ALBERT

Description: We have performed 2,000+ Fluorodeoxyglucose PET (PET) scans for multiple myeloma (MM) staging and restaging at our facility since October 2001. While the usefulness of the PET scan for MM is reported by us and others elsewhere, we have reviewed our list of “incidental” but important findings, some of which are unique or occur more commonly with MM patients and some of which are common to many patients, the more common of which we present below: Occult infection occurs very commonly in patients with MM due to direct tumor effect on the immune system and to medication (especially high dose dexamethasone). Of the 2000+ PET scans for MM done at our facility, 300+ infections (about one half occult) have been detected by PET. These most commonly involve central lines (septic thrombophlebitis), diskitis, lung (either bacterial or fungal), and periodontal abscesses (a source of bacteremia in these patients), though non-catheter related spontaneous septic thrombophlebitis also occurs. While related to MM, extramedullary disease from MM (EMD) is seen more commonly with PET than MRI since PET has a wider field of view. In our series of 172 patients with both baseline PET and MRI, PET detected EMD in 11/11 patients versus MRI detection rate of 7/11. On follow-up, PET detected three times more EMD (29/31 sites) than MRI (9/31sites). Detection of EMD by PET at baseline is a profoundly negative prognostic factor in our series (12 month EFS 20% for EMD+ vs. 47% for EMD−, p = 0.002, and OS 42% for EMD+ vs. 70% for EMD−, p=0.005, n=48). Following tumor response in hypo- or non-secretory disease is difficult by standard prognostic factors (SPFs). FDG PET results were actively used for clinical management in 10 of 11 non-secretory patients at our facility. A major pitfall of PET scanning for MM comes from the use of steroids. Treatment with chemotherapy in general and steroids (i.e. prednisone or dexamethasone) in particular can result in a false negative PET scan by producing a profound but transient suppression of tumor metabolism. In addition, the hyperglycemic effect of the steroids produces competitive inhibition of FDG (a glucose analog) uptake, also causing suppression of FDG tumor uptake that can lead to an underestimation of disease. Second primary malignancies are frequently seen the MM age group. In our population, we have found unsuspected breast cancer, breast lymphoma, thyroid cancer, melanoma, colon cancer and lung cancer. In addition, we have found several functioning thyroid adenomas and premalignant colon polyps. Being a whole body imaging device for metabolism, FDG PET is a powerful though nonspecific tool for imaging that can not only help stage and restage the malignancy under question but which can yield a plethora of additional clinically useful information if used properly and with its limitations understood.

Description: Total Therapy II (TT2), consisting of intensive induction with non-cross resistant VAD, DCEP, CAD (with collection of PBSC) and DCEP.; followed by tandem autotransplants in support of melphalan 200mg/m2; followed by consolidation chemotherapy every 3 months x4 ; followed by interferon maintenance, has complete accrual of 668 patients with upfront randomization to +/− Thalidomide. Collection goal was 20 million CD34+ cells/kg with CAD. This number allows for tandem transplants with support of 〉5 x 106 CD34 cells/kg, additional cells were stored in case of engraftment problems, additional transplants in the future for relapsed disease and in case of development of MDS. Patient characteristics are outlined in Table 1. 319 patients collected 〉20 x106 CD 34+ cells/kg with CAD (median 24.9 million, range 20–127), 2 proceeded to second collection with DCEP (median 5.005 million, range 0–10.01. Of this group 310 patients (97%) underwent single and 234 (73%) tandem autotransplant. 135 patients collected 10 – 20 x106 CD 34+ cells/kg with CAD (median 16.35 million, range 10.09–19.93), 13 patients had second collection with DCEP (median 4.92 million, range 1.23–11.68), 5 patients underwent third collection with growth factor (median 2.61million, range 0.12–12.85). Of this group 130 patients (96%) underwent single and 94 (70%) tandem autotransplant. 46 patients collected 5 – 10 x106 CD 34+ cells/kg with CAD (median 7.82 million, range 5.1–9.97), 20 patients had second collection with DCEP (median 9.785 million, range 0.93–27.14), 8 patients underwent third collection with growth factor (median 2.02million, range 0.3–11.8). No patient underwent a fourth collection attempt. Of this group 44 patients (96%) underwent single and 29 (63%) tandem autotransplant. 63 patients collected 70%) to proceed to autologous transplant. Table 1: Patients depending on initial collection result with CAD Overall 〉20* 10–20* 5–10* 0–5* 0* * (x 106 CD34+ cells collected) N 603 319 (53%) 135 (22%) 46 (8%) 63 (10%) 40 (9% Age 〉65 19% 14% 17% 30% 30% 28% CA 31% 30% 33% 30% 40% 25% Thal Arm 49% 41% 56% 63% 57% 50% CRP 〉4 8% 7% 10% 7% 8% 10% B2M 〉4 31% 26% 36% 33% 37% 43% LDH 〉190 29% 27% 33% 26% 30% 40% BM 〉30% PC 64% 62% 67% 63% 67% 60%

Description: Fluorodeoxyglucose Positron Emission Tomography (PET) scanning can rapidly detect and quantify changes in tumor metabolism. We present early results from the first 47 patients enrolled in Total Therapy III for multiple myeloma using PET scanning and correlating with serum M protein. Since we have previously demonstrated that diffuse uptake patterns do not have prognostic significance, we have limited our assessment of these 47 patients to those with one or more PET-defined focal lesions (FL) at baseline who have had at least one follow-up PET scan (25 patients). Of this group of 25, 10 have had a second follow-up scan. These whole body PET scans were performed at baseline and at time points corresponding to immediately after two cycles of VDT-PACE, landmarks LMK1 and LMK2. Patients with complete resolution of FL on PET are considered to be in PET-CR. We find that 17/25 (68%) had a decrease in number of FL from baseline to LMK1 and 8/10 (80%) had a decrease in number of FL from baseline to LMK2. In addition, 6/25 (24%) were in PET-CR by LMK1 and 3/10 (30%) in PET-CR by LMK2. Similarly, a decrease in serum M-protein was seen in 16/25 (64%) from baseline to LMK1 and in 8/10 (80%) from baseline to LMK2. At baseline, 10 patients had serum M protein levels of 0.01 g/dl or less. Of these, 8 (80%) had PET-defined focal lesions, demonstrating PET’s ability to monitor tumor metabolism in the face of hypo- or non-secretory disease. Of these, 7 have been enrolled sufficient time to have a second PET scan, with 6/7 (86%) demonstrating a continuing decrease in number of FL in response to treatment. At baseline, PET detected 5 of 47 patients with extramedullary disease (EMD), a finding of severe adverse prognostic significance at baseline. Of these patients, all had FL at baseline PET and all had decreasing number of FL on first follow-up PET scan (two have had a second follow-up scan, one improving and one PET-CR). One of these patients with EMD is nonsecretory. From these data, we conclude that functional imaging with PET scanning offers important and unique information about early tumor response that improves patient management by demonstrating rapid change in response to treatment that not only parallels serum M-protein response but which also is reliable in hypo- or non-secretory disease as well as in detection and monitoring of extramedullary tumor, an ominous condition that is particularly difficult to detect in the hypo- or non-secretory patient.

Description: Macrofocal lesions (FL) in the medullary space are a hallmark of mm readily detected by MRI scans, both at diagnosis and at relapse. In this study we evaluated, among 25 patients enrolled in TT II, with FL at relapse and a preceding clinical CR (by M protein and marrow examinations) the relationship to presence and location of FL at diagnosis. On relapse, 17/25 (68%) had fewer FL, 5/25 (20%) had the same number of FL, 3/25 (12%) had more FL than on baseline examination. There was a general trend for the larger FL at baseline to still be present on relapse (Table 1). Despite the trend to fewer focal lesions on relapse compared to baseline, 11/25 (44%) of patients presented with FL in new sites that were not present on baseline. Finally, in relapsing patients, 4/25 (16%) presented with new sites of extramedullary tumor (EMD). These data establish that although there are usually fewer MRI-defined FL on relapse than on baseline examination, the vast majority of FL on relapse represent new areas of macrofocal disease that were not present at diagnosis. MRI FL Persisting from Baseline to Relapse (n=25) Size (cm) Baseline FL Present upon Relapse % 〈 0.5 1/2 50 0.5–1cm 8/10 80 1–2 cm 13/18 72 〉 2 cm 10/16 63 All Sizes 26/36 72

Description: Multiple myeloma results in both diffuse infiltration and/or focal lesions (FL) of the marrow. MRI detects FL long before X-ray. We routinely perform MRI of skull, entire spine, pelvis, sternum and shoulders. MRI-defined number of FL (#FL) is an adverse prognostic variable for both EFS and OS for patients treated according to Total Therapy 2, 2nd in importance only to cytogenetic abnormalities (CA) (univariate and Cox multivariate models using standard prognostic factors, CA, and #FL). Of these patients who also had FDG PET (PET), baseline analysis revealed the same axial skeleton FL as on MRI (r=0.47, p