An overview of the use of FDG-PET in Hodgkin lymphoma and DLBCL

When embarking on treatment for lymphoma, many factors need to be taken into consideration, such as the stage of disease. Moreover, treatment decisions may need to be revisited along the course of disease. Control of the disease is crucial, yet tumor sensitivity to different therapies varies. Therefore, biomarkers are used to assess changes in tumors in response to treatment. Patients with tumors that show poor response may require dose escalation or change of medication, while patients who respond well may be considered for dose reduction. 

18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) has been shown to be useful in assessing changes in tumor metabolism during Hodgkin lymphoma (HL) treatment and guiding clinical decisions. Because of high sensitivity in the detection of nodal and extra-nodal lymphoma, FDG-PET use is also recommended for the staging of diffuse large B-cell lymphoma (DLBCL), the most common subtype of non-Hodgkin lymphoma.

Conrad-Amadeus Voltin recently published an overview of current FDG-PET applications in HL and DLBCL, recent developments, and future directions of imaging-directed therapy in Cancer journal.¹ The article below provides a summary of the original review article.

 Initial staging

Accurate disease staging is key to choosing the optimal treatment. Thanks to the ability to detect involved lymph nodes and extra-nodal disease, FDG-PET has become the standard of care for assessing FDG-avid lymphomas, recommended by the 2014 Lugano criteria.

HL

Several studies reported a higher sensitivity of FDG-PET than biopsy for detecting bone marrow disease, with bone marrow involvement diagnosed by FDG-PET in up to 20% of newly diagnosed patients compared with about 5% detected using biopsy. According to the results of a large head-to-head comparison study, which used a positive biopsy as the reference standard, FDG-PET sensitivity and negative predictive value (NPV) for skeletal involvement were 95% and 99.9%, respectively. The ability to accurately detect focal skeletal involvement is crucial, as it was shown to have a significantly negative impact on progression-free survival (PFS), independent of the chemotherapy regimen used. However, diffusely increased skeletal FDG uptake should not be considered a sign of HL. Based on the existing evidence, there is a general consensus that patients with HL can be staged by FDG-PET rather than biopsy.

DLBCL

With sensitivity and NPV of around 60% and 91%, respectively, FDG-PET is less sensitive in identifying bone marrow involvement of aggressive non-Hodgkin lymphoma compared with HL. This is largely attributed to the often diffusely increased skeletal FDG uptake associated with positive bone marrow biopsy results in patients with DLBCL. However, bone marrow biopsy is not recommended for patients with positive FDG-PET diagnosis and should be used in cases where results would influence the selection of treatment regimen.

 Early response assessment

In order to allow simple and reproducible interpretation of PET results of early response assessment of treatment, the Deauville score was introduced. The score classifies residual tissue on the scale of 1–5, based on comparison of FDG uptake by the lesion with the reference regions of mediastinal blood pool and liver (Table 1).

Table 1. Deauville scale for therapy stratification in patients with FDG-avid lymphomas¹

Score	Criteria	Interpretation*
1	No FDG uptake	CR
2	FDG uptake ≤ mediastinal blood pool
3	FDG uptake > mediastinal blood pool but ≤ liver
4	FDG uptake moderately increased compared to the liver	PR/SD/PD
5	FDG uptake markedly increased compared to the liver and/or new sites of disease
CR, complete metabolic response; PD, progressive disease; PR, partial metabolic response; SD, stable disease *Response to treatment as defined by the Lugano classification.

 

HL

The use of PET to monitor early response to treatment has been investigated in several trials across different disease stages. Studies in early-stage HL focused on whether patients who achieve complete metabolic response after chemotherapy need irradiation. Based on the results of the RAPID (NCT00943423), H10 (NCT00433433), and HD16 (NCT00736320) trials, consolidative radiotherapy in PET-negative individuals after doxorubicin combination with bleomycin, vinblastine, and dacarbazine (ABVD) is still recommended. Furthermore, in PET-positive patients, intensification of therapy to bleomycin plus etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone (eBEACOPP) should be considered.

The ongoing HD17 trial (NCT01356680) should provide an answer for patients with an intermediate-stage disease as to whether irradiation after two cycles of eBEACOPP and two courses of ABVD improves patient outcomes.

FDG-PET results are proven to be important in tailoring treatment intensity in patients with advanced-stage HL. The RATHL (NCT00678327), HD0607 (NCT00795613), and AHL2011 (NCT01358747) studies demonstrate that PET can safely guide therapy in these patients and allow treatment de-escalation in early responders. PET was also shown to be able to guide the number of eBEACOPP therapy cycles in the HD18 trial (NCT00515554), allowing a reduction in the number of treatment-emergent adverse events while maintaining the level of efficacy.

More on interim PET in classic HL is available here, and click here for a general overview of the response-adapted therapies for HL.

DLBCL

Compared with HL, there are fewer randomized trials in DLBCL evaluating the early response to treatment by PET, with the majority focusing on the potential use of PET positivity after 2–4 cycles to guide treatment escalation. Due to inconsistency in image evaluation and the lack of control groups, results did not provide a definite conclusion.

The PETAL trial (NCT00554164) found that, in patients with negative PET after 2 cycles of R-CHOP, therapy could be safely reduced to six cycles without loss of efficacy. Moreover, PET positivity was an independent biomarker for a poor outcome that did not improve after treatment intensification.

A meta-analysis of DLBCL patients from different trials demonstrated the ability of FDG-PET to clearly discriminate between responding and non-responding patients, with optimal timing after two courses of therapy.

Read an update on interim PET as a biomarker for response and PET-directed therapy for limited-stage DLBCL here. You can also watch an interview with Sonali Smith discussing which patients with DLBCL may benefit from PET-directed induction therapy here.

Late response assessment

Historically, computer tomography (CT) was used for end-of-treatment assessment after chemotherapy. However, in the last couple of decades, FDG-PET proved to be more suitable, due to high NPV and the ability to distinguish between fibrotic masses and active residual disease. Although PET can detect early relapses, a proportion of the results are false-positive. There is also a significant number of false-negative results and cost-effectiveness of routine use of PET during follow-up remains to be proved. Therefore, the authors of the article recommend restricting its use to cases of suspected relapse.

HL

Based on the results from recent studies, radiotherapy is recommended to be limited to PET-positive residual tissue in advanced-stage HL after eBEACOPP therapy.

DLBCL

Positive results of PET after R-CHOP therapy are associated with poor prognosis in all stages of DLBCL, with one of the trials (NCT00544219) reporting patient event-free survival of 48% in PET-positive vs 74% in PET negative after six courses of dose-dense R-CHOP. However, studies, including OPTIMAL >60, demonstrated that consolidating radiotherapy can improve the clinical outcome of those patients with PET-positive scans. Available results suggest that in 42–60% of patients with PET-negative bulk, irradiation is not needed.

 Recent advances and future directions

The Deauville score has become an established way to monitor treatment response but, currently, there is no consensus on the interpretation of FDG uptake. The authors of the review article recommend interpretation of the Deauville Scores 4 and 5 as positive.

Quantitative measurements offered by FDG-PET, like change in the maximum standardized uptake value (∆SUV_max), have been demonstrated to add extra value when stratifying response in both HL and DLBCL. Providing there are additional biomarkers, such as the metabolic tumor volume (MTV) or total lesion glycolysis (TLG), PET adds valuable information on tumor burden and disease activity. Those biomarkers outperform the currently used staging systems for baseline risk stratification of patients with early-stage HL into low- and high-risk categories. Moreover, MTV results combined with early PET response assessment has been shown to be an important prognostic factor in HL and DLBCL. However, MTV calculation is often done inconsistently (different approaches used are presented in Table 2). The accurate visual and quantitative response assessment relies on inconsistent PET scanning protocols and image reconstruction methods, and therefore requires standardization.

Table 2. MTV calculation approaches¹

Threshold	Advantages	Disadvantages
Fixed absolute (e.g., SUV 2.5 or 4.0)	High reproducibility	Overestimation if tumor lies adjacent to areas of high physiologic uptake
	Observer independence	Underestimation in tumors that have many voxels with an uptake less than the threshold
Reference regions (e.g., liver or mediastinum) *	Adjusted to patient and scan	More time consuming
		Low availability on commercial software
Fixed relative (e.g., 41% of tumor SUVmax)	Observer independence	Overestimation in case of low lesion-to-background ratio
		Underestimation of tumors with heterogeneous uptake and high SUVmax
Adaptive (e.g., signal-to-background ratio)	Adjusted to patient and scan	More time consuming
		Low availability on commercial software
SUV, standardized uptake value; SUVmax, maximum SUV *Thresholding method proposed by the PET Response Criteria in Solid Tumors (PERCIST)

 

The authors believe that quantitative FDG-PET has great prognostic potential, and metabolic measures are already becoming incorporated into clinical trial designs. However, with novel therapies emerging, there is a growing need for more specific tracers than FDG, as there is a risk of false-positive results caused by therapy-associated inflammatory changes. Such specific tracers, like the radiolabeled programmed cell death ligand 1 (PD-L1) antibodies, the chemokine receptor-targeting ⁶⁸Ga-pentixafor, and the ⁶⁸Ga-labeled fibroblast activation protein inhibitor (FAPI), could expand our knowledge of the dynamic tumor microenvironment and improve checkpoint inhibitor-based therapies.

 In the future, machine learning and artificial intelligence could be used to optimize complex workflows and time-consuming tumor delineation in advanced disease, as well as providing other new opportunities.

 Conclusion

Thanks to its high sensitivity for detecting lymphoma lesions, PET has become the standard tool for diagnosis and assessment of response to therapy (Table 3), removing the need for undirected bone marrow biopsy for patients with HL and those with DLBCL with involvement proven by FDG-PET. However, the overall predictive value of PET is moderate, due to false-positive results caused by non-malignant changes.

Table 3. Evidence-based recommendations on the use of FDG-PET¹

Indication	HL	DLBCL
Staging	+++	+++
Early response	++	++
End of treatment	++	++
Follow-up	+/−	+/−
+ + +, standard modality; + +, standard — depending on therapy protocol; +/−, optional — recommended in selected cases, e.g., suspected relapse

 

The technique allows more personalized treatment approaches, and PET-derived biomarkers and images have increasingly become used to guide escalation, de-escalation, and the need for consolidation radiotherapy. Interim PET is particularly useful in HL, while both baseline and interim PET are used in DLBCL. However, more data from prospective trials are needed to validate quantitative PET measures.