The detection of tyrosinase mRNA in peripheral blood samples is unlikely to aid in the management of patients with localised malignant melanoma.

A number of authors have reported the detection of tyrosinase mRNA in the peripheral blood of patients with malignant melanoma using the reverse transcription polymerase chain reaction (RT-PCR). The precise value of this assay as a prognostic tool, however, remains in doubt. This is particularly so with relation to localised disease, where relatively little data has been accumulated. In this study we analysed the peripheral blood of 50 consecutive patients with primary malignant melanoma referred to a plastic surgical centre with the facility of a pigmented lesion clinic. Samples were analysed from an additional 35 patients with advanced melanoma disease and 35 patients with benign pigmented cutaneous lesions. We were able to identify tyrosinase transcripts in the peripheral blood of only two of 50 patients with localised disease. Of those with more advanced disease, a positive finding was found in three with regional disease and four patients with metastatic spread. Stage of disease was found to correlate significantly with PCR status. No correlation was identified with other prognostic markers or with outcome over a three-year period. This data would support the conclusion that the detection of tyrosinase mRNA in peripheral blood is likely to be of little value as an aid in the management of patients with early malignant melanoma.

Detection of melanoma seeding during surgical procedures--an RT-PCR based model.

AIM: It has long been suggested that malignant cells may be shed into the blood stream during any given surgical procedure for cancer. A number of studies have now reported the detection of occult melanoma cells in peripheral blood using a reverse transcriptase polymerase chain reaction (RT-PCR) based assay. The principal aim of these studies has been to determine a prognostic value for the test and not to evaluate the influence of intervention upon results. METHODS: In this pilot study we aimed to determine whether the assay could be used as a model to detect cells that are seeded during surgery. Peripheral blood samples were obtained pre- and post-operatively on twenty patients undergoing surgery for malignant melanoma - ten with primary disease and ten undergoing regional lymphadenectomy. A further ten patients undergoing surgery for non-melanoma conditions provided controls. RESULTS: Using RT-PCR, it was possible to identify tyrosinase transcripts in the peripheral blood of one of ten patients undergoing excision of local disease and four of ten undergoing surgery for regional metastatic disease. CONCLUSION: It was concluded that this technique does enable detection of a greater percentage of RT-PCR findings post-operatively. This in turn may provide a means for optimizing or comparing surgical techniques and provides a potential guide in the use of adjuvant therapies.

Targeted cytokine delivery to neuroblastoma.

The aim of this study was to construct a fusion protein from the cytokine granulocyte/macrophage colony-stimulating factor (GM-CSF) and a single-chain Fv fragment (scFv D29) and to investigate its potential to activate cells of the immune system against neuroblastoma cells expressing neural cell adhesion molecule (NCAM). Mammalian cell expression of the scFv D29-GM-CSF fusion protein was compared using a number of vectors, including retroviral and adenoviral vectors. The resultant fusion protein, expressed by HeLa cells, was found by ELISA to bind immobilized recombinant NCAM. Moreover, FACS analysis confirmed binding to the human neuroblastoma cell line SKNBE and a murine neuroblastoma cell line engineered to express the glycosylphosphatidylinositol form of human NCAM (N2A-rKNIE). The fusion protein was also found to stimulate the proliferation of the FDC-P1 haemopoietic cell line, which is dependent on GM-CSF (or interleukin 3) for continued growth. In vitro clonogenic assays indicated that scFv-GM-CSF could selectively induce growth inhibition of SKNBE cells by murine lymphoid cells.

Generation of a humanised single chain Fv (Scfv) derived from the monoclonal Eric-1 recognising the human neural cell adhesion molecule.

BACKGROUND: Murine monoclonal antibody (MoAb) ERIC-1 recognises an epitope on the neural cell adhesion molecule (NCAM) whose expression in paediatric and adult tissues is confined mainly to the brain, peripheral nerve, and adrenal medulla. Anti-NCAM antibodies have been used for the treatment and diagnosis of a number of tumours, including neuroblastoma. However, whole antibody exhibits poor penetration into solid tumour deposits and rapid systemic clearance upon repeated administration due to development of a human antimouse antibody (HAMA) response. PROCEDURE: To overcome these problems, recombinant DNA techniques have been used to humanise and assemble the ERIC-1 immunoglobulin variable heavy (VH) and light (VL) chains into a single chain Fv (scFv). RESULTS: Three humanised scFv clones were identified which differ from the predicted humanised sequence by occasional amino acid changes, but these maintain the same specificity as the original ERIC-1 MoAb. CONCLUSIONS: The humanised scFv may prove to be a useful reagents in the treatment and diagnosis of a variety of neuroectodermal tumours and can clearly form a suitable template for the generation of a fully humanised ERIC-1 MoAb.

Treatment of neoplastic meningitis by targeted radiation using (131)I-radiolabelled monoclonal antibodies. Results of responses and long term follow-up in 40 patients.

Between 1984 and 1993, monoclonal antibodies (MAbs) radiolabelled with (131)I were administered into the CSF of 52 patients with neoplastic meningitis (meningosis) with progressive disease despite active conventional therapy. Selection of MAbs was based on immunoreactivity with patients' tumour and lack of binding to normal central nervous system (CNS) tissue. Following full clinical assessment and neuro-imaging which included isotope flow study of CSF pathways, (131)I-MAb was administered via a ventricular access device, lumbar catheter or both. Radioisotope activity varied from 25 mCi to 160 mCi in adults. Dose escalation was carried out and some patients received multiple doses. Distribution of (131)I-MAb and clearance kinetics were derived from serial scintigraphy and CSF/blood sampling. Evidence of localisation to tumour was frequently observed. Toxicity was minimal and easily treated, although one death occurred, possibly due to a seizure. The best results were obtained in primitive neuroectodermal tumour (n=22), where 53% of evaluable cases had responses and 11% had stable disease, adults responding better than children. Three exceptional survivals have been recorded; one patient leads a normal life at 10 years 11 months, one case is alive and normal at 3 years, 2 months. A third case survived in good condition for 8 years. The mean survival of responders was 39 months and non-responders 4 months. In the total series, 50% of patients survived for at least one year with 2 long term survivors. CSF therapy with (131)I-MAb appears to be valuable as a single agent or when used in combination with other modalities. Results of treating leukaemia and carcinoma cases suggest that re-seeding into the CSF compartment from active systemic disease may account for early relapse in the CNS. One carcinoma case with no apparent systemic disease made a remarkable response and survival for 4 years following a single treatment. Neoplastic meningitis generally carries a dismal prognosis. The results obtained in this initial trial are sufficiently encouraging to stimulate further attempts at CSF therapy with (131)I-MAbs.

Dose escalation with repeated intrathecal injections of 131I-labelled MAbs for the treatment of central nervous system malignancies.

We have previously demonstrated a 33% response rate in patients with primitive neurectodermal tumours after the direct injection of 131I-monoclonal antibodies (MAbs) into the cerebrospinal fluid (CSF). Dose-limiting toxicity is myelosuppression due to the passage of the radioimmunoconjugate from the CSF to the blood compartment. This occurs at doses of 2220 MBq of 131I-MAb and above, although this is not seen in all patients studied and appears to be related to the degree of prior therapy received. Rather than attempting to improve the efficacy of this approach to the treatment of disseminated disease within the CSF compartment by dose escalation and haemopoietic rescue, we have explored the possibility of repeatedly administering the radioimmunoconjugate. Eight patients were recruited to the study, two of whom received two and six of whom received three injections of 131I-MAb. After repeated administration of 131I-MAb pharmacokinetic data revealed that, with one exception, the radioimmunoconjugate cleared from the CSF compartment with similar kinetics, while its residence time in the blood decreased with each injection. This was due to the development of an anti-mouse Ig response in the blood. Clearance of 131I-MAb from the ventricular CSF appears to be independent of the presence of an anti-mouse Ig response in this compartment. The differential clearance of the radioimmunoconjugate from the ventricular CSF and from the blood results in a marked increase in the therapeutic index that can be achieved. Up to 5920 MBq of 131I-MAb was administered as the third injection of radioimmunoconjugate and combined doses of up to 12,500 MBq were given without either haematological or neurological toxicity. These data illustrate that dose escalation and thus an increase in the dose rate delivered to tumour cells within the CSF is possible if ways are found to reduce the residence time of the radioimmunoconjugate in the blood compartment. Suggestions as to how this can best be achieved are reviewed in detail.

Direct injection of 90Y MoAbs into glioma tumor resection cavities leads to limited diffusion of the radioimmunoconjugates into normal brain parenchyma: a model to estimate absorbed radiation dose.

PURPOSE: Previously we have demonstrated that radioimmunoconjugates can be injected into glioma resection cavities to deliver a boost of radiation to the cavity edge with little toxicity to the normal brain. In the mathematical models we have previously published to assist in the development of this strategy we assumed that antibody remains associated with the cavity edge and no diffusion occurs. However, moderate diffusion might be beneficial while, if this were excessive, it would decrease the therapeutic index markedly. METHODS AND MATERIALS: Selected individuals with relapsed malignant glioma underwent further surgical debulking; 90Y MoAb radioimmunotherapy; and open biopsy to determine the extent to which the conjugate diffuses from the cavity edge. Samples from these patients were taken in radial tracts and the corrected activity in each sample was plotted against distance from the cavity wall to determine appropriate diffusion constants. RESULTS: Our data indicates that diffusion of radioimmunoconjugate from the edge of a glioma resection cavity appears to be an exponential process. The mean Ro for each patients data set ranged from 0.48-0.63 (overall mean 0.6) cm. A dosimetric model was developed that translates these measurements into estimates of radiation dose. Applying the clinical data to this model indicates that, in each patient, the peak dose is delivered 0.16-0.18 cm below the cavity margin, and the mean dose at 2 cm deep is 5.3% (4.4-5.8%) of the peak. CONCLUSION: The model described can be used to translate diffusion constants measured by any method into estimates of absorbed radiation dose. Assuming similar diffusion kinetics, it can also be used to predict the dose deposited if alternative radionuclides are linked to MoAb, although the effect of dose rate should also be considered. In the future, it may be possible to manipulate diffusion by using either different antibodies or antibody fragments for intracavity radioimmunotherapy. Before this can be done, however, further data are needed and a noninvasive approach to measuring diffusion would clearly be optimal.

Sub-lethal effects of exposing the human melanoma cell line SKmel-23 to 532 nm laser light.

The human melanoma cell line SKmel-23 has been used to investigate the sub-lethal damage that can occur as a result of exposing melanin containing cells to light (532 nm) from a frequency doubled Q-switched (Nd:YAG) laser. A dose response curve was obtained, which indicates that at energy levels of 0.6 J/cm2 and below no effect on either the viability or growth rate of the cell line was observed. Above this, cells rapidly died and at an energy level of 2.0 J/cm2, only approximately 15% of cells survived. This contrasts with the effects on the G361 melanoma line, which contains far less melanosomes, as an LD50 for this cell line was approximately 5.5 J/cm2. Exposing SKmel-23 cells to 0.4 J/cm2 of 532 nm light results in a diminution of the number of melanosomes within cells as well as a marked decrease in melanin content, as determined by spectrophotometric assay and electron microscopy. Using the reverse transcriptase polymerase chain reaction technique, the reduction in melanin content of the cells was accompanied by a selective decrease in mRNA coding for tyrosinase, the first enzyme in the biosynthetic pathway for melanin. No decrease in the mRNA coding for the GAPDH protein was observed. Our finding has implications for understanding the control processes that regulate the melanin content of cells and suggests that the model described can be used to further investigate changes that may occur in cells as a result of their exposure to sub-lethal levels of laser light.

A model to estimate the dose to tumour following intracavity administration of radioimmunoconjugates to patients with malignant gliomas.

Patients who have relapsed following primary treatment for malignant glioma and have undergone further surgical debulking have been treated with an anti-human neural cell adhesion molecule (NCAM) MoAb linked to either Iodine-131 or Yttrium-90. These reagents are introduced into the tumour resection cavity via an Ommaya reservoir. Pharmacokinetic and imaging studies indicate that the radioimmunoconjugate remains within the cavity for a protracted period of time. In this manuscript we develop a dosimetric model to predict the dose delivered to the rim of tissue surrounding the resection cavity. The model takes into account variables such as the diameter of the cavity and the degree of antibody binding which is achieved. Whilst the calculated doses to the wall of the cavity are relatively inaccurate due to our inability to measure factors such as diffusion and heterogeneity in antibody uptake, the model illustrates the potential benefits and pitfalls that can result from targeting the two radionuclides. It is hoped that as increasing interest is shown in this type of "liquid brachytherapy" other groups will find it useful to apply the model to allow comparisons to be made between our targeting strategy and those developed by other individuals.

L1 adhesion molecule on human lymphocytes and monocytes: expression and involvement in binding to alpha v beta 3 integrin.

The L1 adhesion molecule is a member of the immunoglobulin (Ig) superfamily initially identified in the nervous system which contains six Ig-like domains. Besides the known L1-L1 homotypic interaction, L1 was recently shown to bind to very late antigen (VLA)-5 in the mouse and alpha v beta 3 in the human. The sixth Ig domain is critical for this function. We now demonstrate that human CD4+ peripheral blood T lymphocytes, monocytes and B lymphocytes, but not CD8+ T lymphocytes, express L1. When compared to the expression of CD31, another ligand for alpha v beta 3 on T lymphocytes, only a small proportion of cells were CD31+L1+ double positive. L1 was also detected on the surface of human monocytic and lymphoid tumor lines and was shown to have a molecular mass of approximately 220 kDa, similar to the molecule present on neuroblastoma cells. The function of the sixth Ig domain of human L1 as an integrin ligand was also investigated. Using an RGD-containing peptide derived from the sixth Ig domain as well as a fusion protein of the sixth Ig domain of L1 and the Fc portion of human IgG1 (6.L1-Fc), we demonstrated the binding of human MED-B1 (alpha v beta 3hi, alpha 5 beta 1lo) tumor cells and this binding was blocked by alpha v-specific mAb. In contrast, human Nalm-6 cells (alpha v beta 3lo, alpha 5 beta 1hi) did not bind to the 6.L1-Fc fusion protein. MED-B1 cells could also be stained with the 6.L1-Fc fusion protein. Our results suggest that human L1 binds predominantly to alpha v beta 3 and that its presence on leukocytes could be important for adhesion and migration.

Identification of neural crest derived cells within the cellular microenvironment of the human thymus employing a library of monoclonal antibodies raised against neuronal tissues.

The organization of optimal microenvironmental conditions within the developing thymus for lymphatic stem cell migration and their further maturation requires cellular and humoral participation of the neural crest. Recently, the immunophenotypical (IP) heterogeneity of lymphatic cells has become a scientific fact. Monoclonal antibodies (MoABs) produced against the various subpopulations of the reticulo-epithelial cells (RE) demonstrated their heterogeneity. We suggest that with a library of MoABs, raised against normal neuronal tissues, neural tumors, and a medulloblastoma cell line, including UJ13/A, UJ127.11, UJ167.11, UJ223.8, UJ308, J1153, A2B5, 215.D11, 275.G7, 282.1, antineurofilament (NF - med. m.w.), and anti-Thy-1 it is possible to recognize cells of neural crest origin within the postnatal thymic cellular microenvironment. Evidence has been collected concerning such connections between the nervous system and the thymus, such as the production of neuropeptides, oxytocin, and neurophysin by the thymus. Our immunohistochemical study was carried out on quick-frozen sections of human postnatal thymuses removed during open heart surgery, employing an indirect, alkaline phosphatase conjugated streptavidin-biotin technique. The employed MoABs reacted with the subcapsular (outer cortex) thymic nurse cells (TNCs) and with medullary RE cells, in close contact with already mature, immunocompetent T lymphocytes ready to leave the thymic microenvironment and enter the peripheral blood. The thymic medulla's strong immunoreactivity with A2B5, which binds to the GQ ganglioside, is typical for peptide secreting cells often migrated from the neural crest. A2B5+, Thy-1+ IP was demonstrated on the large TNCs. Cortical RE cells showed reactivity with UJ127.11, UJ223.8, and UJ308. Dense expression of neural crest antigens was detected in the Hassall's bodies (HBs) employing MoABs UJ223.8, UJ308, 215.D11, and 275.G7. These results suggest a neural crest origin for TNCs and for 20% to 30% of the cells of thymic microenvironment. The outer (peripheral) part of the HBs contained functionally very active RE cells. These RE cells also expressed antigens characteristic of the neural crest, detectable with MoABs UJ127.11, UJ223.8, UJ308, J1153, 215.D11, 275.G7 and A2B5.

The epidermal growth factor receptor in glioblastoma: genomic amplification, protein expression, and patient survival data in a therapeutic trial.

Human glioblastomas were evaluated for overexpression of the epidermal growth factor receptor (EGFR) in a therapeutic trial with the anti-EGFR antibody EMD 55900. A total of 55 cases were examined by immunohistochemistry using 4 different monoclonal antibodies on frozen or on paraffin sections: EGFR-1 (Amersham), E 62 (Merck), E 30 (Merck), and EMD 55900 (MAB 425, Merck). Definition for inclusion in clinical trials of EMD 55900 was an immunohistochemical overexpression of grade 4+ or 3+ in a scale of 4 grades of staining quality. The use of the 4 different antibodies gave essentially equal results. In 21 cases, the immunohistochemical results were supplemented by molecular genetic analysis of EGFR amplification on the corresponding locus of chromosome 7, using frozen tissue from the same blocks after screening for vital tumor areas. Since no other material was available, the differential polymerase chain reaction technique was applied. Interferon-gamma (IFN-gamma) served as a reference gene. In a preliminary experimental series with cases of known EGFR amplification, a densitometric EGFR/IFN-gamma ratio higher than 3 was determined as indicator for amplification of the EGFR gene. With this experimental approach we were able to identify an amplified EGFR gene in 13 specimens including 2 from recurrent glioblastomas in the same patients. All of these showed an increased immunoreactivity for EGFR protein. The degree of EGFR amplification (EGFR/IFN-gamma ratio as measured by DNA densitometry) showed a positive correlation with the grade of immunohistochemical protein expression, both in regard to the fraction of positive cells and to the overall staining intensity.(ABSTRACT TRUNCATED AT 250 WORDS)

A pilot study of the treatment of patients with recurrent malignant gliomas with intratumoral yttrium-90 radioimmunoconjugates.

A pilot study of the treatment of patients with relapsed malignant gliomas with direct intratumoral injections of yttrium-90 (90Y) radioimmunoconjugates has been completed. Patients were recruited following maximal tumour resection, and received 1-3 injections of 90Y conjugated to a monoclonal antibody designated ERIC-1, which binds the neural cell-adhesion molecule. Data were collected to establish clinical toxicity, pharmacokinetics and radiation doses to the cavity wall and critical body organs. Twenty-three injections were completed in 15 patients, with a mean injected activity of 675 MBq (range 399-921). Early toxicity manifested as cerebral oedema and was readily controlled with dexamethasone. Delayed myelosuppression was observed but no intervention was required. Pharmacokinetic analysis confirmed prolonged retention of isotope in the cavity with correspondingly low activity in the bloodstream. These data were translated into estimates of absorbed radiation dose using the Medical Internal Radiation Dosimetry (MIRD) scheme. Mean doses, and dose rates, to the wall of the cavity, i.e. 'tumour,' were very high in comparison to normal tissue doses, with a further advantage if targeting was achieved.

Pharmacokinetics and dose estimates following intrathecal administration of 131I-monoclonal antibodies for the treatment of central nervous system malignancies.

PURPOSE: Treatment of malignant disease in the central nervous system (CNS) with systemic radiolabeled monoclonal antibodies (MoAbs) is compromised by poor penetration into the cerebrospinal fluid (CSF), limited diffusion into solid tumors, and the generation of anti-mouse antibodies. To attempt to avoid these problems we have treated patients with diffuse neoplastic meningitis with radioimmunoconjugates injected directly into the intrathecal space. METHODS AND MATERIALS: Tumor-specific MoAbs were conjugated to Iodine-131 (131I) (629-3331 MBq) by the Iodogen technique, and administered via an intraventricular reservoir. A clinical response rate of approximately 33% was achieved, with better results in more radiosensitive tumors. Here, we present detailed pharmacodynamic data on patients receiving this intracompartmental targeted therapy. RESULTS: Elimination from the ventricular CSF appeared biphasic, with more rapid clearance occurring in the first 24 h. Radioimmunoconjugate entered the subarachnoid space and subsequently the vascular compartment. From this information, the areas under the effective activity curves for ventricular CSF, blood, and subarachnoid CSF were calculated to permit dosimetry. Critical organ doses were calculated using conventional medical internal radiation dose (MIRD) formalism. Where available, S-values were taken from standard tables. To calculate the doses to CSF, brain, and spinal cord, S-values were evaluated using the models described in the text. CONCLUSION: A marked advantage could be demonstrated for the dose delivered to tumor cells within the CSF as compared to other neural elements.

The potential of targeted radiotherapy in the treatment of central nervous system leukaemia.

We describe the results of clinical studies investigating the role of monoclonal antibody (MoAb) targeted radiotherapy in the treatment of central nervous system (CNS) leukaemia. Seven children, aged 3-16 years, in second or subsequent meningeal relapse of acute lymphoblastic leukaemia (ALL), have been treated. Each patient received a single injection into the cerebrospinal fluid (CSF) of between 629 and 1,702 MBq of 131-Iodine (131I) conjugated to MoAb HD37 (CD19, n = 2), WCMH 15.14 (CD10, n = 4) or both antibodies (n = 1). One patient underwent a course of repeated targeted therapy following his initial treatment. Acute toxicity was manifest in five patients by a transient aseptic meningitis. Myelosuppression was observed in four children. Pharmacokinetic studies investigated whole body, blood and CSF clearance of radioisotope. Progressively more rapid systemic clearance of 131I was noted in the patient receiving repeated therapy, indicating the development of the human anti-mouse Ig (HAMA) response. Dosimetric studies revealed a radiation dose to the red bone marrow of between 0.6 and 2.2 Gy. The dose to the subarachnoid CSF was between 12.2 and 25.3 Gy, over six times higher than that to the surface tissue of the brain and spinal cord and between 40 and 140 times higher than that to the whole brain. In all but one patient, a transient complete response, in terms of disappearance of lymphoblasts from the CSF, was observed. These studies demonstrate the feasibility of targeted radiotherapy in CNS ALL.(ABSTRACT TRUNCATED AT 250 WORDS)

Immunomagnetic colloids for the enrichment of tumor cells from peripheral blood and bone marrow: a model system.

The characterization of tumor cells in bone marrow harvested for autologous bone marrow rescue remains a problem due to the limited sensitivities of the techniques available for their analysis. This complicates the use of purging techniques as their usefulness is debated, specifically because it is not known if they can remove all residual tumor cells from either peripheral blood stem cell harvests or bone marrow. Apart from developing more sensitive techniques for tumor detection, one way of increasing our efficiency at identifying rare malignant cells in normal hematopoietic cells is to develop approaches to enrich the population of interest prior to analysis. Here, we describe a laboratory-based system for tumour enrichment employing panels of monoclonal antibodies (MAbs) and an immunomagnetic colloid. In model systems, tumor cells could be enriched approximately 100-fold with high yield and purity. The adaptations of this technology to permit further cell enrichment and the choice of either an immunological technique or a molecular approach to tumor identification are discussed in detail.

Tissue-specific expression of neural cell adhesion molecule (NCAM) may allow differential diagnosis of neuroblastoma from embryonal rhabdomyosarcoma.

The MSD1 region of neural cell adhesion molecule (NCAM) was originally described as being spliced into the 120-kDa isoform of NCAM isolated from muscle. The 105 bp region is inserted between exons 12 and 13 and actually consists of three separate exons, MSD1a, MSD1b and MSD1c of 15, 48, 42 bp, respectively. In addition, a further exons consisting of a single triplet has been designated MSD1d, making the full insert size 108 bp. As the MSD1 region was originally described as being selectively expressed in muscle tissue, we have investigated whether it is also present on tumours of rhabdoid origins and whether its presence can be used as the diagnostic marker to distinguish other small round cell tumours of childhood, such as neuroblastoma. Using a variety of human tumour cell lines, we demonstrated the presence of the MSD1 region on all rhabdomyosarcomas investigated. However, neuroblastoma cell lines only expressed subcompartments of the MSD1 region. The MSD1c exon was not spliced into the NCAM molecules isolated from any of the neuroblastoma cell lines investigated. On the basis of this finding, it appears that neuroblastoma and rhabdomyosarcoma can be distinguished by the expression of MSD1c mini-exon. Further studies are underway to attempt to define a monoclonal antibody that recognises the region, using mice immunised with synthetic peptides, and to confirm the finding using fresh biopsy material.