Microscopic Transsphenoidal Surgery Research Articles

Comparative study of enodonasal endoscopic surgery vs. microsurgery for pituitary adenoma

Objective To study the enodonasal endoscopic surgery and microsurgery for the treatment of pituitary adenoma. Methods Following the randomized controlled experiment method, the patients were randomly divided into two groups. Forty cases of pituitary adenoma underwent the endoscopic enodonasal transsphenoidal surgery, and in the same period,40 cases of pituitary adenoma underwent the microscopic enodonasal transsphenoidal surgery. The physical data, clinical manifestation, imaging examination, endocrine examination before surgery,bleeding vomume and CSF leak during the surgery,and the postoperative hospital stay, tumor total removal, the relief of clinical symptom, the recovery of hormone and complications were compared between two procedures. Results Endoscopic group was more excellent than microscopy group in bleeding volume,The former is(48.6±6.0)ml and the latter is(78.4±13.0)ml,CSF leak or not during the surgery(χ2= 13.131,P 0.05), the relief of clinical symptom(χ2= 4.310,P> 0.05), the recovery of hormone after operation(χ2= 8.342,P> 0.05),and the difference was not statistically significant. The complications after operation of Endoscopic group were much less than microscopy(χ2= 3.675,P<0.05), and the difference was statistically singificant. Conclusion The endoscopic endonasal transsphenoidal surgery is a safe and effective procedure. Key words: Hypophysoma; Neuroendoscopy; Endonasal transsphenoidal surgery

Repeated transsphenoidal surgery for resection of pituitary adenoma.

To investigate the surgical strategy of repeated microscopic transsphenoidal surgery (TSS) for treatment of pituitary adenoma, surgical techniques and treatment outcomes for 29 patients with pituitary adenoma were reviewed and analyzed. There were 17 patients who underwent TSS 18 times and 12 patients who underwent TSS 13 times. The interval between each TSS ranged from 3 months to 18 years, with a median time of 4 years. The tumor height was 15 to 45 mm on the last surgery. Among the 29 patients, 16 patients underwent total tumor resection, 11 patients underwent subtotal resection, and 2 patients underwent partial resection. Cerebrospinal fluid leak occurred in 10 patients. Among 24 patients who were followed up effectively, 1 patient developed abducens paralysis after surgery, 1 patient had chronic diabetes insipidus, and 1 patient received steroid-dependent alternative treatment. The repeated TSS may present satisfied outcomes in experienced hands. The upper edge of the posterior choanae should be identified to ensure the right orientation. The openings of the anterior wall of the sphenoid sinus and the sellar floor should be appropriately expanded to improve tumor exposure. The artificial materials should be identified and removed carefully. Intraoperative cerebrospinal fluid leakage should be managed well.

Comparison of Sinonasal Quality of Life between Patients undergoing Microscopic and Endoscopic Transsphenoidal Surgery for Pituitary Lesions: A Prospective Cohort Study

Comparison of Sinonasal Quality of Life between Patients undergoing Microscopic and Endoscopic Transsphenoidal Surgery for Pituitary Lesions: A Prospective Cohort Study

There is No Difference in 30-Day Unplanned Readmission Rates between Microscopic and Endoscopic Transsphenoidal Surgery for Pituitary Tumors

There is No Difference in 30-Day Unplanned Readmission Rates between Microscopic and Endoscopic Transsphenoidal Surgery for Pituitary Tumors

Clinical effectiveness of microscopic and neuroendoscopic endonasal transsphenoidal pituitary adenoma surgery in China: systematic review and Meta-analysis

Objective To evaluate the clinical effectiveness of microscopic and neuroendoscopic endonasal transsphenoidal pituitary adenoma resection in China.Methods The case-controlled trials of microscopic and neuroendoscopic endonasal transsphenoidal pituitary adenoma surgery were collected,and the systematic review and Meta-Analysis were conducted by revman 5.2 software.Results A total of 8 case-control trials involving 525 patients were included.As for the total resection rate and postoperative complications,significant differences were found between microscopic and neuroendoscopic endonasal transsphenoidal surgery (OR =0.41,95％ CI:0.25-0.67,P =0.0004; OR =1.90,95％ CI:1.24-2.91,P =0.003) respectively.Conclusions According to the available data,endoscopic transsphenoidal surgery provided favorable results in both tumour resection and control of complications in comparison with microscopic surgery. Key words: Pituitary neoplasms ; Microsurgery; Neuroendoscope ; Total resection rate ; Postoperative complications; Meta-analysis

Long-term results of endonasal endoscopic transsphenoidal resection of nonfunctioning pituitary macroadenomas.

Several studies report early results of endoscopic endonasal transsphenoidal surgery; however, none discuss long-term outcome measures such as tumor recurrence rates and the need for additional surgical procedures. To discuss the long-term outcomes after endoscopic endonasal transsphenoidal surgery for nonfunctioning pituitary macroadenomas. This is a retrospective study. Patients were included only if they had at least 5 years of clinical and imaging follow-up after surgery. Eighty patients met the study criteria. Grossly complete resection was achieved in 71% of patients. Knosp grade 0 to 2 tumors and tumor with volumes <10 cm were significantly more likely to have received a grossly complete resection. There were 7 recurrences (12%) in patients who had received grossly complete resections, with a mean time to recurrence of 53 months. Among the 23 patients who had subtotal resections, 11 (61%) progressed radiographically, and 3 (17%) had symptomatic progression. Knosp score, surgical and radiographic evidence of invasion, and preoperative visual deficits were predictive of recurrence in a univariate analysis, but Knosp grade was the only independent predictor in a multivariate analysis. Kaplan-Meier analysis projected a 10-year progression-free survival rate of 80% and 21% for patients with gross total resections and subtotal resections, respectively. At the long-term follow-up, 12% of patients had recurrent tumors after grossly complete resection. Recurrent or residual tumors were treated with either repeat surgery or Gamma Knife radiosurgery. Rates of complete resection, postoperative surgical and endocrinological complications, and additional surgical procedures are similar to previously published reports after microscopic transsphenoidal surgery.

Endocrinological outcomes following endoscopic and microscopic transsphenoidal surgery in 113 patients with acromegaly

BackgroundTo describe outcomes and complications in patients undergoing transsphenoidal surgery for acromegaly using the 2010 consensus criteria for biochemical remission. MethodsRetrospective review of 113 treatment naïve patients who underwent transsphenoidal surgery with the endoscopic (n=66) and the endonasal microscopic technique (n=47). Cure was defined if the age and sex-adjusted IGF-1 level was normal and either the basal GH was <1ng/ml or the nadir GH was <0.4ng/ml following oral glucose suppression at last follow-up. ResultsThe mean age at presentation was 38.1±7.1 years and 86% of tumors were macroadenomas. Adenoma sizes averaged 21.1±9.7mm, but 56% of all tumors were ≥2cm in size and 43.4% were invasive. Remission rates between endoscopic and microscopic transsphenoidal surgery did not differ significantly overall (28.8% versus 36.2%). On univariate analysis, a preoperative GH level <40ng/ml, adenoma size <20mm and non-invasiveness were predictors of remission at follow-up. Although there were no statistically significant differences in remission rates between the endoscopic and microsurgical groups, surgically induced hypopituitarism was less frequent with the former. ConclusionsWe report our surgical experience with predominantly large, invasive GH adenomas using the 2010 criteria for cure. Patients with smaller, non-invasive tumors with lower preoperative GH levels are most likely to achieve remission. Outcomes with either the microscopic or endoscopic approach do not differ significantly, although the rate of surgically induced hypopituitarism may be higher with the former. Transsphenoidal surgery remains the first line of treatment for patients with acromegaly, but invasive adenomas will frequently require adjuvant therapy.

Comparative inpatient resource utilization for patients undergoing endoscopic or microscopic transsphenoidal surgery for pituitary lesions.

OBJECT.: An increasingly important measure in the health care field is utilization of hospital resources, particularly in the context of emerging surgical techniques. Despite the recent widespread adoption of the endoscopic transsphenoidal approach for pituitary lesion surgery, the health care resources utilized with this approach have not been compared with those utilized with the traditional microscopic approach. The purpose of this study was to determine the drivers of resource utilization by comparing hospital charges for patients with pituitary tumors who had undergone either endoscopic or microscopic transsphenoidal surgery. A complete accounting of all hospital charges for 166 patients prospectively enrolled in a surgical quality-of-life study at a single pituitary center during October 2011-June 2013 was undertaken. Patients were assigned to surgical technique group according to surgeon preference and then managed according to a standard postoperative institutional set of orders. Individual line-item charges were assigned to categories (such as pharmacy, imaging, surgical, laboratory, room, pathology, and recovery unit), and univariate and multivariate statistical analyses were conducted. Of the 166 patients, 99 underwent microscopic surgery and 67 underwent endoscopic surgery. Baseline demographic descriptors and tumor characteristics did not differ significantly. Mean total hospital charges were $74,703 ± $15,142 and $72,311 ± $16,576 for microscopic and endoscopic surgery patients, respectively (p = 0.33). Furthermore, other than for pathology, charge categories did not differ significantly between groups. A 2-step multivariate regression model revealed that length of stay was the most influential variable, followed by a diagnosis of Cushing's disease, and then by endoscopic surgical technique. The model accounts for 42% of the variance in hospital charges. Study findings suggest that adoption of the endoscopic transsphenoidal technique for pituitary lesions does not adversely affect utilization of resources for inpatients. The primary drivers of hospital charges, in order of importance, were length of stay, a diagnosis of Cushing's disease, and, to a lesser extent, use of the endoscopic technique. This study also highlights the influence of individual surgeon practice patterns on resource utilization.

Comparative Resource Utilization in Patients Undergoing Endoscopic and Microscopic Transsphenoidal Surgery for Pituitary Tumors

Comparative Resource Utilization in Patients Undergoing Endoscopic and Microscopic Transsphenoidal Surgery for Pituitary Tumors

Postoperative structural complications after microscopic transsphenoidal surgery of GH secreting pituitary adenomas

Abstract Objective: Analyzing the rate of structural complications after transsphenoidal surgery for removal of growth hormone (GH) secreting pituitary adenomas. Methods: Retrospective study of 58 patients who underwent microscopic transsphenoidal neurosurgery for GH secreting pituitary adenomas exeresis in Curitiba, Parana state, Brazil, between 1998 and 2011 by the same neurosurgeon. The Criteria for diagnosis of complications were clinical. Results: Five (8,6%) of the 58 patients who underwent transsphenoidal surgery developed postoperative anatomical complications, which was due to sixth cranial nerve palsy (3,4%), surgical wound infection (1,7%) and CSF fistula (3,4%). Conclusion: The rate of postoperative complications observed in the present study is likely the literature review. There is a drop in the rate of complications with increasing experience of the neurosurgeon.

Factors Associated with Biochemical Remission after Microscopic Transsphenoidal Surgery for Acromegaly

Objectives To analyze surgical outcomes and predictive factors of disease remission in acromegaly patients who underwent microscopic transsphenoidal surgery (TSS) for a growth hormone (GH)-secreting adenoma. Design A 6-year retrospective review of 86 consecutive acromegaly surgeries. Setting Procedures performed at a single institution by a single surgeon. Participants Seventy acromegaly patients. Main Outcome Measures Demographic information, preoperative laboratory values, tumor imaging data, and morphological and immunohistochemical data were collected. Predictive values using the latest and most stringent biochemical remission criteria were determined using univariate and multivariate statistical analyses. Results Remission rate for 59 (18 males) acromegaly patients meeting the study inclusion criteria was 52.5%. Remission rates for micro- and macroadenomas were 81.8% and 45.8%, respectively. Patients of older age, with a smaller tumor, lower Knosp grade, lower preoperative GH, and insulinlike growth factor 1 levels were more likely to achieve remission. Remission rate decreased significantly with repeat surgeries. Those patients with adenomas that stained positive for somatostatin receptor subtype 2A were less likely to experience tumor recurrence and more likely to respond to medical treatment with persistent or elevated GH hypersecretion. Conclusions Microscopic TSS continues to be a viable means for treating acromegaly patients. Patients should be followed long term.

Endonasal endoscopic transsphenoidal pituitary surgery: early experience and outcome in paediatric Cushing's disease

Selective adenomectomy remains the first-line treatment for Cushing's disease (CD), until recently by microscopic transsphenoidal pituitary surgery. Endonasal transsphenoidal endoscopic surgery (ETES) is emerging as a novel, less invasive treatment for pituitary adenomas and has become the optimal surgical approach. There are no published series for the treatment of paediatric CD by ETES, and we report our centre's preliminary results. Retrospective analysis. Six paediatric patients (median age 15·8 years; range 11·7-17·0 years) fulfilled standard diagnostic criteria for CD. Preoperatively, no abnormality was identified on pituitary MR scanning in 3 (50%) patients, one had a macroadenoma. Bilateral petrosal sinus sampling demonstrated central ACTH secretion (IPS/P ACTH ratio ≥3·0, post-CRH) in 3/6 (50%) patients. The same neurosurgeon and endoscopic nasal surgeon undertook all the operations. Therapeutic outcome and rate of complications. Clinical recovery and biochemical 'cure' were achieved in 5 (83%) patients, and a corticotroph adenoma was confirmed histologically in all cured cases. One case developed post-operative CSF leak requiring lumbar drain insertion and patching. At a mean interval of 4·7 years (0·1-10·8 years) post-operatively, cured patients have shown no recurrence. One patient, with a large diffuse adenoma requiring more extensive surgery, has panhypopituitarism, and another patient has GH and gonadotrophin deficiencies. Our experience shows that ETES for removing corticotroph adenomas in children, in most cases not visualized on MRI, is minimally invasive and gave excellent post-operative recovery/results. In skilled hands, this technique provides an alternative to conventional transsphenoidal microscopic surgery in managing paediatric CD.

Case-specific protocol to reduce cerebrospinal fluid leakage after endonasal endoscopic surgery

Endoscopic transsphenoidal surgery is expanding in acceptance, yet postoperative CSF leak rates remain a concern. This study presents the Cornell closure protocol, which has yielded significantly lower postoperative CSF leak rates compared with prior reports, as an algorithm that can be used by centers having difficulty with CSF leak. A single closure algorithm for endoscopic surgery has been used since January 2010 at Weill Cornell Medical College. A prospective database noting intraoperative CSF leak, closure technique, and postoperative CSF leak was reviewed. The authors used a MEDLINE search to identify similar studies and compared CSF leak rates to those of patients treated using the Cornell algorithm. The retrospective study of a prospectively acquired database included 209 consecutive patients. In 84 patients (40%) there was no intraoperative CSF leak and no postoperative CSF leak. In the 125 patients (60%) with an intraoperative CSF leak, 35 of them with high-flow leaks, there were 0 (0%) postoperative CSF leaks. It is possible to achieve a CSF leak rate of 0% by using this closure protocol. With proper experience, endoscopic skull base surgery should not be considered to have a higher CSF leak rate than open transcranial or microscopic transsphenoidal surgery.

The use of the nasoseptal flap to reduce the rate of post-operative cerebrospinal fluid leaks following endoscopic trans-sphenoidal surgery for pituitary disease

Introduction. Our aim was to evaluate the routine use of a pedicled nasoseptal flap (NSF) as a primary repair for intra-operative CSF leaks compared with fat/fascia lata graft and lumbar drain insertion by performing a retrospective review of elective patients undergoing primary trans-sphenoidal surgery from January 2008 to present. Materials and methods. A retrospective review of consecutive primary elective trans-sphenoidal cases for sellar pathology was undertaken. Prior to September 2009, microscopic trans-sphenoidal surgery (MTSS) was performed in 40 cases and since then endoscopic trans-sphenoidal surgery (ETSS) was performed in 100 cases. Results and analysis. In the MTSS group, intra-operative CSF leaks were repaired with fat graft, tisseal and lumbar drain insertion. Intra-operative leaks in the ETSS group were routinely repaired with a pedicled NSF. Post-operative CSF leak rates were 12.5% in the MTSS and 3% in the ETSS groups, respectively. Three patients in the ETSS group developed a post-op CSF leak. All of these patients underwent surgery for craniopharyngiomas. This group had subsequent endoscopic repairs and were managed with repositioning of the NSF. Conclusion. The advent of the pedicled NSF conveys a significant advantage in preventing post-operative CSF leak, decreasing the morbidity associated with lumbar drain insertion and reduces the length of hospital stay.

The Utility of High Resolution Intraoperative MRI in Endoscopic Transsphenoidal Surgery for Pituitary Macroadenomas: Early Experience in the Advanced Multimodality Image-Guided Operating (AMIGO) Suite

Introduction: Maximal extent of resection for pituitary macroadenomas achieves optimal neural decompression, the best chance at endocrine cure, and lowers the rate of recurrence. Many studies have demonstrated the utility of low resolution intraoperative MRI (iMRI) during microscopic transsphenoidal surgery in uncovering residual tumor. Few studies, however, have investigated the utility of high-resolution iMRI in combination with endoscopy with regard to gross total resection. The Advanced Multimodality Image Guided Operating (AMIGO) suite at the Brigham and Women’s Hospital harnesses the power of a high-resolution 3-T iMRI and integrated neuronavigation. We reviewed the experience of our initial 12 cases of pituitary macroadenomas operated on by an endoscopic transsphenoidal approach in the AMIGO suite.

Comparison of endoscopic and microscopic trans-sphenoidal pituitary surgery: early results in a single centre

Introduction. Pituitary surgery has seen a recent shift from a microscopic to an endoscopic trans-sphenoidal approach. We present our early experience with endoscopic surgery and compare the outcome with our recent microscopic experience. Methods. From January 2008 until present time, 80 consecutive patients underwent trans-sphenoidal pituitary surgery in our institution. Until September 2009, all patients had a microscopic trans-septal approach. After this time, the patients underwent endoscopic trans-sphenoidal surgery. All patients underwent pre- and post-operative MRI and full endocrinological evaluation. Data was collected prospectively including tumour volume, endocrine function, visual function, length of stay and complications. Results. There were 40 patients in each group. In the microscopic group, there were 26 non-functioning tumours and 14 functioning tumours. In the endoscopic group, there were 24 non-functioning and 16 functioning tumours. There were significantly better results in terms of tumour resection (p = 0.002) and remission (p = 0.018) in the endoscopic group. In this group there was also a lower incidence of CSF leaks and a shorter length of stay for secreting tumours (p = 0.005). 1 patient in the endoscopic group died at day 43 post-operatively, having initially presented in a poor clinical state with pituitary apoplexy. Conclusion. Microscopic trans-sphenoidal surgery remains the benchmark for future surgical techniques. Our early results suggest that endoscopic trans-sphenoidal surgery provides favourable results in both tumour resection and control of secreting tumours in comparison with microscopic surgery. Further longer-term evaluation is required to ensure the outcome of endoscopic surgery.

Outcomes of Surgically Treated Giant Pituitary Tumours

To evaluate the outcomes of patients with giant pituitary tumours (GPTs) who underwent a purely binasal endoscopic transsphenoidal surgery (BETS) and compare their outcomes with those achieved through craniotomy and microscopic transsphenoidal surgery (MTS). Seventy-two consecutive patients with GPTs (greater than 10 cm3 in volume) who were treated surgically with BETS, craniotomy, or MTS from October 1994 to July 2009 were reviewed for clinical outcomes, degree of tumor resection, recurrence rates, and surgical complications. The BETS group had significantly better mean reduction of tumor volume (91%) than the craniotomy (63%, p = 0.001), and the MTS (63%, p = 0.010) groups. Gross total resection rates were also higher for BETS patients than for craniotomy patients (p = 0.010). Improvements in vision and headaches were noted in 96% and 100% of patients in the BETS group, respectively; these rates were similar to those in the craniotomy and MTS groups. Of the four patients with hormone-secreting tumours in the BETS group, three remained in remission. The median length-of-stay (four days) for the BETS group was shorter (p = 0.010), and surgical complications were less frequent (p = 0.037) and less severe compared to the craniotomy group. There were no differences in the recurrence rates: 79% percent of patients in the BETS group, 69% in the craniotomy group, and 79% in the MTS group were recurrence free at last follow-up (p = 0.829). Treatment of GPT with BETS offers excellent oncologic and clinical outcomes and can frequently obviate the need for craniotomy in these patients.

Endoscopic endonasal transsphenoidal approach for pituitary adenomas invading the cavernous sinus

In this report, the authors describe their experience with surgical access to the cavernous sinus via a fully transnasal endoscopic approach in 20 cases. Clinical and endocrinological follow-up are discussed. The authors used an endoscopic transsphenoidal approach in 192 patients with pituitary adenomas between September 1997 and January 2008, adding a cavernous sinus approach in 20 patients with invasive tumors during the last 5 years of this period [corrected]. Parasellar extension of the tumor was measured according to the Knosp Scale. Radical tumor removal was achieved in 13 (65%) of 20 patients, and subtotal removal in 7 (35%). The authors used recently defined cavernous sinus approaches in the first 14 cases, including the paraseptal approach in 6, middle turbinectomy in 7, and contralateral middle turbinectomy in 1 case. Combined approaches rather than defined standard cavernous sinus approaches were used in 4 cases and an extended approach in 2. The tumors included nonsecretory adenomas in 5 cases (25%), growth hormone-secreting adenomas in 7 (35%), prolactin-secreting adenomas in 4 (20%), and adrenocorticotropic hormone-secreting adenomas in 4 cases (20%). Normal growth hormone and insulin-like growth factor 1 levels were achieved in 4 patients (57%) with growth hormone adenomas, and remission criteria were obtained in 3 patients with prolactinomas and 3 patients with adrenocorticotropic hormone-secreting adenomas. Compared with transcranial and microscopic transsphenoidal surgery, endoscopic transsphenoidal surgery offers a wide exposure for cavernous sinus medial wall adenomas that enables removal of the adenoma from the medial cavernous sinus wall. Because of the necessity for multidisciplinary treatment to achieve satisfactory results, Gamma Knife surgery and medical therapy should be supplementary treatment options after endoscopic transsphenoidal surgery.

Surgery for clival lesions: open resection versus the expanded endoscopic endonasal approach

Clival lesions pose significant challenges with regard to their surgical management. The expanded endoscopic endonasal (EEE) approach is a promising minimally invasive technique for lesions of the central skull base. The authors' aim in the current paper was to discuss the surgical treatment of clival lesions and to present the technical details, indications, and limitations of the EEE approach. Data from a recent endoscopically treated group will be compared with findings in a previous cohort of patients treated via classic open anterior and lateral approaches. Since June 2005, 17 patients with clival lesions underwent surgery via the EEE approach. Suitable candidates were chosen according to lesion characteristics, clinical parameters, and surgical goals. Neurological outcomes, Karnofsky Performance Scale scores, the extent of lesion resection, and complications were evaluated among these patients. Eighteen percent of the patients in the endoscopic group presented with recurrent disease. Another series of 43 patients, who had undergone resection of clival lesions via an anterior (rhinotomy, maxillectomy, microscopic transsphenoidal surgery, or transoral surgery) or lateral (pterional, frontoorbitozygomatic, or combined suprainfratentorial retrosigmoid) approach, was similarly reviewed. Twenty-three of these patients (53%) presented with recurrent disease and thus had undergone prior surgery. Following the EEE approach, 11 (79%) of 14 patients who had presented with neurological symptoms experienced improvement, and gross-total resection was achieved in 59% of the patients and subtotal removal in 41%. Complications included CSF leakage (24%), tension pneumocephalus (6%), and intracranial hematoma (6%). The patient with the latter complication was the only one who experienced permanent neurological worsening. In the open resection group, neurological worsening occurred in 33% of the patients (14 of 43). Total and grosstotal removals were achieved in 84% of patients and subtotal removal in 14%. The EEE approach has been shown to be a safe and effective technique for the resection of clival lesions with limited lateral extension. The choice of surgical approach must be tailored according to both patient and tumor characteristics. Although the 2 patient series featured in this paper are not comparable-because of a selection bias-higher rates of neurological morbidity and total and gross-total resections were observed in the open resection group. Given the long survival of some patients, the EEE approach should be favored whenever reasonable.

Delayed massive epistaxis following endonasal transsphenoidal surgery

To describe delayed massive epistaxis, a rare but serious complication after transsphenoidal surgery. Two patients underwent microscopic transsphenoidal surgery through a right endonasal approach for nonfunctioning pituitary adenoma. Severe epistaxis suddenly developed on day 8 and on day 13, respectively, after uneventful surgery. In each patient, the epistaxis was due to arterial haemorrhage from the left posterior nasal cavity. Otolarygologists failed to detect the origin of bleeding which was refractory to conservative treatment. Emergency internal carotid or external carotid angiography, performed during tight packing of the posterior nasal cavity, did not show abnormal findings; nevertheless, each patient was treated successfully by endovascular embolisation of the external carotid artery without further complication. When severe delayed epistaxis follows transsphenoidal surgery and damage to the internal carotid artery has been ruled out, endovascular embolisation of the external carotid artery should be considered in patients refractory to conservative treatment.