Role of Surgical Resection in Primary CNS Lymphoma: A Resurrected Discussion

The outcome of immunocompetent patients treated for primary central nervous system lymphoma (PCNSL) has significantly improved with the use of upfront high-dose methotrexate (HD-MTX)-based therapy, and PCNSL is now considered a potentially curable disease. As in other lymphomas, the diagnostic approach has been to forgo surgical resection and to pursue tissue biopsy alone. This is based on the well-studied fact that these tumors are usually diffuse and multifocal and are often associated with ocular or leptomeningeal involvement. In addition, resection can lead to surgical morbidity and delay the start of HD-MTX, which is the mainstay of treatment for these cancers.

The authors of a recently published unplanned subgroup analysis of the randomized controlled German Primary CNS Lymphoma Study Group 1 trial (G-PCNSL-SG-1) challenge this paradigm; their study showed significantly shorter median overall survival (OS) and progression-free survival (PFS) in patients who underwent biopsy only compared with those who had a subtotal or gross total resection.[1,2] These findings are certainly provocative and have led to a lively discussion about a topic that had previously been considered as “written in stone.” The review article by Dr. Bierman in this issue of ONCOLOGY comprehensively discusses the arguments for and against surgical resection in newly diagnosed PCNSL.[3] The overall conclusion of his review article is that even if some patients did benefit from surgical resection, the actual practice pattern is not likely to change, mainly because the diagnosis of PCNSL is mostly made in retrospect.

As the authors of both articles correctly point out, there will likely not be a randomized controlled study to answer the question of a potential advantage for upfront surgical resection over biopsy for a select group of patients with PCNSL. Such a study would need to randomize only patients with potentially resectable PCNSL to either resection or biopsy. However, the diagnosis is usually unknown until the final pathology is available. Repeat resection after initial biopsy would therefore delay systemic therapy. Moreover, the incidence of PCNSL is comparatively low, further complicating the ability to conduct any randomized studies. Given the limitations of the German PCNSL Study Group subgroup analysis, including its unplanned retrospective design, and possibly the selection of patients with a more favorable biology, it is not possible to derive firm conclusions from it. It is worth noting that, although median OS and median PFS were improved, long-term survival data showed virtually no difference between the cohorts at time points beyond the 6-year mark, suggesting that the extent of surgery likely has no effect on long-term survivorship.

Nonetheless, the analysis encourages a re-appraisal of whether surgical cytoreduction could in theory improve outcome in select patients. Considering some of the very basic principles of chemotherapy (and HD-MTX–based chemotherapy is the only treatment that leads to long-term progression-free intervals and potential cures in PCNSL), there is a clear argument for optimal cytoreduction prior to starting systemic treatment. The fewer tumor cells that are present at the time chemotherapy starts, the fewer cycles are needed to induce a complete remission, based on the fractional (“log-linear”) cell killing of cytotoxic drugs. This raises the question of whether surgery might provide some patients who can only tolerate a limited number of cycles of HD-MTX with the chance to achieve a complete response, which is a prerequisite for long-term survival and which they would otherwise not have. Another theoretical advantage of upfront surgical cytoreduction would be a reduced likelihood of chemotherapy-induced mutations, as well as a reduced number of genetically unstable cells that make the development of chemotherapy resistance more likely.

What did the data provided by the subgroup analysis of the German PCNSL trial add to our knowledge, and what conclusions should we derive from this study? Challenging existing paradigms is critically important if we are to eventually improve outcomes for these patients. The data presented by the German PCNSL Group may inspire a re-evaluation of the role of surgical resection in patients with PCNSL if resection is feasible and does not delay systemic therapy. However, since this was a retrospective subgroup analysis, the data are not sufficient to identify extent of resection as an independent prognostic factor. Collecting data on resection status in future prospective trials would certainly be reasonable and might help shed more light on this question. Nonetheless, without a definitive prospective clinical trial that randomizes potentially resectable patients to biopsy vs resection, this question will unfortunately remain unanswered. For the practicing medical or neuro-oncologist, the treatment approach would currently not change, given that systemic therapy should be started as soon as the diagnosis of PCNSL is made and the patient is stable from a neurosurgical perspective. In most cases, one would not refer the patient back to the surgeon for additional debulking.

Financial Disclosure:The author has no significant financial interest in or other relationship with the manufacturer of any product or provider of any service mentioned in this article.

References:

1. Thiel E, Korfel A, Martus P, et al. High-dose methotrexate with or without whole brain radiotherapy for primary CNS lymphoma (G-PCNSL-SG-1): a phase 3, randomized, non-inferiority trial. Lancet Oncol.2010;10:1036-47.

2. Weller M, Martus P, Roth P, et al. Surgery for primary CNS lymphoma? Challenging a paradigm. Neuro Oncol. 2012;14:1481-4.

3. Bierman P. Surgery for primary central nervous system lymphoma: Is it time for reevaluation? Oncology (Williston Park). 2014;28:632-7.