Case planning involves co-ordination with the multidisciplinary team including oncologists, nurse specialists, research nurses, pharmacists and radiologists (including interventional radiology). Several key considerations for case planning will be addressed.

Case selection

Trial protocols vary and some may stipulate that ultrasound-guidance should be used in all cases, whereas others may have a more selective approach. Certain protocols may permit ‘free-hand’ (non-image guided) injection of tumours if they are superficial, visible (or easily palpable), and well clear of major vascular structures. The need for image guidance is decided at a multi-disciplinary meeting (MDT). Some trials inject both the primary site in the head and neck for e.g., hypopharynx or oropharynx (performed by head and neck surgeons endoscopically) and the nodal sites of disease. As part of clinical trials, eligible patients would have had baseline cross-sectional imaging (usually CT, with or without MR) which allows tumour characteristics to be assessed in terms of extent, presence of necrosis, local invasion and vascular encasement. Absolute contraindications (identifiable on imaging) include intracranial invasion and/or tumour encasing the carotid arteries (due to the risk of stroke or carotid blow-out), jugular veins or thoracic great vessels (including the braciocephalic and subclavian arteries and veins). Wenig et al. [11] have reported the occurrence of fatal complications with subsequent modification of trial protocols to exclude tumours directly invading or near the common and internal carotid arteries, with no additional treatment-related cerebrovascular effects.

Using ultrasound guidance confers several advantages. Firstly, targeting deeper lesions that cannot be palpated and for which target identification in the absence of ultrasound would be challenging. Secondly, within the context of necrotic tumours, ultrasound-guidance facilitates targeting of the solid component of the tumour which is likely to increase efficacy of treatment. Thirdly, ultrasound ensures that administration of the drug is performed safely when target lesions are near critical structures including blood vessels. Furthermore, ultrasound-guidance allows assessment of tumour volume and additional parameters, specifically the presence of necrosis (which may suggest treatment response) and doppler flow, and thus can provide a guide with respect to treatment response (although CT is classically used for RECIST, iRECIST and IT RECIST measurement). Regions well suited to US guided injection include the head and neck, scalp, pleura, limbs, and superficial masses in the torso. Generally, tumours located deeper in the torso, for example in the lungs or deep to the peritoneum, require CT guided injection.

Pre-procedure work-up and consent

Some trial protocols will require participants to have pre-procedure blood tests (including full blood count and International Normalised Ratio). Baseline cross-sectional imaging is mandatory and would normally include a CT of the neck, chest and abdomen. MRI may be used in certain cases, for example, to stage tumours that are poorly visualised or obscured by artefact (dental amalgam). Informed consent would need to be obtained from the patient for participation in a trial and, specifically, for the potential complications of the intratumoral injection, which vary in scope, frequency and severity depending on the drug(s). The patient will need to be provided with information about the nature of the trial, drug-specific complications (which should be provided by the pharmaceutical companies), the requirements for further injections and imaging follow-up. Clinical and research nurse specialists have an important role in supporting patients throughout the process, providing information and addressing any concerns that patients may have.

Organisation and planning

Close collaboration with both pharmacy and the research team is important because intratumoral drug comes pre-prepared and delivery can be time-sensitive, such that delays may lead to expiry of the investigational medical product. Most of the drugs have a short time to expiry so are prepared on the day at the pharmacy. In some cases, such as with the drug IXOVEX (an adenovirus), the time to expiry is only thirty minutes, and therefore the drug needs to be prepared in the ultrasound room where the injection is being performed. Availability of the drug needs to be confirmed prior to the patient’s appointment time to ensure that the drug is available, and an ultrasound slot is not wasted with the associated inconvenience caused to healthcare staff and patients. The dates and times of the treatment is pre-determined by trial protocol and, so, generally cannot be easily altered to meet clinician and/or patient preferences or fixed radiology lists, although allowances should be made where possible. Ideally, there should be at least two or three radiologists trained to perform the procedures who are familiar with the trial protocol. This allows for continuous service provision, including annual leave cover, and ensures that appointment times and adherence to trial protocols is maintained.

Ideally, a double ultrasound slot (at our institution, typically fifteen minutes for a diagnostic and thirty minutes for a procedural case) should be booked as this allows time to discuss and formally consent the patient for the procedure, as well as adequate time to perform the injection without haste. It is preferable to have the slot at the end of an ultrasound list to minimise contamination of the local environment by spillage or aerosolization of the injectate.

Pre-procedural preparation

Patients generally prefer a small team of familiar operators. Given that DIT are infrequently performed and currently only as part of a clinical trial, it is preferable to both patients and medical staff that a small team of familiar operators well-versed in trial protocols and technical procedural aspects performs the procedure. These operators should receive formal training of “Hazardous drug handling” as some of the drugs used e.g. Tiganol Tigate are highly irritant. At our institution, a group of three operators perform these injections on a regular basis and are named on trial delegation logs. Other members of the research team who are highly familiar with DIT and their particulars are also routinely present during the procedures, adding a further level of oversight and safety.

Suitable personal protective equipment, specifically gloves, a gown/apron and safety goggles need to be available. Sometimes, trial protocols require that patients have immediate pre- and post-procedure bloods, electrocardiography, and monitoring of blood pressure and other vital signs before and during the injection. Rarely, sedation with midazolam may be required, particularly if the patient is anxious or multiple prior attempts have been performed that have resulted in local inflammation and/or discomfort. If the patient is having radiotherapy treatment, there will need to be consideration of the timing of DIT with respect to radiotherapy planning. Ideally, DIT should be administered before or during the first few weeks of radiotherapy due to increased risks of radiodermatitis and mucositis in the second half of radiotherapy which may increase the risk of ulceration or infection.

Before each injection, an assessment with ultrasound may be performed to assess the target lesion(s), to obtain measurements (length, width and height [cm] and volume [cm3]) and evaluate for any treatment response (if injections have been previously performed). This assessment will inform the operator as to the required dose and volume of drug to be injected, as this should be adjusted if the shrinks. Pertinent data, as in Table 1, is recorded for trial purposes. Where multiple tumours are present, the preferred injection target is agreed between the research team and radiologist; whichever target is regarded as the safest, technically easiest, and largest tumour (in that order of priority). Ideally, measurements should be obtained in three perpendicular dimensions ensuring that the maximum dimension in any plane is recorded. However, response evaluation will usually be better assessed on contrast enhanced CT which enables more objective measurements and staging of all lesions, an assessment for abscopal effects and application of the Response Evaluation Criteria in Solid Tumours version 1.1 (RECIST1.1) [12,13,14]. Identification/land-marking the injection target can be important with clustered/ multifocal disease where the target may be hard to discern and to ensure that the same lesion is injected on subsequent visits (if that is stipulated per protocol). Of note, if there is a response to treatment, the target lesion can become harder to measure or to inject on subsequent visits due to tumour shrinkage.

Table 1 Example of a reporting template and data recorded for trial purposesTechnical procedural aspects

A WHO checklist should be completed, and aseptic technique observed for all procedures. Once the target lesion has been identified, the most appropriate approach will need to be considered and the patient positioning optimised. Ideally, as with any ultrasound-guided procedure, the radiologist, probe and the ultrasound screen should be arranged to make the procedure comfortable and easier to perform for the radiologist which, in turn, will ensure optimal delivery of the injectate. In most cases, local anaesthetic is administered from the skin surface to the tumour edge (5 ml, 1% lignocaine); however, this may not always be required, for example if the target is easily accessible or if the patient can tolerate the procedure and is reluctant to have local anaesthesia. The radiologist is provided with a syringe (typically 5 ml, Leur lock), containing the pre-dosed, pre-prepared injectate, as supplied by pharmacy. Prior to administration, it is important to ensure that no residual air is expelled from the syringe containing the pre-prepared injectate as this could turn the injectate (containing genetically modified microorganisms) into an aerosol. The volume/dose of injectate is predetermined by the research team and pharmacy, and will be governed by parameters including patient size and tumour size. The needle (typically 23 Gauge [blue]) should initially be directed into the deeper aspect of the target lesion, and this aspect injected first, as in Fig. 1, before the more superficial components of the lesion are injected (usually into four quadrants of the tumour, if feasible). This ensures that any initially expelled gas does not obscure the deeper parts of the lesion which could be problematic if the more superficial components of the lesion were injected first. The edges and peripheries of the lesion will need to be injected due to higher immune cell accessibility as the periphery of the tumour often has a higher density of blood vessels compared to its core. This is particularly relevant in the case of large centrally necrotic tumours where the solid and vascular elements of the tumour that are generally at the periphery are targeted at several (up to four) points. This increased vascularity allows better access for the immune cells and therapeutic agents, potentially enhancing the effectiveness of treatment. The periphery generally has a more oxygen rich environment than the core, which can support more active and effective immune responses. The core of a tumour often contains dense extracellular matrix and fibrotic tissue, which can physically prevent the penetration of the drug. Spillover of drug at the periphery of tumour and into healthy tissue may trigger a systemic immune response. These reasons make the periphery of the tumour a critical target for DIT, potentially increasing the overall efficacy of the immunotherapy.

Fig. 1

B-mode ultrasound images of a neck tumour (*) targeted for DIT. (a) Initial pass with needle into the deeper aspect of the target lesion. (b) hyperechoic gas bubbles on injection from pre-prepared injectate. (c, d) Superficial passes into the target lesion

The treated neck can be particularly challenging to address, as patients will often have had previous radiotherapy and surgery which results in fibrosis, scarring and a firm, sclerotic texture to the soft tissues. This may require some force to inject the treatment into the target lesion, and once the lesion is injected and the needle has been removed, the puncture site should be covered with the ultrasound probe to enable dispersion of the injectate in the target and avoid jets of injectate being expelled from the skin.

Post-procedural considerations

Following the procedure, ultrasound should be performed to ensure that there is no haematoma that may be suggestive of vascular injury. The patient should be transferred to a designated clinical area for monitoring in an armchair (or daybed if available) for a pre-specified period (according to trial protocol) and provided with steroid cover to prevent severe inflammatory reactions and oral analgesia, if any pain develops as the local anaesthetic wears off. In some cases, the injections are painful enough to require oral or intravenous opioids for e.g., fentanyl. The ultrasound measurements of the target lesions, the technical aspects of the procedure and the presence or absence of any immediate complications should be clearly documented in a report on the radiology information system. The ultrasound room will need either a standard or a deep-clean depending on how hazardous the exact type of drug injected is.

Directly-injected therapies: our local experience

The approach to image guided DIT described in this article was developed through our early experience at our local cancer centre. For context, in between January 2018 and January 2024, 71 patients have undergone DIT at our institution, in the context of 9 clinical trials. Of these, 21 cases (30%) were performed with ultrasound guidance, 11 (50%) of which were neck masses, with a median number of 7 injections performed under image-guidance (range 2–25), totalling more than 150 procedures.

The procedures were all well-tolerated and there were no immediate technical complications related to the needle insertion and injectate delivery. As several of the trials are ongoing, outcome data, including delayed systemic symptoms related to the oncolytic viral treatments will be published in due course.

Figures 2 and 3 show sequential images of two patients treated for a maxillary adenocarcinoma and a parotid oncocytic carcinoma, respectively.

Fig. 2

Axial and coronal contrast-enhanced CT images of locally advanced adenocarcinoma involving the left maxilla, mandible and orbit. (a, b) Recurrence several years after extended left maxillectomy, before DIT (c, d) tumour shrinkage 6 weeks after initial DIT session. (e, f) further reduction in size of the tumour 3 months after initial DIT session

Fig. 3

Axial contrast-enhanced CT images of a serially injected necrotic right level II lymph node related to a right parotid oncocytic carcinoma. (a) Baseline imaging one month prior to the first DIT session. (b) Reduction in size of the metastatic lymph node after four DIT sessions. (c, d) Ongoing reduction in size after three subsequent DIT sessions (12 months from baseline)