. With this review, we aim to fill a gap in the understanding of primary culture models from head and neck tumors, with potential importance for other tumor types as well.

Organotypic Co-Cultures as a Novel 3D Model for Head and Neck Squamous Cell Carcinoma

“…Conventional mouse xenograft models [ 38 ] as well as patient-derived xenograft (PDX) models [ 39 ] show limitations due to the adaption of human tumor cells to the murine microenvironment and the replacement of human stromal components by murine equivalents [ 40 , 41 ]. The costly and time-consuming implementation of other models such as humanized mouse models [ 42 ] and orthotopic PDX models [ 43 ] also remains sparse for HNSCC [ 44 ].…”

Sponge-supported cultures of primary head and neck tumors for an optimized preclinical model

“…As HNSCC patients need to start their treatment within 5–6 weeks after diagnosis, in vitro screening should be performed preferably within 1–2 weeks to guide decision making. The HDRA assay has led to a culture model most comparable to the in vivo tumor with successful culture rates, with a read-out after 7–8 days, and the best correlation between in vitro and in vivo treatment responses [ 7 ]. Despite these promising results, the HDRA is not taken into routine clinical practice, likely for various reasons.…”

“…In order to potentially use the histoculture as preclinical individual drug-response assay, a larger window in terms of tumor viability and proliferation at day 7 may be required to assess the efficacy of drugs and/or irradiation. We choose a 7-day read-out in line with earlier HDRA studies in HNC [ 7 ], showing good culture success rates and relatively good correlation to the clinic. It could be that shortening of the culture period from 7 to 3–5 days demonstrates higher viability and proliferation rates.…”

“…Our group performed a systematic review on primary HNSCC culture models and their ability to predict clinical response. We found that the most successful culture rates and best clinical correlations are obtained with the sponge-gel-supported histoculture, used as the histoculture drug response assay (HDRA) [ 7 ]. Leighton et al developed this technique in 1951 in an effort to resemble a patient's tumor more accurately [ 8 ].…”

“…Robbins et al ., however, were the first to test the HDRA technique on HNSCC tissue in 1994 [ 18 ]. Later, the HDRA model was adopted by several authors for preclinical chemosensitivity correlations in patients with head and neck cancer [ 7 , 19 – 22 ].…”