Fluorodeoxyglucose-Positron Emission Tomography in Relapsed/Refractory Hodgkin Lymphoma: A Practical Approach

Positron emission tomography (PET) with the use of fluorodeoxyglucose (FDG), implemented subsequently with low-dosage computer tomography (CT), is to be considered as the most important evolution of imaging in the management and assessment of classical Hodgkin lymphoma (cHL) patients [1], improving the identification and localization of FDG-avid lesions [2]. In HL, microenvironment is particularly active because of inflammatory cells that are more representative than Reed-Sternberg cells. In the management of patients with cHL, FDG-PET has been used as a tool for staging, prognostic value during therapy, especially for the definition of chemosensitive patients and monitoring of treatment response. FDG-PET can detect involvement sites with a sensitivity of more than 80% and a specificity of 90% [3, 4]. If the role of FDG-PET in first-line cHL patients is well defined, this is not the same in the subset of relapsed or refractory patients. FDG-PET is mandatory to define metabolic response to frontline therapy and moreover it is important in the definition of nonresponders or refractory disease patients [5].

Refractory disease is reported in about 15% of patients, with some variations based on the choice of first-line chemotherapy, and particularly in advanced stages, up to 40% eventually relapse within 3 years [6, 7]. The aim of this review was to highlight a practical way to use FDG-PET in the subset of HL, with some notes of its use in first-line patients, and particularly focused on relapsed or refractory setting, with a final focus of the evaluation of response by FDG-PET in the new treatment era of immunocheckpoint inhibitors (CPI).

FDG-PET is fundamental in HL for staging and to evaluate response to treatment, chemotherapy, and/or radiation therapy (RT) [8]. Moreover, with the introduction of brentuximab in the frontline therapy of advanced HL (III–IV Ann Arbor stages), the correct initial staging is fundamental, and FDG-PET has a central role [9] to detect involvement in subcentimeter lymphonodes and/or extranodal sites and also, to detect bone marrow involvement with a sensitivity very close to 97% [10].

In the last year, the greatest interest of researchers was focused on the use of FDG-PET in the evaluation of early response assessment as an important prognostic marker and in order to select chemosensitive from chemo-refractory patients, in particular interim PET (iPET) after 2 cycles of chemotherapy, which serves to identify patients for whom the standard ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine) treatment is insufficient and consequently should be intensified. On the other hand, it is useful to identify patients who are early responders and who may benefit from the de-escalation of chemotherapy in the advanced stages of HL; many retrospective studies [11, 12] validated the use of iPET for de-escalation or intensification and nowadays it is the standard way to act for clinicians [3, 4].

The Deauville Score (DS) criteria, which is a standardized method for visually interpreting FDG-PET scans, is commonly used to evaluate results of interim and final PET. FDG-PET scans are rated on a scale of 1–5 with the score 1–3 indicating a favorable response or a complete metabolic response (CMR) and a score 4–5 indicating no response to treatment, condition that may require further treatment or closer monitoring [13]. In a recent study with the aim to analyze the impact of a positive iPET, patients with HL had excellent long-term outcomes regardless of iPET status; for those with DS4, the choice to intensify is debated and the DS should be integrated with other data such as level of FDG avidity [14].

At the end of induction therapy, patients typically undergo another FDG-PET scan to evaluate their metabolic response according to the new Lugano criteria [5]. An FDG-PET negativity at the end of therapy in Hodgkin’s has a markedly favorable prognostic significance, is highly predictive of favorable progression-free survival (PFS) and overall survival (OS), whereas a positivity does not always coincide with disease persistence [15, 16]. The most obvious example of this, is the positivity at the mediastinal level of the thymus area that correlates with physiologic rebound (Fig. 1).

The role of FDG-PET in evaluating the efficacy of first-line treatment is well documented. In a meta-analysis, FDG-PET scans showed high positivity and specificity when used to stage and restage patients with lymphoma [17]. FDG-PET positivity at the end of treatment has been shown to be a significant adverse risk factor that means relapse of disease, in all HL patients (Fig. 2) [18].

An accurate assessment of the presence of residual disease is essential to determine which patients would benefit from additional therapy. In a study by Guay et al. [19], 48 patients with HL underwent FDG-PET after the completion of chemotherapy with a response rate of 12/48 (25%) PET-positive. Eleven out of12 patients with PET positivity (92%) relapsed (biopsy-proven) during a mean follow-up of 197 days. The positive predictive value and the negative predictive value of PET were both equal to 92%, so in conclusion, a positive FDG-PET after the end of therapy in HL patients is a strong predictor of relapse. A negative FDG-PET status is whereas an excellent predictor of good prognosis. A clearly positive FDG-PET after treatment is significant for refractory disease; more challenging situation is a residual metabolically active lesion.

A study realized by Sher et al. [20] was conducted to demonstrate the prognostic significance of post-ABVD PET in patients with HL, 73 patients were included in this study; 13/73 (16%) patients were FDG-PET positive at the conclusion of chemotherapy. Based on bivariable Cox regression, post-ABVD positivity (hazard ratio: 4.8, p = 0.05) was predictive of disease recurrence; patients with residual bulky disease were treated with involved-field radiotherapy obtaining 69% of response, indicating that most cases with residual persistent PET-positive disease [19, 20] could be sterilized by RT.

In a meta-analysis conducted by Zijlstra et al. [21], sensitivity and specificity of PET in the assessment of residual disease in HL were evaluated to be 84% and 90%, respectively. The negative predictive value at end of treatment was higher than positive predictive value due to the post-therapy inflammatory changes. Most of clinicians are favorable to RT in localized HL; more discussed are the cases of bulky in advanced-stage disease.

In the German Hodgkin Study Group’s (GHSG) HL15 trial, advanced HL patients were randomized to receive BEACOPP versus escalated BEACOPP for 6–8 cycles [22]. Results demonstrated that treatment with 6 cycles of escalated BEACOPP followed by PET-guided radiotherapy (only in cases with a persistent mass measuring 2.5 cm or more, and positive on PET scan after chemotherapy) was more effective in terms of freedom from treatment failure and less toxic than eight cycles. On the contrary, the study of Gallamini et al. [23], the HD0607, demonstrated that regarding advanced HL treated with ABVD, RT can be omitted in patients who achieve a negative PET status. In a recent study from Fondazione Italiana Linfomi (FIL), HL patients with a bulky lesion at baseline who have CMR after 2 and 6 ABVD cycles were randomly assigned 1:1 to RT versus observation (OBS); results showed a 2-year event-free survival (EFS) of 87.8% versus 85.8% for RT versus OBS; so, patients in CMR randomly assigned to OBS had a good outcome [24].

In patients with relapsed/refractory (R/R) HL, FDG-PET is used to evaluate the response to treatment before autologous stem cell transplantations (ASCT). In general, a negative PET after salvage treatment is indicative of long-term PFS, and on the contrary, a positive response induces clinicians to modify therapy strategy or use an additional treatment with the aim to reach a CR [25].

In the relapse HL setting, which represents 10–15% of patients, the standard treatment is a second-line therapy followed by consolidation with ASCT; the chemotherapy strategy used are IGEV, DHAP, or more recently BEGEV [26, 27]. In this setting of R/R HL, the role of FDG-PET is useful at different time-points (Fig. 3):

• -

  During an interim evaluation (as first line, not routinely done)

• -

  Before ASCT

• -

  At the end of salvage treatment

In particular, the role of FDG-PET pre-ASCT can predict response after salvage therapy in the R/R HL, making it a very useful prognostic tool. Also in the salvage setting, an iPET may allow response-adapted strategies to tailor treatment strategy to individual’s needs.

The group of Daw et al. [28] on behalf of the EuroNet Paediatric Hodgkin Lymphoma Group gave consensus-based guidelines for pediatric and adolescent patients with first R/R HL, based on literature review and expert consensus and gave a central role of iPET. Panel recommendations were a common starting point for all patients with 2 cycles of re-induction salvage chemotherapy followed by response assessment with an iFDG-PET (PET2). Low-risk patients with CMR on PET2 have non-transplant salvage with only RT of consolidation. All other patients received intensified consolidation with high-dose chemotherapy (HDC) and ASCT. Another very recent study [29] was conducted with the aim to compare PET-based response criteria on the quotient PET (qPET) value between first and second line. qPET value is the quotient of the SUV peak value divided by the SUV mean of the liver. The results showed that qPET values in first-line treatment tended to be higher in patients who relapsed and that the empiric cumulative distribution of the qPET evaluated at iPET (after 2 cycles) between first- and second-line treatments were similar, so the response criteria used in first lines should be the same as in second lines.

In another study conducted by Moskowitz et al. [30], 96 patients with R/R HL received 2 cycles of ICE or augmented ICE (aICE), then underwent an iPET; 58/96 (60%) patients achieved a negative FDG-PET after either ICE/aICE, and went on with ASCT, on the other side 38/96 (40%) had a positive iPET. In this group, the treatment was intensified with gemcitabine-vinorelbine and liposomal doxorubicin (GVD) for 4 cycles. Finally, between these patients, 17 obtained PET-negative status. In term of survival, all patients with PET-negative after 2 cycles of ICE and the 4 of GVD had a PFS of 80% at 51 months versus the group with PET-positive status had a PFS of 30%. On univariate analysis of the intent-to-treat patient population, FDG-PET-positive disease after GVD was the most discriminative in terms of unfavorable outcome. There are actually 2 studies in recruiting phase about the role of iPET in this setting:

• -

  NCT05660993: a study with the aim to evaluate the safety and efficacy of PET-adapted treatment of nivolumab in combination with bendamustine, gemcitabine, vinorelbine (Nivo-BeGEV) in patients with R/R HL. Patients with CR after 2 cycles of Nivo-BeGEV will proceed to ASCT. Patients with <CR at the PET-CT evaluation, will receive additional 2 cycles Nivo-BeGEV. Patients with CR after 4 cycles Nivo-BeGEV will proceed to ASCT.

• -

  NCT04981899: patients will receive 6 cycles of nivolumab at the fixed dose of 40 mg, with subsequent assessment of response by PET-CT. Patients with CR will proceed to ASCT; patients with less than CR, after nivolumab monotherapy, will be treated with 2 cycles of a combination of nivolumab and ifosfamide, carboplatin and etoposide (NICE-40), with subsequent PET-CT assessment (Table 1).

Multiple recent studies demonstrated the role of FDG-PET as a predictive marker for survival outcomes during salvage chemotherapy and before ASCT and now is recognized as the most important predictive tool of outcomes in the R/R HL patients rather than B symptoms, extranodal disease, onset of relapse after first line; parameters used almost in the past [31, 32].

In this setting, Devillier et al. [33] reported a series of 111 consecutive patients with R/R HL who achieved CR or PR evaluated with FDG-PET, after salvage chemotherapy and who underwent random single or tandem ASCT. In patients with R/R HL, FDG-PET response at time of ASCT favorably influences outcome: patients with positive FDG-PET scans had improved 5-year PFS with tandem ASCT compared to single ASCT of 43% versus 0%, so FGD-PET can help identify patients who need more intensified treatment.

Another retrospective study evaluated the role of FDG-PET after salvage chemotherapy and before ASCT. In this group, patients with negative PET before ASCT reached a 3-year EFS of 62% and OS of 78%, with a median follow-up of 38 months. Multivariate analysis demonstrated pre-ASCT FDG-PET as an independent prognostic significance for EFS [34].

Another study conducted by Castagna et al. [35] investigated FDG-PET’s role before ASCT, and after two cycles of salvage chemotherapy. Twenty four HL patients were included. PET before ASCT was negative in 58% and positive in 42% of patients. 9/10 (90%) patients with positive PET relapsed after HDC, while all but 1 patient with negative PET maintained a CR. The 2-year PFS was 93% versus 10% for patients with negative and positive PET, respectively (p < 0.001). This study confirmed the role of PET in this patients’ setting: predicting outcome after HDC.

The group of Moskowitz et al. [36, 37] demonstrated the importance of FDG-PET before ASCT; they conducted another study in which 45 patients were enrolled, subjects were treated with 2 cycles of brentuximab followed by FDG-PET, that resulted positive in 33 patients and negative in 12 patients. Patients who achieved PET-negative status proceeded directly to ASCT; those with persistent abnormalities on PET received two cycles of aICE, of whom 22 were negative and 10 remained positive, then received ASCT. Patients with FDG-PET negative after brentuximab had 2-year EFS of 92% versus 91% for the group who achieved PET negative after aICE. In the third group with positive PET after brentuximab and aICE, EFS after 2 years was 46%. Results from AETHERA trial of brentuximab as a consolidative treatment option for adult patients with cHL at high risk after ASCT, showed 5-year PFS of 59% versus 41% in the placebo group; about criteria for definition of high-risk group, there was a PR or stable disease (SD) at pre-ASCT PET phase [38].

The group of Chen et al. [39] also evaluated the use of brentuximab in the setting of R/R HL and their responses were evaluated after 2 or 4 cycles by FDG-PET; in the pretransplant setting, in which the goal of treatment was to achieve PET negative response in order to proceed with ASCT, if a CR was not achieved after 2 cycles, treatment was intensified with another line chemotherapy, so the treatment decision was PET-oriented.

All these studies (summarized in Table 2) demonstrate that the strongest predictors of outcome for patients with R/R HL is the achievement of PET negative before ASCT. Besides, FDG-PET performed after ASCT could help provide prognostic information in R/R HL patients.

The study of Sucak et al. [40] was realized with the aim to investigate this PET’s role; 43 consecutive patients were enrolled in this study. FDG-PET was performed following salvage chemotherapy and ASCT. FDG-PET positivity was found in 26/43 (60%) patients before ASCT and in 13/43 (30%) patients after ASCT. The patients who had negative PET scan before or after ASCT had significantly better outcomes in terms of OS and PFS.

Another study focused on the role of FDG-PET after ASCT was conducted on 97 patients with R/R HL treated with HDC followed by ASCT. A multivariate analysis for age, sex, disease stage, B symptoms, presence of extranodal involvement, bulky disease, elevated lactate dehydrogenase, number of previous chemotherapy lines, remission status before transplant, 18FDG-PET status before and after transplant was done and result showed remission status before and after ASCT as the most important prognostic factor [41]. After ASCT, FDG-PET could be an important tool in case of positivity to individuate patients who may benefit from additional treatment such as allogeneic SCT or immunotherapy, and in case of negativity, it is associated with improved outcome.

In the setting of ultra-R/R patients at second or more relapse, the role of FDG/PET and its prognostic utility are not well defined. In some setting of patients could be more useful: the group treated with brentuximab as salvage therapy after ASCT, the group who underwent to allogeneic SCT and the one who received CPI as treatment.

In the study of Kedmi et al. [42], 49 HL patients treated with brentuximab for relapse post-ASCT were analyzed. They were evaluated after 4 cycles with FDG-PET. Results showed that for patients who reached CR, median PFS was 47.9 versus 1.5 months for nonresponders group.

In a study with the aim to investigate whether pre-allogeneic transplantation PET status predicted outcome after allogeneic SCT, 80 patients with lymphoma (20 with HL) received a reduced-intensity allogeneic SCT. PET and CT scans were performed before transplantation and up to 36 months after transplantation. Forty-two (52%) patients were PET-positive before transplantation, but they find that pretransplantation PET status had no significant impact on either relapse rate or OS. These findings suggest that, in contrast to ASCT, pretransplantation PET status is not predictive of relapse and survival in allogeneic SCT setting [43]. Another study evaluated the role of PET before allogeneic SCT, in 46 HL and 34 non-HL patients and found that patients who had negative PET studies before undergoing allogenic SCT had significantly better outcomes in terms of 3-year OS (76% vs. 33%) [44].

In a study conducted by Reyal et al. [45] in large cohort of patients with HL undergoing to T-cell-depleted allogeneic SCT, FDG-PET was performed before transplantation. They found that patients with the highest burden of disease based on FDG-PET (DS 5) and those with progressive disease had worse OS. In this setting of patients, the role of PET is yet to be determined and prospective studies are needed.

Recently, the treatment paradigm for patients with R/R HL after second line has shifted to employment of immune CPI. Antibodies anti-PD-1 (pembrolizumab and nivolumab) have recently been demonstrated to be effective with a response rate of 60–87% [46]. Hodgkin itself disease has a microenvironment with high avidity for glucose and this determine high FDG uptake visible on PET. Theoretically, anti-PD-1 could determine an activation of antitumor microenvironment with a consequent increase in glucose metabolic consumption that could determine an increased FDG-PET uptake. For this reason, it is necessary to understand if we need to introduce new diagnostic and response criteria in the use of FDG-PET in the anti-PD-1 era.

About the use of FDG-PET in patients with HL treated with anti-PD-1 new, concepts have been introduced: new patterns of response and progression; new lesions can be observed due to the immune-related reaction [47‒49]. In a centralized review [50], 60 consecutive patients with R/R HL treated with nivolumab and subjected to FDG-PET were retrospectively analyzed and the results showed unconventional immune-related phenomena regarding tumor response or progression, in particular transient progression in lesion size and metabolism; in clinical routine, these immune-related phenomena should be considered as potential differential diagnoses.

In 2016, a consensus panel of experts developed a modification of the existing Lugano response criteria for FDG-PET in the immunomodulatory era known as the Lymphoma Response to Immunomodulatory Therapy Criteria or LYRIC [51]; the group introduced the concept of indeterminate response (IR) as the frame time between the onset of progression or pseudoprogression and the performance of a biopsy or a subsequent imagine to confirm that hypothesis and indicate clinicians to a period of watch and wait; the group also introduced the concept of hyper progression effect of the anti-PD-1: a lesion showing an initial increase in size. An Italian panel of expert agreed that a critical appraisal of the new response criteria highlighted that IR evaluation introduces the need for repeated imaging, which is not yet validated as beneficial for the patient [52].

In order to confirm the applicability of these criteria, Dercle et al. conducted a study [53], in which 16 patients with R/R HL with 6 previous median lines of therapy were included. They were treated with anti-PD-1, and a FDG-PET and a CT scan were acquired every 3 months. After the first 3 months, 9/16 patients (56%) reached a response. FDG-PET reclassified best overall response in 5 patients compared with CT; 5/16 (31%) patients showed new imaging pattern according to the new criteria: 2 had IR and 3 new lesions associated with the immune-related adverse event. In patients with response to treatment, FDG-PET showed a decrease in tumor metabolism and reduction in CT volume at the disease’s sites and an increase in splenic metabolism suggesting a favorable immunologic answer. These data demonstrated that the FDG-PET evaluation in HL patients who are treated with anti-PD-1 is different from classical chemotherapy. In another recent study of Mokrane et al. [54], 45 patients with R/R HL were treated with nivolumab (median 6 lines of previous therapy), then underwent an early evaluation treatment with FDG-PET and CT scan after 2 months of treatment: the use of early FDG-PET reclassified response of 20/45 patients (44%); among this group, 11/20 (55%) obtained a CMR versus CR (by CT evaluation), no pseudoprogression was detected; PET and CT response were associated with OS, so as happened in the previous treatment lines and types of therapy, early FDG-PET predicted OS in R/R HL and was helpful to detect response patients more precisely than CT scan.

Response-adapted treatment strategies in patients with R/R HL treated with immune therapy is a goal to reach as in the previous treatment lines. For this aim, Chen et al. [55] explored whether the first early response assessment evaluated using FDG-PET, may be associated with OS in the setting of patients with R/R HL treated with anti-PD-1.

This retrospective study included 45 patients; the classification at early PET was identical between Lugano and LYRIC. The 2-year OS probability was significantly different in patients according to response to treatment and allows early risk-stratification, suggesting that e-PET may be used to develop risk-adapted strategies.

In order to better define the role of FDG-PET in patients treated with CPI, an ongoing trials are now being conducted with the aim of comparing chemotherapy versus novel immune CPI plus chemotherapy in treating R/R HL. In arm A (chemotherapy regimen), patients who obtain a response at FDG-PET to salvage chemotherapy proceed with ASCT; in arm B, after chemotherapy regimen plus pembrolizumab, patients undergo a PET scan. Patients who achieve a CR or PR proceed to ASCT, in order to define the effect of anti-PD-1 and the role of PET to identify responder patients. All these studies (Table 3) demonstrate the importance to apply the new Lugano criteria for patients treated with anti-PD-1 and also the importance of correct interpretation of FDG-PET imaging and distinguish a real progression versus an IR or a pseudoprogression. In this setting, anyway, the early use of PET is a prognostic tool for evaluating long-term response. In this setting, it is particularly crucial to use the concomitant use of FDG-PET and CT scan to better define response to CPI.

FDG-PET is used as a diagnostic tool in patients with suspected relapsed or refractory HL and it is essential for developing an appropriate treatment plan. FDG-PET is also used to monitor the response to treatment in patients with R/R HL, in particular at time-point strategically useful for prognostic relevance such as early evaluation after 2–3 cycles, or before and after ASCT.

The information obtained is critical for identifying patients who require additional treatment or alternative therapies. The use of FDG-PET should be part of a multidisciplinary approach to manage patients with R/R HL, and decisions about treatment should be made in consultation with an experienced team of specialists.

As a result, FDG-PET is recommended as a standard component of the diagnostic workup and monitoring of patients with R/R HL. Recently, with the use of anti-PD-1 in ultra-R/R HL that induce high response rates, the role of FDG-PET is being redefined regarding the response criteria. Studies demonstrated that early response is predictive of OS, as demonstrated the role of the PET in the previous lines, and that it is also useful to detect progressive disease rather than pseudoprogression and consequently oriented clinicians’ decision to change treatment or go on with OBS in case of response.

The authors have no conflicts of interest to declare.

No funding was obtained for this work.

V.T. and L.R. contributed equally to the conception of the work. They wrote the paper and revised the final version. Both gave their final approval of the version.