FASEB J.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.07-097808.
 (The FASEB Journal. 2008;22:1839-1848.)
© 2008 FASEB
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
fj.07-097808v1
22/6/1839    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kelly, K.
Right arrow Articles by Gil, Z.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kelly, K.
Right arrow Articles by Gil, Z.

Attenuated multimutated herpes simplex virus-1 effectively treats prostate carcinomas with neural invasion while preserving nerve function

Kaitlyn Kelly*, Peter Brader{dagger}, Avigail Rein*, Jatin P. Shah*, Richard J. Wong*, Yuman Fong* and Ziv Gil*,1

* Department of Surgery, and

{dagger} Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, New York, USA

1Correspondence: Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10065, USA. E-mail: gilz{at}mskcc.org or ziv{at}baseofskull.org


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Many cancers can cause disability and pain by invading nerves. In particular, prostate carcinoma has a high propensity for neural invasion (NI) at an early stage. Attempted surgical treatment of tumors with NI often leads to erectile dysfunction and deteriorated quality of life. Therefore, there is a need for novel modalities that will selectively target cancer cells while preserving neural function. Herpes simplex viruses (HSVs) have a natural trophism for peripheral nerves. We hypothesized that oncolytic therapy using HSV engineered to minimize neurotoxicity would be appropriate for this clinical setting. Attenuated HSV (NV1023) injected to sciatic nerves of nude mice had no toxic effect on nerve function (n=30). NV1023 had significant oncolytic effect on prostate carcinoma cells (PC3, DU145, and LNCap) in vitro. An in vivo model of NI was established by implanting prostate carcinoma cells in the sciatic nerves of nude mice. Mice were treated with NV1023 or saline 7 days after establishment of tumors. Significant reduction in tumor size and inhibition of NI was found 6–8 wk after treatment (P<0.005). All animals treated with saline developed complete paralysis <5 wk post-treatment, whereas most NV1023-treated animals had preserved nerve function >12 wk after treatment (P<0.0001). We conclude that oncolytic therapy effectively treats prostate carcinomas with NI in an in vivo murine model while preserving neural function. These findings may hold significant clinical implications for patients with prostate cancer or other neurotrophic tumors.—Kelly, K., Brader, P., Rein, A., Shah, J. P., Wong, R. J., Fong, Y., Gil, Z.—Attenuated multimutated herpes simplex virus-1 effectively treats prostate carcinomas with neural invasion while preserving nerve function.


Key Words: perineural invasion • gene therapy • prostate cancer • oncolytic therapy • erectile dysfunction


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
PROSTATE CANCER IS THE LEADING CAUSE of cancer in men, with more than 200,000 cases diagnosed each year (1) . Prostate carcinomas have the ability to track along nerves and significantly impact the clinical outcome of patients (2 , 3) . Tumor invasion along the cavernous nerve bundle is a major contributor to cancer progression, recurrence, and mortality in patients with prostate carcinoma (4) . Unfortunately, attempted surgical resection of tumors with perineural invasion necessitates nerve resection, ultimately leading to erectile dysfunction. Maintenance of erectile function after surgery is a major concern for prostate cancer patients. When choosing therapy, two-thirds are willing to compromise survival benefits for greater chances of sexual function (5) . Recovery of erectile function after treatment for prostate cancer is directly related to the degree of preservation of the cavernous nerves (6) and is an independent determinant of improved health-related quality of life after therapy (7) . Given the dismal functional prognosis after sacrificing the cavernous nerves, there is a need for alternative cancer therapies that offer preservation of neural function.

Herpes oncolytic vectors are effective gene transfer agents, with attenuated toxicity to normal cells. Unlike other herpes vectors used solely for gene transfer, these replication-competent viruses have potent antitumor activity by inducing direct tumor cell lysis. Previous studies have shown that oncolytic HSVs can effectively treat a variety of human tumor cells, both in vitro and in vivo (8 , 9) . Recently, phase-I clinical trials have demonstrated their safe administration in patients with advanced colorectal and brain cancers (10 , 11) .

Oncolytic herpes vectors were genetically engineered from HSV type 1, a wild-type strain with significant tropism and toxicity to a broad range of neurons (12) . Unlike the wild-type strain, attenuated herpes viruses can be safely administered to nerves and used for gene transfer into the peripheral and central nervous system (13) . Based on the oncolytic properties and neurotropism of herpes viruses, we hypothesized that nerve infiltration by prostate carcinoma cells can be treated with replication-competent attenuated herpes vectors. This treatment has the potential to selectively target cancer cells invading nerves while preserving neural function.

In this study, we evaluated the safety and efficacy of oncolytic HSV therapy against human prostate carcinomas with neural invasion. We show that this therapeutic approach has the potential to improve current treatment of advanced prostate cancer by selectively sterilizing nerves infiltrated by cancer, while preserving function.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cell lines
The human prostate carcinoma cell line PC3 was maintained in F12K with 2 mM L-glutamine supplemented with 1.5 g/L sodium bicarbonate; DU145 was maintained in Dulbecco modified Eagle medium; and LNCap was maintained in RPMI 1640. All media also contained 10% fetal calf serum, penicillin, and streptomycin. Cells were maintained in 5% CO2 in a 37°C humidified incubator.

Viruses
NV1023 is an attenuated, replication-competent oncolytic HSV whose construction was previously described (14) . In brief, the virus was derived from R7020, an HSV-1 (F strain) vector originally designed as an HSV vaccine candidate (15) . The virus carries a 5.2-kb fragment of HSV-2 DNA (containing HSV-2 genes US2-2 through US2-5) inserted in the UL/S junction. Attenuation was achieved by a 15-kb deletion in the inverted repeat region that extends from the 3' end of UL55 to the promoter for ICP4, deleting UL56 and 1 copy of the diploid genes ICP0 and ICP4 and the neurovirulence gene {gamma}134.5. The Escherichia coli β-galactosidase gene (lacZ) was inserted at the US10–12 locus as an infection marker.

Cytotoxicity assays
Cancer cells were plated at 2 x 104 cells per well in 12-well plates in 1 ml medium. After incubation for 6 h, NV1023 (100 µl) was added to each well at multiplicities of infection (MOI, or the ratio of viral particles per tumor cell) of 0, 0.01, 0.1, or 1. Viral cytotoxicity was measured at 24 h intervals. On day 1, 1 ml of fresh medium was added to each well. Cells were washed with PBS and lysed with Triton X (1.5%) to release intracellular lactate dehydrogenase (LDH), which was quantified with a Cytotox 96 kit (Promega, Madison, WI, USA) by spectrophotometry (EL321e; Bio-Tek Instruments, Winooski, VT, USA) at 490 nm. Results are expressed as the ratio of surviving cells determined by comparing the measured LDH of each infected sample relative to control untreated cell samples that were considered 100% viable. All samples were analyzed in triplicate.

In vivo model of neural invasion
Six-week-old male athymic nude mice (National Cancer Institute, Bethesda, MD, USA) were anesthetized with inhalational Isoflurane for all procedures. The left sciatic nerve was exposed deep to the femorococcygeous and biceps femoris muscles. PC3 and DU145 cells were microscopically injected into the perineurium of the sciatic nerve, distal to the bifurcation of the tibial and common peroneal nerves. Slow microinjection of 3–5 µl of cell suspension at a concentration of 1 x 105 cells/µl was performed using a 10 µl Hamilton syringe (Hamilton, Reno, NV, USA) over a 2-min period (16) . Sciatic nerve infiltrating tumors were established, confirmed with a stereoscope at day 7 and then injected with NV1023 at 5 x 107 plaque-forming units (pfu) or saline. For tissue histochemistry, mice were euthanized and the sciatic nerve excised, frozen in Tissue-Tek OCT compound (Sakura Finetechnical, Tokyo, Japan), and cut into sections 5 µm thick.

Measures of sciatic nerve function
Sciatic nerve function was measured weekly as previously described (17) . The sciatic nerve innervates the hind limb paw muscles. Functional measures for monitoring tumor neural invasion included: 1) gross behavior: signs of motor weakness or repetitive biting of the hind limb were monitored for 10 min once a week; 2) limb function: function was graded according to hind limb paw response to manual extension of the body, from 4 = normal to 1 = total paw paralysis; and 3) sciatic nerve function index: calculated as the spread length (in mm) between the first and fifth toes of the mouse hind limbs.

Ultrasound assessment of neural invasion
A dedicated high-resolution (55 MHz) small-animal ultrasound system (Vevo 770TM; VisualSonics, Toronto, ON, Canada) was used to measure the primary sciatic nerve tumor diameter and the proximal sciatic nerve diameter on a weekly schedule. The primary sciatic nerve diameter was measured at the tumor implantation site and reflects the tumor cell burden. The proximal sciatic nerve was measured 4 mm proximal to the cancer cell implantation site, just before the nerve enters the spinal column. The proximal sciatic nerve diameter is a measure of the thickness of the nerve along its course and reflects the degree of neural invasion by cancer cells. Using this method, we were able to monitor neural invasion in vivo, with a spatial resolution of 50–100 µm.

Real-time polymerase chain reaction (PCR)
Sciatic nerves invaded by cancer were injected with NV1023 at 5 x 107 pfu (left side) or with saline for control animals (right side). At 14 days after injection, mice were euthanized by CO2 inhalation, and the brain, spinal cord, and left (ipsilateral) and right (contralateral) sciatic nerves were excised and homogenized separately in 1.2 ml TRIzol reagent (Invitrogen, Carlsbad, CA, USA). DU145 cells infected with NV1023 in vitro were used as positive control (standard) in these experiments. After the addition of chloroform, samples were centrifuged for phase separation. RNA was precipitated from the aqueous phase with isopropanol and treated with DNase (DNA-free; Ambion, Austin, TX, USA) according to the manufacturer’s directions. cDNA was reverse transcribed from tissue RNA using random hexamer priming. For each sample, 250–500 ng of total RNA was added to 4 µl of 5x reverse transcriptase (RT) buffer (Invitrogen), 10 pmol random hexamers, 12.5 µM each of dATP, dTTP, dGTP, and dCTP, 200 U of RT (Invitrogen), and 20 U of RNasin (Promega) to a final volume of 20 µl. Real-time PCR was performed on the total RNA extracts from mice treated with saline or NV1023. Real-time PCR was run in triplicate on an ABI Prism 7700 thermal cycler (Applied Biosystems, Foster City, CA, USA) and contained cDNA or tail DNA, TaqMan universal PCR mix (Applied Biosystems), and target-specific TaqMan dye-labeled primer/probe (Assays by Design; Applied Biosystems). The primers used for RT-PCR were for expression of the early HSV-1 gene ICP6 and late gene LAT. Each sample was measured quantitatively by real-time PCR and standardized to an 18S rRNA control. Probes used for ICP6 included: ATA GCC AAT CCA TGA CCC TGT ATG (forward), GGG TGG AGG CTG GGA GG (reverse), and CAC GGA GAA GGC GGA CGG GA (probe). Probes used for LAT exon included: CCC ACG TAC TCC AAG AAG GC (forward), AGA CCC AAG CAT AGA GAG CCA G (reverse), and CCC ACC CCG CCT GTG TTT TTG TG (probe). Standard curves were generated from serial dilutions. PCR was performed under the following conditions: stage 1: 50°C for 2 min; stage 2: 95°C for 10 min; stage 3 (35 cycles): 95°C for 15 s and 60°C for 1 min; and stage 4: 25°C.

X-gal histochemistry
Sciatic nerve sections were stained with hematoxylin and eosin (H&E) or X-gal for assessment of β-galactosidase expression. Cells were stained for 4 h with X-gal (1 mg/mL) in an iron solution of 5 mM K4Fe(CN)6, 5 mM K3Fe(CN)6, and 2 mM MgCl2, as previously described (18) . Counterstaining of background cells with nuclear fast red was performed. Virally infected cells expressing β-galactosidase were identified histologically as blue-staining cells.

Immunocytochemistry
Sciatic nerve specimens were excised 6 wk (DU145 tumors) and 8 wk (PC3 tumors) after tumor injection, frozen in OCT, and cut into 8 µm thick sections on glass slides. Some of the slides were fixed and stained with H&E. To differentiate between the nerve tissue and cancer cells, the sciatic nerve preparation was immunostained for Ki67 (Santa Cruz, Santa Cruz, CA, USA). Slides were fixed with 2% formaldehyde and 0.2% glutaraldehyde, then quenched, blocked, and incubated with primary antibodies overnight at 4°C. Slides were incubated with a biotinylated secondary antibody and visualized with an avidin-biotin complex kit (Santa Cruz). Slides were counterstained with hematoxylin and reviewed in a blinded fashion by 2 of the authors.

Statistical analysis
A Student’s t test was used for statistical analysis as appropriate (Microcal-Origin; OriginLab, Northampton, MA, USA). Mantel-Haenszel and Fisher exact tests were used for evaluating differences in toxicity between groups (StatCalc, University of Louisiana, Lafayette, LA, USA). Differences were considered significant at P < 0.05. All data are presented as mean ± SD, unless indicated otherwise. All experiments were repeated in triplicate unless indicated otherwise. Data from representative experiments are shown.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Cytotoxic effect of oncolytic herpes viruses on prostate carcinomas in vitro
We first evaluated the oncolytic efficacy of NV1023, a replication component herpes vector, against prostate carcinoma cell lines with a propensity to invade nerves. The effect of NV1023 on PC3, LNCaP, and DU145 human prostate cancer cell lines was studied in vitro by cytotoxic (LDH) assays during a 7-day period. Figure 1 shows the cytotoxic effects of NV1023 against the 3 prostate cell lines when multiplicity of infection (MOI, or the ratio of viral particles to cancer cells) of 0.01–1 was used. At a MOI of 0.1, all LNCaP and DU145 cells died by day 3–5. PC3 cells were more resistant, and showed 100% death by day 5–7 with a MOI of 1. Based on this data, we continued and evaluated the oncolytic effect of NV1023 in vivo using 2 prostate carcinoma cell lines, DU145 and PC3, which had a slight sensitivity profile to the virus.


Figure 1
View larger version (7K):
[in this window]
[in a new window]

  		Figure 1. Cytotoxic effect of oncolytic herpes viruses on prostate carcinomas in vitro. Cytotoxicity assays of 3 human prostate carcinoma cell lines demonstrating sensitivity to NV1023 at multiplicity of infection (MOI) of 0.01, 0.1, and 1. A) PC3 cells; B) LNCaP cells; and C) DU145 cells. All cell lines showed sensitivity to the viruses.

Safety profile of intraneural inoculation of NV1023
We first evaluated the safety profile of NV1023 treatment on nerves by direct left sciatic nerve injection. The left sciatic nerve was injected with saline and served as a negative control. The safety profile of NV1023 (5x107 pfu) was compared with wild-type (F-strain HSV) infection at high dose (5x106 pfu) and low dose (1x106 pfu). High-dose wild-type HSV-1 caused systemic disease, total paralysis, and 22% weight loss by day 5, at which time the mice required euthanasia (n=4). Mice treated with the lower dose had 13% weight loss at day 5. These mice developed 17% weight loss and complete paralysis by day 7 and required euthanasia (n=4). All mice treated with the wild-type virus had pathological findings associated with HSV infections, including skin lesions, dilated colon, and severe spinal deformation. Histopathologic analysis of the ipsilateral sciatic nerve of mice treated with high-dose wild-type HSV-1 showed signs of nerve swelling, Wallerian degeneration, demyelinization, and inflammation (Fig. 2 A). In the low-dose group, virus-induced pathological changes in the nerves consisted of an increase of mononuclear cells and neuronal damage. In contrast, mice treated with a high dose of NV1023 survived up to >60 days, when the study was terminated. We observed no significant toxicity attributable to the virus, such as change in nerve function, body weight, behavior, skin condition, infections, or mortality (n=30). Histological analysis of the nerves performed 7 days after inoculation with high-dose NV1023 revealed only mild Wallerian degeneration (Fig. 2A ; n=4).


Figure 2
View larger version (96K):
[in this window]
[in a new window]

  		Figure 2. Infection of sciatic nerves with wild-type HSV and NV1023. A) The safety profile of wild-type HSV-1 (high dose, 5x106 pfu; and low dose, 1x106 pfu) and NV1023 (high dose, 5x107 pfu) was assessed 5–7 days after direct left sciatic nerve injection. Saline injection served as control. Histopathologic analysis of nerves injected with high-dose wild-type HSV-1 showed signs of nerve swelling, Wallerian degeneration, demyelinization, and inflammation. Nerves inoculated with high-dose NV1023 revealed only mild Wallerian degeneration. Scale bars = 130 µm. B, C) X-gal staining of nerves treated with NV1023, performed 48 h after viral infection. Normal sciatic nerve (B) and sciatic nerve invaded by PC3 cells (C) were injected with NV1023. Prostate cancer cells invading nerves had significant X-gal staining (blue cells), whereas normal nerves (devoid of tumor) had no signal. Scale bar = 120 µm. D) Real-time PCR was performed on RNA extracts from sciatic nerves invaded by tumors 14 days after viral injection (n=4 per group). The primers for the ICP6 and LAT genes were used. SC = spinal cord; LSN = left sciatic nerve; RSN = right sciatic nerve.

Herpes vectors selectively infect cancer cells but not nerves in vivo
Next, a murine model of neural invasion was established by implanting tumor cells in the left sciatic nerve. To evaluate the ability of NV1023 to selectively infect cancer cells invading nerves, we used histochemical staining for X-gal as an infection marker. The gene coding for this protein is an E. coli β-galactosidase (lacZ) gene, which was inserted at the US10–12 locus of the virus and is absent in eukaryotic cells. Histochemical staining for X-gal revealed significant expression of lacZ by NV1023 within tumors at >48 h post-treatment (Fig. 2) . No X-gal staining was evident in tumors that were treated with saline or in nerves without tumors treated with NV1023 (n=4 per group).

Quantitative RT-PCR was performed on RNA extracts from sciatic nerves invaded by tumors that were treated with NV1023 at 5 x 107 pfu. The primers for the HSV-1 ICP6 and LAT genes were used to assess for presence of NV1023. At 14 days after NV1023 viral injection (21 days after tumor cell implantation), high levels of ICP6 and LAT mRNA expression were detected in the left sciatic nerve and to a lesser degree in the spinal cord and contralateral nerve. ICP6 and LAT were not expressed in the brain or in saline-treated nerves (Fig. 2 ; n=4 per group).

Treatment of neural invasion by cancer cells in vivo
We next assessed the propensity of DU145 and PC3 human prostate carcinoma-derived tumors to induce sciatic nerve paralysis in nude mice and the ability of NV1023 to treat the tumor while preserving nerve function. Neurally invasive tumors were established in a distal site of the sciatic nerve of nude mice. One week after tumor implantation, the sciatic nerves were exposed and examined under the dissecting microscope before injection of NV1023 or saline. In all of the treated animals, there was a visible tumor at day 7. Animals were then randomized into 2 groups and treated with intratumoral injection of NV1023 (5x107 pfu) or saline. The function of the sciatic nerve was evaluated weekly after treatment. Measures of nerve function included gross animal behavior, hind limb function, and ability of the mice to spread their paw following manual extension of the body. The diameter of the sciatic nerve proximal to the tumor (i.e., the ability of cancer cells to invade along the nerve) was also evaluated using a dedicated high-resolution small-animal ultrasound system.

In the control group (n=10), mice with PC3-derived tumors began to develop left hind limb paralysis 4 wk after tumor implantation (Fig. 3 ). Most animals developed complete left hind limb paralysis by wk 8, and the animals were euthanized. In contrast, most mice treated with intraneural injection of a single dose of NV1023 (n=14) had normal nerve function. At wk 8, all mice with PC3-derived tumors treated with NV1023 had normal nerve function, whereas all mice treated with saline injection had nerve paralysis. Figure 3 clearly shows the difference in nerve function between the treatment and control groups (P<0.0001). The right hind limb served as control and was intact in all of the animals. Eleven weeks after treatment with NV1023, a decrease in the sciatic nerve score of the NV1023-treated mice (Fig. 3B ) resulted from tumor regrowth, demonstrated by high resolution ultrasound imaging in wk 8–14 (Fig. 3C ). At 14 wk, the study was terminated, the mice were euthanized, and the nerves were analyzed histologically. Similar results were found in the group of mice with DU145-derived tumors (Fig. 4 ). All of the control mice in this group had nerve paralysis at wk 6 and had to be euthanized (n=7). On the other hand, 8 of 14 mice with DU145-derived tumors treated with NV1023 had normal nerve function, and another 6 had mild paresis. Data analysis of sciatic nerve function among the DU145 mice showed significantly better scores in the NV1023 treatment group compared with the controls (P<0.0001).


Figure 3
View larger version (32K):
[in this window]
[in a new window]

  		Figure 3. Treatment of neural invasion by PC3 prostate cancer cells using NV1023. A) Representative images of mice with neural invasion of PC3 prostate tumors. Tumor cells were grown in the left sciatic nerve and treated with intraneural injection of saline (left panel) or NV1023 (right panel). Mice treated with NV1023 (n=14) showed normal hind limb function 8 wk after tumor implantation, while saline-treated mice (n=8) had complete left hind limb paralysis. (Arrows indicate paw function in the tumor side.) The contralateral paw (on the right) served as control and had normal function. B) Mean left sciatic nerve score of saline (solid circles) and NV1023 (open circles) -treated mice was measured weekly. At wk 8, most of the saline-treated mice developed paralysis and had to be euthanized. To allow long-term follow-up of nerve function, half of the NV1023-treated mice were randomly selected and examined for 6 more weeks. C) High-resolution ultrasound measurements of sciatic nerve diameter in the control (solid circles) and NV1023 (open circles) -treated mice. The diameter of the sciatic nerve was significantly smaller in the NV1023 treatment group, even 14 wk after establishment of tumors (P<0.0001). D) In the saline-treated group, there was significant reduction in the left sciatic nerve index (paw span in mm) from wk 1–8 (solid circles). The right paw had normal function (open circles). E) Left sciatic nerve index of NV1023-treated mice showing normal function in most animals (n=12 of 13 mice). F) Left sciatic nerve index of NV1023-treated mice 14 wk after treatment showing preserved nerve function in most of the animals. Each circle represents 1 mouse. Solid circles represent the left paw (tumors treated with NV1023); open circles represent the right paw (normal nerves).


Figure 4
View larger version (34K):
[in this window]
[in a new window]

  		Figure 4. Treatment of neural invasion by DU145 prostate cancer cells using NV1023. A) Representative images of mice with neural invasion of DU145 prostate tumors. Tumor cells were grown in the left sciatic nerve and treated with intraneural injection of saline (left panel) or NV1023 (right panel). B) Mean left sciatic nerve score of saline (solid circles, n=7) and NV1023 (open circles, n=14) -treated mice was measured weekly. C) High-resolution ultrasound measurements of sciatic nerve diameter in the control (solid circles) and NV1023 (open circles) -treated mice. D) Sciatic nerve index (paw span in mm) from wk 1–6 (solid circles) in the control group. The right paw had normal function (open circles). E) Left sciatic nerve index of NV1023-treated mice. F) Left sciatic nerve index of NV1023-treated mice, 12 wk after treatment. Each circle represents 1 mouse. Solid circles represent the left paw (tumors treated with NV1023); open circles represent the right paw (normal nerves).

High-resolution ultrasound images of mice with PC3- and DU145-derived tumors demonstrated significantly smaller left sciatic nerve diameter in the NV1023-treated groups compared with the controls (performed at wk 6 and 8, respectively). Figure 3C shows that on wk 8, the nerve diameter in the PC3 control animals was >4-fold larger compared with those treated with NV1023 (P<0.0001). Similarly, in the DU145 group, the sciatic nerve diameter was significantly larger in the controls compared with the treated animals (Fig. 4C , P<0.0001).

Once >80% of the saline-treated mice developed paralysis, they were euthanized. To allow histopathological comparison, half of the NV1023-treated mice (that were randomly selected) were euthanized (mice with DU145-derived tumors at wk 6 and mice with PC3-derived tumors at wk 8). Tumor diameter was measured at the implantation site (distal to the bifurcation of the tibial and common peroneal nerves). In addition, in order to estimate the degree of neural invasion along the sciatic nerve, we systematically measured the nerve diameter 4 mm proximal to the site of implantation, immediately before its entry to the spinal column (see Fig. 5 ). The mean diameter of the contralateral (right) sciatic nerve was 0.54 ± 0.08 mm (n=7). H&E staining and detailed histopathological analysis showed that the NV1023-treated mice had significantly smaller nerve diameter compared with controls (P<0.005) (Figs. 5 and 6 ). Most importantly, whereas most of the control mice had significant proximal nerve invasion, all of the mice treated with NV1023 had no evidence of proximal neural invasion. This conclusion is demonstrated by the normal proximal nerve diameter in the NV1023-treated mice relative to an increase in proximal nerve diameter in the control group (P<0.03 for PC3 and P<0.0001 for DU145-derived tumors; Fig. 5 ). Immunohistochemical staining with the proliferation marker Ki-67 showed dense staining in sciatic nerves treated with saline (Fig. 6E ). On the other hand, Ki67 staining was rarely positive in nerves treated with NV1023 (Fig. 6F ; n=6).


Figure 5
View larger version (107K):
[in this window]
[in a new window]

  		Figure 5. Representative in situ illustrations and high-resolution ultrasound images of sciatic nerves invaded by prostate carcinoma cells and treated with saline or NV1023. A) In situ image of a nerve invaded by PC3 cells treated with saline. B) In situ image of a nerve invaded by PC3 cells treated with NV1023 (5x107 pfu). C) High-resolution ultrasound imaging performed on a saline-treated nerve at wk 8. D) Ultrasound image of an NV1023-treated nerve. The red line delineates the peripheral outlines of the sciatic nerve. The NV1023-treated animals had significantly smaller nerve/tumor diameter compared with controls. The white arrowheads indicate the area of tumor implantation; arrows indicate the diameter of the proximal nerve; L = lateral, M = medial.


Figure 6
View larger version (69K):
[in this window]
[in a new window]

  		Figure 6. Representative histopathologic photomicrographs of tumors treated with saline or NV1023. A) H&E staining of a sciatic nerve invaded by PC3 cancer cells treated with saline (performed at wk 8). B) Nerve invaded by PC3 cells treated with NV1023. C) Nerve invaded by DU145 cells treated with saline. D) Nerve invaded by DU145 cells treated by NV1023. Scale bars = 200 µm. E) Immunostaining with antibodies against Ki-67 (a proliferation marker) in a nerve treated with saline (the same nerve as in A). F) Ki-67 staining in a nerve treated with NV1023 (same nerve as in B). G) DU145 tumor diameters and the diameters of the proximal nerves in mice treated with NV1023 or saline. H) PC3 tumor diameters and proximal nerve diameters in saline- and NV1023-treated mice. Tumor diameter was measured at the implantation site, whereas the proximal sciatic nerve diameter was measured 4 mm proximal to the implantation site, prior to its insertion to the spinal cord. Each circle represents 1 animal. The horizontal bars represent the mean.

Finally, we evaluated the long-term effect of NV1023 treatment by monitoring nerve function in the remaining NV1023-treated animals. At the time the study was completed (14 wk after tumor implantation), 5 of the 6 mice with PC3-derived tumors had normal nerve function (Fig. 3F ). Ultrasound measurement of the nerves showed small increases in diameter; however, even at wk 14, the nerve diameter was significantly smaller compared with the control group (Fig. 3C ). Histological evaluation of the nerves in the PC3 group revealed normal nerves in 4 of the 6 mice 14 wk after establishment of the tumors. In the DU145 group, 2 mice had intact nerve function and the other 5 had mild paralysis (Fig. 4F ). Two of the nerves in this group were devoid of cancer cells.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Developments in the fields of cancer biology have had a major impact on the prevention, diagnosis, and treatment of patients with prostate cancer (19 20 21) . Nevertheless, means to preserve nerve function after surgical resection of invasive prostate cancer are meager. The aim of this study was to develop a therapeutic modality that will selectively treat cancer while preserving function. The goal of such a treatment would be to improve both survival and quality of life of patients with neurotrophic cancers. Using an in vivo animal model, we demonstrate that a single injection of an attenuated, replication-competent, oncolytic virus can effectively treat nerve invasion by prostate carcinoma cells while preserving its physiological function. Significant reduction in tumor size and inhibition of neural invasion was demonstrated in 2 different human prostate carcinoma cell lines following treatment. Remarkably, whereas all of the control animals developed complete nerve paralysis within a few weeks, most of the treated animals had intact nerve function up to 3.5 months after treatment.

Our study introduces an alternative therapeutic modality to surgery that preserves the anatomy of the peripheral neural bundle, with the potential of maintenance of function. Furthermore, combined with other therapeutic modalities such as conservative surgery, radiotherapy, chemotherapy, and hormonal therapy, oncolytic viruses may offer a chance for cure in advanced prostate cancer while preserving potency and continence. Intraoperative administration of oncolytic herpes vectors to the infiltrated nerves may also prove advantageous by being able to potentially treat anatomic areas inaccessible by surgery. Our PCR data showed viral gene expression in the spinal cords and contralateral nerves of treated mice. Therefore, the PCR data likely reflect HSV gene expression in both human (cancer cell) and mouse (nerve) tissue. Nevertheless, our lacZ expression analysis showed positive staining primarily in cancer cells and not in nerves devoid of tumors (Fig. 2) , suggesting that productive infection occurs mainly in the tumor cells infiltrating the nerve bundles. The ability of NV1023 to infect spinal cord neurons and collateral nerves demonstrates the potential of herpes oncolytic therapy to eradicate extensions of tumor beyond the margins of the prostate. It was shown that injection of HSV-1-derived vectors to sciatic nerves of mice induces LacZ transgene expression in the spinal cord, specifically in motor neurons and dorsal root ganglion cells (22 , 23) . In our experiments, we looked at the expression of LacZ in sciatic nerve bundles (axons) invaded by prostate cancer to show the ability of NV1023 to infect tumor cells. Spinal cord or dorsal root ganglia specimens were not investigated in our study. We, as well as others (22 , 23) , rarely found LacZ expression in sciatic nerves devoid of cancer cells.

The in vitro data show that DU145 was more sensitive to the virus than was PC3; however, our in vivo experiments show that NV1023 was more efficient as a nerve-sparing treatment against PC3-derived tumors. This discrepancy between the in vitro and in vivo results could originate from differences in the biological behavior of the PC3 and DU145 cell lines. Variation in tumor growth, cancer cell migration, or the ability of cells to invade nerves could potentially affect the efficacy of viral treatment in vivo. Indeed, we have found that nerves invaded by DU145 cancer cells had a significantly larger proximal diameter than nerves invaded by PC3 cells, suggesting that DU145 has a greater tendency to invade nerves (Fig. 6) .

Our results showed that intraneural injection of NV1023 is safe in immunosuppressed animals. In contrast to nerve infection with wild-type HSV-1, which causes complete nerve paralysis and death from disseminated disease (24) , a single deletion of the {gamma}134.5 gene appears to be sufficient to preserve peripheral nerve function and to prevent death in nude mice. Infection of nerve tissue by NV1023 was found to be safe, and none of the treated mice had clinically apparent neural toxicity or other side effects such as impaired behavior, weight loss, or death attributable to viral administration. The safety of NV1023 treatment was also demonstrated by our PCR data, showing that the NV1023 virus has a limited ability to spread to the brain. The safety profile of oncolytic HSV therapy was also demonstrated in rodents and primates as well as in patients with advanced cancers (11 , 15 , 25) . In a recent phase I clinical trial, patients with metastatic colorectal carcinoma treated with systemic administration of a similar virus suffered no adverse events attributable to the treatment (10) . In this study, we demonstrated the ability of a single administration of NV1023 to treat prostate cancer invading nerves. In addition to this, we have previously shown that a multiple dosing strategy may enhance therapeutic efficacy without adding observed toxicity in animals (26 , 27) .

HSV-1 vectors have several advantages that support their implementation as clinical grade therapeutics (28) . First, the deletion of nonessential genes can serve to increase tumor selectivity and to confer an improved safety profile. Second, they have the ability to infect cells without integrating their transgenes into the cell’s DNA. Third, effective antiviral drugs against HSV-1, including acyclovir and ganciclovir, can be used as additional safeguards against the virus (29) . Oncolytic herpes viral therapy was found to be synergistic with radiation, most probably because of the distinct mechanisms of action of oncolytic viral therapies and because ionizing radiation also produces cellular changes that favor viral replication (30 , 31) . Similar synergism was reported with chemotherapeutic agents such as cisplatin (32) , vincristine (33) , and mitomycin C (34 , 35) . Oncolytic herpes viruses carrying the murine immunostimulatory cytokine gene IL-12 were shown to incite significant immune response and anti-angiogenic effects in combination with direct tumor lysis in mouse models of cancer (36) . Such an approach may be beneficial in both enhancing therapeutic efficacy and potentially reducing the likelihood of tumor resistance.

Dissemination of cancer cells along peripheral nerves can induce chronic neuropathic pain, a common cause of morbidity in oncological patients (37) . Our results also suggest that oncolytic vectors administered to nerves may serve as palliative treatment against cancer-induced neuropathic pain. Delivery of HSV could be performed via transrectal ultrasound-guided injections to cavernous nerve bundles in patients with prostate cancer refractory to conventional therapy (38) . In addition, oncolytic therapy with herpes vectors may also provide benefits to patients with other neurotrophic tumors such as pancreatic, colorectal, and head and neck carcinomas, which are notoriously known to invade nerves (39 40 41) .

In conclusion, this study demonstrates the advantage of targeted therapy with oncolytic herpes vectors: preservation of function of peripheral nerves infiltrated by cancer. We have shown that nerve invasion by prostate carcinoma can be effectively treated with an attenuated oncolytic herpes virus. Viral therapy may be an alternative to surgical resection of nerves invaded by cancer cells, allowing preservation of erectile function. This modality has potential advantages for improving patient care and quality of life as well as the ability to deliver targeted cancer therapy to remote anatomic areas. Further investigation is warranted to determine the potential impact of oncolytic HSV therapy on patients with locally advanced neurotrophic cancers.


   ACKNOWLEDGMENTS
 
This work was supported by grants from the U.S. National Institutes of Health, the American College of Surgeons, and the Flight Attendant Medical Research Institute. We thank M. Greenblatt for editorial assistance.

Received for publication September 19, 2007. Accepted for publication January 3, 2008.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. American Cancer Society (2007) Cancer Facts and Figures American Cancer Society Atlanta, Georgia, USA.
  2. Quinn, D. I., Henshall, S. M., Brenner, P. C., Kooner, R., Golovsky, D., O'Neill, G. F., Turner, J. J., Delprado, W., Grygiel, J. J., Sutherland, R. L., Stricker, P. D. (2003) Prognostic significance of preoperative factors in localized prostate carcinoma treated with radical prostatectomy: importance of percentage of biopsies that contain tumor and the presence of biopsy perineural invasion. Cancer 97,1884-1893[CrossRef][Medline]
  3. Bonin, S. R., Hanlon, A. L., Lee, W. R., Movsas, B., al-Saleem, T. I., Hanks, G. E. (1997) Evidence of increased failure in the treatment of prostate carcinoma patients who have perineural invasion treated with three-dimensional conformal radiation therapy. Cancer 79,75-80[Medline]
  4. Maru, N., Ohori, M., Kattan, M. W., Scardino, P. T., Wheeler, T. M. (2001) Prognostic significance of the diameter of perineural invasion in radical prostatectomy specimens. Hum. Pathol. 32,828-833[CrossRef][Medline]
  5. Singer, P. A., Tasch, E. S., Stocking, C., Rubin, S., Siegler, M., Weichselbaum, R. (1991) Sex or survival: trade-offs between quality and quantity of life. J. Clin. Oncol. 9,328-334[Abstract]
  6. Tobisu, K. (2006) Function-preserving surgery for urologic cancer. Int. J. Clin. Oncol. 11,351-356[CrossRef][Medline]
  7. Penson, D. F., Feng, Z., Kuniyuki, A., McClerran, D., Albertsen, P. C., Deapen, D., Gilliland, F., Hoffman, R., Stephenson, R. A., Potosky, A. L., Stanford, J. L. (2003) General quality of life 2 years following treatment for prostate cancer: what influences outcomes? Results from the prostate cancer outcomes study. J. Clin. Oncol. 21,1147-1154[Abstract/Free Full Text]
  8. Mineta, T., Rabkin, S. D., Yazaki, T., Hunter, W. D., Martuza, R. L. (1995) Attenuated multi-mutated herpes simplex virus-1 for the treatment of malignant gliomas. Nat. Med. 1,938-943[CrossRef][Medline]
  9. Kooby, D. A., Carew, J. F., Halterman, M. W., Mack, J. E., Bertino, J. R., Blumgart, L. H., Federoff, H. J., Fong, Y. (1999) Oncolytic viral therapy for human colorectal cancer and liver metastases using a multi-mutated herpes simplex virus type-1 (G207). FASEB J. 13,1325-1334[Abstract/Free Full Text]
  10. Kemeny, N., Brown, K., Covey, A., Kim, T., Bhargava, A., Brody, L., Guilfoyle, B., Haag, N. P., Karrasch, M., Glasschroeder, B., Knoll, A., Getrajdman, G., Kowal, K. J., Jarnagin, W. R., Fong, Y. (2006) Phase I, open-label, dose-escalating study of a genetically engineered herpes simplex virus, NV1020, in subjects with metastatic colorectal carcinoma to the liver. Hum. Gene Ther. 17,1214-1224[CrossRef][Medline]
  11. Markert, J. M., Medlock, M. D., Rabkin, S. D., Gillespie, G. Y., Todo, T., Hunter, W. D., Palmer, C. A., Feigenbaum, F., Tornatore, C., Tufaro, F., Martuza, R. L. (2000) Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene Ther. 7,867-874[CrossRef][Medline]
  12. Chen, X., Schmidt, M. C., Goins, W. F., Glorioso, J. C. (1995) Two herpes simplex virus type 1 latency-active promoters differ in their contributions to latency-associated transcript expression during lytic and latent infections. J. Virol. 69,7899-7908[Abstract]
  13. Maguire-Zeiss, K. A., Bowers, W. J., Federoff, H. J. (2001) HSV vector-mediated gene delivery to the central nervous system. Curr. Opin. Mol. Therapeut. 3,482-490[Medline]
  14. Wong, R. J., Patel, S. G., Kim, S., DeMatteo, R. P., Malhotra, S., Bennett, J. J., St-Louis, M., Shah, J. P., Johnson, P. A., Fong, Y. (2001) Cytokine gene transfer enhances herpes oncolytic therapy in murine squamous cell carcinoma. Hum. Gene Ther. 12,253-265[CrossRef][Medline]
  15. Meignier, B., Martin, B., Whitley, R. J., Roizman, B. (1990) In vivo behavior of genetically engineered herpes simplex viruses R7017 and R7020. II. Studies in immunocompetent and immunosuppressed owl monkeys (Aotus trivirgatus). J. Infect. Dis. 162,313-321[Medline]
  16. Mashour, G. A., Moulding, H. D., Chahlavi, A., Khan, G. A., Rabkin, S. D., Martuza, R. L., Driever, P. H., Kurtz, A. (2001) Therapeutic efficacy of G207 in a novel peripheral nerve sheath tumor model. Exp. Neurol. 169,64-71[CrossRef][Medline]
  17. Hu, D., Hu, R., Berde, C. B. (1997) Neurologic evaluation of infant and adult rats before and after sciatic nerve blockade. Anesthesiology 86,957-965[CrossRef][Medline]
  18. Geller, A. I., Breakefield, X. O. (1988) A defective HSV-1 vector expresses Escherichia coli β-galactosidase in cultured peripheral neurons. Science 241,1667-1669[Abstract/Free Full Text]
  19. Chang, S. S., Bander, N. H., Heston, W. D. (1999) Monoclonal antibodies: will they become an integral part of the evaluation and treatment of prostate cancer–focus on prostate-specific membrane antigen?. Curr. Opin. Urol. 9,391-395[CrossRef][Medline]
  20. Cooperberg, M. R., Moul, J. W., Carroll, P. R. (2005) The changing face of prostate cancer. J. Clin. Oncol. 23,8146-8151[Abstract/Free Full Text]
  21. Thompson, I. M., Goodman, P. J., Tangen, C. M., Lucia, M. S., Miller, G. J., Ford, L. G., Lieber, M. M., Cespedes, R. D., Atkins, J. N., Lippman, S. M., Carlin, S. M., Ryan, A., Szczepanek, C. M., Crowley, J. J., Coltman, C. A., Jr (2003) The influence of finasteride on the development of prostate cancer. N. Engl. J. Med. 349,215-224[Abstract/Free Full Text]
  22. Dobson, A. T., Margolis, T. P., Sedarati, F., Stevens, J. G., Feldman, L. T. (1990) A latent, nonpathogenic HSV-1-derived vector stably expresses β-galactosidase in mouse neurons. Neuron 5,353-360[CrossRef][Medline]
  23. Perez, M. C., Hunt, S. P., Coffin, R. S., Palmer, J. A. (2004) Comparative analysis of genomic HSV vectors for gene delivery to motor neurons following peripheral inoculation in vivo. Gene Ther. 11,1023-1032[CrossRef][Medline]
  24. Ramakrishnan, R., Fink, D. J., Jiang, G., Desai, P., Glorioso, J. C., Levine, M. (1994) Competitive quantitative PCR analysis of herpes simplex virus type 1 DNA and latency-associated transcript RNA in latently infected cells of the rat brain. J. Virol. 68,1864-1873[Abstract/Free Full Text]
  25. Ram, Z., Culver, K. W., Oshiro, E. M., Viola, J. J., DeVroom, H. L., Otto, E., Long, Z., Chiang, Y., McGarrity, G. J., Muul, L. M., Katz, D., Blaese, R. M., Oldfield, E. H. (1997) Therapy of malignant brain tumors by intratumoral implantation of retroviral vector-producing cells. Nat. Med. 3,1354-1361[CrossRef][Medline]
  26. Kim, S. H., Carew, J. F., Kooby, D. A., Shields, J., Entwisle, C., Patel, S., Shah, J. P., Fong, Y. (2000) Combination gene therapy using multiple immunomodulatory genes transferred by a defective infectious single-cycle herpes virus in squamous cell cancer. Cancer Gene Ther. 7,1279-1285[CrossRef][Medline]
  27. Yu, Z., Eisenberg, D. P., Singh, B., Shah, J. P., Fong, Y., Wong, R. J. (2004) Treatment of aggressive thyroid cancer with an oncolytic herpes virus. Int. J. Cancer 112,525-532[CrossRef][Medline]
  28. Shen, Y., Nemunaitis, J. (2006) Herpes simplex virus 1 (HSV-1) for cancer treatment. Cancer Gene Ther. 13,975-992[CrossRef][Medline]
  29. Jeyaretna, D. S., Rabkin, S. D., Martuza, R. L. (2007) Oncolytic herpes simplex virus therapy for peripheral nerve tumors. Neurosurg. Focus 22,E4[Medline]
  30. Blank, S. V., Rubin, S. C., Coukos, G., Amin, K. M., Albelda, S. M., Molnar-Kimber, K. L. (2002) Replication-selective herpes simplex virus type 1 mutant therapy of cervical cancer is enhanced by low-dose radiation. Hum. Gene Ther. 13,627-639[CrossRef][Medline]
  31. Jarnagin, W. R., Zager, J. S., Hezel, M., Stanziale, S. F., Adusumilli, P. S., Gonen, M., Ebright, M. I., Culliford, A., Gusani, N. J., Fong, Y. (2006) Treatment of cholangiocarcinoma with oncolytic herpes simplex virus combined with external beam radiation therapy. Cancer Gene Ther. 13,326-334[CrossRef][Medline]
  32. Chahlavi, A., Todo, T., Martuza, R. L., Rabkin, S. D. (1999) Replication-competent herpes simplex virus vector G207 and cisplatin combination therapy for head and neck squamous cell carcinoma. Neoplasia 1,162-169[CrossRef][Medline]
  33. Cinatl, J., Jr, Cinatl, J., Michaelis, M., Kabickova, H., Kotchetkov, R., Vogel, J. U., Doerr, H. W., Klingebiel, T., Driever, P. H. (2003) Potent oncolytic activity of multimutated herpes simplex virus G207 in combination with vincristine against human rhabdomyosarcoma. Cancer Res. 63,1508-1514[Abstract/Free Full Text]
  34. Toyoizumi, T., Mick, R., Abbas, A. E., Kang, E. H., Kaiser, L. R., Molnar-Kimber, K. L. (1999) Combined therapy with chemotherapeutic agents and herpes simplex virus type 1 ICP34.5 mutant (HSV-1716) in human non-small cell lung cancer. Hum. Gene Ther. 10,3013-3029[CrossRef][Medline]
  35. Bennett, J. J., Adusumilli, P., Petrowsky, H., Burt, B. M., Roberts, G., Delman, K. A., Zager, J. S., Chou, T. C., Fong, Y. (2004) Up-regulation of GADD34 mediates the synergistic anticancer activity of mitomycin C and a {gamma}134.5 deleted oncolytic herpes virus (G207). FASEB. J. 18,1001-1003[Abstract/Free Full Text]
  36. Parker, J. N., Gillespie, G. Y., Love, C. E., Randall, S., Whitley, R. J., Markert, J. M. (2000) Engineered herpes simplex virus expressing IL-12 in the treatment of experimental murine brain tumors. Proc. Natl. Acad. Sci. U. S. A. 97,2208-2213[Abstract/Free Full Text]
  37. Rigor, B. M., Sr (2000) Pelvic cancer pain. J. Surg. Oncol. 75,280-300[CrossRef][Medline]
  38. Kravchick, S., Cytron, S., Peled, R., Ben-Dor, D., Kravchenko, Y. (2003) Colour Doppler ultrasonography for detecting perineural invasion (PNI) and the value of PNI in predicting final pathological stage: a prospective study of men with clinically localized prostate cancer. BJU Int. 92,28-31[CrossRef][Medline]
  39. Van der Wal, J. E., Snow, G. B., van der Waal, I. (1990) Intraoral adenoid cystic carcinoma. The presence of perineural spread in relation to site, size, local extension, and metastatic spread in 22 cases. Cancer 66,2031-2033[CrossRef][Medline]
  40. Tanaka, A., Watanabe, T., Okuno, K., Yasutomi, M. (1994) Perineural invasion as a predictor of recurrence of gastric Cancer. Cancer 73,550-555[CrossRef][Medline]
  41. Conlon, K. C., Klimstra, D. S., Brennan, M. F. (1996) Long-term survival after curative resection for pancreatic ductal adenocarcinoma. Clinicopathologic analysis of 5-year survivors. Ann. Surg. 223,273-279[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
P. Brader, K. J. Kelly, N. Chen, Y. A. Yu, Q. Zhang, P. Zanzonico, E. M. Burnazi, R. E. Ghani, I. Serganova, H. Hricak, et al.
Imaging a Genetically Engineered Oncolytic Vaccinia Virus (GLV-1h99) Using a Human Norepinephrine Transporter Reporter Gene
Clin. Cancer Res., June 1, 2009; 15(11): 3791 - 3801.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
fj.07-097808v1
22/6/1839    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kelly, K.
Right arrow Articles by Gil, Z.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kelly, K.
Right arrow Articles by Gil, Z.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS