How to prepare a molecular tumor board session: the minimum data worth bringing

A molecular tumor board has very limited time per case, and that time is consumed very differently depending on how the information arrives. When a case is presented with the data in order, the session is devoted to what adds value: interpreting, comparing and deciding. When basic elements are missing, much of the discussion becomes a reconstruction of the file out loud, and the case is frequently postponed.

The purpose of this review is not to impose a form but to describe the minimum set of information that, in practice, conditions how a molecular profile should be read. Every service has its own protocols and ways of working; what follows is a general checklist, not a treatment recommendation.

The sequencing report should arrive in full, including its technical section. The list of alterations alone is insufficient because it does not allow an assessment of what could be detected and with what reliability. At least these elements are worth having in view:

• Panel used and genes included, to know what falls outside the scope of the study.
• Types of alteration the assay can detect: point variants, insertions and deletions, copy number variations, fusions and rearrangements.
• Sample quality parameters: tissue type, tumour cellularity, coverage and allele fraction of the variants detected.
• Aggregate biomarkers such as TMB and MSI, with the method and units used, since they are not directly comparable between platforms.
• Detail of the fusions detected: genes involved, breakpoint and, where available, whether the reading frame is preserved.
• Classification of the variants and the framework used, stating expressly whether it is AMP/ASCO/CAP, ESCAT or another.
• Date of the report and version of the knowledge bases used.

One item often forgotten: if there are previous molecular studies for the same patient — an earlier biopsy, a liquid biopsy determination — they are worth bringing too. Comparing different moments in time can explain findings that, seen in isolation, are puzzling.

Performance status is one of the data points that most strongly conditions the viability of any therapeutic option, and yet it is frequently missing from the case presentation. The widely used ECOG scale summarises the patient's autonomy in daily activity and is a reference element in most clinical trial eligibility criteria and guideline recommendations.

The value should be recent and dated. An ECOG recorded weeks earlier may not reflect the patient's situation at the time of the session, and that difference can change the conversation entirely. Alongside performance status it helps to have the current clinical situation: relevant symptoms, recent course and significant laboratory results.

Treatment history is essential to interpret the molecular profile. The same alteration reads differently if it appears before any treatment or after progression on a targeted therapy, where it may correspond to an acquired resistance mechanism.

For each previous line it is worth having the regimen given, start and end dates, the reason for discontinuation, the best response obtained and relevant toxicities, particularly those that would limit the use of certain drug families in future. Previous radiotherapy and surgery are also part of the picture.

The presence or absence of central nervous system involvement is a data point with direct consequences, because drugs' ability to reach effective concentrations in the CNS varies considerably and because many clinical trial inclusion criteria address it explicitly.

It is worth being more precise than a yes or no: whether disease is active or treated and stable, which local treatment was given and when, and the date of the most recent imaging supporting that assessment. Where no CNS study has been performed, it is better to state so than to leave the field blank: absence of a study and absence of involvement are not the same thing.

Relevant comorbidities — renal and hepatic function, cardiovascular disease, pulmonary disease, history of autoimmune disease — condition both eligibility and the monitoring required. Equally important, and often more poorly documented, is the complete list of concomitant medication.

The list should include both chronic and occasional medication, and it is worth extending it to products patients do not always mention: over-the-counter drugs, supplements and herbal products. Many targeted therapies are metabolised through pathways susceptible to interaction, and some common drugs affect the QT interval or the absorption of oral treatments.

It also helps to record allergies, documented intolerances and, where relevant to treatment logistics, the patient's social and support situation. A theoretically suitable option may not be viable if it requires a frequency of visits or monitoring that the patient's circumstances do not allow.

No case arrives complete. The difference between good and poor preparation is not the absence of gaps but whether the gaps are identified. Explicitly noting which data point is missing, why, and who can supply it turns an omission into a concrete task, and prevents the session from reaching a conclusion built on an undeclared assumption.

It is equally useful to record the questions the case puts to the board before the session. Stating what is expected to be decided guides the discussion and makes it easier for the conclusion to be recorded verifiably afterwards, when someone reviews the case with hindsight.

Gathering and verifying all this information is, in most services, manual and repetitive work. That is where Sphera OncoCore comes in: it structures the available data, explicitly flags what is missing and documents the reasoning with its sources, so the session can be devoted to the clinical decision.