Gene Therapy (2000) 7, 859–866
 2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00
www.nature.com/gt
ACQUIRED DISEASES
RESEARCH ARTICLE
Toxicity evaluation of replication-competent herpes
simplex virus (ICP 34.5 null mutant 1716) in patients
with recurrent malignant glioma
R Rampling1, G Cruickshank2, V Papanastassiou3, J Nicoll4, D Hadley3, D Brennan3, R Petty3,
A MacLean5, J Harland6, E McKie6, R Mabbs6 and M Brown6
1
Beatson Oncology Centre, Western Infirmary, Glasgow; 2Department of Neurosciences, Queen Elizabeth Hospital, Birmingham;
Institute of Neurological Sciences, Department of Neurosurgery, Southern General Hospital, Glasgow; 4University of Glasgow,
Department of Neuropathology, Institute of Neurological Sciences, Southern General Hospital, Glasgow; 5University of Glasgow,
Division of Virology, Glasgow; and 6University of Glasgow, Neurovirology Research Laboratories, Institute of Neurological Sciences,
Southern General Hospital, Glasgow, UK
3
The herpes simplex virus (HSV) ICP34.5 null mutant 1716
replicates selectively in actively dividing cells and has been
proposed as a potential treatment for cancer, particularly
brain tumours. We present a clinical study to evaluate the
safety of 1716 in patients with relapsed malignant glioma.
Following intratumoural inoculation of doses up to l05 p.f.u.,
there was no induction of encephalitis, no adverse clinical
symptoms, and no reactivation of latent HSV. Of nine
patients treated, four are currently alive and well 14–24
months after 1716 administration. This study demonstrates
the feasibility of using replication-competent HSV in human
therapy. Gene Therapy (2000) 7, 859–866.
Keywords: HSV1716; clinical trial; glioma therapy
Introduction
Glioblastoma remains one of the most formidable problems in cancer medicine. Following conventional therapy
with surgery, radiotherapy1 and chemotherapy2 the
median survival, following diagnosis, remains approximately 1 year. Progression following primary therapy is
associated with short-term survival (average 5 months).3
There is an urgent need for novel treatments and it is
accepted that new regimens can be explored soon after
the diagnosis of relapse.4–6
We have adopted a radical approach to brain tumour
therapy using the selectively replication-competent
mutant 17167 of herpes simplex virus type 1 (HSV1) to
treat recurrent glioma. HSV 1716 is a mutant that lacks
both copies of the RL1 gene8 which encodes the protein
ICP34.5,9 a specific determinant of virulence. The properties of 1716 have been extensively described. It is avirulent in the normal brains of animals.7 Its phenotype is cell
type and state dependent. It replicates in actively dividing but not in terminally differentiated cells.10 It replicates in a range of tumour models and can eliminate
metastatic brain tumours while failing to replicate in normal tissue.11–13 The pathological change induced in normal mouse CNS is minimal.14 In human glioblastoma
cells in vitro, lytic replication of 1716 results in cell
death.15
Correspondence: Dr R Rampling, Beatson Oncology Centre, Western
Infirmary, Glasgow G11 6NT, UK
Received 1 December 1999; accepted 10 February 2000
We have shown that ICP34.5 functions by complexing
with proliferating cell nuclear antigen (PCNA), a protein
involved in DNA replication and repair.16 In so doing an
environment is provided in which the virus can replicate.
In tumour cells, where functional PCNA levels are high,
ICP34.5 is not required for productive HSV replication
whereas in neurons in which PCNA levels are low,
ICP34.5 is an absolute requirement for the production of
infectious progeny virus. In a subset of cell types, preclusion of host cell protein synthesis shut off has been
shown to be a function of ICP34.5.17,18 The host cell protein synthesis shut off induced by removal of ICP34.5 can
be restored by altering the control of expression of the US
11 gene product.19,20 However, the virulence phenotype
conferred by expression of ICP34.5 is separable from the
host cell protein shut off response.19 Therefore, deregulation of US 11 will not restore a virulence phenotype to
1716 and should not pose a safety issue.
We concluded that 1716 had reached the stage where
a study to test its safety in patients was justified. In this
report, we evaluate the toxicity of 1716 in patients with
recurrent high grade glioma.
Most early gene therapy trials have used strategies
adapted from the development of conventional cytotoxic
chemical agents. These assume that a safe starting dose
can be derived from animal experiments and depend
upon the use of dose escalation schemes, (modified
Fibonacci), to obligatory toxicity end-points. Such a
design might not be appropriate for replication-competent HSV. In brain tissue ICP34.5 is an absolute
requirement for HSV virulence. In model systems,
deletion of the gene encoding ICP34.5 totally abrogates
Toxicity study of HSV1716
R Rampling et al
860
virulence and the mutated virus fails to cause encephalitis, no matter how high the input dose.7 HSV 1716 does
not contain the gene and hence ICP34.5 is never
expressed, irrespective of the dose. The question being
asked in this study is whether the avirulent phenotype
of 1716 demonstrated in model systems, pertains in
the context of human therapy up to a dose at which
therapeutic activity is a possibility.
We wished to show that a safe starting dose exists
when 1716 is injected directly into recurrent malignant
glioma. A starting dose of 103 infectious particles (p.f.u.)
was calculated to be approximately equivalent (in p.f.u.
per kilogram of brain) to the dose of wild-type HSV
which has been shown to cause a fatal encephalitis in
mice.7,21 With HSV Glasgow strain 17 (the wild-type parental virus of 1716), 2 p.f.u. will kill a mouse within 5
days when injected directly into the brain.7
As 1716 has the potential to replicate in tumours, the
final titre within the brain could be several orders of magnitude higher than the inoculated dose. Extrapolating
from in vitro data, 20% of infectious particles could result
in a productive infection and each infected cell could give
a burst size of 100 p.f.u. within a single infectious cycle.
An input titre of 105 p.f.u. could produce 2 × 106 p.f.u.
within 12 h.22 It was estimated that 105 p.f.u., delivered
under optimal conditions, could achieve a detectable
level of tumour kill in some patients in a future activity
study. This was the maximum dose allowed by the UK
Gene Therapy Advisory Committee, (GTAC) in this
study.
GTAC agreed that when safety at these dose levels had
been demonstrated, permission could be sought for a
separate trial to examine an activity end-point. At this
stage further dose escalation could be investigated, if
required. This developmental strategy is in line with the
recent National Cancer Institute/European Organisation
for Research and Treatment of Cancer (NCI/EORTC)
workshop on phase 1 drug development, whose recommendations pointed out the need for ‘new (biological)
end-points for new modalities which may produce nonspecific and sporadic toxicities which are not clearly dose
related’.23,24 That severe toxicity can arise when viral
treatments are extended to obligatory toxicity end-points
is demonstrated by a recent report of a toxic death in
a phase I study using adenovirus.25 It adds further
justification to GTAC’s cautious approach.
Results
Patient profiles
Nine patients were treated within the protocol (Table 1).
They were an unremarkable cross-section of high grade
glioma patients who had relapsed following radical treatment. All had previously undergone surgery and radical
radiotherapy. Two patients had received tirapazamine
chemotherapy concomitant with radiotherapy and six
patients had received nitrosourea chemotherapy at
relapse. Patient 5 underwent surgery and repeat
irradiation at his second relapse. All patients were receiving dexamethasone (4–16 mg daily) and five were taking
anticonvulsants. Haematology, renal and hepatic function, chest X-ray and ECG were within study limits. Eight
patients were seropositive for HSV1; only patient 4 was
seronegative. He did not seroconvert. The IgG and IgM
Gene Therapy
titres did not change significantly after 1716 administration in any patient (data not shown).
Eight patients had glioblastomas and one had an
anaplastic astrocytoma. Tumour size is reported as the
volume of enhancing tumour on gadolinium enhanced
MRI (Gd MRI) (Table 2). Size varied between 8.6 and 129
ml at the time of injection. (The volume of brain involved
with tumour necrosis, tumour-associated oedema and
non-enhancing tumour is larger than this figure.) Also
recorded are volumes measured from thallium SPECT
scanning. These are larger, but are also thought to relate
to the volume of actively growing tumour cells.26
The patients’ immune status at the time of injection is
shown in Table 3. Total white cell counts (not shown)
were normal in all patients. Lymphopenia was detected
in all patients, most marked in patients 3 and 8. B cells
were almost undetectable in four patients (Nos 3, 5, 6 and
8). T cell numbers tended to be around the lower end of
the normal range with particular reductions in the CD4+
subset in patients 3 and 5. Cellular proliferative responses
were reduced in all patients except 2 and 7 when compared with normal healthy controls. The phenotypic cytometric data together with the proliferative functional
results indicate a significant degree of immuno-incompetence in this group of patients.
Administration of 1716 and clinical outcome
Nine patients were treated. Three each received 103, 104
and 105 p.f.u. of HSV 1716 by stereotactic injection
directly into the tumour. The procedure was well tolerated with no immediate post-operative complications. All
patients recovered to their preoperative state within 24 h.
A summary of events in the first week is given in Table
4. Patients 3 and 6 were both prone to regular seizures
and those experienced after 1716 were typical and
required no additional action. Patient 4 (who was
seronegative for HSV1) experienced a short-lived, selflimiting pyrexia for which no cause was found. The
worsening condition in patient 8 was due to post-operative swelling and responded promptly to an increase in
steroid dose. The MRIs performed 4–6 days after injection
showed no changes suggestive of active infection in any
patient. The tumour volumes measured on Gd MRI at
this point were similar or slightly greater, as might be
expected following surgery (Table 2). None satisfied Macdonald’s criteria for progression.27 One of the thallium
SPECT scans showed an increase in volume; the remainder were smaller. The significance of this observation is
doubtful.
Patients were seen weekly until week 5. No adverse
clinical effects and no marked improvements which
could have been attributed to the administration of 1716
were recorded. The Gd MRI volumes were measured at
this point (4–9 weeks). Three satisfied the Macdonald criteria for progression.27 One patient (No. 4), although
clinically unchanged, requested resective surgery before
his week 5 scan could be performed. The remaining five
MRI were stable. The thallium SPECT volumes were
smaller in one, stable in two, larger in five and not assessable in one. The HMPAO SPECT scans showed no evidence of hyperperfusion to suggest encephalitis in any
patient.
Buccal swabs and serum were taken on days 2 and 6
and weekly for 4 weeks after 1716 infection and assayed
for HSV. There was no evidence of HSV shedding, either
Toxicity study of HSV1716
R Rampling et al
861
Table 1 Patient characteristics prior to their recruitment into the study
Patient
No.
Sex
Age
Histology
Previous treatment
Location of tumour
Karnofsky
score
Bartel
Index
1
2
3
4
5
6
7
8
9
M
M
M
M
M
M
M
F
F
22
49
62
34
41
62
37
65
56
GB
GB
GB
GB
GB
GB
AA
GB
GB
S, RT, CT
S, RT, CT,
S, RT+CT, CT
S, RT+CT, CT
Bx, RT, CT, S, RT
S, RT
S, RT
S, RT, CT
S, RT
L Basal ganglia
R Fronto/parietal
R Temporal
L Temporal
R Frontal
L Fronto/parietal
R Frontal
L Parietal
L Parietal
60
60
60
90
60
60
80
70
80
19
18
18
21
11
16
19
21
21
M, male, F, female; GB, glioblastoma; AA, anaplastic astrocytoma; S, surgery; RT, radiotherapy; CT, chemotherapy; Bx, biopsy only.
Table 2 Tumour volumes at three time-points during the study period
Patient
No.
MRI
vol
1
2
3
4
5
6
7
8
9
Pre op
(day)
8.6
39.3
53.7
24.5
129.0
31.4
11.4
19.5
60.1
(−1)
(−2)
(−1)
(−1)
(0)
(−3)
(−2)
(−3)
(−2)
vol
Thallium SPECT
Post op
(day)
8.8
48.6
76.0
25.9
162.6
31.9
15.9
21.0
54.8
(6)
(5)
(6)
(6)
(4)
(5)
(5)
(5)
(6)
vol
Post op
(day)
9.4
68
67.7
—
216.0
36.7
12.5
19.2
58.1
(28)
(33)
(32)
(66)
(32)
(27)
(43)
(35)
vol
Pre op
(day)
11.9
82
151.2
53.0
436.5
69.2
11.3
28.1
73.7
(−1)
(−2)
(−1)
(−1)
(0)
(−3)
(−2)
(−3)
(−2)
vol
Post op
(day)
2.5
51.5
115.3
40.3
354.3
46.1
1.9
22.1
122.1
(6)
(5)
(6)
(6)
(4)
(5)
(5)
(5)
(6)
vol
Post op
(day)
2.4
133.8
141.9
—
500.8
120.6
32.9
41.4
91.5
(28)
(33)
(32)
(61)
(32)
(27)
(43)
(35)
Volumes are given in ml. The timing of the scan, in days, with respect to the injection date is given in parentheses. The MRI estimate is
a volumetric measurement of the region of enhancement following injection of gadolinium contrast material. The thallium SPECT volumes
are calculated to be the volume within the brain registering counts above a background threshold determined from an equivalent normal
region in the brain contralateral to the tumour (also see text).
1716 or reactivated endogenous latent HSV, in any
patient. No patient experienced reactivation or recrudescence of HSV in the form of skin lesions.
After completion of the formal study period, patients
were followed with visits at approximately monthly
intervals. The condition of all patients at the end of
November 1999 is given in Table 4.
Five patients have died. Patient 2 underwent further
decompression of tumour at 12 weeks after injection and
remained well for more than 6 months before dying with
recurrent tumour at 9 months after injection. Patients 3,
5 and 6 enjoyed short periods of clinical stability (2–5
months) before dying with clinical and image documented tumour progression. Patient 9 was well following
her procedure and returned to an independent life at
home. She developed a sudden onset, fulminating bacterial pneumonia and septicaemia (Listeria and E. coli)
with total organ failure from which she died at 2 months.
Post-mortem analysis of her tissues showed no evidence
of HSV encephalitis or systemic HSV infection.
Four patients remain alive. Patient 1 has had no further
antitumour treatment and is alive and well at 24 months
after 1716 injection on a replacement dose of steroid.
Patients 4 and 7 are both currently well and in complete
remission (19 and 17 months after 1716 respectively) following further tumour decompressions at 3. weeks and
2 months after injection, respectively. Patient 7 received,
in addition, five courses of lomustine immediately following resective surgery. Patient 8 is clinically and radiologically stable at l4 months after 1716 administration.
She is currently receiving chemotherapy with a single
alkylating agent.
Further CT, MRI and thallium and HMPOA SPECT
scans have been performed according to clinical need in
all patients. On no occasion have there been changes
which would suggest an encephalitis.
Evaluation of biopsy and post-mortem material
Post-inoculation tissue was available from patients 2, 4
and 7 who underwent a subsequent tumour resection at
3 months, 3. weeks and 2 months, respectively, and from
patients 5 and 9 who died at 6 and 2 months, respectively, after inoculation. Histological examination of all
five cases showed high grade glial tumour with no
unusual features. There was no significant immunoreactivity of tumour cells or adjacent brain tissue for HSV1
using two monoclonal antibodies (Z1F11 and Novocastra
HSV1). In addition, there was no evidence of 1716 or
wild-type genomes by PCR. In both of the autopsied
patients the location of the site of 1716 injection was
identified as a cyst on the post-injection CT scan. The
cysts (approximately 1 cm diameter) had persisted from
the time of injection until the autopsy. There was no evidence of encephalitis involving 1716 or wild-type HSV
Gene Therapy
Gene Therapy
0.2–
0.45
—
0.27
0.28
0.18
0.27
0.46
0.1
0.29
0.9–
1.8
—
1.5
1.7
1.3
1.3
1.8
0.7
1.7
0.8–
4.0
2.3
1.8
2.1
1.4
0.53
0.94
2.4
0.74
1.4
5.3–
16.5
7.9
6.5
4.9
5.8
2.7
6.6
6.5
7.2
7.9
IgG
0.5–
2.0
0.8
0.8
0.56
0.76
0.16
0.7
0.80
0.59
1.6
IgM
IgA
C3
C4
Immunoglobulin levels in g/l
Complement
1.5–
4.0
1.2
1.1
0.53
0.77
0.86
1.1
1.4
0.48
1.2
Total
0.63–
3.0
0.80
—
0.41
0.53
0.64
0.84
0.92
0.33
0.74
CD3
0.43–
1.8
0.36
—
0.14
0.23
0.19
0.52
0.56
0.26
0.52
CD4
0.25–
1.2
0.54
—
0.25
0.30
0.50
0.39
0.29
0.08
0.29
CD8
0.12–
0.57
0.14
—
0.01
0.18
0.03
0.00
0.23
0.03
0.19
CD19
Lymphocyte numbers ×109 per litre
0.060
−0.24
0.26
—
0.03
0.02
0.12
0.21
0.14
0.06
0.26
CD56
SIC
SIP
Con A
Normal ranges not available.
See text for interpretation.
57
192
25
65
105
28
62
548
14
53
81
—
91
658
35
118
732
40
434
738
380
70
307
63
65
457
2
SIP
Phytohaemagglutinin
136
25
584
—
206
409
452
306
136
SIC
87
162
10
20
13
59
160
29
14
260
177
302
821
370
675
635
634
382
Pokeweek
mitogen
SIP
SIC
Proliferative response to mitogens
Immunoglobulins are measured with a routine nephalometric assay. Total number of lymphocytes and absolute numbers of T and B subsets are given with the normal ranges for our
laboratory. A whole blood lymphocyte assay was used (text). Lymphocyte function was assessed using a triple stimulation assay (text). SIP is the Stimulation Index (SI) at maximum
stimulation for the patient specimen. SIC is the SI of the control.
Normal
range
1
2
3
4
5
6
7
8
9
Patient
No.
Table 3 Routine immunology profiles for all study patients at the time of injection with HIV 1716
Toxicity study of HSV1716
R Rampling et al
862
Toxicity study of HSV1716
R Rampling et al
863
Table 4 The condition of the patients during the study and at November 1999
Patient No.
Viral dose
Condition during first week
1
2
103 p.f.u.
103 p.f.u.
Entirely well
Seizure day 3 otherwise well
3
103 p.f.u.
Entirely well
4
5
104 p.f.u.
104 p.f.u.
6
104 p.f.u.
Headache day 4, T 37.0°C for 12 h
Worsening of pressure symptoms otherwise
stable
Seizure day 4 otherwise well
7
8
105 p.f.u.
105 p.f.u.
9
105 p.f.u.
Entirely well
Worsening of dysphasia and ataxia. One
episode of vomiting. Improved with increased
dexamethasone
Neurologically stable
in peritumoral brain or in limbic structures by histology,
immunocytochemistry, culture or PCR analysis. Tissue
cultures for virus were negative.
Discussion
Most ‘gene therapy’ studies for brain tumours have
involved genetically modified, replication-deficient
viruses expressing a transgene capable of delivering a
tumour killing product. We have taken the alternative
approach of using a selectively replication-competent
HSV. The high proliferative activity in the tumour
against a background of quiescent normal cells and the
lack of systemic spread of gliomas make them good
candidates for selective killing by such a virus.
Herpes simplex virus is neurotropic and introduction
directly into the brain inherently carries the risk of major
toxicity. A prerequisite for its development for human
therapy is a demonstration of the avirulent phenotype of
1716 in its natural human host. Thus, a limited study was
designed solely to look for evidence of toxicity when 1716
was injected directly into brain tumours in patients. In
particular, no patient developed HSV encephalitis and no
patient required the anti-herpes drug aciclovir. As a
group they have lived at least as long as would be
expected for a population of this type.3 Four surviving
patients all received 1716 over 1 year ago.
Contrary to their previous findings28 Kesari et al
showed that 1716 can replicate in ependymal cells29 and
that intraventricular injection of high doses in nude mice
is followed by death in some animals.30 They concluded
that these studies should serve as a warning for the use
of 1716 in immunosuppressed humans. The patients in
the present study were immunocompromised by previous antitumour therapy, contemporary steroids, and
disease effects. However, none showed evidence of replication of 1716 in normal brain. Moreover, one was
seronegative for HSV and eight seropositive and there
was no difference in the response to 1716 inoculation. The
seronegative patient did not seroconvert and the seropositive patients did not demonstrate any elevation of either
IgG or IgM antibody titres after 1716 administration. No
patient demonstrated reactivation, recrudescence or
shedding of endogenous latent HSV or shedding of 1716.
Where examination of brain tissue (tumour and adjac-
Current condition (Nov 99)
Remains well at 24 months
Further surgery for tumour growth. Died of
tumour progression at 9 months
Tumour stable for 2 months then tumour
progression: died at 3 months
Further surgery. Alive and well at 19 months
Tumour stable for 4 months then tumour
progression: died at 6 months
Tumour stable for 5 months then tumour
progression: died at 6 months
Further surgery. Alive and well at 17 months
Stable at 14 months
Died from bacterial sepsis at 8 weeks
ent brain) has been possible following either tumour
resection or post-mortem, no evidence of encephalitis or
ongoing lytic replication was found. Immunohistochemistry using two HSV-specific antibodies failed to show
immunoreactivity in either tumour or normal brain. Similarly, PCR analysis failed to detect any HSV genomes. It
may be that there were no 1716 genomes in the tissue and
no residual HSV antigen. As we were unable to analyse
material any earlier than 3. weeks after injection, this
would not be surprising. Alternatively, it may represent
a sampling problem in that the injected tumours were
large and only a proportion of the cells would be actively
dividing and thus able to support 1716 replication. Without the expression of a marker gene (eg ␤-gal) it is
impossible to detect virus migration and select infected
tissue samples. A less likely explanation is that the assays
used were not sensitive enough to detect either HSV
DNA or protein at levels present in the samples.
Without doubt this study demonstrates that 1716 can
be delivered intratumorally to patients with growing
intracranial neoplasms to doses of l05 p.f.u. without detriment to their health. Experiments in animals have shown
that selective replication-competent viruses at doses of
105 p.f.u. are most effective when applied to small
tumours11,12,31 but may fail to control bulky tumours. It
is unlikely that obvious antitumour effects such as size
reduction, will be apparent in large heterogeneous
tumours using a single inoculation introduced by a simple delivery system to one site when evaluation methods
such as MRI and SPECT scans are used. The present
study was designed to test the hypothesis that 1716 could
be introduced into the human brain without inducing
encephalitis. It was not designed to test efficacy. This
report provides the first evidence in support of the safety
of 1716 for human use, at least in this form of cancer, and
justifies further clinical research to test activity. Proposed
studies will take into consideration stage of disease at
which to introduce virus, mode of delivery, dose of virus
and methodology required critically to assess efficacy.
The animal evidence suggests that for glioma the most
likely use of selectively replication-competent HSV will
be as part of a combination therapy regimen in minimal
residual disease.
Concurrently with our study, another clinical study has
been started in glioma patients using the HSV1 strain F
Gene Therapy
Toxicity study of HSV1716
R Rampling et al
864
mutant G207 which has mutations in both ribonucleotide
reductase (RR) and ICP34.5.32 As the mutation in RR also
impairs virulence of HSV although not abolishing it, use
of this double mutant provides a further safety dimension. However, the replication competence of G207 in
tumour cells is markedly reduced compared with that of
1716 and hence to achieve tumour cell killing, substantially higher doses will be required. Consequently doses
of G207 are being given which are several orders of
magnitude higher (eg 109 p.f.u.) than those of 1716.
Patients and methods
Virus production
A low passage laboratory grade stock of 17167 was the
starting material for production of GMP grade virus for
human use. Production of 1716 in BHK-21/C13 cells from
a Master Seed Cell Bank under cGMP conditions was carried out by Q-One Biotech Ltd, Glasgow, UK. The final
clinical grade virus preparation was titrated for infectivity, tested for contamination and stored at −70°C in
identical pairs of vials. Each vial contained 1716 at the
appropriate concentration (103–105 p.f.u.) in 1 ml total
volume of human serum albumen, HSA 0.5–1.0% in isotonic PBS. One week before the intended injection date,
the titre of virus in one of the paired vials was confirmed
on BHK21/C13 cells.
Titration of 1716 was carried out by standard methods
on BHK21/C13 cells.33 Viral titres were expressed as concentration of p.f.u. in comparison to the TCID50 values
obtained from Q-One Biotech.
Trial procedure
The study protocol was approved by the UK Gene Therapy Advisory Committee (GTAC) and the Medicines
Control Agency (MCA), the Glasgow Southern General
Hospital Research Ethics Committee and the Health and
Safety Executive Committee.
Patient population
Patients were considered for study if they had biopsy
proven high grade glioma, previously treated with surgery and radiotherapy and had exhausted all other conventional treatments at relapse. They must have had no
chemotherapy or radiotherapy within 6 weeks. Proof of
recurrence was on the basis of deteriorating clinical
symptoms and unequivocal worsening in two successive
scans of the same type (CT or MRI) with further confirmation using thallium SPECT. Subjects were aged
between 18 and 70 with Karnofsky status ⬎60 and a life
expectancy ⬎8 weeks. Their haematological, renal,
hepatic and coagulation functions were within normal
limits for our laboratory and their general condition
satisfactory.
Trial subjects were required to be neurologically stable
on an appropriate dose of steroid and able to give informed written consent. Consent was given by both the
patient and a chosen relative or carer to an independent
physician who was not otherwise concerned with the
design or immediate conduct of the study.
Patients considered for the study underwent a comprehensive work-up including general clinical, and neurological evaluation, microbiological, immunological,
haematological and biochemical assessments. NeuroradiGene Therapy
ology included CT scanning, gadolinium-enhanced MRI
and thallium-201 and Tc99mHMPOA SPECT scans.
Administration of 1716
Three patients were treated at each dose level (103, 104,
105 p.f.u.). The study was to be terminated if any patient
experienced symptoms or signs of HSV encephalitis or if
any grade 3 or 4 toxicity was encountered. Stereotactic
injection of 1716 was carried out using a specially
designed inner needle with a curved distal end that was
passed down a standard sedan outer biopsy needle. The
curve allowed the needle to penetrate laterally into the
tissues. The extent of penetration into the tissue was controlled by a stop set on the needle. The procedure was
carried out in a designated CT imaging suite. The
patients underwent premedication and routine anaesthesia with alfentanyl and propofol. The recurrent
tumour was localised using a ‘Lecksell’ stereotactic frame
and an initial biopsy was taken. An immediate cytological smear was prepared and reported within 10 min to
confirm positioning of the needle within active tumour.
Following confirmation, the biopsy trocar was replaced
by the delivery needle and multiple radial injections (6–
10) at different levels within the tumour were performed
as the probe was withdrawn in a stepwise fashion after
a single pass. Between 5 and 10 aliquots of virus to a total
of 1 ml were delivered into the tumour. An immediate
post-operative CT scan without contrast was taken to
ensure no internal haemorrhage and to confirm delivery
of the virus within the tumour.
Patient evaluation
Patients were monitored in a high dependency unit for
2–3 h and then in an isolated room in a routine ward.
During the 6 day inpatient stay, patients were assessed
every 4 h for blood pressure, heart rate, peripheral O2
saturation, Glasgow Coma Scale and focal neurological
features. Neurological and general physical examinations
were performed twice daily. Haematological, biochemical
and immunological status were monitored routinely. Full
protocol neuroradiology was repeated on day 6.
Additional scanning was performed only if there was a
deterioration in neurological status. Evidence of HSV
encephalitis was sought through detailed assessment of
neurological condition and vital signs. Other toxicities
were routinely assessed and judged according to the NCI
Common Toxicity Criteria (CTC) checklist. An anti-HSV
protocol based on the delivery of aciclovir was available
in event of concern over encephalitis. Patients were discharged on day 7 provided they remained well. They
were followed weekly for 4 weeks at which times clinical,
haematological, biochemical, and virological assessments
were performed. At the end of week 5, a full radiological
assessment was performed with MRI and SPECT.
Tumour response was monitored using the patient’s
clinical condition and radiological response by volumetric analysis and reported according to conventional
criteria.27
Investigations on patient samples
Immunological investigations: A routine nephalometric
(Behring BNA II, Milton Keynes, UK) assay was used to
measure the immunoglobulins IgA, G and M. Lymphocyte populations and subsets were measured in whole
Toxicity study of HSV1716
R Rampling et al
blood incubated with dual-labelled (fluorescein and
phycoerythrin) monoclonal antibodies. Lysed specimens
were analysed on a Becton Dickinson FACScan.
Lymphocyte function was measured in a whole blood
stimulation assay using serial doses of phytohaemagglutinin, concanavalin A and pokeweed mitogen. Stimulated
cells were pulse labelled with tritiated thymidine for 4 h
and counted by liquid scintillation counting. Results are
expressed as stimulation index (SI) – the ratio of the
stimulated counts to the nonstimulated count at any dose
of mitogen. Each assay was controlled using whole blood
from volunteers. Normal ranges are not available. Stimulation is unequivocally reduced if the SI is less than 10.
Values above 10 are difficult to interpret on individual
samples and are best compared directly with counts
obtained from a healthy control run in parallel. It is also
important to note that where lymphocyte numbers are
low the accuracy of this assay declines.
ELISA for HSV-1 IgG and IgM was carried out using
a Virotech HSV-1-specific kit. Values are calculated for
IgG and IgM levels in Virotech (Russelsheim, Germany)
enzyme (VE) units where ⬎11 VE units is classified as
positive.
Immunohistochemistry for HSV was performed on
paraffin sections using monoclonal antibodies recognising the HSV1 65K DNA binding protein (Z1F11, a gift
from Howard Marsden) and an unspecified epitope of
HSV1 strain Stoker (Novocastra, Newcastle upon Tyne,
UK). Sections were pre-treated by microwaving for antigen retrieval. The primary antibodies were titrated using
sections from wild-type HSV1 encephalitis and were both
subsequently used at a concentration of 1:1000. Bound
antibody was visualised using a multilink kit (Menarini,
Wokingham, UK) and diaminobenzidine. Sections from
wild-type HSV1 encephalitis were used as positive
controls. Omission of primary antibody constituted the
negative control.
PCR assay: DNA was extracted from biopsy and post
mortem samples using a Nucleon Biosciences ST
(Coatbridge, UK) (soft tissue) Genomic DNA Extraction
Kit. The DNA pellet was resuspended in TE buffer and
the concentration measured by agarose gel electrophoresis and spectrophotometry before being used for
the PCR assay. DNA extracted from BHK cells infected
with strain 17 or 1716 was used as controls. On the
assumption of a burst size of 100 p.f.u. per cell and a
particle:p.f.u. ratio of 100:1, the level of sensitivity was 4
× 105 HSV genomes/ml of total DNA as estimated from
reconstruction experiments using DNA from infected
BHK cells. (Using diagnostic primers which amplify gB
of HSV1 and 2,34 the level of sensitivity was the same.)
The primers chosen to distinguish between HSV1
strain 17 and the mutant 1716 were F1 (20-mer) and R2
(18-mer). Sequences are F1 (CAG GCA CGG CCC GAT
GAC CG) and R2 (CTT TAA AGC GGT GGC GGC).
Primer F1 is homologous to the region 125170–125189 of
the HSV-1 genome. Primer R2 is complementary to the
region 125993–125976 of the same genome. These primers
amplified a 64 bp fragment which was diagnostic for 1716
and absent from wild-type 17. Primers were supplied
by Genosys (Cambridge, UK) and made-up as 1 nm
solutions.
PCR reagents used per reaction were 1.5 ␮l 10 mm
dNTPs, 5 ␮l Pfx Amplification buffer; 0.5 ␮l Platinum Pfx
DNA polymerase (Gibco BRL, Paisley, UK); l ␮l 50 mm
Mg2SO4; 1 ␮l each of a 0.1 nm stock of primers F1 and
R2 and 2 ␮l DNA template, made up with H2O to 50 ␮l
total volume.
PCR conditions were 94°C for 2 min; 30 cycles × {94°C
for 15 s; 72°C for 2 min; 72°C for 2 min}; 72°C for 10 min;
store at 4°C.
865
Assay for infectious HSV in patient samples: This was carried out by adding the growth medium in which swabs
were placed to monolayers of BHK21/C13 cells which
were then examined over 7 days for evidence of cytopathic effect and/or HSV plaques. Serum samples were
assayed similarly.
Tumour volume measurement using MRI
MRI data sets were acquired on a 1.5 Tesla Siemans Magnetom. Tumour enhancement volumes were measured
using a threshold method applied to co-registered, preand post-contrast, 3D, T1-weighted MR data sets. Subtraction of these sets resulted in images showing only
enhancement. The background noise in the subtraction
images was estimated using regions of interest (ROI)
from the opposite hemisphere to the neoplasm. The
lower value of the threshold was set to 2.5 standard deviations above this average noise level. A large ROI was
drawn around the general region of the tumour to
exclude other enhancing structures, such as veins, and all
voxels within this ROI that were above the previously set
threshold were counted. As the voxel size is known the
volume of enhancement was calculated by multiplying
the number of enhancing voxels by the number of partitions which showed them, to give an overall enhancing
tumour volume.
Tumour volume measurement using thallium SPECT
Since thallium SPECT is less affected than MRI by nonneoplastic processes it may be considered a more specific
indicator of tumour growth. A semi-quantitative
measurement of tumour volume was obtained using a
slice by slice ROI area measurement technique (volume
= area × slice spacing). Tumour edges were defined using
a thresholding technique. Normal brain uptake was
assessed in the contralateral hemisphere in at least three
slices at the same anatomical level as the tumour. A threshold level was empirically determined as being 1.7 × normal brain uptake. All areas in the brain with an uptake
exceeding the threshold level were outlined. The total
area above threshold was calculated for each slice,
excluding areas considered to represent ‘normal’ high
uptake regions, eg choroid plexus.
Acknowledgements
This work was supported by the UK Medical Research
Council project grant No. G9539438N to RR, GC and MB.
GMP grade 1716 was generously supplied without
charge from Q1 Biotech, Glasgow and we are indebted
to David Onions and Gillian Lees for their support. We
are grateful to Howard Marsden, MRC Virology Unit,
Glasgow who kindly supplied the ZlF11 antibody. Dr A
Farrell and Mr E Galloway gave invaluable help and
advise with respect to the immunology measurements.
We also thank the GTAC committee and the MCA for
allowing this study to proceed.
Gene Therapy
Toxicity study of HSV1716
R Rampling et al
866
References
1 Rampling R. Modern aspects of radiation therapy for glial
tumours of the brain. Forum 1998; 8: 289–301.
2 Rampling R. Chemotherapy. Determining the appropriate treatment. In: Hopkins A, Davies EH (eds) Improving Care for Patients
with Malignant Glioma. Royal College of Physicians: London,
1997, pp 63–74.
3 Rajan B. et al. Survival in patients with recurrent glioma as a
measure of treatment efficacy: prognostic factors following
nitrosourea chemotherapy. Eur J Cancer 1994; 30A: 1809–1815.
4 Gutin P, Leibel S. Stereotaxic interstitial irradiation of malignant
brain tumours. Neurol Clin 1985; 3: 883–893.
5 Mbidde EK et al. High dose BCNU chemotherapy with autologous bone marrow transplantation and full dose radiotherapy for
grade IV astrocytoma. Br J Cancer 1988; 58: 779–782.
6 Lucas GL et al. Treatment results of stereotactic interstitial brachytherapy for primary and metastatic brain tumours. Int J Radiat Oncol Biol Phys 1991; 21: 715–721.
7 MacLean AR et al. Herpes simplex virus type 1 variants 1714
and 1716 pinpoint neurovirulence related sequences in Glasgow
strain 17+ between immediate early gene 1 and the ‘a’ sequence.
J Gen Virol 1991; 72: 631–639.
8 Dolan A, McKie E, MacLean A, McGeoch D. Status of ICP34.5
gene in herpes simplex virus type 1 strain 17. J Gen Virol 1992;
73: 971–973.
9 Ackermann M et al. Identification by antibody to a synthetic
peptide of a protein specified by a diploid gene located in the
terminal repeats of the L component of the herpes simplex virus
genome. J Virol 1986; 58: 843–850.
10 Brown SM et al. Cell type and cell state determine differential
in vitro growth of non virulent ICP34.5 negative herpes simplex
virus. J Gen Virol 1994; 75: 2367–2377.
11 Randazzo RB et al. Treatment of experimental intracranial
murine melanoma with neuroattenuated herpes simplex virus
type 1 mutant. Virology 1995; 211; 94–101.
12 Kesari S et al. Therapy of experimental human brain tumors
using a neuroattenuated herpes simplex virus mutant. Lab Invest
1995; 73: 636–648.
13 Kucharczuk J et al. Use of a ‘replication-restricted’ herpes virus
to treat experimental human malignant mesothelioma. Cancer
Res 1997; 57: 466–471.
14 McKie EA, Brown SM, MacLean A, Graham DI. Histopathological responses in the CNS following innoculation with a nonneurovirulent mutant (1716) of herpes simplex virus type 1
(HSV1): relevence for gene and cancer therapy. Neuropath Appl
Neurobiol 1998; 24: 367–372.
15 McKie EA et al. Selective in vitro killing of primary CNS tumours
using herpes simplex virus type1 (HSV-1), ICP34.5 null
mutants – a potentially effective clinical therapy. Br J Cancer
1996; 74: 745–752.
16 Brown SM, MacLean A, McKie E, Harland J. The herpes simplex
virulence factor ICP34.5 and the cellular protein MyD116 complex with proliferating cell nuclear antigen through the 63
amino acid domain conserved in ICP34.5, MyD116 and
GADD34. J Virol 1997; 71: 9442–9449.
17 Chou J, Roizman B. The gamma 34.5 gene of herpes simplex
virus type 1 precludes neuroblastoma cells from triggering total
Gene Therapy
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
shut off of protein synthesis characteristic of programmed cell
death in neuronal cells. Proc Nat Acad Sci USA 1992; 89: 3266–
3270.
Chou J, Roitzman B. HSV1 34.5 gene function which blocks the
host response to infection maps in the homologous domain of
genes expressed during growth arrest and DNA damage. Proc
Natl Acad Sci USA 1994; 91: 5247–5251.
He B et al. The carboxyl terminus of the murine MyD116 gene
substitutes for the corresponding domain of the gamma 34.5
gene of herpes simplex virus to preclude the premature shutoff
of total protein synthesis in infected human cells. J Virol 1996;
70: 84–90.
Mulvey M, Poppers J, Ladd J, Mohr I. A herpes virus ribosome
associated RNA binding protein confers a growth advantage
upon mutants deficient in GADD 34 related function. J Virol
1999; 73: 3375–3385.
Valyi-Nagi T et al. The HSV-1 strain 17+gamma 34.5 deletion
mutant 1716 is avirulent in SCID mice. J Gen Virol 1994; 75:
2059–2063.
Harland J, Brown SM. HSV growth preparation and assay. In:
Brown SM, MacLean AR (eds). Herpes Simplex Virus Protocols.
Methods in Molecular Medicine. Humana Press: 1997.
Arbuck SG. Workshop on phase 1 study design Ninth
NCI/EORTC New Drug Development Symposium, Amsterdam, March 1996. Ann Oncol 1996; 7: 567–573.
Steward WS. Phase 1 studies–alternative endpoints: when is toxicity not appropriate? In 9th NCI/EORTC Symposium on New
Drugs in Cancer Therapy, Amsterdam, 1996.
Lehrman S. Virus treatment questioned after gene therapy
death. Nature 1999; 401: 517–518.
Cruickshank GS. The use of SPECT in the analysis of brain
tumours, In: Duncan R (ed). SPECT Imaging of the Brain. Kluwer
Academic Publishers: Dordrecht, 1997, pp 161–178.
Macdonald DR, Terrance LC, Cascino S, Schold C. Response
criteria for phase II studies of supratentorial malignant glioma.
J Clin Oncol 1990; 8: 1277–1280.
Kesari S et al. Selective vulnerability of mouse CNS neurons to
latent infection with neuroattenuated HSV1. J Neurosci 1996; 16:
5644–5653.
Kesari S et al. A neuroattenuated ICP34.5-deficient herpes simplex type 1 replicates in ependymal cells of the murine central
nervous system. J Gen Virol 1998; 79: 525–536.
Lasner TM et al. Toxicity and neuronal infection of a HSV-1
ICP34.5 mutant in nude mice. J Neurovirol 1998; 3: 100–105.
Chambers R et al. Comparison of genetically engineered herpes
simplex viruses for the treatment of brain tumours in a scid
mouse model of human malignant glioma. Proc Natl Acad Sci
USA 1995; 92: 1411–1415.
Markert JM et al. Initial report of phase 1 trial of geneticallyengineered HSV-1 in patients with malignant glioma. In: 23rd
International Herpesvirus Workshop. York, 1998.
Brown SM, Ritchie DA, Subak-Sharpe JH. Genetic studies with
Herpes Simplex virus type 1. The isolation of temperature sensitive mutants, their arrangement into complementation groups
and recombination analysis leading to a linkage map. J Gen Virol
1973; 18: 329–346.
Cone RW et al. Extended duration of Herpes Simplex Virus
DNA in genital lesions detected by polymerase chain reaction.
J Infect Dis 1991; 164: 757.