International consensus on the initial diagnostic workup of cancer of unknown primary

BACKGROUND

Although the incidence of Cancer of Unknown Primary (CUP) is estimated to be 1-2 % of all cancers worldwide, no international standards for diagnostic workup are yet established. Such an international guideline would facilitate international comparison, provide adequate incidence and survival rates, and ultimately improve care of patients with CUP.

METHODS

Participants for a four round modified Delphi study were selected via a CUP literature search in PubMed and an international network of cancer researchers. A total of 90 CUP experts were invited, and 34 experts from 15 countries over four continents completed all Delphi survey rounds.

FINDINGS

The Delphi procedure resulted in a multi-layer CUP classification for the diagnostic workup. Initial diagnostic workup should at least consist of history and physical examination, full blood count, analysis of serum markers, a biopsy of the most accessible lesion, a CT scan of chest/abdomen/pelvis, and immunohistochemical testing. Additionally, the expert panel agreed on the need of an ideal diagnostic lead time for CUP patients. There was no full consensus on the place in diagnostic workup of symptom-guided MRI or ultrasound, a PET/CT scan, targeted gene panels, immunohistochemical markers, and whole genome sequencing.

INTERPRETATION

Consensus was reached on the contents of the first diagnostic layer of a multi-layer CUP classification. This is a first step towards full consensus on CUP diagnostics, that should also include supplementary and advanced diagnostics.

HGG-59. Pediatric high-grade gliomas and the WHO classification on CNS Tumors - Different perspectives of pediatric neuro-oncologists and neuropathologists in the light of recent updates

BACKGROUND: The WHO Classification of Tumors of the Central Nervous System has undergone major restructuring following rapid advances in brain tumor genomics and epigenomics. The most significant changes resulted from the introduction of molecularly defined diagnostic criteria in 2016 (revised 4th edition). In 2021 (5th edition), further essential molecular criteria were incorporated. In the present study, we sought to investigate potential differences between specialists in perception of these newly defined molecular subtypes of pediatric high-grade gliomas (pedHGG). METHODS: We designed a 22-question survey studying the impact of the revised 4th edition of the WHO classification on pedHGG. Data were collected and statistically analyzed to capture the spectrum of viewpoints and possible differences among neuro-oncologists and neuropathologists. RESULTS: 465 participants from 53 countries responded, of which 187 pediatric neuro-oncologists (40%), 160 neuropathologists (34%) and 118 experts in other related fields (neurosurgeons, radiotherapists, neuroradiologists and others; 26%). Neuro-oncologists reported having issues with the introduction of new molecular entities, such as the abolishment and renaming of established tumor entities. Neuropathologists did not define these problems to the same extent. However, both groups felt that in the 2016 version, less relevant or insufficient diagnostic definitions were available for pedHGG. Within the 2021 WHO classification, a substantial improvement was perceived regarding the definition of pedHGG entities. However, some issues of high clinical relevance, like the definition of clinical phenotypes such as diffuse intrinsic pontine glioma (DIPG) and gliomatosis cerebri, are yet to be addressed. CONCLUSIONS: Within the WHO classification of pediatric brain tumors, such as high-grade gliomas, rapid changes in nomenclature have been introduced because of substantial improvement in molecular characterization. This study highlights that ongoing cross-talk between advancing classification of pedHGG subtypes and its biological relevance and clinical impact is essential.

DIPG-51. Hydrocephalus Treatment and the Effect on Survival in Diffuse Intrinsic Pontine Glioma

BACKGROUND: Diffuse intrinsic pontine glioma (DIPG), can cause hydrocephalus and if symptomatic, leads to rapid changes in consciousness requiring surgical intervention. The effect of cerebrospinal fluid (CSF) diversion on overall survival and the clinical factors influencing outcome remain unclear. The aim of this study was to evaluate the impact of the treatment of hydrocephalus on survival in DIPG patients. METHODS: The study was retrospective in design using data from the SIOPE-European Society of Pediatric Oncology DIPG Registry. Hydrocephalus was determined based on a centrally reviewed diagnostic MRI. The Kaplan-Meier method was used for survival statistics. Clinical prognostic factors including: duration of symptoms, age and cranial nerve palsy at diagnosis were evaluated for confounding and effect modification. The effect of hydrocephalus treatment (CSF diversion) on survival was examined using Cox regression. RESULTS: Among 582 patients from the SIOPE-DIPG Registry, 86 (14%) had hydrocephalus at diagnosis. Median OS for hydrocephalus patients treated with CSF diversion (n=43) was 13 months (95% CI, 10.2-17.7) and 9 months (95% CI, 7.4-10.6) for hydrocephalus patients without a CSF diversion (n=43). Survival rates were not significantly different (p=.41). On adjusted Cox regression, correcting for duration of symptoms, hydrocephalus patients with signs of cranial nerve palsy at diagnosis and a CSF diversion had a hazard ratio 0.476 (p=0.004). CONCLUSION: Survival among DIPG patients presenting with hydrocephalus at diagnosis was not influenced by CSF diversion. Hydrocephalus patients with signs of cranial nerve palsy at diagnosis, had a significantly reduced risk after undergoing CSF diversion. There is an indication this subgroup of DIPG patients may benefit more from CSF diversion, although the relationship between hydrocephalus and cranial nerve palsy requires further investigation.

DIPG-77. TREATMENT EXTENT AND THE EFFECT ON SURVIVAL IN DIFFUSE INTRINSIC PONTINE GLIOMA

BACKGROUND

Front line radiotherapy for diffuse intrinsic pontine glioma (DIPG) remains the only standard of care. Is this still appropriate?

PATIENTS AND METHODS

We examined survival outcomes across six treatment modalities including I) no treatment (n=19), II) radiotherapy alone (n=38), III) radio-chemotherapy (n=101), IV) radiotherapy and relapse chemotherapy (n=35), V) radio-chemotherapy and relapse chemotherapy (n=163), and VI) radio-chemotherapy and relapse chemotherapy, plus reirradiation (n=54). Data were collected retrospectively using the Society of Pediatric Oncology and Hematology (GPOH) and the SIOPE DIPG Registry. 410 patients were included with radiologically centrally reviewed DIPG, mostly unbiopsied. Of note, the untreated patients and radiotherapy only cohorts chose limited treatment voluntarily.

RESULTS

Median overall survival (MOS) of the whole cohort was 11 months and progression free survival (PFS) 7 months. PFS was not significantly different between the treatment groups. OS and post-progression survival (PPS) were significantly different between cohorts. For the respective treatment groups, median OS was 3 months (I), 7 months (II), 8 months (III), 13 months (IV), 13 months (V), and 15 months (VI). For only front line vs at least one second line therapy, MOS was 8 months vs 14 months and PPS 2 months vs 5 months.

CONCLUSIONS

Although subject to biases to some extent, it seems that additional therapies beyond radiation therapy are of benefit to extending survival in DIPG patients. This is at least partially caused by the introduction of reirradiation regimens. To what extent other therapies contribute to survival and quality of life is subject to further investigation.

DIPG-84. COMPLEMENTARY AND ALTERNATIVE MEDICINE IN DIFFUSE INTRINSIC PONTINE GLIOMA

INTRODUCTION

Diffuse intrinsic pontine glioma (DIPG) is a rare and aggressive childhood brainstem malignancy with a two-year survival rate of ≤10%. In this international survey study we aim to evaluate the use of complementary and alternative medicine (CAM) in this patient population.

METHODS

Parents of-, and physicians treating DIPG patients were asked to participate in a retrospective online survey with questions regarding CAM use during time of illness.

RESULTS

120 parents and 75 physicians contributed to the online survey between January and May 2020. Physicians estimated that <50% of their patients used CAM, whereas 69% of the parents reported to have used CAM to treat their child during time of illness. Cannabis was the most widely used form of CAM, followed by vitamins and minerals, melatonin, curcumin and boswellic acid. CAM was mainly used to actively treat the tumor. Other motivations were to treat side effects of chemotherapy, or to comfort the child. Children diagnosed ≥2016 were more likely to use CAM (χ2=6.08, p=0.014). No significant difference was found between CAM users and non-users based on ethnicity (χ2=4.18, p=0.382) and country of residence (χ2=9.37, p=0.154). Almost 50% of the physicians do not frequently ask their patients about possible CAM use.

CONCLUSION

This survey demonstrates that worldwide a considerable number of DIPG patients use CAM. Physicians should be more aware of potential CAM use and more actively discuss the topic. More research is needed to gain knowledge about possible anticancer effects of CAM and their interactions with conventional therapies.

PATH-04. THE BLOOD-BRAIN BARRIER IN DIFFUSE MIDLINE GLIOMA AND ITS IMPLICATIONS FOR DRUG DELIVERY

INTRODUCTION

Chemotherapy has been unsuccessful for pediatric diffuse midline glioma (DMG) most likely due to an intact blood-brain barrier (BBB). However, the BBB has not been characterized in DMG and therefore its implications for drug delivery are unknown. In this study we characterized the BBB in DMG patients and compared this to healthy controls.

METHODS

End-stage DMG pontine samples (n=5) were obtained from the VUmc diffuse intrinsic pontine glioma (DIPG) autopsy study and age-matched healthy pontine samples (n=22) were obtained from the NIH NeuroBioBank. Tissues were stained for BBB markers claudin-5, zonula occludens-1, laminin, and PDGFRβ. Claudin-5 stains were used to determine vascular density and diameter.

RESULTS

In DMG, expression of claudin-5 was reduced and dislocated to the abluminal side of endothelial cells. In addition, the expression of zonula occludens-1 was reduced. The basement membrane protein laminin expression was reduced at the glia limitans in both pre-existent vessels and neovascular proliferation. PDGFRβ expression was not observed in DMG but was present in healthy pons. Furthermore, the number of blood vessels in DMG was significantly (P< 0.01) reduced (13.9 ± 11.8/mm2) compared to healthy pons (26.3 ± 14.2/mm2). Markedly, the number of small blood vessels (< 10µm) was significantly lower (P< 0.01) while larger blood vessels (> 10µm) were not significantly different (P= 0.223). The mean vascular diameter was larger for DMG 9.3 ± 9.9µm compared to 7.7 ± 9.0µm for healthy pons (P= 0.016).

CONCLUSION

Both the BBB and the vasculature are altered at end-stage DMG. The reduced vascular density might have implications for several drug delivery methods such as focused ultrasound and convection enhanced delivery that are being explored for the treatment of DMG. The functional effects of the structurally altered BBB remain unknown and further research is needed to evaluate the BBB integrity at end-stage DMG

New trial of convection enhanced drug delivery (CED) in DIPG- applying the SIOPe DIPG survival prediction model for power calculation

Background

In diffuse intrinsic pontine glioma (DIPG) drug resistance is in part due to inadequate penetration of blood-brain barrier (BBB) by systemically administered drugs. Convection-enhanced drug delivery (CED) techniques have been established to bypass BBB. Trial design to measure efficacy requires evidence to justify power calculation.

Aims

To apply SIOPe DIPG registry survival prediction tool to pilot cohort of children with DIPG treated with CED of carboplatin and sodium valproate.

Methods

Case note and imaging review of 9 children with typical DIPG on imaging who were treated on compassionate basis with CED intra-tumoural infusions of carboplatin (0.12 mg/ml) and sodium valproate (14.4 mg/ml), after radiotherapy (n=9) and chemotherapy (n=4). Each had Renishaw device placed with 4 micro-catheters located within tumour mass. Up to 8 treatment cycles of CED infusions delivered through 2 pairs of catheters on 2 days to encompass pontine tumour volume. Survival prediction was performed using clinical criteria: age, sex, duration of symptoms, prior chemotherapy; and radiological criteria: absence of distant metastases; disease involving more than 50% of, and confined to, pons, ring enhancement at diagnosis.

Results

Cases were categorized as intermediate or high-risk using SIOPe risk scoring with predicted median overall survival (OS) of 9.7, and 7.0 months, respectively. Four patients, categorized as high-risk, had median overall survival (OS) of 14.2 months. Five children, categorized as standard-risk, had median OS of 16.0 months.

Conclusions

Results justify a phase 2 trial of CED carboplatin and sodium valproate powered to detect at least 4-month prolongation of survival.

The international diffuse intrinsic pontine glioma registry: an infrastructure to accelerate collaborative research for an orphan disease

Diffuse intrinsic pontine glioma (DIPG), a rare, often fatal childhood brain tumor, remains a major therapeutic challenge. In 2012, investigators, funded by the DIPG Collaborative (a philanthropic partnership among 29 private foundations), launched the International DIPG Registry (IDIPGR) to advance understanding of DIPG. Comprised of comprehensive deidentified but linked clinical, imaging, histopathological, and genomic repositories, the IDIPGR uses standardized case report forms for uniform data collection; serial imaging and histopathology are centrally reviewed by IDIPGR neuro-radiologists and neuro-pathologists, respectively. Tissue and genomic data, and cell cultures derived from autopsies coordinated by the IDIPGR are available to investigators for studies approved by the Scientific Advisory Committee. From April 2012 to December 2016, 670 patients diagnosed with DIPG have been enrolled from 55 participating institutions in the US, Canada, Australia and New Zealand. The radiology repository contains 3558 studies from 448 patients. The pathology repository contains tissue on 81 patients with another 98 samples available for submission. Fresh DIPG tissue from seven autopsies has been sent to investigators to develop primary cell cultures. The bioinformatics repository contains next-generation sequencing data on 66 tumors. Nine projects using data/tissue from the IDIPGR by 13 principle investigators from around the world are now underway. The IDIPGR, a successful alliance among philanthropic agencies and investigators, has developed and maintained a highly collaborative, hypothesis-driven research infrastructure for interdisciplinary and translational projects in DIPG to improve diagnosis, response assessment, treatment and outcome for patients.