How and when to use temozolomide to treat aggressive pituitary tumours

How and when to use temozolomide to treat aggressive pituitary tumours

Temozolomide is an oral chemotherapy which has an established role in treating glioblastoma multiforme (GBM) and has since 2006 been used to treat aggressive and malignant pituitary tumours (O’Reilly et al. 1993, Raverot et al. 2012, Syro et al. 2018). The clinical experience has grown over that time and there has been evolving evidence for temozolomide from case reports to case series and most recently a meta-analysis (Raverot et al. 2018) and an international survey of clinical practice (McCormack et al. 2018).

Mechanism of action

Temozolomide is an alkylating agent. It works by covalently attaching a methyl group to guanine bases within the DNA of a tumour. This causes mispairing of the DNA during cell replication and, through the action of the mismatch repair system, leads to tumour cell apoptosis and tumour regression (Syro et al. 2018).

The intracellular enzyme methyl guanine methyltransferase (MGMT) acts as a repair mechanism to remove the methyl group from guanine bases and to therefore counteract the effects of temozolomide. MGMT is sometimes referred to as a suicide enzyme because in the process of removing the methyl group, the molecule of enzyme is destroyed. MGMT is ubiquitously expressed in cells but sometimes cells express low levels of this enzyme.

Evidence of clinical efficacy

There have been several case series of temozolomide-treated pituitary tumours published in the past decade (Bush et al. 2010, Raverot et al. 2010, Hirohata et al. 2013, Bengtsson et al. 2015, Bruno et al. 2015, Ceccato et al. 2015, Losa et al. 2016, Lasolle et al. 2017). They demonstrate that temozolomide is an effective treatment. If we define a clinically successful outcome as achieving either a complete response, partial response or stable disease, then the studies, overall, demonstrate a 69% success rate for temozolomide (Table 1). The recent international survey of temozolomide use for pituitary tumours reported a success rate of 70% by these criteria (McCormack et al. 2018).

If we use a more stringent standard for clinical success and include only cases which have shown clear tumour volume reduction (complete response or partial response), then the success rate from the studies is about 40%. Aggregating figures from nine recent case series (Table 1) gives an overall response of 42%. There was a recently published meta-analysis of a selection of cases which concluded that 47% of cases show significant tumour volume reduction (Raverot et al. 2018). The international survey showed 37% of cases demonstrated tumour volume reduction (McCormack et al. 2018).

Data from the French pituitary series was used to analyse overall survival in patients treated with temozolomide (Lasolle et al. 2017). The patients who responded to temozolomide had a median survival of 44 months, whereas non-responders had a median survival of 16 months. This finding emphasises that temozolomide treatment, when associated with a tumour response, greatly increases the chance of a significantly improved outcome.

Side effects

Clinically relevant side effects occur in about 20% of patients taking temozolomide (McCormack et al. 2018). The main side effects are fatigue, nausea, with or without vomiting, and cytopenia, including leukopenia, thrombocytopenia or combination. The cytopenia is normally transient and can often be managed by delaying the next cycle and/or using a dose reduction. Cessation of temozolomide due to severe cytopenia is reasonably rare but most case series report at least one example of this occurring (Bruno et al. 2015, Losa et al. 2016, Lasolle et al. 2017).

The effect of temozolomide on fertility is uncertain but, as with other chemotherapy agents, there is some risk of gonadal toxicity which is difficult to quantify. If clinically relevant, fertility consultation and sperm or oocyte/embryo preservation should be considered (Raverot et al. 2018).

Severe side effects from temozolomide are rare and have not been reported in the group of pituitary tumours treated. They include aplastic anaemia, acute liver injury, Stevens–Johnson syndrome (Syro et al. 2018). There is thought to a very low long-term risk of haematological malignancy from temozolomide use, with the risk estimated as 1 in 10,000 persons treated (Kim et al. 2009, Chou et al. 2014).

Case selection: can we identify which tumours will respond?

The published data indicate that aggressive pituitary tumours and pituitary carcinoma have the same response rate to temozolomide treatment (Raverot et al. 2018). This is surprising since one might expect carcinoma to be more difficult to treat and to be less responsive.

The case series show that functioning tumours, especially prolactinomas and corticotroph tumours, have a better response to temozolomide (Halevy & Whitelaw 2017). Conversely non-functioning tumours are less likely to respond to temozolomide.

From an understanding of the mechanism of action of temozolomide and from parallel clinical experience in GBM we would expect pituitary tumours which are depleted of MGMT and which have normal functioning mismatch repair (MMR) proteins to show an optimal response to treatment. This has been investigated by several studies with some partially contradictory results.

It is agreed that MGMT status of pituitary tumours needs to be assessed by direct immunohistochemistry (Raverot et al. 2018). The indirect method of assessment using promoter methylation status, while useful in GBM, is not helpful in pituitary tumours (McCormack et al. 2011).

The initial case reports and analysis of small series do indicate an association between low% MGMT immunostaining and response to temozolomide (Kovacs et al. 2008, McCormack et al. 2009). This is also the finding of a large Swedish study (Bengtsson et al. 2015) and is supported by both a review of 99 cases from the literature (Raverot et al. 2018) and also an international survey of temozolomide cases (McCormack et al. 2018).

Conversely the French series did not find an association between MGMT status and temozolomide response (Lasolle et al. 2017). Only a subgroup of patients in that series 13/43 had MGMT immunohistochemistry assessed; nevertheless, it has raised doubt over the clinical use of MGMT to predict temozolomide response.

The 2018 guidelines recommend that MGMT immunohistochemistry is assessed in patients in whom temozolomide therapy is being considered (Raverot et al. 2018). This may enable them to be stratified into a category of more likely or less likely to response; however, there are exceptions and, at present, patients should not be denied a trial of temozolomide therapy solely on the basis of unfavourable MGMT status. Future treatment algorithms may suggest treatment adjustment on the basis of MGMT status.

The other histological markers which might predict response to temozolomide are the proteins which form the mismatch repair (MMR) system, particularly MSH2 (mutS homolog 2) and MSH6 (mutS homolog6). It would be predicted that an intact MMR system would be needed for temozolomide to have an effective response. A Japanese case series demonstrated that MSH6 immunohistochemistry predicted temozolomide treatment response (Hirohata et al. 2013). However the Swedish case series did not replicate this finding and so it remains an open question as to whether MSH2 and MSH6 are useful for predicting temozolomide response.

Case selection: when in the clinical pathway should temozolomide be used?

If we attempt to describe the phases of treatment of a pituitary tumour – the early phase would include initial presentation and initial management: often medical management, initial surgery and perhaps radiotherapy. For many pituitary tumours there will be no need for tumour management beyond this first phase.

A second phase (or middle phase) could be where a tumour, after initial treatment(s), demonstrates progression and/or recurrence requiring additional treatment which currently is often additional surgery and/or radiotherapy.

The late phase is the management of an advanced tumour when the pituitary tumour has perhaps shown metastatic spread or is refractory to multiple surgery and radiotherapy. This includes salvage therapy when all other standard treatments are felt to be exhausted.

Many of the initial case reports relate to this late phase. They explicitly use the term ‘salvage treatment’ to convey that temozolomide was being used for progressive disease after all conventional treatments have been tried and failed (Losa et al. 2010). Temozolomide has be shown to be effective in this context and now has an established role in the management of advanced pituitary tumours, the late phase of management in the scheme outlined above. This is reflected in some recent guidelines: including the European guidelines on aggressive pituitary tumours which name temozolomide as the first-line chemotherapy for pituitary carcinoma (Raverot et al. 2018). Temozolomide is also suggested as a treatment for malignant prolactinoma in the Endocrine Society guidelines on prolactinoma management (Melmed et al. 2011). Temozolomide is recognised to be ‘the best therapy of last resort’ (Syro et al. 2018).

The use of temozolomide earlier in the treatment pathway has recently been suggested (Lin et al. 2016). European Society of Endocrinology clinical practice guidelines defined an aggressive pituitary tumour as one which shows clinically relevant growth despite optimal treatment (Raverot et al. 2018). They offer the alternative definition: radiologically invasive tumour with rapid growth. The same guidelines indicate that temozolomide is the first-line chemotherapy for treating this type of tumour, but they do not make the indication for treatment totally explicit (Raverot et al. 2018). However, it is reasonable to offer temozolomide treatment on the basis that a tumour has demonstrated aggressive behaviour, meaning growth despite optimal initial treatment. Recent expert reviews propose the use of temozolomide in tumours which show growth after surgery and radiotherapy (Syro et al. 2011, 2018, Lin et al. 2016).

Using temozolomide even earlier in the treatment pathway is rare and is not currently part of standard clinical practice. There are reasons it might be considered advantageous but also potential disadvantages (see below). The potential indication for early use of temozolomide is anticipated future aggressive behaviour based on radiological and histological assessment.

The grading system for pituitary tumours, developed by Trouillas and colleagues, offers a method of evaluating both invasion and proliferation to predict future recurrence or progression (Trouillas et al. 2013). Invasion is primarily evaluated by MRI (defined as significant cavernous sinus or sphenoid sinus invasion). Proliferation is evaluated by increased Ki67, p53 expression and/or mitoses. The overall grade is determined by whether the tumours are invasive (1 or 2) and/or proliferative (a or b). The combination of both invasion and proliferation (grade 2b) strongly predicts recurrence or progression (Asioli et al. 2019).

This grading system can identify a high-risk subgroup of pituitary tumours with anticipated future aggressive behaviour. A high-grade tumour (e.g. grade 2b) with a significant residual after initial surgery should be considered for additional treatment, normally radiotherapy but under some circumstances temozolomide might be considered.

Case examples of early (anticipatory) treatment with temozolomide are relatively rare. One case (case 1 from series) shows a corticotroph adenoma with very elevated Ki67 (50%) treated with surgery and both radiotherapy and temozolomide in quick succession (Bengtsson et al. 2015). It appears, in that case, that temozolomide is offered on the basis of anticipated aggressive future behaviour. Another series shows two prolactinoma tumours, resistant to dopamine agonist therapy, treated with first surgery and then temozolomide due to a need for tumour control, in preference to radiotherapy and in anticipation of aggressive future behaviour (Whitelaw et al. 2012).

The largest published case series (Lasolle et al. 2017) reports that a short interval between diagnosis and temozolomide initiation is associated with a positive response to temozolomide. This finding relates to a heterogeneous group of 43 cases but in general suggests earlier temozolomide treatment may be associated with better outcomes and therefore is not harmful.

One of the advantages of considering temozolomide early in the treatment pathway is the possibility of combining treatment with radiotherapy (such as in the Stupp protocol). There is some evidence that concurrent use of temozolomide and radiotherapy gives better outcomes (Raverot et al. 2018).

Conversely it can be argued that the early use of temozolomide is potentially unnecessary and/or harmful. By intervening early, on the basis of anticipated future progression and recurrence, there are a proportion of patients who will have received temozolomide chemotherapy but might never have required it, had they been managed more conservatively. Temozolomide is well tolerated but its use exposes patients to the small risk of rare serious side effects.

There is no indication in the literature that temozolomide promotes aggressive tumour behaviour and the extent to which temozolomide promotes tumour epigenetic methylation or tumour mutation is uncertain (Barciszewska et al. 2015, Lin et al. 2018). Theoretically the early use of chemotherapy (or radiotherapy) treatment agents could encourage mutation or hypermethylation of a tumour which might have an adverse effect on their clinical behaviour and response to future treatments. This would be an argument for reserving temozolomide treatment for later, rather than early, use in the treatment pathway.

There are some other rarer situations in which temozolomide might be considered. A frail patient with co-morbidity who was thought to be unsuitable for surgery and radiotherapy but who was perhaps losing vision and independence due to tumour progression could be offered temozolomide as a first-line therapy with the aim of restoring or protecting vision and promoting independence (Halevy & Whitelaw 2017).

In summary, there are four broad areas of potential indication for temozolomide treatment (shown in Table 2). The late stage indication of salvage and/or metastases (C) is well established. The earlier indication of demonstrated aggression (B) is becoming accepted as an indication, which is a welcome development and means patients do not need to wait to a salvage phase (all other treatment exhausted) before being offered temozolomide.

Anticipated (but not yet demonstrated) aggression (A) remains a controversial potential indication which is not currently part of general clinical practice. It could be argued that, in selected cases, where a tumour showed both invasion and proliferation, temozolomide could be offered early to avoid losing tumour control and to potentially allow for combined treatment with radiotherapy. Conversely, there is no evidence base for this early intervention at present. Surveillance for aggressive clinical behaviour is the current standard practice and would minimise iatrogenic harm.

The fourth group of other indications (X) represent special cases such as problematic hormone hypersecretion and/or frail patients with co-morbidity.

What is the relapse rate?

Despite a good initial response to temozolomide therapy some case series have demonstrated a high relapse rate (Losa et al. 2016). For example 10/22 cases show relapse in a large French series with 14–16 months median follow-up (Lasolle et al. 2017). However the same series shows that, despite the relapse rate, temozolomide responders overall have a significantly better median survival of 44 months, compared with 16 months for non-responders. Conversely other series have not identified high relapse rates (Bengtsson et al. 2015).

Practical aspects of temozolomide use

Temozolomide is an oral medication. The standard dose is 200 mg/m² per day for 5 days. This delivers one cycle of temozolomide and the cycles are normally repeated every 28 days. This dose is recommended in the ESE guidelines (Raverot et al. 2018). During each cycle safety blood tests are performed to assess for cytopenia or other side effects or toxicity. Clinical assessment is made of side effects and patients are normally given antiemetics to take at least during the 5 days that they take temozolomide. If patients develop cytopenia then they normally have a delay in taking the next cycle of temozolomide and sometimes a dose reduction is considered.

The initial tumour response after 3 months of temozolomide is strongly predictive of overall response (Lasolle et al. 2017). Therefore after three cycles, repeat neuroimaging (MRI scan) is performed to assess the radiological response (Raverot et al. 2018). If prior to this there is clear clinical progression or new symptoms of progression a decision may be taken to scan earlier and/or to discontinue treatment. If the initial scan shows either stable disease or a response then further cycles will be given, normally completing 12 cycles in total.

The ESE guidelines advise giving at least six cycles of treatment (Raverot et al. 2018) but internationally there is a preference to give 12 or more cycles in patients who are tolerating the medication and who appear to be showing some response (Lasolle et al. 2017, McCormack et al. 2018). The evidence for this is not strong. Some retrospective analysis indicates that longer term treatment (12 or more months) is associated with a trend towards better progression-free survival (Ji et al. 2016). However, it has been pointed out that this type of retrospective data is subject to survival bias, since non-responders and those who rapidly progress will discontinue temozolomide early (Lasolle et al. 2017).

Practice has also been influenced by the observation that re-challenge with temozolomide tends to be ineffective, encouraging a tendency to ensure the first course of temozolomide is substantial and increase the number of cycles offered during the primary course of treatment (Raverot et al. 2018).

The use of temozolomide concurrently with radiotherapy is appealing since there may be synergistic effect. This would involve either commencing temozolomide within 4 months of radiotherapy or possibly using the Stupp protocol (daily low-dose temozolomide given during 6 weeks of radiotherapy) (Raverot et al. 2018).

Who should prescribe temozolomide?

Temozolomide should be prescribed by a doctor who has the knowledge, competence and experience to appreciate the clinical indications for its use and who understands the side effect profile and toxicity. Because it is a chemotherapy it should be prescribed within the appropriate oncology governance structures. However the fact that it is oral and well tolerated means that it does not need to be administered by a chemotherapy specialist unit.

Within the United Kingdom, it is prescribed by both oncologists and endocrinologists provided the doctor has the appropriate training and experience. The decision to give temozolomide should be taken by a pituitary multidisciplinary team who have the experience to consider both whether temozolomide is an appropriate treatment and to consider the alternative surgical, radiotherapy and medical treatments (Raverot et al. 2018).

Training should be offered to endocrinologists and/or oncologists to ensure that they feel confident to consider temozolomide treatment in both the early and later phases of tumour management. Regional or national expert networks should be developed to allow for support with challenging cases.

Beyond temozolomide

If a tumour fails to respond to temozolomide or recurs after treatment, the next line therapy is unclear. Further surgery and radiotherapy will always be considered if appropriate and not exhausted. Repeat treatment with temozolomide has a generally disappointing clinical effect, review of the published cases showed 13 of 16 cases treated showed progression (Raverot et al. 2018).

Temozolomide in combination with capecitabine showed a good response in a case series of four patients with aggressive corticotroph adenomas (Zacharia et al. 2014). Three of the four cases (75%) showed complete or partial response to the treatment. However, this good clinical response may have been entirely due to temozolomide. Other case reports have shown no response to treatment with this combination and it is unclear whether there is any true advantage in adding capecitabine (Bengtsson et al. 2015). The current guidelines indicate temozolomide monotherapy remains first line (Raverot et al. 2018).

Knowledge of the MGMT and MSH status of the tumour might in the future allow for modification of the temozolomide regime to increase effectiveness. For example a tumour with high expression of MGMT might be treated with a dose-dense regime of temozolomide to deplete MGMT. The addition of an agent, such as withaferin A, to deplete MGMT has also been proposed but evidence for this is at a pre-clinical phase only (Grogan et al. 2014).

The current range of proposed treatments for temozolomide-resistant tumours has been recently reviewed and is summarised in Table 3 (Dworakowska & Grossman 2018).

The most promising recent case report concerns a 35-year-old woman with pituitary carcinoma (Lin et al. 2018). She had Cushing’s syndrome and was treated with two surgical procedures, radiotherapy and multiple medical therapies. Temozolomide was used as monotherapy and then in combination with cabecitabine; however, there was progression of the primary tumour and liver metastases. Cushing’s was controlled with bilateral adrenalectomy. The patient was then treated with two combined checkpoint inhibitors (ipilimumab and nivolumab) and showed a dramatic reduction in tumour volume. ACTH fell from 45,550 pg/mL to 66 pg/mL. Other studies have shown immunotherapy targets in pituitary tumours (Wang et al. 2018). Although very promising, the use of combination immunotherapy will need replication to see if it is consistently effective.

Conclusions

Temozolomide has established itself as the chemotherapy of choice for treatment of advanced pituitary tumours. It is effective and well tolerated. It is important to ensure pituitary MDTs feel confident to consider and initiate temozolomide treatment and are supported to do this. The point at which temozolomide treatment should be actively considered is when a tumour has demonstrated aggressive behaviour (meaning growth despite standard treatment). Often this time point will be when growth is seen after radiotherapy. It is unnecessary to wait until the point of salvage therapy when all other options are exhausted.

Anticipated (but not yet demonstrated) aggressive behaviour can be regarded as a potential indication for temozolomide, but there is currently insufficient evidence to recommend this. Ideally a trial should assess this potential indication. Treatment could be considered in selected cases when high levels of proliferation and invasion were demonstrated, causing significant clinical concern. One advantage of early identification of the potential need to use temozolomide is that concurrent use with radiotherapy becomes more of an option.

There is a need to develop and agree standardised methods of reporting MGMT and MSH tumour status, ideally on a prospective basis, to allow for the development of greater clinical understanding of the significance of these markers.

Clinical trials are needed and pending there establishment, national and international registries should continue to be promoted to facilitate our understanding of clinical practice and progress.

Declaration of interest

The author declares that there is no conflict of interest that could be perceived as prejudicing the impartiality of this review.

Funding

This work did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.