• diffuse large cell lymphoma;
  • Hodgkin disease;
  • chemotherapy;
  • FDG-PET;
  • therapy response;
  • early response


  1. Top of page
  2. Abstract


Early prediction of response to therapy may offer the potential to identify patients who will benefit from standard conventional therapy. The objective of this study was to determine the predictive value of FDG-PET as an early response indicator after 1 cycle of chemotherapy for progression-free survival (PFS) in diffuse large cell lymphoma (DLCL) and classic Hodgkin disease (HD).


FDG-PET was performed before, after 1 cycle, and after completion of chemotherapy in 47 patients. The patients were followed with a median follow-up of 21 months (range, 3–47 months). PFS was compared between PET-positive and PET-negative patients after 1 cycle and after completion of therapy.


All PET-negative patients after 1 cycle (n = 31) had sustained complete remission with a median follow-up of 28 months. Fourteen of 16 PET-positive patients after 1 cycle had refractory disease or relapsed (median PFS, 5.5 months). There were 2 false-positive results, 1 with an active infection at the biopsy site and the other in a patient who had been in remission after radiation therapy. There was good agreement between the results obtained after 1 cycle and at completion of therapy (kappa, 0.80); however, the negative predictive value was higher for FDG-PET after 1 cycle than after completion of chemotherapy (100% vs 91.4%), although not statistically different (P = .40).


FDG-PET had a high prognostic value after 1 cycle of chemotherapy, thus it can be a valid alternative for posttreatment evaluation of DLCL and HD and may offer the potential for change in treatment paradigms. Cancer 2006. © 2006 American Cancer Society.

The current paradigm of therapy in both classic Hodgkin lymphoma (HD) and diffuse large cell lymphoma (DLCL) is a risk-adapted approach, based primarily on established pretreatment prognostic factors. Recent advances in dynamic imaging technology have motivated attempts to determine an individual's risk early during therapy, thereby allowing development of a patient response-based therapeutic management strategy.1

With current treatment regimens, at least 80% of patients with HD can be cured by the use of single or combined modality therapies.2 DLCL is a biologically heterogeneous lymphoproliferative malignancy with inconsistent response to therapy, hence the existing therapeutic approach is successful in only 50% to 70% of patients.3, 4 Prognostic models, based on clinical and biological parameters (International Prognostic Index [IPI] for DLCL and international prognostic score [IPS] for HD),4, 5 provide risk stratification to predict outcome. Nonetheless, there is some variability in outcome within the individual IPI and IPS risk groups as these models do not identify the true genetic makeup of these diseases. Hence, there is a risk of treating low-risk patients more intensively than needed with potential unnecessary adverse effects.4 Alternatively, a substantial proportion of patients who are falsely stratified in the low-risk category may theoretically be undertreated. This issue is particularly problematic with DLCL, because of the significant overlap with the intermediate-risk category. Although microarray techniques have further refined the prognostic groups identified by the IPI, their routine use is limited by the requirement for fresh, cryopreserved samples and high cost.6

There is convincing evidence that imaging with fluorodeoxyglucose positron emission tomography (FDG-PET) can provide timely information about response when used early during therapy.7–12 The metabolic information obtained from FDG-PET may lead to improved patient management by determining rapid response, which could theoretically be used as a basis for therapy changes.

The purpose of the current study was to use FDG-PET as an early indicator of therapy response after 1 cycle of chemotherapy to determine its prognostic value in patients with previously untreated classic HD and DLCL.


  1. Top of page
  2. Abstract


Forty-seven consecutive patients with newly diagnosed, histologically proven DLCL (n = 24) or HD (n = 23) who had undergone FDG-PET scans at baseline and after 1 cycle of chemotherapy were included in the study. The data were retrospectively collected but prospectively analyzed for outcome. The patient group comprised 23 women and 24 men with a mean age of 48.2 + 18.7 years (range, 18–76 years; Table 1). Exclusion criteria included secondary malignancy, active infection and granulomatous disease at baseline, presence of human immunodeficiency virus, primary cerebral lymphoma, and hyperglycemia (>200). The institutional review board allowed an exempt retrospective review of a PET database for this study. The study was in compliance with the Health Insurance Portability and Accountability Act.

Table 1. Patient Demographic Data
 Early PET negative (n = 31)Early PET positive (n = 16)Overall (n = 47)
  • BM indicates bone marrow; IPI, international prognostic index; IPS, international prognostic score.

  • *

    Ann Arbor staging.

Mean age ± SD46.2± 18.852.1± 18.348.2± 18.7
Histology, n (%)
 DLCL14 (45)10 (62.5)24 (51)
 HD17 (55)6 (37.5)23 (49)
Stage,* n (%)
 I-II22 (71)6 (37.5)28 (60)
 III-IV9 (29)10 (62.5)19 (40)
Bulky disease, n (%)3 (10)5 (31)8 (17)
Extranodal disease >1 site, n (%)4 (13)3 (19)7 (15)
BM involvement, n (%)2 (6)1 (6)3 (6)
IPS score (n = 23), n (%)
 Low13 (76)3 (50)16 (69)
 Intermediate2 (12)3 (50)5 (22)
 High2 (12)02 (9)
IPI score (n = 24), n (%)
 Low9 (64)7 (70)16 (67)
 Intermediate†5 (36)3 (30)8 (33)


All patients underwent a baseline FDG-PET before induction treatment, after 1 cycle, and after completion of therapy. Sixteen scans were obtained on a stand-alone PET scanner, and 31 studies were obtained using an integrated PET-CT scanner (Advance or Discovery LS; GE Medical Systems, Fairfield, CT). At baseline, the interval between FDG-PET studies and therapy initiation was 1–4 weeks. After 1 cycle of therapy, all PET studies were obtained within 0–7 days before the second cycle, although the majority were imaged within 0–3 days of the second cycle. FDG-PET studies were acquired 3–6 weeks after completion of the chemotherapy.

Blood glucose concentrations were similar between pretherapy and posttherapy scans. Blood glucose was <140 mg/dL for all patients. The interval between F-FDG injection time and imaging was 60 min. The difference in intervals among PET scans ranged between 5–20 min.

Therapy Protocol

Diffuse large cell lymphoma

All DLCL patients received CHOP (cyclophosphamide [Cytoxan], hydroxydaunomycin, vincristine sulfate [Oncovin], and prednisone) or rituximab plus CHOP treatment for 6–8 cycles every 3 weeks for 4–6 months. For DLCL, 1 treatment was 1 cycle.

Hodgkin disease

All HD patients received ABVD (doxorubicin [Adriamycin], bleomycin, vinblastine, and dacarbazine) for 12 treatments every 2 weeks (6 cycles) over 6 months. For HD, 2 treatments were considered 1 cycle.

Response evaluation and follow-up

All results obtained after 1 cycle and completion of therapy were confirmed by repeat FDG-PET studies or long-term clinical and laboratory data including biopsy (n = 5). The response status was assessed using standardized guidelines.13, 14

Analysis of FDG-PET images

All images were analyzed by 2 experienced physicians with no clinical knowledge of the patient. However, after therapy the anatomic location of the primary or original tumor was known to the interpreting physicians.

Qualitative analyses

All FDG-PET scans were interpreted in a dichotomous fashion as positive or negative. A positive lesion was defined as the presence of FDG activity that exceeded the activity seen on the contralateral site or in the background in a location incompatible with normal anatomy or physiologic variants (Table 2). For patients with contralateral site involvement and for asymmetrical sites, a positive lesion was defined as the lesion > background activity. The background was the highest activity excluding pathological and physiologic sites of uptake. For lesions in the head and neck, the background was taken within the jugular vessels; for those in the chest, in the mediastinum around the aortic arch region; and those in the abdomen and pelvis, in the mesentery or abdominal vessels, whichever had the higher activity. A negative study was defined as one having no pathological FDG uptake at any site with uptake < the uptake on the contralateral site or the background (Table 2). Pretherapy and posttherapy evaluation of disease was performed on a site-by-site analysis.

Table 2. Qualitative Interpretation Guidelines
 Symmetrical sitesAsymmetrical sites
  • CS indicates contralateral site; Bkg, background.

  • *

    Occurred in lesions with subtle uptake where the lesion was near high physiologic activity.

  • No lesion fulfilled this criterion.

PositiveUptake>CS and/or BkgUptake>Bkg
Uptake>CS and <Bkg*
NegativeUptake<CS and/or BkgUptake<Bkg
Uptake<CS and >Bkg
Semi-quantitative analyses

Semi-quantitative analyses were performed only for measurable nodal disease sites in those who had an integrated PET-CT study (n = 33). The patients with extranodal and nonmeasurable disease13 were evaluated using only qualitative analysis. In large lesions (>2 cm), 2 maximal standardized uptake values (SUVmax) were obtained at 2 different sections through the lesion to factor in tumor heterogeneity, and an average SUVmax was calculated.

FDG-PET findings were correlated with the PFS. PFS was defined as the time interval without progression of disease from the start of treatment.

Statistical analysis

Descriptive frequencies and percentages were calculated to characterize demographic and prognostic characteristics of the study population. Sensitivity, specificity, negative-predictive and positive-predictive values were estimated for FDG-PET after 1 cycle of therapy and after completion of therapy. Ninety-five percent confidence intervals (95% CI), assuming binomial distribution, were calculated to assess precision of the estimates. The McNemar test and the kappa statistic were used to assess the concordance between FDG-PET after 1 cycle and after completion of therapy.

Kaplan-Meier survival analysis was used to evaluate progression-free survival (PFS), and the log-rank test was used to compare PFS between PET-negative and PET-positive patients. The area under the receiver operator curve (ROC) was measured to assess the accuracy of post SUVmax as a discriminator of recurrence or no recurrence status (gold standard) of individual lesions. The optimal post SUVmax cutpoint to maximize sensitivity and specificity was estimated from the ROC curve. All P-values were 2-sided with an alpha of .05.


  1. Top of page
  2. Abstract

Results are shown in Tables 3–5.

Table 3. Post 1 Cycle PET Negative Results Versus PET After Completion of Therapy
DiseaseAgeStageEarly PETLate PETPFS, mIPIIPSOther
  1. PFS indicates progression free survival; IPI, international prognostic index; IPS, international prognostic score; HD, Hodgkin disease; DLCL, diffuse large cell lymphoma.

HD40IINegativeNegative34 0bulky
HD39IINegativeNegative34 0 
HD30IINegativeNegative38 0 
HD22IVNegativeNegative29 2 
HD33IINegativeNegative20 3 
HD55IINegativeNegative27 1 
HD18IINegativeNegative20 0 
HD35IINegativeNegative23 0 
HD71IINegativeNegative29 2 
HD22IIINegativeNegative29 0 
HD30IINegativeNegative20 1 
HD36IIINegativeNegative21 0 
HD32IINegativeNegative18 1 
HD28IINegativeNegative30 1 
HD40IINegativeNegative20 0 
HD20IINegativeNegative17 1 
HD75IVNegativeNegative20 4 
DLCL75IINegativeNegative341 bulky
DLCL67IINegativeNegative301 bulky
Table 4. Post 1 Cycle PET Positive Results Versus PET After Completion of Therapy
PatientDiseaseAgeStageEarly PETLate PETConventional restagingPFSIPIIPSOtherTherapy after 1st lineCurrent status
  • PFS indicates progression-free survival; IPI, international prognostic index; IPS, international prognostic score; Prog, progression; PR, partial response ; CR, complete response; FN, false-negative; FP, false-positive; Ex, died of disease; AWD, alive with disease; HDT/ASCT, high dose therapy with autologous stem cell transplant.

  • *

    Progressed before any decision was made for alternative therapy.

  • In remission following radiotherapy.

  • Infection at the site of biopsy.

1HD76IIIPositivePositiveProg32  None*Ex
2HD76IIIPositiveNegative-FNCR7 2 2nd LineAWD
3HD23IIPositivePositivePR3 1bulky2nd LineAWD
4HD23IIPositive-FPPositive-FPCR36 1 RadiotherapyCR
5HD31IVPositivePositivePR7 2 2nd LineAWD
6HD50IIIPositivePositivePR8 1 2nd LineEx
7DLCL59IIIPositivePositiveProg32  None*Ex
8DLCL51IIIPositivePositivePR41  HDT/ASCTAWD
9DLCL46IPositive-FPNegativeCR261  NoneCR
10DLCL62IIPositivePositiveProg31 bulkyNone*Ex
11DLCL23IIIPositivePositivePR51 bulkyHDT/ASCTAWD
12DLCL69IIIPositivePositiveProg32  None*Ex
13DLCL63IIPositivePositivePR41 bulkyNone*Ex
14DLCL59IIIPositiveNegative-FNCR222  2nd Line and HDT/ASCTAWD
15DLCL68IIPositivePositivePR61  2nd Line and HDT/ASCTAWD
16DLCL55IIIPositiveNegative-FNCR151 bulky2nd Line and HDT/ASCTAWD
Table 5. Statistical Analysis of Early FDG-PET Versus Late FDG-PET
%95% CI%95% CI
  1. NPV indicates negative predictive value; PPV, positive predictive value.


PFS After 1 Cycle

The median interval between the first treatment and progression or date of last follow-up, if no progression, was 21 months (range, 3–47 months). The patients with negative FDG-PET findings after 1 cycle (n = 31) had a median follow-up of 28 months; in this group, median PFS was not reached (NR). There were 16 PET-positive patients after 1 cycle of therapy with a median PFS of 5.5 months (95% CI = 3–8 months; Table 3).

PFS After Completion of Therapy

For PET-negative patients after completion of therapy (n = 35), the median PFS was not reached. There were 3 patients who had false-negative results and whose disease recurred at 7, 15, and 22 months, respectively (Table 4; Patients 2, 14, and 16).

The median PFS for the 12 FDG–PET-positive patients was 4 months (95%CI = 3.0–6.0 months; Table 4). There was 1 (8.3%) patient who had a false-positive result. This patient received consolidative radiotherapy after completion of chemotherapy and remains in clinical remission after 36 months (Table 4; Patient 4).

Qualitative Evaluation

FDG-PET after 1 cycle of therapy

Post-1 cycle FDG-PET demonstrated complete resolution of disease in 31 patients (17 HD,14 DLCL; Table 3), all of whom had a sustained complete response after completion of therapy. In this group of patients, the median follow-up was 28 months (range, 17–47 months). There were 14 patients who were followed up for <2 years, and of these patients, 10 were followed up for 20 to 24 months, and 4 were followed up for 17 to 20 months after completion of therapy (Table 3).

Of the 16 PET-positive patients (10 DLCL and 6 HD; Table 4), 14 (87.5%) had refractory disease or had relapsed. PET was false-positive in 2 patients who were in remission at the end of the follow-up period (Table 4; Patients 4 and 9). In these patients, 1 patient subsequently received consolidative radiotherapy for a persistently positive hilar lymph node. The other patient was known to have infection at the biopsy site (axilla). The median PFS for the 16 PET-positive patients was 5.5 months (95% CI = 3.0–8.0 months). More specifically, in the PET-positive group (n = 16), restaging procedures, including biopsy (n = 5) demonstrated refractory disease in 11 patients; 4 patients progressed before completion of chemotherapy (Table 4; Patients 1, 7, 10, 12); 7 patients showed a partial response after completion of chemotherapy (Table 4; Patients 3, 5, 6, 8, 11, 13, 15). In the remaining 5 patients, conventional restaging showed a complete response with first-line chemotherapy; however, 3 of 5 patients relapsed at 7, 15, and 22 months, respectively (Table 4; Patients 2, 14, 16). The remaining 2 patients have been in remission during a follow-up period of 26 months and 36 months, respectively. In all patients who relapsed or progressed, the sites of recurrence or progression remained the same as in those who were PET-positive after 1 cycle, except in 1 patient. In this particular patient, FDG-PET was positive in an iliac lymph node after 1 cycle; however, the disease recurred in cervical lymph nodes that were PET-negative after 1 cycle (Table 4; Patient 16).

Of the 14 patients with recurrence or progressive disease, 4 (1 HD and 3 DLCL) had a rapid progression and were too frail for the second-line therapy to be instituted. The remaining 10 patients received either a second-line therapy or autologous stem cell transplantation (ASCT; Table 4).

All patients who died during the follow-up period (n = 6) had a positive early interim FDG-PET (Table 4). No deaths occurred in the PET-negative group (P = .0007; Fisher exact test). However, given an excellent overall short-term survival and few fatal events, we shall not further consider overall survival as an endpoint. A survival analysis could not be performed due to the small number of patients.

Two of 16 PET-positive and 31 of 31 PET-negative patients had a satisfactory remission after first-line chemotherapy (P < .0001; Fisher exact test).

Bulky disease.

After 1 cycle, in the PET-negative group, there were 3 patients with bulky disease (>10 cm), all of whom remained in remission with follow-up periods of 34, 34, and 30 months, respectively (Table 3). In the PET-positive group, there were 5 patients with bulky disease (Table 4). Of these patients, 1 received radiation therapy and remains in remission; 2 underwent ASCT, both with failure; and the other 2 died of progressive disease (Table 4).

Quantitative Evaluation

Quantitative SUVmax measurements were obtained for 330 lesions from 33 patients. Thirty-six (11%) of the 330 lesions recurred, and 294 (89%) did not recur. An ROC analysis based on using post-1 cycle SUVmax as a discriminator of the recurrence or no recurrence status yielded an area under the ROC curve of 91.5% (95% CI = 86.6–96.4; P < .0001; Fig. 1). The optimal post SUVmax cutpoint on the ROC curve that maximized both sensitivity and specificity was ≤1.75 vs >1.75. By using this post-SUVmax cutpoint, the sensitivity was 91.7% (95% CI = 77.5–98.2), and the specificity was 83.7% (95% CI = 78.9–87.7; Fig. 1). However, because the 1.75 SUV cutpoint was derived from the current data, the reported estimates of sensitivity and specificity may be biased. The 1.75 SUV cutpoint needs to be used in another data set to validate the reported sensitivity and specificity estimates. By using a cutpoint of <2.0 vs >2.0, the sensitivity was 80.6% (95% CI = 64.0–91.8), and the specificity was 85.0% (95% CI = 80.4–88.9).

thumbnail image

Figure 1. SUVmax is a discriminator of recurrence. The optimal post SUVmax cutpoint that maximized sensitivity and specificity was ≤1.75 vs >1.75; sensitivity 91.7% (95% CI = 77.5–98.2), specificity 83.7% (95% CI = 78.9–87.7).

Download figure to PowerPoint

In the nonrecurring lesions (n = 294), there were 45 (15.3%) lesions whose SUVs exceeded 1.75 in 15 patients. In the remaining 249 (84.7%) lesions, the SUVs were <1.75. Among those lesions whose SUVs were >1.75, SUVs were between 1.75 and 2.0 in 20 lesions and >2.0 in 25 lesions. In the latter group, SUVs were between 2.0 and 2.5 in 13 lesions and >2.5 in 12 lesions. The highest SUV obtained was 4.4 in this group. Two of 4 patients whose SUVs exceeded 3.0 had either brown adipose tissue or infectious process near lymph node stations. Among lesions with an SUV of >1.75, the majority were located in the mediastinum and hila (n = 22; range, 1.8–3.8), 9 in the neck (range, 2.0–2.4), 5 in the axilla (range, 1.9–4.6), 4 in the abdomen (SUVs 1.8–3.4), and 5 in the pelvis (range, 1.8–4.4). Except in 2 patients who had a false-positive FDG-PET (range, 1.7–4.6) and the patient who had metabolically active pelvic lymph nodes (range, 2.3–4.4) but recurred in the cervical region; all other nonrecurring lesions were visually indiscernible from the background.

Statistical Analysis

The sensitivity, specificity, negative and positive predictive value for FDG-PET, after 1 cycle of therapy and after completion of therapy are summarized Table 5. Descriptive frequencies and percentages were calculated to characterize demographic and prognostic characteristics of the study population. Sensitivity, specificity, negative-predictive (NPV) and positive-predictive values (PPV).

After 1 cycle of therapy

The 2-year progression-free survival for PET-negative patients after 1 cycle of therapy was 100.0%, compared with only 12.5% (95% CI = 2.1–32.8) in those with a positive result (Fig. 2).

thumbnail image

Figure 2. Progression-free survival is shown for PET-positive results after 1 cycle of therapy. Note that survival was 100% (no events) in the PET-negative group, thus a curve can be generated only for the PET-positive group.

Download figure to PowerPoint

When patients with DLCL were separated from those with HD, the prognostic impact of FDG-PET on progression-free survival remained the same (P < .0001 for both, log-rank test).

After completion of therapy

The 2-year progression-free survival for PET-negative patients after completion of therapy was 90.0% (95% CI = 71.4–96.8), compared with only 8.3% (95% CI = 1.0–31.1) in PET-positive patients (P < .0001, log-rank test; Fig. 3). When patients with DLCL were separated from those with HD, the prognostic impact of FDG-PET on progression-free survival remained the same (P < .0001 for both, log-rank test).

thumbnail image

Figure 3. Progression-free survival is shown after completion of therapy.

Download figure to PowerPoint

FDG-PET after 1 cycle vs after completion of therapy

There was excellent agreement between the findings of FDG-PET after 1 cycle and after completion of therapy (kappa statistic = 0.80; 95% CI = 0.61–0.98). Forty-three (91.5%) patients were concordant on FDG-PET after 1 cycle and after completion of therapy, and 4 (8.5%) patients were discordant (P = .05, McNemar test).

All patients who had a true-negative FDG-PET result after 1 cycle (n = 31) also had a true-negative FDG-PET after completion of therapy. There were 12 patients who had a positive FDG-PET result after completion of therapy compared with 16 patients after 1 cycle. Of these 4 patients who were positive after 1 cycle but who converted to negative after completion of therapy, 3 (2 DLCLs, 1 HD) have relapsed with times-to-progression of 7, 15, and 22 months, respectively (Table 4; Patients 2, 14, 16). By IPI and IPS criteria, 2 of 3 patients had intermediate-risk disease, and 1 had low-risk disease.

Of the 2 false-positive FDG-PET results after 1 cycle of therapy, 1 of 2 patients converted to true-negative after completion of therapy with resolution of infection at the biopsy site (Table 4; Patient 9). This patient has been in remission after 26 months. The other unproven positive result obtained after 1 cycle persisted in the posttherapy FDG-PET study (hilar lymph node; Table 4; Patient 4). The latter patient received consolidative radiotherapy to the chest and has been in remission after 36 months.

IPI (DLCL) and IPS (HD) vs FDG-PET after 1 cycle.

Because of the small number of patients low-intermediate and high-intermediate categories are lumped into 1 intermediate IPI category. After 1 cycle, among 17 PET-negative HD patients, 13 had low-risk, 2 had intermediate-risk, and 2 had high-risk disease by IPS criteria. Among 14 PET-negative DLCL patients, 9 had low-risk, and 5 had intermediate-risk disease (no high-risk patients) by IPI criteria. Among 6 PET-positive HD patients, 3 had low-risk, and 3 had intermediate-risk disease (no high-risk patients). Among 10 PET-positive DLCL patients, 7 had low-risk, and 3 had intermediate-risk disease. There were 2 false-positive FDG-PET results in this group, and both had low-risk disease. The 2-year PFS was 58.3% (95% CI = 29.3–78.9) for high-risk to intermediate-risk disease compared with 75.0% (95% CI = 56.2–86.6) for low-risk disease (P = .32, log-rank test). The median PFS was not reached for either risk group. Because of the 100% NPV of FDG-PET, a proper multivariate analysis of PFS including all known prognostic factors was not possible.


  1. Top of page
  2. Abstract

Increasing success of lymphoma therapy has now placed the prevention of long-term toxicities at the forefront of clinical management. In this context, a risk-adapted, individually tailored treatment for patients with DLCL and HD may provide an effective strategy for prevention of major late complications.

FDG-PET, as an indicator of treatment efficacy, has demonstrated encouraging results when used during or after chemotherapy, often despite contrary clinical or biochemical evidence of disease.15–21 In this study, we found a NPV (100%) and PPV (87.5%) for early FDG-PET obtained after 1 cycle of chemotherapy in both DLCL and HD. The vast majority of relapses occur within the first 2 years after therapy. In our group, the follow-up was >20 months in more than 85% of the patients in the early PET-negative group. Thus, our findings suggest that FDG-PET after 1 cycle offers a robust means for defining a patient population who is likely to benefit from treatment.

The results of this current study are in congruity with previous studies that were designed to predict therapy outcome early during chemotherapy.8–12 However, compared with prior studies, our study is unique in its design for evaluating response early after 1 cycle, whereas other studies were performed after 2 or more cycles of chemotherapy. The timing of treatment assessment may be critical, particularly, for patients who prove refractory to first-line therapy or relapse at a later time. These patients may benefit from early alternative therapy avoiding the complications of continued ineffective therapy. Likewise, the identification of patients who are likely to be cured by the first-line therapy may offer the potential of shortening the duration or intensity of treatment. Spaepen et al. reported that in 70 patients with aggressive non-Hodgkin lymphoma (NHL), FDG-PET at midtreatment was correlated with PFS and overall survival.9 None of the patients with persistent FDG uptake achieved a durable complete response (CR), whereas 84% of PET-negative patients remained in CR with a median follow-up of 37 months. In our study, the results were slightly different from this prior study as our data yielded no false-negative results. This difference may be due to the superior resolution of the PET scanner, the attenuation correction applied, and the higher specificity added by the integrated PET-CT system.

A clear definition of positive and negative lesions is crucial to obtain consistent results with less observer variability when using imaging probes as surrogate markers. In this regard, the present study offers a more finely tuned and strict set up definitions of the qualitative end points compared with other comparable studies in the literature.8–12 Mikhaeel et al. has recently reported that early interim FDG-PET after 2–3 cycles of therapy is an independent predictor of PFS in aggressive NHL.10 The estimated 5-year PFSs were 89%, 59% and 16% for the PET-negative group, the minimal residual uptake group, and the PET-positive groups, respectively. However, minimal residual uptake is a vague definition and is highly conducive to subjectivity. Likewise, other studies have not defined a sharp interpretation criteria in their study design.11, 12, 15 In our study, we used a dichotomous and well defined reading scale to avoid inconsistent readings.

Although various studies demonstrated that qualitative evaluation is sufficient for differentiating responders from nonresponders,7–12, 15, 16, 19 we endeavored to perform a ROC analysis based on using post 1 cycle SUVmax as a discriminator of the recurrence from remission status. The optimal post 1 cycle SUVmax cut point was determined to be 1.75 yielding a sufficiently high sensitivity and specificity for discriminating responders from nonresponders. There are currently only limited data using SUVs as a means to evaluate therapy response.17, 22 It has been reported that a SUVmax cutoff of 2.5 yielded a sensitivity and a specificity of 86% and 100%, respectively in the differentiation of sterile masses from active lymphoma using integrated PET-CT.22 In another study, the corresponding values obtained with a SUVmean cutoff of 3.0 were 100% and 93%, respectively.17 These cutoff values, however, may be different depending on the patient population and therapy modality. In our set-up, not only were all FDG-PET studies obtained before radiation, but also only 1 patient received radiation therapy after chemotherapy. Thus, higher cutoff values may have decreased the specificity of FDG-PET in the prediction of outcome. Among the nonrecurring lesions, the majority of lesions that exceeded the cutoff value were in the mediastinum or hilum where the SUVmax is usually between 2.0 and 2.5.

Although there was excellent agreement between the findings of FDG-PET after 1 cycle and after completion of therapy, in those instances of discordance, imaging after 1 cycle proved more accurate than after completion. This finding was similar to our previous results obtained on a dual-head imaging system7 as well as to those reported by a recent study.12 Hutchings et al. evaluated 77 HD patients after 2 and 4 cycles and after completion of chemotherapy with a median follow-up of 23 months. In the prediction of PFS, FDG-PET after 2 cycles was as accurate as that obtained later. Our results demonstrated a slightly higher NPV after 1 cycle compared with after completion of therapy. The explanation for a more favorable NPV obtained with the early scan would be that rapidity of achieving a complete response implies chemosensitivity of the tumor. Because high chemosensitivity usually translates into higher CR, this finding is probably a harbinger of sustained CR.23–25 In a seminal study of DLCL, Armitage et al. reported when therapy was adjusted according to the response after 3 cycles in NHL, 73% of the patients achieved a CR.23 The durability of remission in the rapidly responding patients was significantly longer than those who required more cycles to achieve CR. Similarly in HD, the rapidity of response to initial few cycles of chemotherapy was found to serve as a surrogate for ultimate outcome, reflecting both tumor burden and activity.26 However, these prior studies were based on tumor volume reduction measurements determined by CT, which can underestimate response rates, especially early during therapy. In a recent article, the response accuracy using International Workshop Criteria (IWC) was enhanced by the integration of FDG-PET into the IWC in patients with aggressive NHL.21

In a regression analyses, early interim FDG-PET was found to be a stronger predictor than established prognostic factors.9, 12 Similarly, in our study, it appeared that FDG-PET predicted PFS better than IPI and IPS, although we could not compare the results statistically because of the 100% NPV. In the early PET-negative group, there were 9 patients whose disease was in the intermediate or high-risk lymphoma by IPS and IPI, respectively. Likewise in the PET-positive group, of 10 patients who were in the low-risk group, 8 had a relapse. A major drawback of the IPI is that approximately half of the patients can be allocated into the intermediate-risk category, the group in which the treatment selection is the most challenging. Furthermore, 30% of the patients fit into the low-risk category and 20% to 40% do not survive for more than 5 years.4, 27

Although the high negative value obtained in this study is compatible with prior studies,8–12 it is important to understand that FDG-PET cannot exclude minimal residual disease due to the finite resolution limits of the PET systems. It is possible a small number of patients can still develop relapse during a longer follow up notwithstanding a median of 28 month follow up in our study when most relapses occur.

Ideally DLCL and HD should be evaluated in separate groups as these are 2 different disease entities with different biologic behavior and response profiles. Both, however, share the potential for cure. In our study, when patients with DLCL were separated from those with HD, the prognostic impact of FDG-PET after 1 cycle on PFS remained the same.

An early assessment of response with FDG-PET could provide the basis for selection of patients for alternative therapeutic strategies. Even if we were able to predict early that patients with a positive FDG-PET are destined to fare poorly with first-line therapy, it is possible that early alternative treatments, such as transplantation, may not result in any survival benefit. Indeed, an early positive FDG-PET may simply be a sad harbinger of a dismal prognosis reflecting inherent drug resistance. Nevertheless, an early negative FDG-PET may provide the potential to shorten therapy for curable lymphomas. In addition, a negative FDG-PET may allow the avoidance of radiation in patients with bulky disease. Conceivably, FDG-PET scanning may ultimately prove to be the most robust means for altering current treatment paradigms. Only extensive randomized studies will confirm the true potential of dynamic FDG-PET scanning.


  1. Top of page
  2. Abstract
  • 1
    Coleman M, Kostakoglu L Early 18F-labeled fluoro-2-deoxy-D-glucose positron emission tomography scanning in the lymphomas: changing the paradigms of treatments? Cancer. 2006; 107: 14251428.
  • 2
    Mauch PM, Armitage JO, Diehl V, Hoppe RT, Weiss LM. Hodgkin's disease. Philadelphia, PA: Lippincott Williams & Wilkins; 1999.
  • 3
    Miller TP, Dahlberg S, Cassady JR, et al. Chemotherapy alone compared with chemotherapy plus radiotherapy for localized intermediate- and high-grade non-Hodgkin's lymphoma. N Engl J Med. 1998; 339: 2126.
  • 4
    Shipp MA, Harrington DP, Anderson JR, et al. A predictive model for aggressive non-Hodgkin's lymphoma. The International Non-Hodgkin's Lymphoma Prognostic Factors Project. N Engl J Med. 1993; 329: 987994.
  • 5
    Hasenclever D;Diehl V. A prognostic score for advanced Hodgkin's disease. International Prognostic Factors Project on Advanced Hodgkin's Disease. N Engl J Med. 1998; 39: 15061514.
  • 6
    Shipp MA, Ross KN, Tamayo P, et al. Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning. Nat Med. 2002; 8: 6874.
  • 7
    Kostakoglu L, Coleman M, Leonard JP, Kuji I, Zoe H, Goldsmith SJ. PET predicts prognosis after 1 cycle of chemotherapy in aggressive lymphoma and Hodgkin's disease. J Nucl Med. 2002; 43: 10181027.
  • 8
    Jerusalem G, Beguin Y, Fassotte MF, et al. Persistent tumor 18F-FDG uptake after a few cycles of polychemotherapy is predictive of treatment failure in non- Hodgkin's lymphoma. Haematologica. 2000; 85: 613618.
  • 9
    Spaepen K, Stroobants S, Dupont P, et al. Early restaging positron emission tomography with (18)F-fluorodeoxyglucose predicts outcome in patients with aggressive non-Hodgkin's lymphoma. Ann Oncol. 2002; 13: 13561363.
  • 10
    Mikhaeel NG, Hutchings M, Fields PA, O'Doherty MJ, Timothy AR. FDG-PET after 2 to three cycles of chemotherapy predicts progression-free and overall survival in high-grade non-Hodgkin lymphoma. Ann Oncol. 2005; 16: 15141523.
  • 11
    Haioun C, Itti E, Rahmouni A, et al. [18F]fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) in aggressive lymphoma: an early prognostic tool for predicting patient outcome. Blood. 2005; 106: 13761381.
  • 12
    Hutchings M, Loft A, Hansen M, et al. FDG-PET after 2 cycles of chemotherapy predicts treatment failure and progression-free survival in Hodgkin lymphoma. Blood. 2006; 107: 5259.
  • 13
    Cheson BD, Horning SJ, Coiffier B, et al. Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas.NCI Sponsored International Working Group. J Clin Oncol. 1999; 17: 12441253.
  • 14
    Lister TA, Crowther D, Sutcliffe SB, et al. Report of a committee convened to discuss the evaluation and staging of patients with Hodgkin's disease: Cotswolds meeting. J Clin Oncol. 1989; 7: 16301636.
  • 15
    Spaepen K, Stroobants S, Dupont P, et al. Prognostic value of positron emission tomography (PET) with fluorine 18 Fluorodeoxyglucose ([18F]FDG) after first-line chemotherapy in non-Hodgkin's lymphoma: is [18F]FDG-PET a valid alternative to conventional diagnostic methods? J Clin Oncol. 2001; 19: 414419.
  • 16
    de Wit M, Bohuslavizki KH, Buchert R, et al. 18FDG-PET following treatment as valid predictor for disease-free survival in Hodgkin's lymphoma. Ann Oncol. 2001; 12: 2937.
  • 17
    Naumann R, Vaic A, Beuthien-Baumann B, et al. Prognostic value of positron emission tomography in the evaluation of post-treatment residual mass in patients with Hodgkin's disease and non-Hodgkin's lymphoma. Br J Haematol. 2001; 115: 793800
  • 18
    Weihrauch MR, Re D, Scheidhauer K, et al. Thoracic positron emission tomography using 18F-fluorodeoxyglucose for the evaluation of residual mediastinal Hodgkin disease. Blood. 2001; 98: 29302934.
  • 19
    Cremerius U, Fabry U, Neuerburg J, et al. Prognostic significance of positron emission tomography using fluorine-18-fluorodeoxyglucose in patients treated for malignant lymphomas. Nuklearmedizin. 2001; 40: 2330.
  • 20
    Zinzani PL, Fanti S, Battista G, et al. Predictive role of positron emission tomography (PET) in the outcome of lymphoma patients. Br J Cancer. 2004; 91: 850854.
  • 21
    Juweid ME, Wiseman G, Vose JM, et al. Response assessment of aggressive non-Hodgkin's lymphoma by integrated International Workshop Criteria and fluorine-18-fluorodeoxyclucose positron emission tomography. J Clin Oncol. 2005; 23: 46524661.
  • 22
    Freudenberg LS, Antoch G, Schütt P, et al. FDG-PET/CT in re-staging of patients with lymphoma. Eur J Nucl Med Mol Imaging. 2004; 31: 325329.
  • 23
    Armitage JO, Weisenburger DD, Hutchins M. Chemotherapy for diffuse large-cell lymphoma—rapidly responding patients have more durable remissions. J Clin Oncol. 1986; 4: 160164.
  • 24
    Engelhard M, Meusers P, Brittinger G. Prospective multicenter trial for the response-adapted treatment of high-grade malignant non-Hodgkin's lymphomas: Updated results of the COP-BLAM/IMVP-16 protocol with randomized adjuvant radiotherapy. Ann Oncol. 1991; 2: 177180.
  • 25
    Haq R, Sawka CA, Franssen E. Significance of a partial or slow response to front-line chemotherapy in the management of intermediate-grade or high-grade non-Hodgkin's lymphoma: A literature review. J Clin Oncol. 1994; 12: 10741084.
  • 26
    Carde P, Koscielny S, Franklin J, et al. Early response to chemotherapy: a surrogate for final outcome of Hodgkin's disease patients that should influence initial treatment length and intensity? Ann Oncol. 2002; 13 suppl 1: 8691.
  • 27
    Moller MB, Pedersen NT andChristensen BE. Factors Predicting Long-Term Survival in Low-Risk Diffuse Large B-Cell Lymphoma Am J Hematol. 2003; 74: 9498.