Intraoperative Alcoholization of the Pituitary Gland Does Not Reduce the Recurrence Rate of Growth Hormone Secreting Pituitary Adenomas

Abstract

Objective  Although the role of intraoperative alcoholization of the pituitary gland has been examined for the management of malignant tumor metastases and Rathke's cleft cysts, no such studies have been conducted relating to growth hormone (GH) secreting pituitary tumors, despite the high rate of recurrence in this cohort of patients. Here, we sought to understand the impact of adjunctive intraoperative alcoholization of the pituitary gland on recurrence rates and perioperative complications associated with resection of GH-secreting tumors.

Methods  This is a single-institution retrospective cohort study analyzing recurrence rates and complications among patients with GH-secreting tumors who received intraoperative alcoholization of the pituitary gland postresection versus those that did not. Welch's t -tests and analysis of variance (ANOVA) analyses were employed to compare continuous variables between groups, whereas chi-squared tests for independence or Fisher's exact tests were used for comparing categorical variables.

Results  A total of 42 patients ( n  = 22 no alcohol and n  = 20 alcohol) were included in the final analysis. The overall recurrence rates did not significantly differ between the alcohol and no alcohol groups (35 and 22.7%, respectively; p  = 0.59). The average time to recurrence in the alcohol and no alcohol groups was 22.9 and 39 months, respectively ( p  = 0.63), with a mean follow-up of 41.2 and 53.5 months ( p  = 0.34). Complications, including diabetes insipidus, were not significantly different between the alcohol and no alcohol groups (30.0 vs. 27.2%, p  = 0.99).

Conclusion  Intraoperative alcoholization of the pituitary gland after resection of GH-secreting pituitary adenomas does not reduce recurrence rates or increase perioperative complications.

Introduction

Limited evidence has demonstrated that tumor devascularization can occur with intratumoral alcohol injection during surgery. ¹ In the early 1960s, multiple groups described the use of alcoholization in addressing prostate and other malignant tumors ^{2 3 4} ; later, in the 1980s, several investigators noted alcoholization of the pituitary gland could be used to manage chronic pain in the setting of metastatic tumors. ^{5 6 7 8} From these data, we learned that alcoholization could decrease the hormone-producing function of the pituitary gland. Using this rationale, surgeons began incorporating alcoholization of the pituitary gland for the management of hormone-secreting pituitary adenomas into their practice. Despite anecdotal usage, the literature contains only one study that has examined the role of alcoholization in optimizing recurrence rates in hormone-secreting pituitary adenomas, ⁹ demonstrating potentially lower recurrence rates, but failing to provide significant values.

Given that growth hormone (GH) secreting tumors have high rates of recurrence (up to 50% in macroadenomas and 20% in microadenomas), ^{10 11 12 13} it may be beneficial to know whether incorporation of intraoperative alcohol treatment could decrease recurrence in this patient population. With this in mind, we sought to compare recurrence rates between patients with pituitary microadenomas and macroadenomas who received adjunctive, intraoperative alcoholization after a transsphenoidal approach to resection of GH-secreting pituitary tumors. We also sought to compare postoperative complications across the two groups to better assess risks and benefits. Our group hypothesized that although intraoperative, postresection injection of alcohol into the surgical bed would help decrease the recurrence rate of GH-secreting adenomas, it would also result in higher rates of diabetes insipidus compared with those who did not receive intraoperative alcohol.

Materials and Methods

Institutional Review Board (IRB) approval was obtained at the University of Utah for this retrospective cohort study. The recurrence rates among a cohort of patients with GH-secreting pituitary adenomas were evaluated through a retrospective chart review between the years 2002 and 2017 at a single academic institution. Recurrence was compared between patients who received intraoperative alcohol injections into the surgical bed versus those who did not. Intraoperative alcohol injection was performed only among patients with a GH-secreting pituitary adenoma and no intraoperative cerebrospinal fluid (CSF) leak (of note, three patients who received alcohol were found to have postprocedure leaks). Patients were determined to have a GH-secreting pituitary adenoma if they showed elevated GH and insulin growth factor 1 (IGF-1) levels. ¹⁴ The oral glucose GH suppression test was used in select patients with unclear clinical morphology or laboratory values. Patients who underwent transsphenoidal surgery (TSS) with or without intraoperative surgical bed alcohol injection, with a minimum postoperative follow-up of 1 year were included in this study. All patients underwent TSS via a microscopic approach. Patients were excluded from this study if they were found to have a pituitary lesion other than a GH-secreting adenoma on pathology or if they did not meet the minimum follow-up period of 1 year. A total of 20 patients were excluded based on the aforementioned criteria.

Surgical Technique

All of the TSS included in this series were performed by a senior skull base neurosurgeon using microscopic microsurgical technique. ¹¹ After medialization of the ipsilateral middle turbinate, a Hardy self-retracting speculum was placed. Enlargement of sphenoid os, removal of the sphenoid rostrum, and removal of the anterior sellar bone were performed. Tumor resection was performed with bayoneted ringed curettes. Among those that underwent alcoholization of the pituitary gland, a 1-mL tuberculin syringe was used to spray 1 mL of sterile dehydrogenated alcohol (98% ethyl alcohol) under direct visualization into the resection cavity. In the presence of an intraoperative CSF leak, reconstruction was accomplished using abdominal fat grafting.

Growth Hormone Remission

After completion of the tumor resection, patients underwent postoperative day 1 GH level evaluation and were subsequently followed up with 3-month and 1-year postoperative GH and IGF-1 levels. ¹¹ Remission was defined as a normalized IGF-1 and GH level <0.4 ng/mL after glucose suppression testing or <1.0 ng/mL on random screening. ¹¹ Additionally, a fasting pituitary laboratory panel is routinely obtained in all patients undergoing TSS on postoperative day 2 and includes cortisol, free T4, and thyroid-stimulating hormone (TSH).

Statistical Analysis

Data from the retrospective chart review were tabulated and analyzed. Quantitative variables were analyzed using Welch's t -tests and one-way analysis of variance (ANOVA). Categorical variables were analyzed using chi-squared tests for independence or Fisher's exact tests. Statistical significance was set at p  < 0.05 for all tests. A post hoc power analysis was run using the pwr package on the recurrence rates within inter-op alcohol and no intra-op alcohol treatment groups. All analysis was performed using R statistical software version 4.0.3 (R Core Team; 2020; Vienna, Austria).

Results

Demographics

A total of 42 patients ( n  = 22 no alcohol and n  = 20 alcohol) were included in the final analysis. There were no significant differences in age and gender across the two cohorts ( Table 1 ). The no alcohol group demonstrated a longer operative time (3.47 vs. 2.93 hours, p  = 0.02) but no difference in estimated blood loss (138.0 vs. 47.0.3 mL, p  = 0.11) compared with the alcohol group. The no alcohol group also required a larger number of fat grafts (81.18 vs. 45.0%, p  = 0.03) than the alcohol group, despite there being no difference in the number of CSF leaks between the two groups ( p  = 0.55). Additionally, there were no significant differences in complication rates across the alcohol and no alcohol groups ( p  = 0.99). In addition, no differences in gross total resection (GTR) rates (75 vs. 64.5%, p  = 0.65), or requirement for postoperative radiation (15.0 vs. 13.4%, p  = 0.99), were seen between the alcohol and no alcohol groups ( Table 1 ).

Table 1. Demographics.

	Alcohol ( n  = 20)	No alcohol ( n  = 22)	p -value
Calculated age in years, mean (SD)	42.4 (12.8)	45.2 (16.1)	0.53
Female gender	8 (40%)	9 (40.9%)	0.99
Room time in hours, mean (SD)	2.9 (0.6)	3.5 (0.8)	0.02
Estimated blood loss in milliliter, mean (SD)	47.0 (26.0)	138.0 (239.1)	0.11
Re-resection surgery, total count	3	0	0.10
Fat graft performed, mean (SD)	9 (45.0%)	18 (81.8%)	0.03
GTR	15 (75.0%)	14 (63.6%)	0.65
Radiation	3 (15.0%)	5 (13.4%)	0.99

Abbreviations: GTR, gross total resection; SD, standard deviation.

Table 2 presents a comparison of preoperative tumor volumes across the two groups based on Hardy's classification system. Of the 20 tumors that received adjunctive intraoperative alcohol injection, 5 were microadenomas (<10 mm), 14 were macroadenomas (1–4 cm), and 1 was a giant adenoma (>4 cm). Of the 22 tumors that did not receive adjunctive intraoperative alcohol injection, 2 were microadenomas (<10 mm), 19 were macroadenomas (1–4 cm), and 1 was a giant adenoma (>4 cm). Ultimately, there was no significant difference in preoperative tumor volume across the two groups ( p  = 0.38).

Table 2. Tumor size based on Hardy's classification.

	Microadenoma (<10 mm)	Macroadenoma (1–4 cm)	Giant (>4 cm)	p -value
Ethanol	5	14	1	0.375
No ethanol	2	19	1

Abbreviations: cm, centimeter; mm, millimeter.

Hormone Levels

The presenting IGF-1 level for patients in the alcohol group (696 ng/mL) was not significantly different compared with the no alcohol group (737 ng/mL, p  = 0.72), as was the presenting GH level (14.9 vs. 51.3 ng/mL, respectively; p  = 0.10; Table 2 ). Postoperative IGF-1 ( p  = 0.86) and GH ( p  = 0.78) levels were lower in both groups and did not significantly differ between the alcohol and no alcohol groups ( Table 3 ).

Table 3. Patient hormone levels.

	Alcohol ( n  = 20) Mean (SD)	No alcohol ( n  = 22) Mean (SD)	p -value
Presenting IGF-1 level (ng/ml)	695.8 (326.2)	737.1 (328.4)	0.72
Presenting GH level (ng/ml)	14.9 (15.6)	51.35 (87.5)	0.10
Post-op IGF1 (ng/ml)	425.9 (301.6)	446.3 (198.7)	0.86
Post-op GH (ng/ml)	3.0 (6.4)	3.6 (4.8)	0.78
Last measured IGF1 (ng/ml)	245.8 (134.8)	210.8 (115.5)	0.423
Last measured GH (ng/ml)	0.8 (0.8)	0.9 (0.9)	0.787

Recurrence

The overall recurrence rates did not significantly differ between the alcohol and no alcohol groups (35.0 vs. 22.7%, respectively; p  = 0.59), or by gender ( p  = 0.50; Table 4 ). Radiographic recurrence was present in 10.0 and 4.5% of the alcohol and no alcohol groups, respectively, whereas hormonal recurrence was detected in 20.0 and 9.1% of cases ( Table 5 ). Both hormonal and radiographic recurrence was present in 5.0 and 9.1% of cases. The average time to any recurrence in the alcohol and no alcohol groups was 22.9 and 39 months, respectively ( p  = 0.63), with a mean follow-up of 41.2 and 52.5 months ( p  = 0.34; Table 6 ). Recurrence was most commonly managed with octreotide treatment only in the no alcohol group, and with either octreotide treatment only or a combination of radiation, surgery, and octreotide in the alcohol group ( Supplementary Table S1 , available in the online version only). ¹¹ Specifically, three cases of recurrence (5–6 years after the initial surgery) in the alcohol cohort were managed with re-resection. No additional treatment modality was incorporated after re-resection in two of the three cases, whereas one patient underwent gamma knife radiation 18 months following re-resection for an additional recurrence.

Table 4. Overall disease recurrence rate.

	Alcohol ( n  = 20)	No alcohol ( n  = 22)	p -value
Recurrence	7 (35.0%)	5 (22.7%)	0.59
	Male ( n  = 17)	Female ( n  = 25)
Recurrence	6 (35.3%)	6 (24.0%)	0.50

Table 5. Hormonal vs imaging recurrence.

	No recurrence	Imaging only recurrence	Hormonal only recurrence	Both hormonal and imaging recurrence	p -value
Alcohol ( n  = 20)	13 (65%)	2 (10%)	4 (20%)	1 (5%)	0.61
No alcohol ( n  = 22)	17 (77.2%)	1 (4.5%)	2 (9.1%)	2 (9.1%)

Table 6. Time to recurrence and follow-up.

	Alcohol ( n  = 20) Mean (SD)	No alcohol ( n  = 22) Mean (SD)	p -value
Months to any recurrence	22.9 (23.7)	39 (68.2)	0.63
Months to chemical recurrence	18.9 (25.0), n  = 7	35.0 (70.1) n  = 5	0.64
Months to imaging recurrence	47.5 (29.2), n  = 6	53.3 (92.4) n  = 3	0.92
Follow-up months	41.2 (26.1)	52.5 (52.9)	0.34

Postoperative Hormone Treatment

Of the patients in the group that received alcohol injection, 35.0% also received postoperative hormone treatment, most commonly with octreotide (25.0%). Of the patients in the no alcohol group, 27.3% received postoperative treatment, also most commonly with octreotide (18.2%). These observations were not significantly different between the two groups ( p  > 0.05).

Complications

The overall complication rate did not significantly differ between the two groups (30.0% in the alcohol group and 27.2% in the no alcohol group, p  = 0.99; Table 1 ). Similarly, there were no significant differences in the individual complications across the two groups ( Table 7 ). Importantly, 15.0% of patients experienced diabetes insipidus in the alcohol group versus 13.6% in the no alcohol group ( p  = 0.99).

Table 7. Complications by treatment group.

	Alcohol	No alcohol	p -value
Total	6 (30.0%)	6 (27.3%)	0.99
CSF leak	3 (15.0%)	1 (4.5%)	0.33
Diabetes insipidus	3 (15.0%)	3 (13.6%)	0.99
Meningitis/infection	0 (0%)	1 (4.5%)	0.99
Post-op epistaxis	0 (0%)	1 (4.5%)	0.99
Readmission or reevaluation in ED within 30 d	0 (0%)	0 (0%)
Death	0 (0%)	0 (0%)
ICA injury	0 (0%)	0 (0%)
Visual deterioration	0 (0%)	0 (0%)
Cranial hematoma	0 (0%)	0 (0%)
Abdominal hematoma	0 (0%)	0 (0%)

Abbreviations: CSF, cerebrospinal fluid; ED, emergency department; ICA, internal carotid artery.

Discussion

GH-secreting pituitary adenomas are known to have high recurrence rates, averaging 10 to 20% in microadenomas and 25 to 50% in macroadenomas ^{10 11 12 13} ; these values appear to be fairly consistent across both endoscopic and microscopic approaches. ^{10 15} Reasons for the difficulty in achieving long-term remission in this patient population include extent of tumor resection, presence of dysmorphic features, and coexisting medical comorbidities. ¹⁶ Despite high recurrence rates, few investigations have examined adjunctive techniques to help reduce this burden. Limited data exist on management of malignant tumor metastases to the pituitary gland, as well as treatment of Rathke's cleft cysts with adjunctive intraoperative pituitary alcoholization. ^{2 3 17 18} These latter investigations have demonstrated no significant difference in recurrence rates in the management of Rathke's cleft cysts with versus without alcoholization. ^{17 18}

Based on data demonstrating devascularization of nonpituitary, intracranial tumors, ¹ and decreasing pituitary gland function, our group hypothesized that intraoperative, postresection injection of alcohol into the surgical bed would help decrease the recurrence rate of GH-secreting adenomas. However, data from the present study did not demonstrate this to be the case; rather, recurrence rates in both the alcohol and no alcohol groups were not significantly different from one another (35.0 vs. 22.7%, respectively; p  = 0.59). Thus, similar to Rathke's cleft cysts, intraoperative alcoholization of the pituitary gland after GH-secreting tumor resection did not decrease recurrence rates. Furthermore, although prior data demonstrated an association between alcoholization of the pituitary gland and postoperative complications, such as diabetes insipidus and visual complaints, our data did not demonstrate any difference in major complications between the two groups ^{5 8} ; specifically, 15.0% of patients experienced diabetes insipidus in the alcohol group versus 13.6% in the no alcohol group ( p  = 0.99).

Interestingly, our results differ from the only other investigation to examine alcohol treatment in the management of hormone-secreting adenomas. In a 2014, single-institution retrospective review, Fraioli et al ⁹ noted that patients who received postresection, intraoperative absolute alcohol treatment (with micro-patty soaked in alcohol placed in the surgical bed for 2 minutes) had an overall hormone recurrence rate of 14.8% at 5 years of follow-up, compared with 28.8% among patients who did not receive this treatment ( p  < 0.05). When broken down by the type of hormone-secreting tumor, the authors again noted lower recurrence rates among GH-secreting microadenomas when alcohol treatment was used (14.3%) versus no alcohol (27.5%); however, no accompanying p -value was reported, so the significance is unclear.

Potential reasons for the difference in results may include surgical technique (i.e., use of alcohol-soaked micro-patty in the case of Fraioli et al vs. intraoperative spraying of alcohol through a syringe in the present study), or longer follow-up in the former study (60 vs. 41 months), allowing the authors to capture two additional recurrences in each group. The decision to spray the alcohol, rather than soak cotton patties in alcohol and place them in the cavity, was based on the premise that spraying would allow the surgeon to coat a wider surface area and perhaps even reach the cavernous sinus, where residual tumor tends to be located (especially since the cases were not done with the assistance of an endoscope). Importantly, Fraioli et al only examined pituitary microadenomas , whereas our study also included macroadenomas, as well as giant adenomas (defined as >4 cm), with the literature consistently demonstrating that GH-secreting macroadenomas demonstrate lower long-term remission rates (50–75%) compared with microadenomas (80–90%). ^{10 11 12 13}

Limitations of the current study include its retrospective design, which can lead to selection bias, as well as a relatively short mean follow-up period. Additionally, despite the 16-year period of the study, the sample size was not large, and it is possible that if we had a larger sample size, we would have observed a significant relationship between intraoperative alcohol use and postoperative DI. Despite this, the sample size was similar to the only other study that has examined this topic. ⁹ This study was not designed to delineate the impact of surgical technique of alcoholization (i.e., injection vs. soaked micro-patties). Additionally, many of the surgeries in this study were performed before such adjuncts as intraoperative stereotactic navigation became commonplace at our institution, which may account for a higher rate of recurrence if GTR was not achieved. Additional prospective and comparative studies are needed to examine the impact of surgical technique of alcoholization on recurrence rates in this setting. Finally, our available data for hormone levels at time of recurrence were limited and criteria for defining remission evolved similarly to many other centers over this time period. Chen et al also did not demonstrate clear and agreed criteria for GH-secreting adenoma remission. ¹⁰ Future studies should incorporate objective hormonal measures of recurrence and link those back to use of intraoperative alcohol treatment.

Conclusion

Intraoperative alcoholization of the pituitary gland after resection of GH-secreting pituitary adenomas does not reduce hormonal or radiographic recurrence rates or increase perioperative complications compared with no alcohol.