The current study analyzed the disease course of 63 patients that had any surgical treatment as part of pediatric chiasmatic hypothalamic glioma management. We observed that surgical therapy was an integral part of management to achieve safe tumor control. During a mean follow up period of 108 ± 72months, a mean number of 1.85 ± 1 tumor-surgeries per patient and 1.1 ± 1 medical anti-tumoral therapies per patient were applied while 14.3% of patients received radiation therapy. In addition, 42.9% of patients needed a VP shunt with a mean number of 2.1 ± 1.8 shunt surgeries and 43.1% of patients had a mean number of 1.9 ± 1.8 isolated cyst surgeries during follow up. These multiple and diverse steps of treatment underscore the importance of a multidisciplinary team approach when following patients and making individual decisions for if an intervention is necessary and if so which is the best option. Communication with patients and parents needs to stress right from the very beginning that a CHG diagnosis represents a chronic disease most likely necessitating multiple steps of treatment in years to come.
The HIT-1996 study report focused on management of in LGG patients included the CHG cohort in [5] and did not focus on surgical therapy since the grade of resection was exclusively mentioned alongside with other patient’s characteristics such as age, sex, NF status and tumor localization. Nevertheless, 72.2% of the patient received some kind of tumor surgery and almost half (33.8%) received biopsy alone during a median follow up time of 42.1 months. In five patients of that cohort a complete resection was documented while in another five patients a second tumor surgery was mentioned. The NF-1 ratio in our cohort was relatively small (12.7%) compared to the HIT 1996 study cohort (27.3%) since children were only included when surgical therapy was applied during the course of treatment.
The surgical approach of the HIT 1996 and HIT SIOP 2004 study was quiet different to our recent strategy proposal [3, 5, 18]. Firstly, complete resection would never be the goal of surgery in CHG, secondly repeated surgeries should be more liberally suggested as possible treatment strategies in balanced decisions towards other non-surgical treatment options. A similar approach was recently described technically by Kim et al., however without offering outcome data to their cohort [11]. This is in line with Goodden et al., who proposed to reintegrate surgical strategies into the multimodal management of CHG [6]. In their cohort, 21 patients were treated surgically of whom 10 received surgery alone and survived with SD and 11 patients received surgery together with chemotherapy of whom 10 remained in SD status during a median follow up time of 77 months. One patient received an additional surgery. This cohort presents a smaller group of patients and a shorter follow up time. Nevertheless, surgery was confirmed a relevant part of therapy in CHG. More modern surgical techniques such es pure endoscopic debulking techniques as also utilized in our cohort may further reduce the invasiveness in future approaches [15].
Repeated surgery in pLGG has been recently reported in the huge cohort of patients from the SIOP LGG 2004 study [18]. From 1271 surgically treated patients, 25.5% received at least a second surgery, while this proportion was higher in supratentorial midline tumor location (29.8%) that included CHG. This was relevant for the relative number of possible subsequent surgeries (up to 6 surgeries/patient) with a median follow up time of 9.2 years. Thus, the acceptance of repeated but well calculated surgeries in the long-term management has increased.
Surgical complications in supratentorial midline low grade glioma were more often seen after resection surgeries when compared to biopsy interventions [23]. In resection surgery 41.2% experienced any post-operative neurological deficit in comparison to 15.3% after biopsy. Infection rate was relatively low (5.4 vs. 2.5%) and hemorrhages were seen in 15 vs. 4%, but none of the patients needed additional surgery. Seizures were seen in 8.1 vs. 2.5% which may be linked to the direct approach related brain irritation or more likely to electrolyte concentration shifts due to pituitary hypothalamic manipulation. No data are available on long-term antiseizure medication. It was striking in this study, that complication rates were lower in big volume centers when compared to high volume centers, emphasizing the relevance of pediatric neurosurgical experience.
None of the study so far included possible surgical interventions for tumor cysts or total number of surgeries for CSF circulation disturbance. Goodden et al. reported on a shunt rate of 52.4%, which is higher compared to our cohort with a shunt rate of 43%. Among our cohort the shunt rate also differed between the two centers (Berlin: 39.5% vs Tübingen: 50%). This may be explained by the amount of endoscopic interventions (3 stented ETV) being performed in Berlin using a stent placement for establishing the communication between internal and external CSF spaces [14]. However, in the Tübingen cohort a significant part of children came to surgery from external units already with a shunt in place. On the other hand, the indication for tumor debulking inside the third ventricle to free the CSF spaces was one of the goals for surgical interventions to specifically avoid the implantation of a VP shunt.
Non-surgical treatment is key in LGG patients who experience continuous progressions and are mostly incurable by surgery which is often the case in CHG. In the HIT 1996 study 63.5% of CHG were treated with chemotherapy and 16.6% with radiation [3]. In a more detailed analysis 123 of 198 patients received vincristine/ carboplatin treatment with a response rate of 92%. After a median follow up time of 44.1months, 44 patients still suffered from progression with the need for additional therapy [5]. That portion was generally similar in the SIOP LGG 2004 study with a portion of 70.9% (234/330 optic pathway glioma) receiving non-surgical treatment, of these 44.5% received one non-surgical treatment, 11.8% received two and 14.5% received three or more non-surgical treatment lines, each line yielding a lower PFS when compared to the first chemotherapy approach [9]. Systemic treatment in LGG patients is usually for 18 months which is a significant burden for patients and families. Likewise, radiation therapy is complicated by high rate of neurocognitive impairment. Functional deficits are also observed after surgery. Risks and efficacy of each intervention should be wisely balanced to avoid long term neurological sequalae. This is even more so in patients who need multiple treatment lines as each therapy has its own risks and benefits.
In recent years targeted therapies for pLGG have become available [2, 7, 8] which are beyond the focus of this paper Certainly, efficacy reports on MAPK inhibitors open new perspectives of tumor volume reduction [10, 20]. However, treatment related adverse events were reported in a relevant portion of patients (42% ≥grade 3) and 18.2% discontinued treatment due to adverse events or any kind of progression. Yet, it is a matter of debate, whether these therapies need to be given continuously. This adds another reason why it is important that the MDT decisions include all disciplines to decide about indication of novel drug treatment options. The number of targeted therapies applied in the current cohort was limited and future experience will define the role of each therapy in the framework of CHG management.
In terms of outcome measures we were able to show that three quarters of patients reached SD and 84% of patients were independent measured by the Karnofsky performance score. Visual acuity was preserved in 58% of patients with complete amaurosis in 16.2% of patients. Hormone substitution of any kind was seen in 30% and 40% of children were overweight/obese. These outcomes results underline the challenge of the chronic condition of CHG on the one hand, but also open a perspective that these children participate in age adequate daily activities. Our data are difficult to compare since most studies do not give details on outcome measures, difficult to retrieve after long periods of follow up or all patients are included also those who were not treated surgically [1, 12, 13]. Our results seem comparable to the Liverpool cohort, in which 39% of patients showed any kind of hypothalamic pituitary dysfunction and 74% were reported to have stable vision.
Limitations
This study is limited by a first glance relatively low number of 63 cases and its retrospective nature. On the other hand the two center cohort offer a decent surgical experience for patients with CHG with a long follow-up and will add to the existing literature on the topic. More accurate objective measurements on surgical resection such as volume changes after debulking and more detailed imaging data analysis postoperatively would be preferable. Similarly, outcome in future studies will be more detailed. Instead of reporting detailed surgical complications we therefore focused in this report on long term outcome which presents the combined long-term consequences of the natural course of the disease and all treatments applied. Furthermore, we were able to give detailed information on tumor surgical interventions and also on hydrocephalus and cyst management, which are relevant cofounders to disease and outcome in this group of patients.