Monitoring response to treatment in melanoma patients: Potential of a serum glycomic marker
Article first published online: 24 OCT 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 122, Issue 6, pages 1374–1383, 15 March 2008
How to Cite
Selvan, S. R., Dillman, R. O., Fowler, A. W., Carbonell, D. J. and Ravindranath, M. H. (2008), Monitoring response to treatment in melanoma patients: Potential of a serum glycomic marker. Int. J. Cancer, 122: 1374–1383. doi: 10.1002/ijc.23155
- Issue published online: 21 JAN 2008
- Article first published online: 24 OCT 2007
- Manuscript Accepted: 14 AUG 2007
- Manuscript Received: 27 JUN 2007
- Hoag Hospital Foundation
- autologous vaccine therapy;
- dendritic cells;
- glycomic markers;
- serum total gangliosides;
- antiganglioside IgM antibodies
A mechanistic marker correlating with tumor progression and clinical response is useful for assessing therapeutic response and determining the course of therapy. Since serum-total-ganglioside (sTG) and antiganglioside-IgM antibody levels reflected tumor progression, the feasibility of utilizing sTG for assessing the response to immunotherapy of metastatic-melanoma was tested. Patients (n = 34) were immunized with dendritic cells cocultured with irradiated, IFNγ-treated autologous tumor cells admixed with GM-CSF. Levels of sTG and antiganglioside-IgM antibody titers were measured in sera of vaccine recipients at 0, 4 and 24 weeks of treatment. Based on sTG-level, whether lower (L) or higher (H) than the mean + 1 SD of normal and healthy volunteers on weeks 0, 4 and 24, patients were categorized into cohorts-I (LLL, n = 16), II (HHL/HLL, n = 4), III (LLH/LHH/LHL, n = 7) and IV (HHH/HLH, n = 7). The cohorts were regrouped as sTG- downregulators (sTG-DR; n = 20) and upregulators (sTG-UR; n = 14). These two cohorts differed significantly in their overall (p < 0.012) and progression-free (p = 0.0001) survival post-treatment. 43% sTG-UR died within 39 months, with a median survival of 39 months, whereas 61% of the sTG-DR survived for 48 months. Both endogenous and vaccine-induced antiganglioside-IgM antibodies appeared to regulate sTG levels. Nonresponders had increased sTG with no or low IgM antibody response. The sTG level is regulated within 24 weeks post-treatment and therefore, may serve as an ideal biomarker for assessing therapeutic responses in patients. Clinical correlations of sTG indicate that sTG-downregulating therapy may be an effective treatment strategy for melanoma. © 2007 Wiley-Liss, Inc.
The National Cancer Institute's SEER (Surveillance, Epidemiology and End Results) data has estimated that 59,940 new cases of invasive cutaneous melanoma will be diagnosed and 8,110 patients will die of melanoma in 2007.1 Over the past 50 years, there has been a 619% increase in annual diagnosis and 165% increase in annual mortality for invasive cutaneous melanoma.2 Patients with metastasized and recurrent melanoma are at a higher risk with a need for personalized therapy. There is also a compelling need to identify specific and logistic serum marker(s) for monitoring therapeutic response that correlates with disease status and clinical response (overall survival, OS or progression-free survival, PFS).
The need for personalized therapy for melanoma is recognized by the complexity in the behavior of a heterogeneous tumor cell population,3 diversity in antigens or class of antigens of individual biopsies,4 selective release or shedding of immunosuppressive tumor antigens in tumor microenvironments,5 orchestration of immunosuppression,6 sites of primary lesion and metastasis, varying patterns of tumor recurrence and diversity in hormonal and immune status, and in the HLA profiles of the patients.7 A reliable therapeutic response marker would enable changing therapeutic strategies for nonresponders.
Glycomics of human cancer encompasses gangliosides, a family of antigens which orchestrates suppression of cell-mediated immunity.6 The level of gangliosides in circulation is commonly referred to as serum total gangliosides (sTG). The sTG comprises free gangliosides (present as micelles and aggregates), bound gangliosides (present as immune complexes and ganglioproteins, and bound to albumin and lipoproteins) and membranous vesicles or exosomes.5, 8, 9, 10, 11 The sTG could be a potential marker of tumor load and progression in patients with melanoma,12 sarcoma,13 prostate,14 ovarian,15, 16 colon,17 and pancreatic18 cancers. Furthermore, necrosis induced by cryoablation of liver metastasis of colon carcinoma17 resulted in an increase in the level of sTG, followed by augmentation of antiganglioside IgM antibodies and subsequent decline in the level of sTG. In stage III melanoma patients (n = 70), the sTG levels increased after treatment with an allogeneic melanoma cell vaccine.12 Following augmentation of antiganglioside IgM antibodies, the sTG levels in vaccine recipients returned to pretreatment levels. Strikingly, these patients survived significantly longer than those whose sTG levels remained high.12 These observations led to the hypothesis that sTG levels could serve as a therapeutic response marker of tumor progression and the levels of sTG together with the levels of antiganglioside IgM antibodies may provide a mechanistic understanding of the response to therapy.
In addition, cancer patients exhibit innate endogenous13, 14, 15, 16, 17, 18 and vaccine-induced12 IgM antibody response to one or more gangliosides, found in tumor cells, which can also serve as a complementary marker to sTG in order to evaluate tumor progression in patients. We hypothesized that a direct glycomic marker like sTG, with immunosuppressive potential, and an indirect counter-marker, namely IgM antibodies against tumor-associated gangliosides, would be ideal “combinatorial markers” to monitor the response to therapy, provided one or more of these markers unequivocally correlate with OS and/or PFS. This retrospective investigation determined the feasibility of using sTG as a marker of clinical response to predict the outcome of an autologous melanoma cell vaccine therapy.
Material and methods
An outline of the active-specific immunotherapy for metastatic melanoma patients
The treatment protocol (clinical trial no., NCI-V01-1646) involves pulsing DC, generated from each patient, with an irradiated pure tumor cell line established from biopsies of the respective patients.19 Tumor cell-pulsed DC were suspended in 500 μg of GM-CSF for subcutaneous administration, weekly for 3 weeks, then monthly for 5 months (Fig. 1). Sera collected before and after immunization were aliquoted and frozen at −80°C.
Clinical characteristics of patients
The patients had (i) distant metastatic melanoma with or without measurable disease based on RECIST20; (ii) ECOG performance status: 0–2; (iii) standard medical criteria for entry into phase II trials: hematologic, renal, hepatic, etc.; (iv) adverse events determined by the toxicity criteria (CTC 2.0, NCI, 2003); (v) age range between 21 and 79; (vi) sera collected before and after immunization on weeks 0, 4 and 24; and (vii) informed consent for pheresis and autologous tumor cell vaccine therapy. The trial protocol was approved by Hoag Hospital Review Board. The total number of patients enrolled in this trial was 53 but this investigation was restricted to 34 patients who had serum samples on weeks 0, 4 and 24 for analyses.
Developing cell lines from tumor biopsies of individual patients
The tumor biopsies were collected and processed in RPMI-1640 medium with iron-supplemented calf serum (7.5%, v/v) and fetal bovine serum (7.5%, v/v) as described earlier,21 and the tumor cell lines (TC) were established as detailed elsewhere.21, 22, 23, 24 Melanoma cell lines were characterized by determining the expression of a panel of antigens including S-100, HMB45/gp100-cl, Melan-A/MART-1, MAGE-1, Tyrosinase, Mel-5 (TRP-1 and TRP-2), HLA-1 and HLA-2.23, 25 Once tumor cell lines were established and expanded to 150 × 106 cells, they were treated with IFNγ for 3 days with (1000 U/ml; ACTIMMUNE, InterMune, Brisbane, CA). The dose was selected based on empirical assessment of dose required for optimal expression of HLA antigens as determined in our earlier investigation.26 This was a part of the stipulated treatment of melanoma cell lines per Investigational New Drug (IND) protocol approved by FDA. The treated cells were harvested, irradiated (100 Gray) to arrest 100 % growth,27 and cryopreserved until pulsing with autologous DC. Before incubating with dendritic cells, irradiated tumor cells had an average cell number of 7.9 × 107 ± 1.7 × 107 and percent viability of 77 ± 12.
Generation of DC
DC were generated after density gradient centrifugation (Ficoll) of pheresis product and placed into T-225 flasks for monocyte enrichment using the adherence technique, as described earlier.19, 28 Final preparation of the TC-DC vaccine involved incubation (overnight at 37°C) of irradiated tumor cells, obtained from each patient, with autologous DC at a ratio of 1:1 and cryopreservation into aliquots. Just prior to each vaccination, an aliquot of TC-pulsed DC were suspended in GM-CSF (500 μg). Patients received an average TC-DC dose of 1.6 × 107 ± 0.8 × 107 cells with a percent viability of 78 ± 11.
Estimation of sTG
The levels of sTG were measured as lipid-bound sialic acid in 100 μl of serum, as described in detail earlier.18 For routine analyses, pooled sera from untreated melanoma stage IV patients were used as the positive control, and pooled and aliquoted sera of age- and sex-matched, normal healthy volunteers, along with varying concentrations of human serum albumin (Instituto Grifols, S.A. Barcelona, Spain), were used as negative controls.
Definition and terminologies used for patient cohorts
For categorization of patient cohorts in this investigation, the sTG levels on weeks 0, 4 and 24 of vaccine administration were compared to those of the normal and healthy population. The sTG levels in age-matched, healthy volunteers (n = 42), who never had any form of neoplastic disease, ranged from 12.10 to 22.00 mg/dl, with a mean and SD of 16.57 ± 3.13 mg/dl.12 The level of sTG was used to denote the cohort of vaccine recipients. Low (L) was used to designate the sTG level of patients on weeks 0, 4, or 24, if the values were same as or lower than the normal mean (16.57) plus SD (3.13) (=20.70 mg/dl) of healthy volunteers. Patients who had a low level of sTG on weeks 0, 4 and 24 were designated as LLL. Patients who had high levels of sTG (if the value was 10% higher than the mean plus SD (20.7 mg/dl) on weeks 0, 4 and 24) were designated as HHH in order to signify that the sTG level was high (H) on weeks 0, 4 and 24.
Measurement of antiganglioside IgM titers
Serum antiganglioside IgM antibodies were measured against melanoma-associated gangliosides GM2, GD3, GD2, GD1a and GD1b. The incidence of these gangliosides in human melanoma tumor biopsies was found to be 100% for GM2, GD2 and GD3.29, 30 The ELISA protocol was reported earlier.31, 32 The homogeneity of the gangliosides was ascertained before coating onto microtiter plates.32 Aliquoted frozen sera from patients and from healthy, age-matched controls were used after diluting (1/100) with 4% HSA (Grifols Biologicals, Los Angeles, CA) in PBS, pH 7.4 and incubating at 37°C for 30 min.31 Sera were further diluted to 1/200 to 1/12800 using a detergent-free blocking buffer with 4% HSA in PBS. Serum aliquots were prepared from the pooled sera of patients who had high antiganglioside IgM titers (Positive control) and normal and healthy volunteers older than 50 years (Negative control). Using Microsoft Excel software, the absorbancy values of the microtiter plates were calculated.32 In healthy volunteers, the titers of the antibodies were as follows: GM2 (<3000); GD2 (<600); GD3 (<1000); GD1a (<400); GD1b (<1600). If prevaccine patients had antibody titers higher than that reported for normal, they were designated as “high preimmune titer”33 or “endogenous IgM response.”14 The vaccine-induced response in vaccine recipients was calculated as follows: If the value obtained after week 4 or 24 postvaccine was 3-fold or more than the prevaccine titer for that patient, then the antibody titer was referred to as “induced-antibody titer”. If the titer remained the same as or increased but was less than 3-fold of the preimmune value after vaccine, then it was designated as “no change” or “no IgM response.”
Survival curves were estimated by the Kaplan-Meier method. Statistical analyses of sTG levels and antiganglioside antibodies were performed by Wilcoxon rank-sum test. Univariate analysis by log rank test was used to determine the survival differences among patient cohorts. All tests were two-sided. Results were considered significant if the probability value was less than 0.05. Progression-free survival (PFS) was the interval between the end of vaccination and recurrence or last follow-up. Overall survival (OS) was the interval between serum collection (week 0) and death or last follow-up. Serum collection coincided with treatment protocol (Fig. 1).
Characteristics of patient population
The melanoma patient population (n = 34; 21 males and 13 females; age 21–79 years) included those with measurable (MD; n = 8) or nonmeasurable (NMD; n = 26) disease (based on RECIST criteria) at the start of vaccine therapy. All of the patients were surgically treated and most of them had also previously received one or more of the following therapies: radiation, radiofrequency ablation, gamma knife, chemotherapy and biotherapy (interferon, GM-CSF, IL-2, BCG, peptide vaccine). One of the minimum requirements for receiving autologous melanoma cell vaccine therapy was that the patients had stopped all therapies for at least 4 weeks. No clinically significant adverse toxicity was observed related to the vaccine product other than local skin reactions in most patients, and rash or hypersensitivity reaction to GM-CSF in few patients.
Categorization of cohorts based on serum total gangliosides levels
To determine the response to TC-DC vaccine with GM-CSF in individual patients, sTG levels (mg/dl) were measured in patients on weeks 0 (prevaccine), 4 and 24 (postvaccine). Earlier, we have noted that the sTG levels in age-matched, healthy volunteers (n = 42) who never had any form of neoplastic disease ranged from 12.10 to 22.00 mg/dl, with a median of 16.30 mg/dl, and a mean and SD of 16.57 ± 3.13 mg/dl.12 In comparison, untreated stage III melanoma patients (n = 70) had significantly (p < 0.05) higher levels of sTG than that of the normal sTG.12 The untreated stage IV melanoma patients (n = 79) (data unpublished) also had significant (p < 0.0001) levels of sTG compared to those of the normal volunteers. The values ranged from 13.8 to 36.6 mg/dl with a median of 19.25 mg/dl and a mean ± SD of 20.70 ± 5.40 mg/dl.
Based on LOW or HIGH values of sTG on weeks 0, 4 and 24, the vaccine recipients were categorized into 8 groups (LLL, HLL, HHL, LLH, LHH, LHL, HHH and HLH), classified into 4 cohorts, and divided into 2 major categories as sTG-downregulators (sTG-DR) and upregulators (sTG-UR) (Table I). sTG-DR category includes those patients with sTG levels low or high at the start of vaccine but declined on week 4 and/or week 24 post-treatment. If the patients with low sTG levels at the start of vaccine showed high levels after week 4 and/or week 24, they were included under sTG-UR category. In addition, sTG-UR category had sTG levels either high throughout the course of treatment or increased from low levels to high levels on week 4 and/or week 24. Table II shows the sTG values, mean ± SD and median of different cohorts.
Cohort I: sTG levels remained low from prevaccine (week 0) to postvaccine (weeks 4 and 24) (LLL);
Cohort II: sTG levels were high at week 0 (prevaccine) and became low either at week 4 and 24 (HLL) or at week 24 (postvaccine) (HHL);
Cohort III: sTG levels were low at prevaccine (week, 0) but became high either at week 4 (LHL) and/or week 24 (LHH and LLH);
Cohort IV: sTG levels remained high at all weeks (HHH) or decreased to L at week 4 and returned to high levels on week 24 (HLH).
|Cohorts||No. of patients||Postvaccine sTG Level (weeks)||sTG regulation|
|Cohort||No. of patients||Definition||sTG levels in serum|
|0 week||4 week||24 week|
|Mean ± SD||Median||Mean ± SD||Median||Mean ± SD||Median|
|I||16||LLL||16.8 ± 2.1||16.7||17.7 ± 2.2||17.7||16.2 ± 2.1||15.6|
|II||4||HLL/HHL||24.0 ± 2.9||23.2||21.0 ± 2.1||20.2||18.0 ± 3.1||19.0|
|III||7||LLH/LHH/LHL||17.5 ± 2.0||17.3||22.0 ± 5.1||22.5||21.6 ± 3.9||21.7|
|IV||7||HHH/HLH||26.8 ± 4.8||26.7||26.7 ± 6.5||24.9||27.1 ± 11.1||24.5|
Table III describes the sTG levels in sTG-DR and sTG-UR categories on weeks 0, 4 and 24 for each patient as LLL, HLL, HHL, LLH, LHH, LHL, HHH and HLH. Furthermore, the disease status of each patient was recorded as measurable (MD) or nonmeasurable (NMD) at the start of the vaccine, and progressed disease (PD) or not (NPD) during vaccine (6 month duration) and after vaccine (7 to 12 or 18 months from vaccine start date). Table III shows that in the sTG-DR group, more patients had nonmeasurable disease (NMD) (18/20) at the start of the vaccine, and 13/20 did not progress during 12 months after vaccine start date. The sTG-UR group was different in that 6 out of 14 had measurable disease (MD), and 11 out of 14 had disease progression during 12 months after the commencement of the vaccine. Information presented in Table III was the basis of a broader classification of the entire patient population based on sTG levels as sTG-DR and sTG-UR. This categorization primarily distinguishes the efficacy of the treatment in bringing down the sTG level in vaccine recipients, which is the central theme critical to this investigation. In order to appreciate the efficacy of the vaccine in lowering the sTG levels in some patients, a detailed clinical history as well as characteristics of the sTG-DR and sTG-UR patients are presented in Table IV, parts A and B, respectively. These two tables show the disease (site/date of diagnosis) and treatment prior to, during and after vaccine therapy of each major category of patients. It should be noted that serum was collected from each patient 4 weeks after “no treatment window” period. The sTG levels at this point (week 0) of some patients were L while that of others were H. Vaccine treatment either maintained the low level of sTG as in LLL or failed to control the sTG levels as in LLH or LHH. Similarly vaccine treatment failed to bring down the sTG level as in HHH or did effectively bring down the sTG level as in HHL or HLL. This information provides necessary background for personalized therapy for each patient and the effect of the vaccine either to bring down or to keep the sTG levels high during the course of treatment.
|S. no.||Patient ID||Age/sex||sTG category||sTG levels (weeks 0, 4, 24)||Disease status at start of vaccine||Disease status|
|Within 12 months since vaccine start||Within 12 months since vaccine completion|
|S. no.||Patient ID no.||Vaccine date||Disease and treatments (site/date of Dx and Tx)|
|A: Recipients who are sTG downregulators|
|1||HH1763||7/03-12/03||Arm (1998): surgery, biochemo; LN and neck (8/01): surgery, peptide vaccine||None||None|
|2||HH1822||5/03-11/03||Thigh (1998): surgery, IFN; thigh (12/02): surgery; thigh and LN (1/03): surgery, XRT||None||None|
|3||HH1845||3/03-8/03||Calf (9/95); groin and LN (5/00): surgery, GM-CSF; neck (3/01): GM-CSF; LN (5/02): surgery||None||Soft tissue (10/03): surgery|
|4||HH1852||4/03-11/03||Back (1998); Back and LN (7/02): Surgery; LN (9/02): Surgery, IL-2, XRT||Eye (6/03): Surgery; Brain (8/03): GK||Brain (6/06): GK; Brain (2/07): GK|
|5||HH1878||3/03-8/03||Scalp and LN (6/97); parotid (2000): biochemo; small bowel (10/02): surgery, GM-CSF||None||None|
|6||HH1897||5/03-11/03||LN (8/02): surgery, IFN, chemo; LN (2/03): surgery; arm and LN (2/03): XRT||None||None|
|7||HH1952||1/04-7/04||Scapula and LN (5/01); Shoulder (5/02): Surgery; LN (5/02): Surgery, IFN; LN (12/02): Surgery; LN (6/03): Surgery; LN (8/03): Surgery; LN (12/03): Surgery, GM-CSF||LN (4/04): Surgery||Sub-cutaneous and Abdomen (2/05): Celebrex, GM-CSF; Scapula (6/05): Surgery, Biochemo; Back (11/05): Surgery, Biochemo, Chemo|
|8||HH1984||4/04-10/04||Chest (4/00): surgery; LN (5/02): surgery; spleen and liver (6/02): surgery, biochemo, IL-2; liver (7/03): surgery||None||None: (maintenance leukine)|
|9||HH2004||6/05-12/05||Ankle (9/01): surgery; LN (10/01): surgery, IFN; ankle (6/03): surgery; liver (7/03): questionable; LN (12/03): surgery, peptide vaccine||None||None|
|10||HH2030||11/04-5/05||Arm and back (7/00): surgery; regional recurrence (7/02): chemo; brain (10/02): craniotomy, chemo; LN and arm (3/03): biochemo, surgery, dendritic cell vaccine; regional met (3/04): surgery; 2nd sub-cutaneous lesion (3/04): surgery, MEL-43 trial; 2 new lesions (7/04): surgery||LN, soft tissue (2/05): surgery||None: (MEL-41 vaccine, leukine)|
|11||HH2036a||4/05-9/05||Neck (1993): Surgery; LN (9/01): Surgery; LN (04/04): Surgery, GM-CSF; Neck (8/04): Surgery; Neck (11/04): Surgery, XRT||Lung and LN (6/05): Surgery||Scalp (5/06): Surgery; Lung (6/06): XRT; Hip (12/06): XRT|
|12||HH2057||8/05-2/06||neck (4/03): surgery; LN (5/04): surgery, GM-CSF||None||None|
|13||HH2134||9/05-2/06||Scapula (6/95): surgery; back (11/04): surgery; LN (5/05): surgery; bone (8/05): questionable||None||None|
|14||HH2136||12/05-6/06||Bicep (8/97): surgery; LN (3/01): surgery; LN (7/01): surgery, GM-CSF; LN (8/01): surgery; spleen and liver (10/04): IL-2, chemo, RFA; spleen (5/05): surgery; brain (12/05): GK||None||brain, lung and abdomen (9/06): craniotomy, IL-2, chemo|
|15||HH1992||1/05-6/05||Back and LN (1998): Surgery, IFN; LN (2001): Surgery, Chemo; LN (11/03): Surgery, Biochemo||None||None|
|16||HH2034||3/05-9/05||LN (9/03): Surgery, IFN; Retroperitoneum (1/04); Mediastinum (4/04): Surgery, IL-2, Chemo; Brain (1/05): Craniotomy||Lung (7/05)||Brain and Adrenal (12/05): Chemo, Craniotomy; Progressive Disease (10/06): Nexavar|
|17||HH1766a||12/02-6/03||Arm (12/97); Arm (1998-2001): Biochemo; Arm and Mesenteric Mets (11/01): Surgery, GM-CSF||None||Liver (5/04): RFA, GM-CSF; Liver and LN (4/05): Chemo|
|18||HH1922||12/03-6/04||Lung (2/02): BCG, Surgery; Small Bowel (5/03): Surgery; Mesenteric Mass (6/03): Surgery, BCG Re-induction; Liver (9/03): Surgery, RFA||None||None|
|19||HH1987||1/05-7/05||Shoulder (5/98): Surgery; LN (6/03): Surgery; LN and Soft Tissue (10/03): Surgery, Biochemo; 2nd 1o Toe (6/04): Surgery, BCG||None||None|
|20||HH1953||12/03-6/04||Scalp (9/02): Surgery, Chemo; Scalp (11/02): Surgery; Lung and Scalp (7/03): Thoracotomy, Surgery, GM-CSF||None||None|
|B: Recipients who are sTG upregulators|
|21||HH1918||10/03-3/04||Irregular Mole Arm (1996); Arm (9/00): Surgery; LN (10/00): Surgery, IFN; LN (1/01): Surgery, GM-CSF; Thigh (11/01): Surgery, Chemo; Retroperitoneum (2/02): Surgery; Retroperitoneum and Groin LN (4/03): Surgery; Renal Bed (7/03): Cyberknife||Kidney (12/03): Cyberknife||Brain (6/04): GK, Temodar, Craniotomy|
|22||HH2009a||11/04-5/05||Scapula and Shoulder (10/94): Surgery; LN (1/04): Surgery, IFN, Chemo||LN (3/05): Surgery||Abdomen (8/05): Surgery, IFN; Widespread Disease in Retroperitoneum (11/05): Biochemo, GM-CSF|
|23||HH1868||4/03-10/03||Cheek (10/94); LN (1/97): Surgery, IFN; Neck (5/99): XRT, Chemo; Liver and Femur (8/01): Liver-RFA, Femur-XRT, Chemo; Femur (8/02): Surgery, GM-CSF; Brain (3/03): GK; Liver Spots (3/03)||Cutaneous and Brain (5/03): GK; Liver, Bone and Lung (6/03)||Brain and Bone (8/04): GK, XRT, Chemo; Liver, Bone, Abdomen and LN (9/05): IL-2; Bone (4/07): Surgery, IL-2|
|24||HH1972a||9/04-3/05||Calf (1999); Groin and LN (N/K): Surgery; Groin (1/02): Surgery, XRT; Pelvis (11/02): Surgery, BCG; Thigh (2/03): Surgery, BCG; Thigh and Abdomen (7/03); Anal Verge (7/03): Colonoscopy; Small Bowel (8/03): Surgery; Thigh and Groin (8/03): Surgery, Biochemo; Thigh (2/04): Surgery, IL-2, GM-CSF||Retroperitoneum (11/04): Surgery||LN (4/05): Biochemo; Widespread Disease (7/05)|
|25||HH2135||11/05-5/06||Irregular Mole Chin (1998); Chin (11/00): Surgery; Neck LN (7/01): Surgery, XRT, Chemo; Regional Recurrence (4/05): Surgery, Biochemo||None||None|
|26||HH1898a||10/03-7/04||Mediastinum (10/01): Biochemo, Surgery, Biochemo; Mediastinum (12/02): Surgery||Brain (2/04): Craniotomy||Brain (8/04): GK|
|27||HH2147||4/06-10/06||Scalp and LN (8/04): Surgery, IFN; Scalp and Neck (11/04): Surgery; Spleen (6/05): Surgery, Biochemo; LN Minimal Disease (12/05)||None||Lung Minimal Disease (5/07): No Tx|
|28||HH1858||2/03-8/03||Scalp (1989); Scalp (1991): Surgery; Scalp (1993): Surgery; LN (8/99): Surgery, Biochemo; Neck and Chest (4/01): Surgery, Allogenic Vaccine, GM-CSF; Thigh and Calf (8/02): Surgery||Subcutaneous (6/03): Surgery||Thigh (9/03): Surgery; Thigh (11/03): Surgery; Subcutaneous Sternum (1/06): Research Tx; Back (2/06): Surgery; Breast (4/07): MDX10, Autologous Vaccine|
|29||HH1937a||1/04-7/04||Chest Wall (5/98): Surgery; LN: Surgery; Chest and LN (6/03): Surgery||None||Brain (4/05): GK, Temodar, GM-CSF, Craniotomy; Abdomen (7/06): IL-2, Chemo; Liver (10/06): Chemo; Carcinomatosis (3/07): Chemo|
|30||HH1896a||5/03-11/03||Calf (5/02); Leg and Groin (1/03): Surgery; Sub-Cutaneous Lesions (2/03): Surgery||Leg (6/03): Surgery; Numerous sub-cutaneous recurrences; Lung (9/03): Stable||Leg (10/04):XRT, Chemo|
|31||HH1945a||3/04-9/04||Buttock (2/02); Eye (3/02): Surgery, Biochemo, IL-2, GM-CSF; Brain (6/03): GK; Neck, LN and Sub-Cutaneous Lesion Rt. Flank (7/03): Surgery||Brain, Lung and Neck (6/04): Palliative||Progressed rapidly multiple sites: XRT|
|32||HH2019||8/04-2/05||Back (1997); LN (1997): Surgery; LN (3/98): Surgery; Neck (4/98): Surgery; Arm (7/98): Surgery; Neck (2/99): Surgery; Elbow (4/99): Surgery; Back (4/00): Surgery; Back (11/00): Surgery; Brain (3/03): GK, Biochemo; Brain and Mesenteric Nodes (5/03): GK, Chemo, IFN, GM-CSF; Brain (6/04): GK||Brain (1/05): Craniotomy||Abdomen (3/05): IFN; Brain (10/05): GK; LN (5/06): Surgery, Chemo, IFN; Neck (12/06): Tomotherapy, Chemo|
|33||HH2071||2/05-8/05||Knee (6/01): Surgery; LN (6/02): Surgery, IFN; Liver (8/03): Surgery; Back (11/03): Surgery; Neck (3/04): Surgery, IFN; Lung (7/04): IL-2; LN (8/04): Surgery, IFN||Lung and Tibia (5/05)||Bone (11/05): NCI trial (TIL, IL-2, MART-1 Vaccine); Arm pit and Rib (2/06): Surgery; Lung (8/06): Surgery, Chemo|
|34||HH1957||8/04-1/05||Leg (1999): Surgery; Leg (5/02): Surgery; Thigh and Knee (8/03): Surgery; Liver (6/04): Questionable||None||Thigh and Knee (2/05): Surgery, Vaccine; Thigh (3/06): Surgery|
Clinical relevance: sTG-DR versus sTG-UR
As indicated above, the sTG-DR group showed whether sTG level remained low throughout vaccine treatment [LLL], or reduced consequent to vaccine treatment [HHL and HLL]. In the sTG-UR group, the sTG levels remained high throughout vaccine treatment with a fall on week 4 [HLH], or without a fall on week 4 [HHH], or increased in spite of vaccine treatment [LLH, LHH and LHL]. The statistical significance of the data is summarized in Table V. Table V (part A) points out the differences in the disease progression within 12 months since vaccine start. The difference between sTG-DR and sTG-UR groups was significant (p < 0.01). However, there was no significant difference between prevaccine-NMD and prevaccine-MD, irrespective of whether they belong to sTG-DR or sTG-UR. Table V (part B) shows the differences in the disease progression within 12 months since completion of the vaccine. The difference between sTG-DR and sTG-UR groups was highly significant (p < 0.001), but there was no significant difference between prevaccine-NMD and prevaccine-MD. Interestingly, sTG-DR and sTG-UR cohorts differed significantly in patients with prevaccine-NMD. Such a difference between sTG-DR and sTG-UR was not observed in the prevaccine-MD group.
|Patient characteristics||Patient cohorts||Disease progression within 12 months||Fisher exact test p value|
|A: Disease progression since vaccine start|
|sTG-DR (n = 20)||LLL/HLL/HHL||13||7||0.017|
|sTG-UR (n = 14)||LLH/LHH/LHL/HHH/HLH||3||11|
|PreVac NMD (n = 26)||N/A||14||12||NS|
|PreVac MD (n = 8)||N/A||2||6|
|B: Disease progression since vaccine completion (i.e. postvaccine)|
|sTG–DR (n = 20)||LLL/HLL/HHL||14||6||0.0004|
|sTG-UR (n = 14)||LLH/LHH/LHL/HHH/HLH||1||13|
|PreVac NMD (n = 26)||N/A||13||13||NS|
|PreVac MD (n = 8)||N/A||2||6|
|PreVac NMD||LLL/HLL/HHL (n = 18)||13||5||0.0016|
|LLH/LHH/LHL/HHH/HLH (n = 8)||0||8|
|PreVac MD||LLL/HLL/HHL (n = 2)||1||1||NS|
|LLH/LHH/LHL/HHH/HLH (n = 6)||1||5|
In view of the striking difference between the sTG-DR and the sTG-UR groups, the overall survival (OS) of these two groups was examined since the start of the vaccine and postvaccine progression-free survival (PFS), respectively. The results were striking, as presented in Figure 2a, which shows OS of vaccine recipients, and Figure 2b, which shows PFS of vaccine recipients. Indeed, the log rank test revealed that both OS (p = 0.012) and PFS (p = 0.0001) are highly significant, indicating the survival relevance of postvaccine sTG down and upregulation in the patients with metastatic melanoma. In the sTG-DR groups, 2 out of 20 patients died, whereas, in sTG-UR group 6 out of 14 died. In all patients who had died, progression occurred within 12 months after completion of the vaccine. Since the start of the vaccine, 7 out of 20 patients in the sTG-DR group showed disease progression, whereas, 11 out of 14 patients showed progression within 12 months in the sTG-UR group (Table III).
Serum TG and antiganglioside IgM antibodies
In order to understand whether there was any correlation between sTG level and serum antiganglioside IgM antibodies, which have been implicated to play a major role in downregulating sTG level,12, 17 the titers of IgM antibodies against major melanoma gangliosides were measured. Table VI compares antiganglioside IgM antibody response with changes in the profile of sTG in various cohorts. High preimmune IgM antibody titers were observed in 7 out of 16 patients in cohort I (LLL), 3 out of 4 patients in cohort II (HLL/HHL), 3 out of 7 patients in cohort III (LLH/LHH/LHL), and 1 out of 7 patients in cohort IV (HHH/HLH). The preimmune titer of the anti-GM2 IgM antibody ranged from 100 to 4,400; high preimmune titers (>3,000) in 3/34 patients; anti-GD2 IgM antibody ranged from 100 to 6,200; high preimmune titers (>600) in 1/34 patients; anti-GD3 IgM antibody ranged from 100 to 3,900; high preimmune titers (>1000) in 0/34 patients; anti-GD1a IgM antibody ranged from 100 to 6,400; high preimmune titers (>400) in 11/34 patients; anti-GD1b IgM antibody ranged from 100 to 3,900; high preimmune titers (>1,600) in 5/34 patients. Presence of high preimmune antiganglioside IgM antibody titers in some melanoma patients is well known.12, 33 Some of the patients showed both high preimmune and vaccine-induced IgM antibody titers for one or more gangliosides. For example, HH1992 had induced-antibody against all gangliosides studied. In cohort I, 11/16 patients; in cohort II, 3/4 patients; in cohort III, 2/7 patients; and in cohort IV, 5/7 patients had vaccine-induced IgM antibodies. Table VI also compares the IgM response rate in sTG-DR and sTG-UR groups. In sTG-DR group (n = 20), 50% had high preimmune IgM and 70% had induced-IgM antibodies against gangliosides, whereas, in sTG-UR group (n = 14) only 29% had high preimmune IgM and 50% had induced-IgM antibodies.
|IgM antibody response||sTG-DR||sTG-UR|
|Cohort I: LLL (n = 16)||Cohort II: HLL/HHL (n = 4)||Cohort III: LLH/LHH/LHL (n = 7)||Cohort IV: HHH/HLH (n = 7)|
|(A) High preimmune Ab only||1||1||1||1|
|(B) Induced Ab only||7||1||2||5|
|(C) Decreased Ab only||0||0||0||0|
|(D) High preimmune and induced Ab||4||2||0||0|
|(E) High preimmune and decreased Ab||2||0||2||0|
|(F) No response||2||0||2||1|
|Total high preimmune Ab (A + D + E)||10/20 (50%)||4/14 (29%)|
|Total induced Ab (B + D)||14/20 (70%)||7/14 (50%)|
|Total no response (F)||2/20 (10%)||3/14 (21%)|
Table VII provides the details of the nature of antiganglioside IgM antibodies observed in each cohort of sTG-DR and sTG-UR categories. High preimmune and vaccine-induced IgM antibodies against GM2, GD2, GD3, GD1a and GD1b are presented for each cohort. Total number of high preimmune IgM antibody response in four cohorts ranked as follows: I > II > III > IV. Although similar ranking was observed for vaccine-induced IgM antibodies (I > II > III), cohort IV showed more number of induced-antibodies than cohort II and III.
|Antiganglioside IgM antibody response||sTG-DR||sTG-UR|
|Cohort I: LLL (n = 16)||Cohort II: HLL/HHH (n = 4)||Cohort III: LLH/LHH/LHL (n = 7)||Cohort IV: HHH/HLH (n = 7)|
Proof of concept
Of the four cohorts categorized in this study, cohort II (HHL and HLL) is unique in that the high level of prevaccine sTG did decline effectively to normal range by week 24, associated with immunization, suggesting that the vaccine protocol impacted this cohort. Supportingly, in this group, vaccine has markedly induced antiganglioside IgM antibody response. Only 4 out of 34 patients belong to this cohort of vaccine responders. Cohort I has also benefited by the vaccine treatment. Sixteen out of 34 patients belong to this cohort. In this cohort I (LLL), the prevaccine sTG levels were low but maintained a steady state until the end of treatment (24 weeks). Either the vaccine would have kept the sTG levels low, or shedding of the gangliosides may not be as high as in cohort IV. The first possibility is strongly supported by the observations that cohort I had highest number of vaccine-induced IgM antibody response than other groups; some of the patients in this group showed induced-IgM antibodies for all (e.g., HH1992) or four (e.g., HH1852) gangliosides. Since IgM antibodies are known to be potent clearing agents for sTG,12, 17 the induction of the antibodies is obviously a result of patients' response to the vaccine. In contrast, cohort IV (HHH and HLH) shows patients who failed to respond to vaccine treatment in that the prevaccine levels of sTG remained high and the treatment did not bring down the levels, suggesting tumor progression. Similarly, cohort III (LLH, LHH) did not respond to treatment in that the sTG levels increased, indicative of tumor progression and failure of treatment strategy. Correlatively, the number of induced-antiganglioside IgM antibody response is the lowest for this group.
In cohorts III and IV, the titers of antibodies have not increased sufficiently to overcome the high level of sTG. Although cohort IV showed induced-IgM antibody response against all the 5 gangliosides, the antigen-excess state suggests that the induction is not sufficient to bring down the levels of sTG. Two nonresponders were identified by lack of IgM antibody response as well as by the increase in the levels of sTG at week 24. The paucity of free IgM antibody levels paralleled with antigen (sTG) excess state may reflect neutralization of the antibodies by the antigens resulting in the formation of immune complexes. This state may denote the tumor burden and the disease status of the patients. Therefore, sTG levels might be useful in signifying the total tumor volume in melanoma patients, i.e., a decrease in sTG levels might indicate clinical response after cancer vaccine treatments.
The members of cohorts III and IV might benefit from additional antimelanoma therapy that would augment the production of antiganglioside IgM antibodies, and reduce the levels of sTG. Evidently, both cohorts III and IV require an alternate therapeutic strategy to bring down the sTG levels. Either another adjuvant-based autologous or allogeneic vaccine modality that can induce more antibody response or passive immunotherapy or intravenous IgM antibody therapy to bring down sTG level may be beneficial. Any kind of alternate therapy or chemotherapy that could boost antibody production may also be beneficial for these patients. Periodic monitoring of the sTG levels is critical for nonresponders to therapies.
Clinical correlation and the proof of concept
This study establishes a clinical correlation between the 2 cohorts, namely sTG-downregulators (sTG-DR) and sTG-upregulators (sTG-UR), in overall or progression-free survival. The cohort of sTG-UR had a median survival rate of 39 months, and their survival rate during 48 months was significantly lower than that of sTG-DR (p < 0.012). Six out of 14 sTG-UR died within 39 months. Progression-free median survival of sTG-UR was only about 2 months, whereas 70% of the sTG-DR group survived for 24 months and 61% survived for 48 months. Therefore, the overall and progression-free survival rates differ significantly between sTG-DR and sTG-UR. This study documents that sTG levels could be a potential tool to monitor up or downregulation within 24 weeks post-treatment and sTG could be an ideal biomarker of therapeutic response for melanoma patients. Such a biomarker would be valuable to monitor efficacy of a therapy for patients and would provide a scope to change the therapy after 24 weeks. Any therapy that decreases sTG levels and acts as a sTG downregulator would be the best therapeutic modality for melanoma. If sTG levels do come down 24 weeks after a therapy, intravenous antiganglioside human IgM antibodies can be contemplated as a passive immunotherapy. Several antiganglioside IgM monoclonal antibodies were developed in the past34 and their potential use is envisioned in light of the present findings.
Incidence of melanoma gangliosides
Human melanoma-gangliosides befit the definition of neoantigens since gangliosides are poorly expressed by melanocytes, the progenitor cells of melanoma, but overexpressed after neoplastic transformation. These include GD3, GM2, GD2, GD1a and GD1b. Their incidence based on chromatographic analyses of 52 tumor biopsies varied from 100% (GD3 and GM2), to 71% (GD2) and to 20% (GD1a and GD1b).29 Culturing tumor cells from tumor biopsies further enhances their expression, particularly GM2 and GD2.4 We are planning to characterize the ganglioside profiles of the autologous cell lines. Our unpublished observations concur with earlier reports on the incidence of gangliosides in human melanoma.
Evolving concept on sTG as a biomarker of therapeutic response
Shedding of melanoma-associated GD3 into circulation is well known.35, 36 Such shedding of tumor gangliosides with tumor progression or with stages of disease was observed in sarcoma13 and after cryoablation-induced tumor necrosis in colorectal carcinomas.17 In this investigation, we have extracted gangliosides from the sera by a procedure modified to recover total gangliosides.18 In prostate and ovarian cancer patients,14, 15, 16 qualitative analyses of sTG reflected the gangliosides found in the tumor of patients, suggesting that sTG represents gangliosides shed from the tumor.
The gangliosides released from tumor cells enhance tumor formation.37, 38, 39 Increasing concentrations of sTG may impact tumor growth and patients' immune functions. Different gangliosides found in sTG are capable of suppressing a variety of immune functions (for details see Ref.40). Therefore, the success of any personalized therapy depends on decreasing sTG levels. In an earlier investigation on melanoma patients, it was noted that the mean sTG level of the intransit melanoma patients (n =17) increased from 18.57 ± 3.18 mg/dl pretreatment to 23.70 ± 5.50 mg/dl between weeks 2 and 16 after initiation of treatment (p2 < 0.0001). By week 24, levels had returned to prevaccine levels in seven clinical responders (18.10 ± 2.30 mg/dl versus 20.40 ± 3.20 mg/dl; p < 0.05), but remained higher than prevaccine levels, in 10 nonresponders (23.30 ± 5.10 mg/dl versus 17.20 ± 2.70 mg/dl). Similarly, the sTG levels of the patients with nodal metastases (n = 70) increased between weeks 2 and 16 after the first vaccine treatment; by week 24, they had returned to pretreatment levels in patients who survived more than 56 weeks, but remained significantly elevated (p < 0.01) in patients who survived less than 56 weeks. The study pointed out the potential of sTG levels as a tool for monitoring response to therapy in melanoma patients by week 24. We have extended the same protocol for stage IV melanoma patients (n = 34) in this study.
Clinical correlation and causation
“Clinical correlation does not imply causation” is a well-known dogma. This study is specifically designed to test the hypothesis that downregulating sTG would restore immune competence, improve survival and stabilize the disease, since tumor gangliosides released into circulation are immunosuppressive.40 It is important to note that the earlier observations made from stage III melanoma patients12 emphasize monitoring down or upregulation of sTG by week 24 post-treatment. Based on such monitoring, this study establishes a statistically significant difference in overall survival (p = 0.012) and progression-free survival (p = 0.0001) between sTG-DR and sTG-UR. Indeed, this correlation suggests causation. However, though the data based on 34 patients is not sufficient to substantiate that causation is indeed the basis for the correlation observed between sTG-DR and sTG-UR, it emphasizes the need to expand the study to a larger cohort. Such an approach would strengthen the discovery of sTG as a novel glycomic therapeutic response marker for melanoma.
We thank Ms. Kanoe Allen, Ms. Carol De Priest, Ms. Cheryl Peterson, Ms. Cristina De Leon and Ms. Robin Ellis for their assistance in conducting the clinical trials and data collections; Dr. Patric M. Schiltz, Ph.D., Hoag Cancer Center, Newport Beach, CA, for his role in preparing dendritic cells and the final vaccine product; Ms. Xing Ye, JWCI, Santa Monica, CA for her assistance with statistical analysis; Mr. Dana Ollestad, Hoag Hospital, Newport Beach, CA for his assistance with graphics; Dr. Donald L. Morton, M.D. and Dr. Sakunthala Muthugounder, Ph.D., JWCI, Santa Monica, CA and Ms. Carolyn Hendrix, R.N., Hoag Cancer Center, Newport Beach, CA, for their critical comments.
- 21Establishment of multiple tumor cell lines from a patient with melanoma: a simple method to control fibroblast growth. Clin Biotechnol 1991; 3: 237–42., .
- 25Characterization of autologous melanoma cell lines for vaccine therapy: pattern of expression of tumor-associated antigens. Proc Am Assoc Cancer Res 2004; 45: 286., , , .
- 28Cryopreserved dendritic cells maintain phenotype and ability to elicit an allogeneic response post recovery. J Immunother 2000; 23: 601., , , .
- 34Gangliosides as targets for monoclonal antibody therapy of cancer. In: BorrebaeckCAK,LarrickJW, eds. Therapeutic monoclonal antibodies. New York: Stockton, 1990. 75–94., .