Patients with metastases of the cerebellum are often included in the same study population, even though posterior fossa lesions typically have different presenting symptoms, clinical outcomes, and complications.
Patients with cerebellar metastases have more distinct clinical presentations and outcomes than patients with non-cerebellar lesions 1).
In adults with only cerebellar masses, cerebellar hemangioblastoma and metastasis are the 2 most important differential diagnoses.
DWI and Dynamic susceptibility contrast MR perfusion (DSC-PWI) are helpful in the characterization and differentiation of hemangioblastomas from brain metastases. DWI appears to be the most efficient MR imaging technique for providing a distinct differentiation of the 2 tumor types 2).
High b value DWI reflects diffusion more accurately than does regular b value. Results showed that ADC calculation by high b value (b = 4000) DWI at 3-T magnetic resonance imaging is clinically useful for differentiating hemangioblastomas from brain metastases 3).
Cerebellar metastases are considered a negative prognostic factor in patients with systemic cancers.
Based on limited literature, an at risk group of patients for development of neoplastic meningitis (NM) are those with resected cerebellar parenchymal metastases.
The incidence of NM following resection of cerebellar metastases (3%: Group 1) was no greater than that seen following treatment of cerebellar metastases with radiotherapy only (2%; Group 2). In addition, in a large data base of patients with NM (40% with parenchymal metastases), cerebellar metastases occurred at a proportion (10%; Group 3) expected based on proportional brain volume and cerebral blood flow 5)
One hundred and forty patients underwent resection for cerebellar metastases. Most common primary tumor sites were lung, colon/rectum, and breast in 45%, 19%, and 14%, respectively. None were prostate cancers. Melanoma metastases were significantly underrepresented, and colorectal cancer metastases significantly overrepresented in cerebellum, compared to the overall proportion of cerebellar/supratentorial metastases surgically resected (P < 0.05). Thirty-day post-operative mortality rate was 4.3%. Median OS was 7.7 months (95% CI 6.0-9.5 months) irrespective of post-operative adjuvant therapy. Median OS was 51.8, 8.4, and 3.4 months, respectively, for recursive partitioning analysis class 1(n = 11), 2 (n = 78) and 3 (n = 34). Significant negative prognostic factors were age ≥65 years, Karnofsky performance score (KPS) <70, extracranial metastases and uncontrolled systemic disease.
Melanoma metastases were significantly underrepresented in cerebellum, whereas colorectal cancer metastases were significantly overrepresented. Surgical mortality and OS after surgical treatment of cerebellar metastases were similar to the results of supratentorial metastases 6).
Ghods et al. performed a retrospective analysis involving 50 patients with cerebellar metastases who underwent surgical resection. Ventriculo-peritoneal shunts were placed in patients necessitating permanent CSF drainage.
The review included 21 males and 29 females, 29 to 82 years of age. Primary tumors included lung (48%), breast (14%), GI (14%), endometrial/ovarian (6%), melanoma (6%), sarcoma (4%), lymphoma (4%), laryngeal (2%), and other (2%).
Clinical symptoms at presentation commonly were those secondary to elevated intracranial pressure and were the initial complaint in 34% of patients. Preoperatively, 29 patients were noted to have hydrocephalus. Importantly, 76% of these patients were able to avoid placement of a ventriculo-peritoneal shunt following surgery. Only two complications were noted in the series of 50 patients, including a symptomatic pseudomeningocele and a wound infection. No symptomatic postoperative hematoma developed in any surgical case.
A review of the literature has shown a high complication rate in patients undergoing surgical resection of cerebellar metastases. We have shown that surgical resection of cerebellar metastases is a safe procedure and is effective in the treatment of hydrocephalus in the majority of patients harboring cerebellar lesion 7).
In 109 patients the primary site of carcinoma was lung in 43 patients, breast in 24 patients, and gastrointestinal in 19 patients. Seventy-three (67.0%) patients underwent surgical resection; and of those patients, 27 (24.8%) also received WBRT. In 68 patients with high performance status (KPS >70), the median survival time (MST) after the diagnosis of brain metastasis was 22.1 months compared to 13.4 months in 41 patients with pretreatment KPS of <70. The MST for 38 patients who had surgical resection alone was 20.5 months compared to 35.5 months for those (n = 27) who had surgical resection followed by radiotherapy 8).
Of 93 patients, the median survival was 12.9 months for RPA class I, 11 months for class II and 8 months for class III. On multivariate analysis, RPA class was an important predictor for overall survival. However, SRS with WBRT or surgery with WBRT or a combination of SRS, surgery and WBRT, was more favorable than surgery or SRS alone within RPA class II patients.
Survival of patients with cerebellar brain metastasis is comparable to that of patients with supratentorial brain metastasis using RPA classification. Aggressive multimodality therapy has a favorable impact on survival 9).
In 11 cases the primary tumor was lung cancer in five cases and breast carcinoma in two cases; the remaining three cases had colon cancer, nasopharyngeal carcinoma (NPC) and Ewing's sarcoma, respectively. All patients underwent craniectomy and gross total excision of the tumor. Seven patients survived less than one year, two cases died in the second year, and one case of NPC survived for more than two years. The only survival is a case of Ewing's sarcoma who underwent surgery 14 months ago. The symptoms and signs of all patients improved satisfactorily after surgery. Four patients received postoperative irradiation to the posterior fossa and two cases of lung cancer had a thoracotomy for the primary lung lesion; however, the survival period was not prolonged. We suggest that a cancer patient or a patient in the fifth to seventh decades of life presenting headache, gait disturbance and vomiting should promptly undergo a computed tomography (CT) scan of the head. In selected cases, surgical intervention for solitary metastatic tumors in the tiny posterior fossa may be the best initial treatment. Adjuvant therapies should then be added according to the type of tumor 10).
59 patients with cerebellar metastases (26 solitary) with respect to clinical presentation, diagnosis, and natural history. Eighty-seven percent of patients initially complained of headache, gait disturbance, and/or dizziness. At time of diagnosis, 92% of patients with solitary cerebellar metastases and 74% of the overall series complained of headache and/or difficulty walking. In three of four cases, magnetic resonance imaging (MRI) was superior to x-ray computed tomography (CT) in detecting the cerebellar lesions. Several patients acutely deteriorated during evaluation or at the initiation of radiation therapy. We conclude that a cancer patient presenting with headache and gait difficulty with or without nausea/vomiting and dizziness should promptly undergo head CT scanning, and that MRI is useful even if CT is negative. In addition, we recommend that patients with documented cerebellar metastases receive high-dose glucocorticoid therapy for 48 to 72 hours before beginning radiation therapy. The presence of symptomatic hydrocephalus or failure to respond to glucocorticoids initially are particularly ominous features that may be best managed by early neurosurgical consultation before beginning radiation therapy 11).
Seventeen patients with computed tomographic (CT) evidence of a solitary cerebellar metastasis were studied. In 11 of 17 cases, neurologic symptoms preceded systemic evidence of carcinoma. Initial neurologic symptoms included gait instability (13 cases) and headache and vomiting (four cases). All patients had evidence of gait or limb ataxia on neurologic examination. Fourteen patients underwent craniotomy and subsequent irradiation, and three had radiotherapy without initial surgical biopsy. One patient with lung carcinoma had clinical and CT evidence of intracranial recurrence 14 months later but no evidence of widespread systemic carcinoma. Fifteen patients later showed evidence of systemic carcinoma but then died without subsequent development of recurrent cerebellar dysfunction or other neurologic abnormalities. Furthermore, in six of these patients with widespread systemic carcinoma, scans taken two to six months after completion of surgery and/or radiation therapy for the solitary metastasis showed no evidence of recurrent intracranial disease 12).
The case of a 56-year-old man who presented with brain secondaries as an initial manifestation of carcinoma of oesophagus is presented and the unique characteristics highlighted.
Patient survival was for 12 months after initial presentation and treatment.
Oesophageal carcinoma is a recognised cause of intracerebral metastatic tumours, though this is an uncommon occurrence 13).
A 67-year-old man presented with worsening headaches and gait disturbance, and with minimal clinical signs. A brain MRI showed multiple solid cerebellar metastases. The pathological diagnosis of metastatic carcinoma of the prostate was further suggested by an elevation in prostate specific antigen, and was pathologically confirmed following the neurosurgical removal of the tumors 14).
A case of cerebellar metastatic disease occurred following combination chemotherapy and a negative second-look operation for a stage III poorly differentiated adenocarcinoma of the ovary. The single metastatic tumor was resectable, and the patient had resolution of her symptoms following surgery. This is the fourth report of an epithelial ovarian carcinoma metastatic to the cerebellum 15).