Abstract
Background/Aim: Craniopharyngiomas involving the ventricular system are rare but pose significant surgical challenges. We systematically reviewed the literature on craniopharyngiomas invading the ventricles (CP-V). Materials and Methods: PubMed, EMBASE, Scopus, Web of Science, and Cochrane were searched to include studies reporting clinical data of patients with CP-Vs. Clinico-radiological features, management, and treatment outcomes were analyzed. Results: We included 73 studies encompassing 407 patients. Patients were mostly male (61.5%), presenting with headache (57.9%) and/or endocrine disorders (52.1%). CP-Vs mostly involved the third ventricle (96.3%), followed by the lateral ventricles (2.9%), and the fourth ventricle (1%). Tumors had cystic components in 59% of cases and were mostly adamantinomatous (70.8%). Open resection was performed in 232 cases (57%), mostly with trans-lamina terminalis (36.6%) and trans-callosal (31.9%) approaches. Endoscopic resection was performed in 169 cases (41.5%), mostly with trans-sphenoidal (74.6%) and transventricular (24.9%) approaches. Gross-total tumor resection was obtained in most cases (62.9%). Adjuvant radiotherapy was delivered in 22.8% cases. A total of 178 patients experienced persistent complications, mostly including diabetes insipidus (47.1%) and panhypopituitarism (12.7%), not significantly different after open versus endoscopic resection (p=0.117). Symptom improvement was obtained in 88% of cases. CP-Vs recurrences were reported in 94 patients (23.1%), with median progression-free survival of 13.5 months (range=0.5-252.0 months). Fifty-nine patients died (14.5%), with median overall survival of 32.0 months (range=0.5-252.0 months), significantly longer after endoscopic resection than open resection (p=0.019). Conclusion: CP-Vs are uncommon and challenging entities. Surgical resection is feasible, but patient-tailored selection of open/endoscopic approaches is necessary to achieve optimal outcomes and minimize complication risks.
Craniopharyngiomas comprise a heterogeneous group of benign neoplasms accounting for 1.2-4.6% of all intracranial tumors and showing two peaks of age incidence, in children or young adolescents (5-15 years) and in middle-age adults (45-60 years) (1, 2). Craniopharyngiomas are presumed to originate from the cell remnants of the craniopharyngeal duct (i.e., Rathke’s pouch) within the pars tuberalis, which migrate upward during hypophysis-genesis into close contact with the infundibulum-tuber cinereum (3, 4). This embryological origin explains the location of craniopharyngiomas within the sellar/suprasellar regions, with classical clinical presentation characterized by endocrine and/or visual impairments (5). Brain magnetic resonance imaging (MRI) assists the pre-operative differential diagnosis, and current treatment strategies primarily include surgical tumor resection, with or without ventricular shunting, and/or upfront or adjuvant radiotherapy (6).
Due to their origin along the hypothalamic-pituitary axis, craniopharyngiomas may involve the ventricular system (7). In particular, infundibulo-tuberal craniopharyngiomas comprise tumors originating at the floor of the third ventricle (8), which may extend into the ventricular cavity after breaking through the ependyma, while the “strictly third-ventricular craniopharyngiomas” define subependymal tumors growing exclusively within the third ventricle (9). In rarer cases, craniopharyngiomas may invade the lateral ventricles (10) and/or occur at the cerebellopontine angle with involvement of the fourth ventricle (11). The surgical and/or radiation management of these tumors appears particularly difficult due to their close relationship with the hypothalamic system and critical neurovascular structures, with significant risks of severe post-treatment complications and major impairment of patient’s quality of life (4).
Recently, Prieto et al. (9) systematically reviewed studies reporting strictly third-ventricular craniopharyngiomas, but they excluded craniopharyngiomas secondarily invading the ventricular system, which characterize challenging entities whose management remains unclear. In this systematic review, we comprehensively summarized the literature on clinico-radiological presentation, treatment strategies, and outcomes in all patients with craniopharyngiomas involving the ventricular system (CP-Vs).
Materials and Methods
Literature search. A systematic review was performed upon the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (12). PubMed, EMBASE, Scopus, Web of Science, and Cochrane were searched from inception to May 25, 2022, using the combination of the Boolean operators “OR” and “AND” and the search terms: “ventricle*”, “intraventricular”, and “craniopharyngioma*”. Studies were exported to Mendeley, and duplicates were removed.
Study selection. Inclusion and exclusion criteria were defined a priori. Studies were included if they: 1) involved ≥1 patients with histologically confirmed craniopharyngiomas involving the ventricular system, as reported by the authors; 2) reported data on clinical characteristics, treatments, outcomes; 3) were written in English. CP-Vs were defined as craniopharyngiomas developing subependimally and exclusively within the ventricular system (i.e., strictly intraventricular craniopharyngiomas) and/or craniopharyngiomas originating below the floor of the third ventricle (i.e., infundibulo-tuberal) and extending into the ventricular system. Studies were excluded if they were: 1) literature reviews, autopsy reports, conference abstracts, videos, editorials, or books; 2) studies with unclear distinction between patients with CP-Vs and patients with different tumors or with craniopharyngiomas not involving the ventricles.
Two reviewers (P.P. and C.O.) independently screened titles and abstracts of all collected studies, and then assessed full texts of articles that met inclusion criteria. A third reviewer (A.S.H.) settled any disagreements at both stages of screening. Eligible studies were included based on the predefined criteria and references were screened to retrieve additional relevant articles.
Data extraction. Data were extracted by three reviewers (K.Y., M.O., and O.Y.) and then confirmed by one additional reviewer (P.P.). Missing data were not reported by the authors. Extracted data included: authors, year, sample size, age, sex, presenting symptoms, ventricle involved, radiological appearance, histological features, surgical approach, extent of surgery, adjuvant radiotherapy, post-treatment complications, clinical improvement, recurrence, progression-free survival (PFS), overall survival (OS), survival status. Extent-of-surgical resection was defined as gross total resection (GTR) for 100% tumor resection, subtotal resection (STR) for 80-99% resection, and partial resection (PR) for <80% resection. Complications were divided into “transient”, if self-resolving or treated with only medical therapy, and “persistent”, if untreatable or requiring re-operation. Clinical improvement was assessed at last available follow-up and was defined as improvement in symptoms (e.g., vision disturbances and/or endocrine deficits) reported at baseline pre-treatment evaluation.
Data synthesis and quality assessment. The primary outcomes of interest were the clinico-radiological features, management, and post-treatment outcomes of patients with CP-Vs. For each article, level of evidence was appraised in accordance with the 2011 Oxford Centre For Evidence-Based Medicine guidelines (13). A study-level meta-analysis was precluded because all included studies had levels IV of evidence and hazard ratios could not be deducted. However, individual patient data were extracted for individual patient data meta-analyses (14). Risk of bias of each study was independently evaluated by two reviewers (P.P. and C.O.) using the JBI checklist (15). The overall risk of bias for this review was determined by considering the risk of bias of all included studies in aggregate.
Statistical analysis. SPSS V.25 (IBM Corp, Armonk, NY, USA) was used for all statistical analyses and a two-tailed p-value <0.05 was considered significant. Continuous variables are reported as medians and ranges, while categorical variables as frequencies and percentages. The Chi-square test was used to compare complication rates after open versus endoscopic tumor resection. The time intervals between surgery and recurrence (PFS curve) or death (OS curve) were estimated using the Kaplan–Meier method, and the survival analyses were conducted using the log-rank test.
Results
Study selection and overview. The initial search yielded 2,409 citations (PubMed: 640; EMBASE: 939; Scopus: 595; Web of Science: 233; Cochrane: 2) (Supplementary File 1). A total of 30 case series and 43 case reports were finally included, categorized as level IV and V of evidence (Supplementary File 2). Critical assessment returned low risk of bias for all included articles (Supplementary File 3), predisposing this review to a low overall risk of bias.
Demographics and clinico-radiological characteristics. Table I summarizes the demographics and clinico-radiological features of all 407 included patients. Patients were mostly male (61.4%) with a median age of 45 years (range=1.5-84 years). Patients most commonly presented with headache (57.9%), endocrine disorders (52.1%), hydrocephalus (46.9%), and visual acuity impairment (46.4%), with 158 patients experiencing intracranial hypertension (40.5%). Xu et al. (16) reported 3 asymptomatic patients who were diagnosed incidentally with CP-Vs at imaging follow-ups. All tumors were suprasellar and mostly involving the third ventricle (96.3%), while 12 tumors (2.9%) involved the lateral ventricle and 4 (1%) involved the fourth ventricle. At imaging, 59% of tumors showed cystic components and 13.5% were partly calcified. Among patients with available histopathological data, 136 tumors were adamantinomatous (70.8%), and 56 were papillary (29.2%).
Treatment strategies. Management strategies are reported in Table II. Patients mostly underwent open tumor resection (57%), followed by endoscopic resection (41.5%), or combined open and endoscopic resection (1.5%). Among patients undergoing open tumor resection, the trans-lamina terminalis approach was the most used (36.6%), followed by trans-ventricular (31.9%), trans-callosal (24.6%), and trans-cortical (5.2%) approaches. Posterior fossa approaches were performed in 4 patients (1.7%) with CP-Vs involving the fourth ventricle (17-20). Among patients undergoing endoscopic tumor resection, the trans-sphenoidal approach was the most used (74.6%), followed by trans-ventricular (24.9%), and trans-callosal (0.5%) approaches. As regards the extent of tumor resection, GTR, STR, and PR were performed in 62.9%, 16.2%, and 20.4% patients, respectively. Post-surgical adjuvant radiotherapy was delivered in 80 patients (22.8%). Sharma et al. (21) and Davies et al. (22) reported 2 patients undergoing tumor biopsy with aspiration of the cystic component.
Outcomes, complications, and survival. Table II summarizes outcomes and complications. Patients were followed-up for a median of 32.0 months (range=0.5-252.0 months). Post-treatment symptom improvement occurred in 88% cases. Transient complications were described in 27 patients (15.8%), including new self-resolving diabetes insipidus (6.9%) and mutism (2.9%), with no significant differences after open vs. endoscopic tumor resection (p=0.511). Persistent complications were reported in 178 patients (47.1%), mostly new untreatable diabetes insipidus (27%) and hypopituitarism (12.7%), with no significant differences after open vs. endoscopic tumor resection (p=0.117). A total of 23 patients (6.1%) experienced cerebrospinal fluid (CSF) leak requiring surgical revision. CP-Vs recurrences were reported in 94 patients (23.1%), with a median PFS of 13.5 months (range=0.5-252.0 months). Most patients were alive at last follow-up (85.5%), with median OS of 32.0 months (range=0.5-252.0 months). The log-rank tests showed that OS was significantly higher only in patients undergoing endoscopic tumor resection compared to patients undergoing open tumor resection (p=0.019) (Figure 1).
Discussion
CP-Vs are uncommon entities characterized by debilitating clinical presentation, complex surgical management, and high risks of negative impact on long-term patient’s functional status. Current treatment strategies primarily focus on relieving life-threatening symptoms, especially in patients with intracranial hypertension and/or severe endocrine/vision impairments, but related complications should be considered on a case-by-case basis. In this review, we present a comprehensive summary of the current literature on CP-Vs, further analyzed within the context of the most common craniopharyngiomas not involving the ventricular system.
CP-Vs comprise a broad group of tumors with variable degrees of involvement of the ventricular system. Although some authors have tried to define distinct categories of intraventricular craniopharyngiomas, mostly referring only to the third ventricle, a standardized and universally accepted classification is currently lacking (7-9). These tumors are largely heterogeneous and often present with few overlapping imaging features, posing many difficulties in accurately categorizing each individual case on routine clinical settings (23). By analyzing all these lesions into a single category of craniopharyngiomas extending into the ventricular system, we aim to inform specialists about the most common clinical presentations and available therapeutic options with related outcomes, which may provide valuable information for optimizing patient-tailored treatment planning.
The origin of CP-Vs remains unclear. The involvement of the third ventricle may derive from craniopharyngioma’s development along the migration of the Rathke’s pouch close to the infundibulum and attached to the third ventricle floor, or from residual squamous epithelium of the stomodeum embedded within the ventricular system at early stages of pituitary gland development (24). Aggressive growth may also lead to tumor’s invasion into the lateral ventricles through the Monro’s foramina, with only 12 cases reported in the current literature (25-29). A total of 4 cases of fourth ventricle CP-Vs were described, defined as ectopic craniopharyngiomas invading the cerebellopontine angle (11, 18-20), with some authors suggesting a potential relationship with Gardner’s syndrome (30). From a histological perspective, adamantinomatous CP-Vs were more common than papillary CP-Vs, in line with the overall craniopharyngioma population (6). The rarity of CP-Vs makes difficult to clearly differentiate their origin as compared to the most common non-ventricular craniopharyngiomas. Detailed genetic and molecular analyses may be warranted to define unique differences of these entities and evaluate the embryogenic role of paracrine cell interactions between Rathke’s pouch and craniopharyngioma’s precursors, with the goal to identify potential therapeutic targets (31, 32).
As found in this cohort, CP-Vs share similar demographic characteristics with the overall craniopharyngioma’s population, prevalent in male patients and peaking in children/adolescents or in middle-age adults (33). The variability in clinical presentation primarily reflects tumor’s location within the suprasellar region, with impairment of the endocrine hypothalamic-pituitary axis and/or optic pathways, and within the ventricular cavities, with frequent obstruction of the Monro’s foramina and consequent hydrocephalus and/or increased intracranial pressure. These non-specific symptoms are largely similar to the other non-ventricular craniopharyngiomas, showing higher rates of early diagnoses in case of tumors causing intracranial hypertension and more frequent late diagnoses in case of lesions causing progressive hormone and/or visual disturbances (5, 34). In the rare events of fourth ventricle and/or cerebellopontine CP-Vs, the clinical presentation may also include typical cerebellar symptoms in addition to hydrocephalus with intracranial hypertension, making the diagnosis of CP-Vs even more difficult (11, 17).
Advances in imaging techniques have led to higher rates of pre-operative CP-Vs diagnoses in recent years as compared to historic cohorts of tumors diagnosed only intraoperatively or at post-mortem autopsy (9, 35). Imaging features of CP-Vs are largely similar to those of non-ventricular craniopharyngiomas, frequently characterized by solid parts with or without cystic components and in some cases presenting with intratumor calcifications (36). Although these features may support non-ventricular craniopharyngiomas’ differential diagnosis with different central nervous system tumors, such as low grade gliomas and germ cell tumors in the pediatric population, the frequent occurrence of solid sellar/suprasellar lesions without intratumor cysts nor calcifications may pose some difficulties (4). In contrast, the ventricular involvement of CP-Vs, which most frequently originate from the suprasellar region, offers additional information to guide the pre-operative diagnosis, with distinct features as compared to other intraventricular tumors (37). In CP-Vs, pre-operative MRI assessment is also important to characterize the tumors’ relationship with the hypothalamus and critical neurovascular structures, adapting surgical approaches and extent of tumor resection on each individual case.
In view of the complex anatomy of CP-Vs, surgical management requires multidisciplinary discussion between neurosurgery, otolaryngology, and radiation oncology (6, 9, 38). Cyst aspiration with biopsy and/or placement of Ommaya reservoir may be an option when tumor resection is deemed not feasible, with the purpose to confirm the diagnosis while reducing tumor’s size and counteracting the continuous production of intra-cystic colloid fluid (39–41). Similarly, urgent shunting may be recommended in patients with obstruction of the ventricular system, aimed at promptly decreasing the intracranial pressure, stabilizing the patient, and planning the next treatment strategies (11, 42–44). When feasible, tumor resection remains the gold-standard also in CP-Vs to restore the visual function and/or the hypothalamic-pituitary axis, with GTR being preferred to reduce the risk of recurrence (4, 25). However, the high risk of post-surgical complications, primarily involving hypothalamic and/or pituitary stalk damage, may favor the completion of subtotal or partial tumor resection to reduce tumor size and ease adjuvant radiation planning (26, 45). Surgical access to CP-Vs somewhat differs from access to non-ventricular craniopharyngiomas, with the selection of the best corridors largely depending on surgeon’s preference, tumor’s epicenter, and extension within the ventricular system (38, 40, 46). In early series, only open surgical approaches were performed, in some cases reaching the target lesion through interhemispheric transcallosal corridors or transcortical transventricular routes (29, 47). The trans-lamina terminalis route represented the preferred open surgical corridor, allowing easy access to the floor of the third ventricle, where the tumor is attached, through pterional or subfrontal approaches, with minimal risk of injuring the optic chiasm and the pituitary stalk (48–50). Of note, posterior fossa approaches, such as the retrosigmoid and the tonsillo-telovelar approaches, were performed in patients with fourth ventricle CP-Vs to access the cerebellopontine angle in the same fashion of other posterior fossa lesions (11, 18, 51). The introduction of endoscopic routes for sellar/suprasellar craniopharyngiomas and intraventricular tumors has led to its large adoption also for CP-Vs (52). Extended trans-sphenoidal and trans-ventricular endoscopic corridors offer adequate exposure of both suprasellar and intraventricular components of target CP-Vs, also allowing minimal brain retraction and wide angle of view (25, 38). However, the risks of inadequate control of adjacent neurovascular structures may pose some challenges in inexperienced hands, and should be carefully considered during pre-operative planning (46). Finally, combined open and endoscopic approaches constitute an additional option for large lesions, safely tackling intraventricular and extraventricular parts of the tumors to achieve maximal resection (40, 53).
The main therapeutic goals in CP-Vs focus on improving patients’ functional status, prolonging local tumor control, and minimizing the risks of treatment-related complications. As found in this cohort, surgical resection led to high rates of symptom improvement (88%), with low rates of tumor recurrence (23.1%), comparable to the other non-ventricular craniopharyngiomas (54). However, both open and endoscopic tumor resection correlated with high rates of persistent post-surgical complications, mainly characterized by the disruption of hypothalamic-pituitary axis and similar to the adverse events after radiation treatment for other sellar/suprasellar lesions (55). The negative impact of these complications on patients’ quality of life is partly limited by the optimal multidisciplinary medical management of hormone deficiencies available across most institutions (6). However, the narrow balance between surgical benefits and drawbacks deserves accurate consideration at pre-treatment stages. In terms of prognosis, CP-Vs show favorable OS rates, comparable to those of non-ventricular craniopharyngiomas (4, 9). Although pooled OS rates were significantly longer after endoscopic resection than those after open resection, we ascribe our results to the differences in years of publication in studies reporting the use of different routes, as open surgery was mostly performed in early series whereas endoscopic resection was preferred in more recent series (9, 29, 40, 47). These findings, coupled with the limited resources in terms of multidisciplinary patient management available in early series, may account for the shorter OS in patients who underwent open tumor resection.
Limitations. Our review has some limitations. All included studies were retrospective, likely exposed to selection bias and published within a 51-year time-period encompassing major changes in surgical protocols, which may have introduced some confounding variables into our analyses. The evaluation of baseline clinico-radiological presentation and post-treatment symptom improvement was subjective in most studies. Due to lack of granular clinical and imaging data found in the literature, we could not differentiate patients with infundibulo-tuberal and strictly intraventricular CP-Vs. Also, we could not analyze the impact of each variable on patient’s functional and survival outcomes. Finally, as we included only patients with histologically confirmed diagnosis of CP-Vs, we could not compare the role of tumor resection versus stand-alone radiotherapy.
Conclusion
CP-Vs comprise heterogeneous and uncommon lesions with often debilitating clinical presentation and complex multidisciplinary management. Surgical resection, with or without ventricular shunting, allows prompt symptom improvement and favorable long-term tumor control, but causes high rates of post-treatment hypothalamic-pituitary axis disturbances. Several open and endoscopic surgical approaches are available, which should be pre-operatively considered and tailored to each patient to offer optimal tumor access while minimizing the risks of severe complications. Further molecular and genetic studies are warranted to accurately define the unique entities of CP-Vs, characterize their differences with the more common non-ventricular craniopharyngiomas, and, eventually, identify potential therapeutic targets.
Footnotes
Authors’ Contributions
Paolo Palmisciano: Conceptualization, methodology, data analysis, writing – original draft preparation; Kurtis Young: Resources, writing – reviewing and editing; Maya Ogasawara: Resources, writing – reviewing and editing; Omid Yousefi: Resources, writing – reviewing and editing; Christian Ogasawara: Resources, writing – reviewing and editing; Gianluca Ferini: Resources, writing – reviewing and editing; Othman Bin-Alamer: Resources, writing – reviewing and editing; Mayur Sharma: Resources, writing – reviewing and editing; Giuseppe E. Umana: Resources, writing – reviewing and editing; Kenny Yu: Resources, writing – reviewing and editing; Aaron A. Cohen-Gadol: Resources, writing – reviewing and editing; Tarek Y. El Ahmadieh: Resources, writing – reviewing and editing; Ali S. Haider: Methodology, resources, writing – reviewing and editing, supervision.
Supplementary Material
Available at: <https://www.dropbox.com/sh/3ro5iiq3216rnf3/AABghPV5HbXCPkCtHogb3svYa?dl=0>
Conflicts of Interest
The Authors have no relevant financial or non-financial interests to disclose.
- Received July 1, 2022.
- Revision received July 12, 2022.
- Accepted July 13, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).