Introduction
Up until the last decade, the treatment options for metastatic renal cell carcinoma
(mRCC) patients were limited. mRCC is resistant to systemic cytotoxic chemotherapy
[1] and cytokine-based therapies like IFN-α and IL-2 resulted in modest response rates
and little survival benefit [2]. Over the past decade, the treatment of mRCC has changed
considerably with the introduction of targeted therapies and, more recently, immune
checkpoint inhibitors (ICI) [3]. Although the introduction of targeted therapies has
markedly improved patient outcome, they rarely induce complete responses, and most
patients eventually develop resistance to these therapies. Clinical trials with ICI
nivolumab (anti PD-1) and ipilimumab (anti CTLA-4) in mRCC reconfirmed the relative
tractability of this tumor type to immunotherapy. However, the objective response
rate of mRCC patients who received combination treatment of nivolumab and ipilimumab
is still only 42% and comes at the cost of substantial (although often manageable)
toxicity [4]. Therefore, further exploration of immunotherapeutic combination approaches
is warranted for the treatment of mRCC.
Recent insights have linked responses to immune checkpoint blockade to mutation burden
and the frequency of neo-antigens [5]. Vaccines aimed at priming or boosting T cell
responses to neoantigens may thus increase response rates to ICI [6]. Unfortunately,
the highly individualized nature of these neoantigens makes them hard to leverage
through therapeutic vaccination. Autologous tumor cell vaccination (ATV) is a strategy
to induce a specific immune response against tumor cells and their particular antigens,
including neoantigens, without the need for prior identification of actionable T cell
epitopes. Whole tumor cell vaccines have shown clinical and immunological activity
in mRCC patients [7–9], as well as in patients with other tumor types [10–12]. To
increase the immune response against autologous tumor cells (ATC), the whole cell
vaccine can be combined with adjuvants. We have demonstrated in the past that ATV
and BCG prolonged disease free survival in stage-II colorectal cancer and improved
survival in stage-III/IV melanoma patients, which correlated significantly with a
positive post-vaccination DTH response [13, 14]. Unfortunately, BCG is relatively
toxic as it can cause ulcerations [13–15]. The discovery that unmethylated cytosine-phosphate-guanine
oligodeoxynucleotides (CpG ODN) are the active elements in bacterial DNA and can directly
activate and induce maturation of B cells and plasmacytoid dendritic cells (pDC) has
led to the development of CpG ODN as treatment modality and vaccine adjuvant for infectious
diseases and cancer [16, 17]. Indeed, B-class CpG ODN (CpG-B) has been demonstrated
to enhance vaccine responses to hepatitis B, malaria and cancer [18–23].
We conducted a phase II clinical trial with the primary objective of investigating
whether the treatment with ATV, CpG-B and IFN-α was feasible and tolerable and resulted
in higher clinical response rates than IFN-α alone (by historical controls). Secondary
objectives were to assess progression-free survival and overall survival of treated
patients compared to historical data. Here, we report on the biological and clinical
efficacy of this experimental treatment.
Materials and methods
Patients
Patients with bi-dimensional measurable metastases of histologically proven RCC, and
in whom progression before or after nephrectomy had been demonstrated, were eligible
for this trial. Furthermore, a WHO performance status of 0 or 1 was required and patients
were only eligible when sufficient numbers of tumor cells were available for the production
of a minimum of three vaccines. Patients with a history of autoimmune- or antibody-associated
disease, prior malignancy, patients who were using immune suppressive drugs, or who
had undergone prior immunotherapy for metastatic disease (e.g., IL-2 or IFN-α treatment)
were excluded.
During the first month of therapy, the patients were seen bi-weekly. Thereafter, follow-up
visits started at E3 (see Fig. 1) and were scheduled every 12 weeks or at treatment
discontinuation due to disease progression. At each follow-up visit, the patients
were subjected to a physical examination including WHO performance status, blood panels
and a tumor measurement to define response which was assessed on the basis of a set
of “target lesions” chosen before the first vaccination. Response (at E3) was defined
with computed tomography (CT) scans according to the WHO criteria for response.
Fig. 1
ATV treatment scheme. E1, E2, E3 Evaluation 1, 2 and 3 (i.e., time of heparinized
blood collection), DTH delayed type hypersensitivity, Vac 1, Vac 2, Vac 3 vaccination
1, 2 and 3, IFN-α interferon alpha, s.c. subcutaneous, CpG
-B cytosine-phosphate-guanine Class B
The primary endpoint of the study was tumor response compared to historical data.
Secondary endpoints included toxicity, progression-free survival, overall survival,
the relation between DTH responses against tumor cells and clinical responses, and
the value of pre-vaccination tumor-specific T cell reactivity as a predictor of successful
immunotherapy.
Of note, IFN-α was the only treatment option for potentially eligible patients at
the start of this trial. However, the enrolment of this study was halted in 2006 when
sunitinib became available. Therefore, enrolment stopped at 15 patients instead of
41 as originally planned.
Vaccine preparation
Patients underwent a total nephrectomy or, if the kidney was already removed before
inclusion, a metastasectomy. The tissue that was not used for pathologic diagnosis
and staging was transferred to our vaccine production laboratory within 48 h of the
surgery. The tumor tissue was then dissociated as previously described [13, 24]. Briefly,
tumor tissue was cut into small pieces and subsequently incubated in a 0.1% DNase
I, 0.14% collagenase (Boehringer) solution. After 45 min incubation at 37 °C, single
cells were harvested and remaining tumor fragments again suspended in a DNase/collagenase
solution; this cycle was repeated 3–4 times, after which single cells were harvested
through a 100 µm gauze, a sample for bacteriology control was taken and viability
was tested using trypan blue exclusion. Cells were aliquoted (at 15–20 × 106 viable
cells per vial) and cryopreserved using a linear freezer. Vials were stored in liquid
nitrogen until vaccination. Prior to vaccination, the frozen tumor cells were irradiated
(20,000 rad), thawed, counted and assessed for viability. For each patient, we aimed
to produce as many vaccines as possible.
Vaccination procedure
Vaccination started 4–6 weeks after nephrectomy or metastasectomy. For every vaccine
0.7–1.3 ×107 viable autologous tumor, cells were used and 100 µg GM-CSF (Leukine;
Berlex Laboratories Inc.), 1 mg CpG-B (CPG 7909, Coley Pharmaceutical Group, inc.
Wellesley, MA 02481 USA) and 50 µg Keyhole Limpet Hemocyanine (KLH) (Calbiochem) were
added. Vaccines were administered intradermally. For individual evaluation of the
injection sites, different locations were used for each administration as previously
described for ATV in patients with colon carcinoma [13]. The local injection site
reactions were monitored and documented at each visit. All patients received 3 weekly
intradermal injections of the vaccine, followed by booster vaccinations every 3 months
for as long as the vaccines lasted and the disease did not progress. After the first
three vaccinations, patients were treated bi-weekly with 8 mg of CpG-B s.c. and 6
MU IFN-α s.c. three times per week for at least 3 months to enhance both innate immunity
and the effector phase of the specific immunity. To prevent additive toxicity and
to enable a separate observation of toxicity of CpG-B and IFN-α, these two compounds
were never administered on the same day. IFN-α was administered for a maximum of 1 year
and CpG-B for a maximum of 2 years or until disease progression, grade III/IV toxicity,
or death.
Sampling of peripheral blood
For immune monitoring, heparinized blood samples were taken from the patients before
start of therapy [evaluation 1 (E1)], at the day of the third vaccination [evaluation
2 (E2)] and 5–12 weeks after the third vaccination [evaluation 3 (E3)] (see Fig. 1
for treatment and evaluation scheme). PBMC were isolated by density centrifugation
(Nycomed AS, Oslo, Norway) and subsequently cryopreserved for later analysis as previously
described [25].
Antibodies and four-color flow cytometry
Peripheral blood lymphocyte, monocyte, peripheral blood dendritic cell (PBDC) and
monocytoid myeloid-derived suppressor cell (mMDSC) frequencies and activation status
were assessed before and during treatment by four-color flow cytometry staining. Cell
surface antibody staining of PBMC was performed in PBS/0.1% BSA/0.02% Sodium-Azide
for 30 min at 4 °C. The following antibodies were used: FITC, PE, PerCP-Cy5.5 or APC-labeled
Abs directed against human CD3, CD4, CD8, CD11c, CD14, CD15, CD16, CD19, CD25, CD27,
CD33, CD45, CD45RO, CD45RA, CD56, CTLA4, CD123, HLA-DR, PD-1 (all BD Biosciences),
CD11b, FoxP3 (eBioscience, San Diego, CA), CD40 (Beckman Coulter, Marseille, France),
Fab-M-FITC (Southern Biotec, Birmingham, AL), and blood DC antigens BDCA1, BDCA2,
BDCA3 (all from Milteny Biotec, Bergisch Gladbach, Germany) and MDC8 (a kind gift
from Dr. E.P. Rieber, Dresden, Germany) and matching isotype control antibodies. Intracellular
FoxP3 and CTLA-4 staining was conducted with the anti-human FoxP3 staining kit (eBioscience,
San Diego, CA) according to the manufacturers’ protocol. Stained cells were analyzed
on a FACScalibur (BD Biosciences) using Cell Quest software.
T cell subset and differentiation state definitions
Naive CD4+ or CD8+ T cells (Tn) were defined as CD27+CD45RO− cells, effector T cells
(Teff) as CD27−CD45RO+, central-memory CD4+ T cells (Tcm) as CD27+CD45RO+ cells and
effector-memory cells (Tem) as CD27−CD45RO+ [26]. Tregs were defined as CD3+CD4+CD25hi,
and FoxP3+. As FoxP3 has also been described to be transiently up-regulated on dividing
(activated) effector T cells [27–29], we also analyzed FoxP3 expression within these
activated (effector-like) T cells, which we defined as CD4+CD25intermediate (CD4+CD25int)
cells. For Treg gating procedures, we refer to Huijts et al. 2017 [30].
Myeloid subset definitions
PBDC frequencies were determined on the basis of expression of BDCA or MDC-8 markers:
cDC1 was detected as CD11c+CD14− BDCA3+ [31]. DC belonging to the so-called conventional
DC2 (cDC2) subset was identified as: CD11chiCD19−CD14−BDCA1/CD1c+; non-classical monocytes
[32] were detected as CD11c+CD14loMDC8+ (also previously known as 6-sulfo LacNAc+
or SLAN-DC [33, 34]) and pDC were detected as CD11c−CD14−CD123hiBDCA2+ [35]. Classical
monocytes were defined as CD14hi and mMDSC were defined as Lin−CD14+HLA-DRneg/lo cells
[36]. Activation status of the above-mentioned cDC and pDC subsets was determined
by calculating the median fluorescence index (MFI) of CD40 expression by dividing
the median fluorescence (med. fl.) of the CD40 antibody by the med. fl. of the isotype-control
antibody. For detailed gating procedures, we refer to Santegoets et al. [37], including
its supplementary materials.
Tumor-specific T cell reactivity and IFN-γ ELISA
Tumor reactivity of T cells in peripheral blood before, during and after ATV was assessed
by IFN-γ secretion. To this end, ATC suspensions were used as stimulator cells in
an overnight stimulation assay. Tumor cell suspensions were thawed and resuspended
in IMDM medium (Lonza, Verviers, Belgium) supplemented with 10% FCS (Hyclone, Amsterdam,
The Netherlands), 100 I.E./ml sodium penicillin (Yamanouchi Pharma, Leiderdorp, The
Netherlands), 100 µg/ml streptomycin sulphate (Radiumfarma-Fisiopharma, Naples, Italy),
2.0 mM l-glutamine (Invitrogen, Breda, The Netherlands) and 0.01 mM 2-mercapoethanol
(Merck, Darmstadt, Germany; hereafter referred to as complete medium). Next, 50,000
ATC and 100,000 PBMC were cultured either alone or together for 20 h in complete medium
in a 96-well round-bottom plate, after which supernatants were harvested and frozen.
IFN-γ levels were determined by ELISA (sensitivity 1 pg/ml) according to manufacturer’s
instructions (M1933, Sanquin, Amsterdam, The Netherlands). IFN-γ levels are given
as the mean IFN-γ concentration in pg/ml per 1 × 10e6 PBMC/ml ± SD of triplicate wells.
Responses were considered positive when the amount of IFN-γ produced by PBMC in response
to ATC was at least twice the amount of the sum of IFN-γ detected in overnight unstimulated
PBMC or ATC mono-cultures, and was at least 10 pg/ml. Mean ATC-specific IFN-γ concentration
was calculated by subtracting the IFN-γ levels from ATC + PBMC alone from the IFN-γ
levels from ATC:PBMC co-cultures.
Delayed type hypersensitivity (DTH) response assessment
Several studies have demonstrated that the size of a DTH response after autologous
tumor cell vaccination strongly correlates with recurrence and survival of cancer
[7, 14, 39]. The presence of a DTH response to tumor cells is a measure of immunogenicity
and reflects the efficacy of the vaccination and the general immune status of the
patient. In our study, DTH skin tests were performed prior to the first vaccination
and at the time of the third vaccination. To this end, 2 × 106 ATC and 5 µg KLH were
injected intradermally into separate sites and 48 h later the DTH response was evaluated
by measuring the induration by the “Sokal pen method”[38]. In brief, a line was drawn
with a pen 1–2 cm away from the margin of the skin test reaction towards the lesion.
The pen was held at a 45° angle and the pen was advanced with moderate pressure until
resistance was met. This procedure was repeated four times. Next, induration was measured
between opposing points by centimeter ruler. Total induration was calculated as an
ellipse (πab) and given in mm2.
Statistics
Sample size
The response rate for IFN-α treatment, taken from historical data, was assumed to
be 10% [2]. A response rate of 25% for the combination of ATV, CpG-B and IFN-α was
expected and initially planned as primary outcome of this study. To detect this increase
with a two-sided test (α = 0.05) and 80% power, 41 evaluable patients had to be enrolled
in the study. Abortion of the enrolment after 15 patients disabled us to perform a
reliable response rate analysis. However, the sample size proved sufficient to obtain
significance levels in the immune monitoring analyses.
Statistical analyses
Differences between immune parameters before (E1) and during treatment (E2 or E3)
were analyzed with a two-tailed paired t test. A two-tailed unpaired t test was used
for the analysis of the difference in induration area caused by ATC and KLH between
patients with clinical benefit (CB) (CR, PR and SD) and patients without clinical
benefit (NCB) (PD) and for the analysis of the difference in tumor-specific T cell
reactivity between CB and NCB before and during treatment. Microsoft Excel (version
2010) and GraphPad Prism (Version 6.02) were used for all graphs, tables and analyses.
Differences were considered significant when p ≤ 0.05.
Results
Patient characteristics
Between April 2004 and April 2006, 30 mRCC patients were assessed for eligibility
to enter this single-centre, single-arm phase II trial. 90% (n = 27) of the patients
had sufficient tumor material for vaccine preparation. 15 patients were eventually
enrolled and 12 patients were excluded due to death prior to the first vaccination
(n = 5), additional malignancies (n = 3), the absence of progression or metastasis
(n = 3) or poor renal function (n = 1). We refer to Table 1 for detailed characteristics
of the 15 enrolled patients.
Table 1
Patient characteristics
Patient
Age
Sex (M/F)
ECOG status
MSKCC risk factors
Histologic subtype
Metastatic sites
Response
Response (in months)
1
61
M
1
3
ccRCC
Lungs, liver, MLN, RLN
PD
–
2
74
M
0
2
ccRCC
Lungs, liver, MLN
PD
–
3
45
M
0
2
ccRCC
Lungs, MLN
SD
3.6
4
69
M
1
3
ccRCC
Liver, MLN, RLN
PD
–
5
56
M
1
2
ccRCC
Lungs, bone, MLN, RLN
PD
–
6
52
M
1
2
ccRCC
MLN
PD
–
7
54
M
0
1
ccRCC
Lungs, MLN
PR
7
8
60
F
1
4
ccRCC/ sRCC
Lungs, bone, MLN, RLN
PD
–
9
57
M
0
2
ccRCC/ sRCC
Lungs, Skin, MLN, intramuscular
PD
–
10
60
F
0
1
ccRCC
Lungs
SD
5.1
11
62
M
0
2
ccRCC
Lungs, MLN, RLN
PR
6.5
12
71
M
0
2
ccRCC
Lungs, MLN
CR
146+
13
57
F
0
2
ccRCC
Lungs, liver, MLN, RLN
PD
–
14
73
F
0
2
ccRCC
Lungs, bone, RLN
SD
4.4
15
59
M
0
3
ccRCC
Lungs, bone, MLN
PD
–
M Male, F Female, ECOG Eastern Cooperative Oncology Group, MSKCC Memorial Sloan-Kettering
Cancer Center, ccRCC clear cell Renal Cell Carcinoma, sRCC sarcomatoid Renal Cell
Carcinoma, MLN Mediastinal Lymph Node, RLN Retroperitoneal Lymph Node, CR Complete
Response, PR Partial Response, SD Stable Disease, PD Progressive Disease
Clinical results
All 15 patients received the first three vaccines and were, therefore, included in
the current analysis. Objective clinical responses occurred in three patients, including
one CR and two PR. The CR is still ongoing for more than 12 years and PR lasted 6.5
and 7 months. The patient with the CR was alive at the last moment of follow-up (October
2017) and has not shown any signs of disease since the experimental treatment and,
therefore, never received any other form of therapy for RCC (See Fig. 2 for pre- and
post-treatment lung window CT scans). Three patients had SD which lasted between 3.6
and 5.1 months. The six patients with CR, PR and SD were designated patients with
CB. Nine patients developed PD and were designated patients with NCB (Table 1) and,
therefore, did not receive any booster vaccines. From the six patients with CB, four
patients received one and two patients received two booster vaccinations after which
there was either no ATV available anymore or disease progression occurred.
Fig. 2
Pre-treatment lung window CT scan of the patient that had a complete response shows
multiple solid nodules in both lungs (a). In the most recent lung window CT scan (approximately,
5 and a half years later), no solid lung nodules are identified (b)
Adverse events
Adverse events to the experimental treatment were relatively mild with grade one and
two fever and fatigue being the most common events. Elevated gamma-glutamyl transferase
and alkaline phosphatase (in six and three patients, respectively) were the most prominent
grade three or four adverse events that might have been related to the experimental
treatment. Furthermore, two cases of grade three or four anaemia were recorded and
grade three or four melena, fatigue, dizziness, disturbed balance, sensory neuropathy,
aphasia, hemiplegia were all recorded once (Supplemental table 1) which were all considered
to be unrelated to the experimental treatment (e.g., neurological adverse events were
found in one patient who suffered from brain metastases). Common local toxicity consisted
of induration and edema at the vaccination site. We did not observe any ulcerations
at the vaccination sites.
Delayed type hypersensitivity (DTH) response
None of 15 patients demonstrated a positive skin test in response to ATC before vaccination,
but 13 of them showed a positive skin test upon treatment. 12 of 15 patients demonstrated
a positive skin test in response to KLH after vaccination. The median induration in
response to ATC was 169.6 mm2 and to KLH was 201.1 mm2. Interestingly, we observed
a significant difference in the size of the DTH response between patients with CB
(i.e., stronger DTH reaction) and patients with NCB for ATC (p = 0.038), but not for
KLH (Fig. 3a).
Fig. 3
Delayed type hypersensitivity (DTH) response against autologous tumor cells (ATC)
and keyhole limpet hemocyanine (KLH) at the time of the third vaccination and DTH2
(E2) in mm2 (a). ATC-specific IFN-ɣ production (pg/ml) determined by ELISA at baseline
(E1), third vaccination (E2), and follow-up (7–14 weeks, E3) for patients with clinical
benefit [CB: stable disease (SD), partial response (PR) and complete response (CR)]
and patients with no clinical benefit [NCB: progressive disease (PD)] (b)
Tumor-specific T cell reactivity and IFN-γ ELISA
Similarly to our findings for DTH responsiveness to ATC, ATC recognizing circulating
T cells were revealed to be more frequent in patients who clinically benefitted from
the therapy compared to patients with NCB (i.e.three of four in CB versus two of seven
in NCB patients; p = 0.061 with two-sided Fisher’s exact test; data not shown). We
also found that the magnitude of the IFN-γ response was significantly higher in patients
with CB at E1 (p = 0.046) and at E3 (p = 0.042) but, interestingly, not at E2 (Fig. 3b).
B and T cell activation
No significant differences were observed in overall frequencies of circulating CD3+
T cells, CD19+ B cells, CD3−CD56+ NK cells, CD14hi classical monocytes and CD4+CD25hiFoxP3+
Tregs, nor in frequencies of circulating CD4+ T cells, CD8+ T cells and CD4+ or CD8+
T naive, Teff, Tcm or Tem cells following treatment (Supplemental figure. 1). However,
ATV delivery induced activation of B cells and T cells, as reflected by significantly
increased percentages of CD19+CD86+ B cells and of CD4+PD-1+, CD8+ PD-1+ and CD8+CTLA-4+
T cells following treatment (see Fig. 4). Yet, no correlations between baseline levels
or increases in activated B or T cell rates and treatment response were observed (data
not shown).
Fig. 4
Activation markers at baseline (E1), third vaccination (E2), and follow-up (7–14 weeks,
E3) on T cells (HLA-DR) (a), B cells (CD86) (b) and monocytes (CD40) (c). Activation
markers CTLA4 and PD1 are shown separately on CD4 + T cells (d and e) and CD8 + T
cells (f and g)
Treatment-induced changes in myeloid subset frequencies and activation states
Frequencies of cDC2, non-classical monocytes and pDC, but not of cDC1, decreased during
treatment (Fig. 5a–d). Maximal and significant decreases of cDC2 and non-classical
monocytes subsets were reached after three vaccinations and multiple injections of
CpG-B and IFN-α at E3. Significant decreases in pDC frequencies were observed at E2,
and were shown to be only transient as pDC levels were restored at E3 after the 3rd
vaccination and repeated CpG-B and IFN-α injections. Decreases in non-classical monocytes
and pDC frequencies were paralleled by increases in their activation status at E3,
as indicated in Fig. 5g, h by significant up-regulation of CD40 expression. In line
with this, decreases in rates of non-classical monocytes and pDC at E2 were only observed
in patients with CB, suggesting that indeed early decreases in frequencies of these
antigen presenting cell subsets were associated with their activation and the patients’
response to treatment (See Supplementary Fig. 3a–d). Finally, after a transient non-significant
increase at E2, mMDSC were significantly decreased at E3 in the CB group of patients
(Supplementary Fig. 2e).
Fig. 5
Frequencies and median fluorescence index (MFI) of the activation marker CD40 on cDC1
(a and e), cDC2 (b and f), non-classical monocytes (c and g) and pDC (d and h) over
treatment: baseline (E1), third vaccination (E2), and follow-up (7–14 weeks, E3)
Discussion
This phase II study in 15 patients with mRCC demonstrates that i.d. delivery of ATV
with ATC/CpG-B/GM-CSF and systemic CpG-B/IFN-α is feasible, immunogenic and clinically
active. All patients with CB developed a DTH response against ATC during the treatment
and the induration area in the skin of patients with CB was significantly larger than
in the patients with NCB, a difference that was not seen for KLH DTH suggesting an
association of the DTH response against ATC with clinical activity as we have demonstrated
before in stage III/IV melanoma patients [14]. We found relatively mild toxicity and
no treatment-related deaths. Importantly, CpG-B as intradermally administered adjuvant
(combined with GM-CSF) was shown to be safe and did not cause the ulcers that we observed
in our previous studies where we used BCG instead [13, 39]. Unfortunately, we were
not able to complete the pre-calculated enrollment level of 41 patients due to the
approval of sunitinib as standard treatment for the patient group with mRCC. This
has left us unable to perform a reliable assessment of the response rate (which was
one of our primary objective) and overall survival. Nevertheless, a response rate
of 20% (3/15 patients) was reached in this small cohort which is in line with our
expectations prior to the start of the trial.
We found that ATC-specific IFN-ɣ production before ATV was related to clinical outcome
(Fig. 3b). Interestingly, this difference in ATC responsiveness between patients with
CB and patients with NCB was no longer detectable in the peripheral blood at E2 due
to equally low IFN-ɣ levels in both patient groups, but reappeared at E3 with no detectable
response to ATC in the patients with NCB. This may be due to a previously described
phenomenon where ATC-specific T cells, upon their activation in the circulation, acquire
the ability to migrate to the effector sites (the vaccination and tumor sites) under
the influence of the immunotherapy and, thus, are transiently less abundant in the
peripheral blood [40, 41]. This is corroborated by the relative (increase in) size
of DTH to ATC at E2, which was elevated over E1 and significantly higher in patients
with CB, further supporting the presence of an anti-tumor response that could localize
to the site of tumor cell presence in the patients who responded to ATV administration
(Fig. 3a). Our data thus suggest that ATC-specific IFN-ɣ production in vitro before
treatment may be a predictive biomarker for treatment response, whereas the size of
the DTH to ATC may be a first indicator of effective immunization against ATC (and
treatment response) as early as 2 weeks after the first ATV.
Further immune monitoring revealed a decrease over treatment in cDC2, non-classical
monocytes and pDC frequencies, which may also reflect recruitment of these myeloid
effector subsets to effector sites. The frequencies of pDC (which are directly targeted
by CpG-B via TLR-9) returned to baseline levels at E3, whereas cDC2 and non-classical
monocytes frequencies further declined. Interestingly, the decline in non-classical
monocytes and pDC frequencies was associated with a significant increase in their
activation status (measured by CD40) at E3 which may be attributed to the bi-weekly
CpG-B injections that started 1 week after E2. Also, mMDSC frequencies were significantly
decreased by E3 in patients with CB.
This study shows that i.d. administration of ATV has clinical activity in a subset
of patients but it may be even more interesting to look at possible combinations of
this vaccination approach with other treatment modalities for mRCC. It was found for
example that sunitinib has the ability to modulate the anti-tumor immune response
by reversing MDSC accumulation and Treg elevation in RCC [42]. In a randomized study
in metastatic renal cell carcinoma, it was demonstrated that the combined treatment
with avelumab (anti-PDL1) and axitinib resulted in an improved progression-free survival
as compared to sunitinib alone, suggesting that tyrosine kinase inhibitors have at
least an added effect to immunotherapy (Abstract ESMO LBA6_PR ‘JAVELIN Renal 101).
Therefore, when combined with therapeutic vaccines, sunitinib may help to overcome
tumor-induced immune escape leading to increased numbers of tumor-infiltrating lymphocytes
and tumor-specific CD8+ T cells, as well as enhanced tumor eradication and improved
survival, as was previously shown in murine models [43, 44]. In patients, however,
a recent randomized controlled phase III trial showed that there was no clinical benefit
from the addition of a multipeptide cancer vaccine to sunitinib and even though T
cell responses and monocyte counts were only assessed in a subset of the patients
in the combination arm (which left the authors unable to compare these parameters
between the combination and sunitinib monotherapy), the authors actually found evidence
for a potential immune inhibitory role of sunitinib [45]. This seems to be in stark
contrast with previous findings by us and others showing favorable immune modulation
in patients with mRCC [35, 46], but may be due to differential effects of sunitinib
in peripheral blood versus the tumor microenvironment [47]. As depicted in Fig. 4,
we also observed an increase in the number of activated B cells (by CD86) and T cells
(by CTLA-4 and PD-1). CTLA-4 and PD-1 are up-regulated when T cells become activated,
which in the case of PD-1 has also been linked to neo-antigen specificity [48], and
since the introduction of ICI, we know that selectively inhibiting these immune checkpoints
can result in unprecedented anti-tumor activity. Moreover, it has become clear that
response to ICI relies on T cells reactive to highly individualized neo-antigens [49].
In contrast to allogeneic or peptide-based vaccines, autologous vaccines cover all
the personal (neo-) antigens that the tumor may express. In this light, ATV approaches
may be able to enhance the sensitivity to ICI. Obstacles to further clinical development
of ATV, however, are the fact that they do not qualify as a pharmaceutical product
which makes funding of further development challenging, and the fact that vaccine
production is laborious and dependent on the availability of tumor material. Nonetheless,
90% of the patients that were assessed for inclusion in this trial had enough tumor
material available for successful vaccine production.
Patients received IFN-α after the third vaccination (E2) and since the clinical response
evaluation (E3) was 12 weeks after the start of this active drug, it is possible that
the clinical responses were (in part) the result of this treatment. Unfortunately
we are unable to discriminate between the clinical effects of the ATV and the IFN-α
in this trial. However, clinical responses in mRCC were previously demonstrated by
others in vaccine-based clinical trials (without IFN-α) with cultured CD83 + blood
DC loaded with autologous tumor cell lysates [50], DC pulsed with MUC1-derived peptides
[51], a multipeptide cancer vaccine [52] and RNA coding for tumor-associated antigens
[53]. As for the immunomonitoring data, all the findings obtained at E1 and E2 are
the result of the ATV and that the findings at E3 can be attributed to either ATV
or IFN-α, or both.
In conclusion, our data show that our ATV approach combined with IFN-α in mRCC is
feasible, well tolerated and clinically active. Moreover, this treatment approach
induced DTH responses against ATC and systemic activation of circulating PBDC and
T cells in mRCC patients. In addition, preexisting ATC responsiveness of circulating
T cells may be predictive for clinical outcome following treatment. Based on our observations
in these 15 patients, further investigation of our ATV approach and current treatment
modalities is indicated to improve response rates in this patient group.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Supplementary material 1 (PDF 276 KB)