The outlook for large B-cell lymphoma that fails to respond to chemoimmunotherapy was, until recently, very poor. The pooled SCHOLAR-1 analysis found that only a minority of such patients responded to the next line of conventional treatment and that median survival was about six months (Crump et al., 2017). Autologous T cells engineered to express a chimeric antigen receptor (CAR) against CD19 changed that picture. Three products, axicabtagene ciloleucel, tisagenlecleucel and lisocabtagene maraleucel, each produced complete remissions in a substantial fraction of patients with refractory disease in single-arm registration trials (Neelapu et al., 2017; Schuster et al., 2019; Abramson et al., 2020).

The field began with a single case. Kochenderfer et al. (2010) reported regression of advanced follicular lymphoma and prolonged loss of normal B cells in one recipient of anti-CD19 CAR T cells, and the report helped to justify the trials that followed. Case reports retain that function. They are the natural vehicle for a first observation in a rare lymphoma subtype, an unexpected late toxicity, a response after an earlier cellular therapy has failed, or experience with a locally manufactured product. Their weaknesses are equally well known (Vandenbroucke, 2001; Nissen and Wynn, 2014).

This article offers a method for writing one such report well. It concerns the case in which a patient with refractory lymphoma enters a remission that lasts after a salvage cellular therapy, and it sets out what must be documented for readers to accept each element of that description. No case is presented. The procedure is proposed, and it has not been validated against published reports or applied by the author to a particular patient.

Salvage cellular therapy in lymphoma covers autologous CAR T cells, which are now the most frequent form, and allogeneic hematopoietic cell transplantation, which remains an option after CAR T-cell failure or where no product is available. Both rest on immune cells as the active agent, and both are followed by a period in which the timing of events is essential to interpretation (June and Sadelain, 2018). A report must therefore be built around dates and intervals.

The general standard for case reports is the CARE guideline, a consensus checklist of thirteen items running from title to informed consent (Gagnier et al., 2014), accompanied by an explanation and elaboration paper with worked examples (Riley et al., 2017). CARE is generic. It says that diagnostic assessment, intervention and outcome should be described, and leaves the specialty to say how. For lymphoma treated with cellular therapy, the content can be supplied from four published standards: the World Health Organization classification for diagnosis (Swerdlow et al., 2016), the Lugano classification for staging and response (Cheson et al., 2014), its companion imaging consensus (Barrington et al., 2014), and the consensus grading of cytokine release syndrome and neurotoxicity (Lee et al., 2019). Table 1 maps these onto the CARE items.

The procedure follows the clinical course in order. Figure 1 shows the phases that the report's timeline should cover, and the six steps below describe what to record in each.

Publication of a case requires the written consent of the person described, or of the next of kin where that is not possible, and journals following the recommendations of the International Committee of Medical Journal Editors will ask for confirmation of it (ICMJE, 2023). Rare lymphomas treated at a small number of centers make individuals easy to recognize. Exact dates can be replaced by intervals from infusion, and details that do not bear on the clinical argument, such as occupation or place of residence, should be left out. CARE also invites a short account in the patient's own words (Gagnier et al., 2014), which is seldom included and is especially apt after a treatment as demanding as this one.

Presentation follows from the procedure. A timeline figure, a table of prior therapies and a panel of paired images carry most of the evidence, and the text can then be brief. The discussion should place the case against the registration trials without suggesting that it adds to their estimates of efficacy.

The procedure set out here is a proposal assembled from reporting guidelines and disease-specific standards. It has not been applied to a set of published case reports to test whether it discriminates well-documented from poorly documented ones, nor has agreement between different users been examined. Both would be straightforward to study. The standards on which it depends are themselves revised from time to time, and authors should check for current versions of the classification and grading systems.

Two broader cautions apply. Cases with good outcomes are more likely to be written up and accepted than those with poor outcomes, so the published record of single cases overstates what a treatment achieves. And a remission in one patient, however well documented, gives no estimate of the probability of remission in the next. Readers who need that estimate should turn to the trials and to registry data. The value of the case report lies elsewhere, in showing that something can happen and in describing precisely how it did.

A report of lasting remission after cellular therapy for refractory lymphoma makes four claims: that the disease was refractory, that a defined product was given, that a complete response followed, and that it has lasted. The six steps proposed in this guide attach to each claim the evidence a reader needs, using the CARE checklist as the frame and the Lugano and consensus toxicity scales as the measures. Followed with care, they should yield reports that are shorter, more comparable and more useful to clinicians who face a similar patient. Whether they do so in practice has still to be shown.

To be declared by the author(s) before publication.

To be declared by the author(s) before publication.