Clinical and Biological Prognostic Factors in Follicular Lymphoma Patients During Treatment

Introduction: Follicular lymphoma (FL) is an indolent yet heterogeneous B-cell lymphoproliferative disorder. Most people respond to treatment well. However, a particular group of patients has a poor prognosis, and these patients are dicult to dene. Patients and methods: We retrospectively analyzed FL patients treated at the University of Debrecen in the past 20 years. We investigated prognostic factors that may inuence the survival of FL patients. Results: We found a standardized uptake value (SUV)max cut-off value of 9.85 at the staging PET/CT to signicantly separate FL patients’ progression-free survival (PFS) (p=0.0003, HR: 0.2560, 95%CI: 0.1232-0.5318). Lymphocyte/ monocyte (Ly/Mo) ratio of 3.45 drawn at diagnosis also signicantly predicted PFS (p=0.0324, HR: 1.806, 95% CI: 1.051-3.104). Combining patients’ with staging SUVmax >9.85 and Ly/Mo < 3.45 a high-risk group of FL patients can be identied (p<0.0001, HR: 0.1033, 95%CI: 0.03719-0.2868). Similarly, a signicant difference was shown with a SUVmax cut-off of 3.15 at the interim PET/CT (p<0.0001, HR: 0.1535, 95%CI: 0.06329-0.3720). Combining patients with staging SUVmax >9.85 and interim SUVmax >3.15, a high-risk group of FL patients can be identied (p<0.0001, HR: 0.1037, 95%CI: 0.03811-0.2824). The PFS difference is translated into overall survival advantage (p=0.0506, HR: 0.1187, 95%CI: 0.01401-1.005). Discussion: Biological prognostic factors, such as the Ly/ Mo ratio, may improve the prognostic assessment of staging PET/CT. Nevertheless, into when using a combination of staging and interim SUVmax. We investigating additional biological prognostic factors while currently highlighting


Introduction
Follicular lymphoma (FL) is an indolent, germinal center B-cell-derived lymphoproliferative disease. [1] In general, FL is associated with the undue function of the proto-oncogene BCL2, which is activated by chromosome translocation of t(14; 18) (q32; q21). [2] FL is the most common non-Hodgkin lymphoma (NHL) in the Western world [3], representing 35% of all NHLs.
FL is a biologically and clinically heterogeneous disease with a wide variation in the outcomes of individual patients. Results for FL treatment have improved signi cantly, thanks to the introduction of anti-CD20 antibody rituximab, median overall survival (OS) of FL is approaching 20 years [4], but most patients eventually relapse. The ability to provide individualized treatment based on risk assessment of individual patients is the subject of ongoing research.
Classically, histological grade, tumor mass, Follicular Lymphoma International Prognostic Index (FLIPI) 1 (involvement of > 4 lymph node regions, elevated LDH, > 60 years of age, advanced stage and < 120 g/L hemoglobin) and − 2 (elevated beta-2 microglobulin (B2M), largest diameter lymph node greater than 6 cm, bone marrow involvement, < 120 g / L hemoglobin, > 60 years of age) are the parameters by which low-and high-risk patients are distinguished. [5] Several reports con rmed the unfavorable survival of FL patients who progress early after stopping treatment. Twenty percent of patients are expected to progress within 24 months. [6][7][8] Unfortunately, the range of these patients cannot be determined in advance.
Hence, we need to precisely predict patients' outcomes.

Patients And Methods
We retrospectively investigated our FL patients' prognostic factors treated between 2000 and 2020 at the University of Debrecen (UD), Department of Hematology. Factors that may have in uenced survival: histology, age, stage, sex, staging-, interim-and restaging standardized uptake value (SUV) max, presence or absence of B symptoms, bone marrow involvement, ECOG performance status, hemoglobin (Hgb), lactate dehydrogenase (LDH), B2M, absolute lymphocyte (Ly), and monocyte (Mo) count, lymphocyte/ monocyte (Ly/Mo) ratio, the progression of disease within 24 months (POD24). POD24 was calculated from the time of diagnosis until progression.
This retrospective analysis was approved by The Regional and Institutional Research Ethics Committee of the University of Debrecen.
The patients were treated according to the current guideline of the Hungarian Society of Hematology and Transfusion. The patients consented before treatment initiation to collect and publish their data retrospectively according to the Declaration of Helsinki. PET/CT was used routinely as an imaging modality at the UD since 2008. However, it was not routine in the interim and restaging setting. A staging PET/CT was done for every patient after a histological diagnosis of disease, unless there was a clinical urgency or the patient was treated primarily in another institution with no access to PET/CT scans.
Interim PET/CT was performed after 3 cycles of immune-chemotherapy, while restaging PET/CT was done after 4-6 weeks of completion of induction treatment.
Patient outcome was analyzed by progression-free survival (PFS) and overall survival (OS). PFS was calculated from diagnosis to June 2020, progression, histological transformation, or death, while OS was calculated from diagnosis to June 2020 or death. Survival was estimated based on the Kaplan-Meier method. Factors that could affect survival were evaluated based on a multivariable Cox regression model with a stepwise backward variable selection approach to obtain hazard ratio (HR). Comparison of the survival curves was based on the Log-rank test.

Results
We investigated 211 FL patients with a median age of 53 years. The majority of them had B symptoms. Grade 1 and 2 histology occurred in about -of the cases. The majority of the patients were diagnosed with advanced-stage disease. POD24 cases were found in less than of the cases. Out of FL patients diagnosed since 2008, we found 115 accessible staging PET/CT-s. Sixty-ve of them had an interim, while 80 had a restaging scan. First line treatment was dominated by anti-CD20 antibody, rituximab, whereas the major chemotherapy backbone was -CHOP (cyclophosphamide, doxorubicin, vincristine, prednisolone)(-like), -CVP (cyclophosphamide, vincristine, prednisolone), or bendamustine. Table 1.
The interim PET/CT, Ly, Mo were excluded because they are related to another variable. We put the rest into a multivariable COX regression; staging PET/CT and Ly/ Mo ratio remain an independent prognostic factor for PFS. Table 2. A total of 115 patients had PET/CT as staging imaging modality since 2008. We investigated SUVmax values, and with the ROC curve, we found a cut-off value of 9.85 (range: 5.6 -28.9) to divorce survival signi cantly. When illustrating the patients' PFS, a signi cant difference was found when using the de ned SUVmax cut-off value (p=0.0003, HR: 0.2560, 95%CI: 0.1232-0.5318). The median PFS of FL patients with SUVmax> 9.85 was 51 months, while median PFS with SUVmax <= 9.85 was not met. Fiveyear PFS was 85.29 vs. 48.25%. When presenting these patients' overall survival, the difference is not signi cant (5-year OS: 96.87 vs. 85.73%). Figure 1. Histology was done from the most accessible site since the histology result is needed to order a PET/CT imaging. Patients with a relatively high staging SUVmax or with grade 3a histology received R-CHOP protocol.
A total of 169 patients had an accessible blood count in the medical record system. The cut-off value for Ly/ Mo was found to be 3.45 with the ROC curve. When illustrating the PFS of these patients using the de ned value, we found a statistically signi cant difference (p=0.0324, HR:

Discussion
The weakness of FLIPI is that it has been determined on the one hand using retrospective data. On the other hand, it does not de ne a treatment indication and, like International Prognostic Index (IPI), it represents few high-risk patients. FLIPI2 is designed to protect against all this. Treatment is still determined based on high tumor mass according to the criteria of the Groupe d'Etude des Lymphomes Folliculaires (GELF) [9] or the British National Lymphoma Investigation (BNLI) [10,11].
Initial total metabolic tumor volume (TMTV) measured by PET/CT was strongly correlated with survival in FL patients who received R-CHOP without maintenance treatment. [12] Patients with TMTV> 510 cm3 had a signi cantly less favorable 33% 5-year PFS compared with 65% 5-year PFS in patients with TMTV <510 cm3. The 5-year overall survival (OS) was 85% and 95%, respectively. We found that staging SUVmax was prognostic for PFS with a cut-off of 9.85 based on 115 FL patients' data. The survival advantage, however, did not translate into signi cant OS differences. Our 5-year PFS (85.29 vs. 48.25%) and OS (96.87 vs. 85.73%) results with staging SUVmax were similar to that found with TMTV.
A low lymphocyte count is an adverse prognostic factor not only in Hodgkin lymphoma (HL) [13] but in FL, also [14] and may relate to the patient's immunity. In contrast, monocyte count could relate to the tumor microenvironment. [15] Elevated monocyte count was associated with a poor prognosis. Ly/ Mo count is also reported to be a prognostic factor in HL. [16,17] In our multivariate analysis Ly/ Mo ratio was prognostic in our FL dataset. The cut-off value of 3.45 was similar to that found in our HL population [18] and found in an Italian [19] and Hong Kong FL dataset. [20] Ly/Mo ratio is not standardized since the results are scattered, however when combined staging SUVmax with Ly/Mo ratio p-value improved both for PFS and OS. Further, larger studies are warranted to specify prognostic value of Ly/ Mo ratio in FL.
The PET/CT is a standard imaging method for response evaluation in FDG-avid lymphomas, including FL. However, prognostic evaluation in FL by PET/CT is not widely established. In a study published in 2019, a survey of 33 (!) FL patients found that interim PET/ CT done after 3 or 4 cycles of rst-line treatment were predictive of progression-free survival (PFS). [21] A meta-analysis published in 2016 found one trial that reported a positive correlation between positive or negative interim PET/CT results and PFS, while two studies reported a negative correlation. [22] Our results con rm a positive correlation.
The interim PET/CT scan of 65 FL patients, performed after three cycles of immunochemotherapy, and a cut-off of 3.15 SUVmax showed signi cant PFS survival bene t. However, the difference was not signi cant in terms of OS, which may be explained by extensive and effective relapsing treatment options. [23] A recent paper, published in 2019, reported 84 FL patients, of whom 59 had a baseline and 24 an interim PET/CT scan. Similarly to our results, they found a positive correlation between the baseline SUVmax value of 10.44 and PFS. However, survival difference was not signi cant in terms of OS. Interim PET/CT results interpreting them as "positive /negative", "Deauville score 1-3 and 4-5" and "ΔSUVmax (change of SUVmax from baseline to interim point)" were neither prognostic for PFS nor OS. [24] Nevertheless, when we combined staging and interim SUVmax, a patient group with a signi cantly poor prognosis could have been identi ed. The signi cant survival disadvantage in PFS is translated into OS difference, also. Half of these patients progress within 21 months, determining POD24 patients after three cycles (and practically three months) of treatment. POD 24 also translates to OS disadvantage. We believe this is an unmet medical need. Patients belonging to this group should change therapy with a more aggressive therapeutic approach -if possible -(compared to the indolent clinical characteristic of FL) after getting an unfavorable interim PET/CT scan result. It would be fortunate to predict even earlier these adverse cases, possible at the time of diagnosis. Therefore, these patients could get more aggressive treatment or vice versa: should the good prognostic group get more "permissive" therapy e.g., leaving maintenance therapy at the time of ongoing COVID pandemic and thus moderating B cell depletion [25] or just because the patient requires continuous granulocyte colony stimulating factor support?
Restaging PET/CT results are reported in several ways to predict PFS. PET/CT done three months after completion of induction treatment was also an independent prognostic factor. [26,27] A meta-analysis of large multi-center trials veri ed in 2014 that a negative PET/CT done after six cycles of induction treatment was prognostic for both PFS and OS. [28] Our results of restaging PET/CT scans were also prognostic for PFS at a cut of 2.45 SUVmax. However, the results were not signi cant again in terms of OS.
Based on these, we believe that although biological prognostic factors, such as the Ly/ Mo ratio, are essential because they may improve the prognostic assessment of staging PET/CT. We can demonstrate the difference in overall survival using a combination of staging and interim SUVmax. Besides, this can provide more signi cant help than the individual use of staging or interim SUVmax. For all these reasons, we consider it necessary to investigate additional biological prognostic factors while currently highlighting PET/CT' s role in FL.