Clinicopathologic Factors Related to Post-Operative Outcome of Pituitary Adenoma: a Meta-Analysis

Purpose: About 30% to 40% of patients with pituitary adenoma require surgery. About 25% to 40% of those who have had surgery will have an aggressive outcome. The purpose of this study was to see if certain clinicopathologic factors such as size, type/subtype, invasiveness, proliferative (ki-67, mitotic rate, and p53), and grade inuenced the aggressiveness of postoperative pituitary adenomas. Methods: The factors mentioned in the research objectives were examined as independent variables. Ten studies out of 736 were chosen. The 10 studies had 2727 participants and 632 cases. The monitoring lasted 3-11 years. The studies' quality ranged from fair to excellent. Results: The results of the meta-analysis were: size ≥ 10mm OR 1,79 (1,29-2,48), corticotroph OR 1,91 (1,41-2,58), invasive OR 3,67 (1,95-6,90), proliferative OR 4,78(3,61-6,32), Ki-67 ≥ 3% OR 4,13 (2,94-5,81), mitotic rate > 2 OR 3,91 (2,74-5,57), p53 positive OR 1,92 (1,28-2,90), and grade 2b OR 4,56 (3,0-6,91). Conclusions: Size, type/subtype, invasiveness, proliferative (ki-67, mitotic rate, and p53), and grade of postoperative pituitary adenoma inuenced the postoperative aggressiveness outcome. and an incidence of 26.1 cases per 100,000 people, according to a population-based study using data from the National Health Insurance. 6 Data from the English city of Banbury shows 63 cases of pituitary tumors out of a total population of 89,334 people, with a population prevalence of 77 cases per 100,000. cohort Raverot at grade 2b et in their study prognostic pituitary prognosis. The higher the grade, the more positive prognostic factors there are, so this assessment becomes objective and means that the postoperative outcome is quite good.


Introduction
Pituitary adenoma, also known as pituitary neuroendocrine tumor (PitNET) 1,2 is a type of intracranial tumor that typically arises from the anterior lobe of the pituitary gland and accounts for 10-15% of all intracranial tumors. 3,4 According to the Japan brain tumor registry center, 19.2% of 13,431 cases of brain tumors were histopathologically con rmed as pituitary adenoma. 5 Pituitary adenoma was found to be prevalent in South Korea, with a prevalence of 62.9 cases per 100,000 people and an incidence of 26.1 cases per 100,000 people, according to a population-based study using data from the National Health Insurance. 6 Data from the English city of Banbury shows 63 cases of pituitary tumors out of a total population of 89,334 people, with a population prevalence of 77 cases per 100,000. 7 About 60-70% of pituitary adenomas are prolactinomas that respond well to medical treatment, though about 10-15% require surgery. With indications of macroadenoma invading surrounding tissue accompanied by visual disturbances or clinically due to hormone hypersecretion that does not respond well to medical therapy, approximately 30-40% of patients with pituitary adenoma require surgical management. 8,9 About 25-40% of patients who had surgery had unfavorable outcomes, such as local invasion, resistance to conventional therapy, a high recurrence rate, and metastases. 10,11 Pituitary adenomas are malignant in about 0.2-0.5% of cases. While mortality occurs in approximately 28% of aggressive pituitary adenomas and 42.5% of pituitary carcinomas over the median of 11-12 years. 12,13 It is challenging and necessary to predict the nature of pituitary adenomas because early detection of potentially aggressive adenomas will lead to aggressive treatment, which is expected to improve patient outcomes.
The size of the adenoma, the type/subtype of the adenoma, the invasive nature of the adenoma (imaging), the proliferative nature of the adenoma (immunohistochemical examination (IHC) on Ki -67, mitotic activity, and p53), adenoma grade, and novel biomarkers are some of the factors that are thought to be involved in the outcome of pituitary adenoma analyzed in the past studies. These factors require healthcare providers to have adequate imaging, surgery, and anatomical pathology (in this case, IHC examination) modalities to support pituitary adenoma patients' speci c management. The European Society of Endocrinology (ESE) guidelines for the management of pituitary adenoma cases, on the other hand, include recommendations for IHC examination with a very low level of evidence. 14 Previous studies that looked at the prognostic factors above came up with varying signi cance values. According to Trouillas J et al. 15 and Asioli S et al. 16 , when adenomas are classi ed as micro or macroadenomas, the odds ratio (OR) for an aggressive outcome is 3.2 consecutively (con dence interval (CI) = 1.64-6.26) and 1.81 (CI=1.099-2.979), respectively in macroadenomas. However, Raverot G et al. 17 concluded that it lacked prognostic signi cance, with a hazard ratio (HR) of 1.34 (CI=0.65-2.76).
Atypical pituitary adenomas were classi ed as a special classi cation by the World Health Organization (WHO) in 2004 solely based on the adenomas' proliferative nature. 18 However, it was later removed from the 2017 WHO classi cation since it did not provide patients with different prognostic values. 19 This is because, as several studies have shown, assessing the level of aggressiveness of pituitary adenomas solely based on proliferative properties is insu cient; other clinical evaluations such as mass size and invasiveness are also necessary. [15][16][17][18][19][20] Existing studies have found con icting results regarding the prognostic value of pituitary adenoma type; according to Asioli S et al 16 , lactotrophs and corticotrophs have a worse outcome than other types, with HR of 2,968 (CI=1,672-5,27) and 2,336 (CI=1,233-4,426), respectively, whereas other studies found no difference in prognostic value in the type of adenoma. 15,17,20 Although the data available to date is still limited, WHO emphasizes the high risk of aggressive pituitary adenoma in the ve subtypes of pituitary adenoma. Sparsely granulated somatotroph adenoma, lactotroph adenoma in men, Crooke's cell adenoma, silent corticotroph adenoma, and pituitary-speci c transcription factor 1 (Pit-1) positive adenoma are among the ve subtypes. 21 Another point of debate is the lack of consensus on the Ki-67 proliferative marker's threshold value. The percentages proposed range from 1.3-10%. 22,23 The majority of studies, however, use a 3% threshold. 14 While a value of more than 10% is considered to be a sign of malignancy. 23 All of these gures are only based on a limited number of case studies, a short observation period, or even expert opinions.
According to the ndings, there are still few studies examining the importance of assessing prognostic factors in postoperative pituitary adenoma. Also, no systematic study has been attempted to collect all the studies out there. The main aim of this study was to see if clinicopathological factors in uence the aggressiveness of postoperative pituitary adenomas.

Method
This research is a systematic examination that collects data from both electronic and manual sources. This study's inclusion criteria were as follows: 1. Adult participants in observational studies 2. The study included subjects with pituitary adenoma who had undergone surgery and had histological data. 3. Studies looked at adenoma size, invasiveness, proliferative qualities, adenoma type, adenoma subtype, ki-67, p53, mitotic rate, and adenoma grade as independent variables. 4. The study looked at the relationship between independent variables and aggressiveness (recurrence or advancement to conventional therapy (surgery, medicine, and/or radiation)) of pituitary adenoma as early as one year after surgery.
The following were excluded: 1. Only certain patients with pituitary adenoma were included in the study.
2. The study only assessed less than three independent variables.
The study protocol was registered using PROSPERO.With the help of the Covidence device, the search was carried out. PubMed, CINAHL, Scopus, ProQuest, and Science Direct were used to conduct a literature search using electronic-based data sources. Manual searches are also conducted using reference lists and citations from related journals (snowballing), as well as other search engines such as Google Scholar and Global Index Medicus, the results of theses and dissertations through the online libraries of Indonesia's major universities, and contacting endocrinologists, particularly in the eld of pituitary adenoma, through related semi-structured interviews.
The literature search employs keywords to represent each PICO element and is tweaked with MESHterms, truncation, and proximity signs to match the desired data source. Keywords that were used were ( Data extraction for each study was completed with the help of Covidence and Excel, then grouped according to each independent variable, type, measurement estimation/conclusion, and study design, in which the information is presented in a tabular format later. The Newcastle Ottawa Scale (NOS) is used to determine the quality of a study and the possibility of bias. Examined biases are baseline confounding, study participant selection bias, searching for missing data, and reported bias. The results of the assessments are tabulated and grouped according to the study's design. The study component's assessment results are then added together and concluded the quality of the study according to the score : 1. Between 7 and 9: good 2. Between 4 and 6: adequate 3. Between 1 and 3: poor This evaluation will be used in concluding the study's ndings. The study selection bias was assessed regarding how much data the independent variables had from the list of excluded studies. Two researchers worked independently on each stage of this systematic study. Any disagreements are worked out in collaboration with the third researcher. RevMan 5.4 was used to carry out the quantitative synthesis. If the I 2 value is greater than 60%, heterogeneity is considered signi cant. A narrative review is performed if data is unavailable or cannot be extracted, or heterogeneity is too high.

Result
The literature review took place between July 21 and August 9, 2020. In total, 736 studies were found, with limitations applied to each database in the form of studies in humans and adults, duplication of studies eliminated, and titles and abstracts screened. 568 studies were agreed to be excluded, leaving 22 studies. The search was continued by reading the full text of 22 studies. Both researchers agreed that 12 studies should be eliminated for the reasons listed in table L1. In the end, ten studies were analyzed qualitatively and quantitatively. Figure L1 illustrates the detailed search ow.
The studies took place in Europe, Asia, and the United States. There were a total of 2727 people in the study, with 632 people in the case group. The majority of studies had a monitoring period of 3 to 5 years, but one study had a monitoring period of up to 11 years. Table 1 lists the characteristics of each selected study.
All of the studies were of a good to an excellent standard. The poor-quality studies are:   The corticotroph type was more aggressive than the other types, while the somatotroph type was less aggressive. Figure L7 depicts the results of other types of meta-analysis. There are currently no guidelines that could determine which type of pituitary adenoma is more aggressive than the other. However, two of the ve high-risk subtypes highlighted by WHO, namely crook's cell adenoma and silent corticotroph adenoma, are found to be part of the corticotroph type. 19 The small number of studies evaluating this subtype could be due to the WHO's relatively recent recommendation in 2017 or certain di culties in the examination, which is quite detailed and costly. However, as recommended by WHO, this should still be considered when evaluating patients with pituitary adenoma. 19,21 In the case-control meta-analysis, the invasive trait variable showed a signi cant difference in aggressiveness. The signi cance of invasiveness could be seen through the assessment of grade 2a, which showed signi cant differences in aggressiveness compared to grade 1a in the Jang et al. 29  aggressiveness. 18 Experts at the time were outraged, and this drew much criticism. The WHO abolished the atypical classi cation in 2017 after it was proven that many patients do not have proliferative traits but have an aggressive outcome. 19,21 On the aggressiveness, the proliferative trait variable showed signi cance in the meta-analysis result. Because it requires a combination of two positives from three IHC tests (Ki-67, p53, and mitosis), each of these components is predicted to have a prognostic value for the aggressiveness of pituitary adenoma, this proliferative property has a fairly strong prognostic value.
The variable Ki-67 showed a signi cant meta-analysis result on the aggressiveness. This is also supported by the meta-analysis results of the mean aggressiveness difference, which shows a signi cant difference of 1.05 (0.87-1.22)% ( Figure L4). No meta-analysis could be performed in the cohort study group; however, Jang et al. 27 found that aggressiveness was associated with higher Ki-67 values. This procedure has long been used to diagnose endocrine tumors, particularly pituitary adenomas. Guadagno E et al. 31 conducted a fairly large study on the signi cance of Ki-67 in pituitary adenomas, with mixed results. Ki-67 is one component in assessing the proliferative nature of pituitary adenoma, according to WHO guidelines, so it must be combined with other markers to provide a better prognostic value. 18,19,21 In one of the three studies that looked at it, the variable of mitosis rate showed a signi cant difference in aggressiveness. Three studies with signi cant results on the aggressiveness were meta-analyzed.
In one of the four studies that looked at it, the p53 variable showed a signi cant difference in aggressiveness. A positive p53 was found to signi cantly impact the aggressiveness in the meta-analysis.
The variable of adenoma's grade showed signi cant aggressiveness at grade 2b. The ndings of a meta-analysis comparing grades 2b and 1a with signi cant differences in aggressiveness back this up ( gure L5). Although no meta-analysis could be performed in a cohort study, Raverot et al. 17 found aggressive aggression at grade 2b versus grade 1a. Trouillas et al. 15 proposed this grade in their study evaluating prognostic factors for pituitary adenoma postoperative outcome. This grade is based on pituitary adenomas' invasive and proliferative nature, both of which have long been used to predict its' prognosis. The higher the grade, the more positive prognostic factors there are, so this assessment becomes objective and means that the postoperative outcome is quite good.
Because it has been extensively studied in the selected literature, the postoperative residue is an additional component that is quite important in assessing the aggressiveness of postoperative pituitary adenomas. In studies that looked at the tendency to occur in adenomas with postoperative residues, this component showed signi cant differences in aggressiveness. This component is commonly used to assess postoperative outcomes in other organ malignancies. On patients where gross total resection cannot be performed, the unclean removal of the mass tissue will result in future recurrence. The meta-analysis found that aggressiveness was signi cantly increased in the presence of postoperative residues, with an OR of 2.86 (1.32-6.19). (Fig. L6).
The limitations include: 1. Wide range of study designs 2. The number of independent variables studied and research subjects.
3. The study's quality 4. Differences in de nitions and variable data types resulted in a separated meta-analysis of each variable between studies with case-control and cohort designs.

The limited number of studies.
According to an assessment of case-control study quality, four studies had poor quality, and ve studies had an unclear de nition of control, indicating a selection bias possibility from the included studies. Meanwhile, the cohort study's quality assessment revealed a high dropout rate for no apparent reason, indicating the possibility of study outcome bias. The proliferative trait variable is one of the components of adenoma grade, where an anatomical pathology examination relies on the subjective assessment of anatomical pathologists-only one anatomical pathologist is examining in the studies, so the chance of assessment bias is there. Because pituitary adenoma has a low incidence, there may be unpublished studies. This publication bias could also be due to limited access to large databases like Embase.

Conclusion
The aggressiveness of postoperative pituitary adenomas is in uenced by clinicopathological factors such as adenoma size, adenoma type/subtype, invasiveness, proliferative nature (Ki-67, mitotic rate, and p53), and adenoma grade.

Recommendations
To compare high-risk and low-risk adenoma subtypes to postoperative pituitary adenoma aggressiveness and develop a scoring system to predict postoperative pituitary adenoma aggressiveness. Figure 1 Forest plot of adenoma size ≥10mm with adenoma size <10mm on aggressiveness (case-control study) Figure 2 Forest plot of adenoma size ≥10mm with adenoma size <10mm on aggressiveness (cohort study)   Forest plot of mitotic rate >2 with mitotic rate ≤2 on aggressiveness  Forest plot of grade 2b (invasive and proliferative) with grade non 2b on aggressiveness

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Attachment.docx