In neoplasia, telomere maintenance mechanisms (TMMs) can be prognostic and may direct therapy in the future. Two types of TMM, telomerase and recombination-based alternative lengthening of telomeres (ALT), result in four prognostic tumor groups when they occur individually, in combination, or in mutual absence. Correct designation of the TMM therefore requires an assessment of telomerase activity and for ALT telomere length distribution and ALT associated promyelocytic leukaemia protein (PML) bodies (APBs). The four groups are associated with differing prognoses that are dependent on the tumor type. As TMM inhibitors are developed, oncologists will require that pathologists determine the TMM, and the treatments will differ accordingly. Furthermore, any anti-TMM therapy administered has the potential to selectively change the TMM used by a tumor, necessitating reassessment of the therapeutic strategy. Herein, we review the telomere maintenance mechanisms, the current diagnostic measures and their utility as prognostic markers in the clinical setting.
A predictable pattern of carcinogenesis is beginning to emerge based on a multitude of growth factors, cytoplasmic receptors, signal transducers, gene mutations and epigenetic changes that have been described in the cellular pathways of cancer [
Evidence that telomere maintenance mechanisms exist, in addition to telomerase, came from investigations of telomerase-null yeast cells that carry out many divisions [16,17]. Two mechanisms for survival were identified, termed type I and type II. It is unknown whether a mechanism resembling that observed in the type I survivors exists in human tumors, but it is likely given that a similar mechanism has been identified in a human cell line that lacked both telomerase activity and the Werner syndrome protein [
Combinations of the two telomere maintenance mechanisms exist [
TelTMM is the principal mechanism in carcinomas of the breast (86%), colon (89%), prostate (88%), and pancreas (95%), as well as melanoma (91%) [
In mesenchymal tumors, ALTTMM is found with frequency in osteosarcoma (35% - 47%), leiomyosarcoma (62%), liposarcoma (24% - 33% overall and 44% in grade 3), malignant fibrous histiocytoma (77%), diffuse malignant pleural mesothelioma (26%), and uterine sarcoma (46%), but it is rare in rhabdomyosarcoma (5%), and it is not found in Ewing’s sarcoma [
Examples of other tumors in which all TMM groups exist include Wilms’ tumor [
Methods to detect TelTMM and ALTTMM are routinely used in the research laboratory but are not always readily applicable to the diagnostic medical laboratory (reviewed in detail elsewhere) [10,28,46]. A combination of two methods is currently required, one that provides information on TelTMM and another that detects the ALTTMM status, to differentiate all four TMM groups. Each method has advantages and disadvantages, and their suitability to the clinical setting is summarised below. The key aspects of five assays (TRAP, TRF, APB, FISH, c-circle and TERRA) are given in
TRAP assay: The determination of telomerase activity by the TRAP (telomere repeat amplification protocol) assay is a specific, well-established protocol (reviewed in) [
Although, the TRAP assay is currently the gold standard for detecting TelTMM, it requires fresh or frozen tissue and has aspects that make it problematic for the diagnostic laboratory. Most adaptions of the TRAP assays allow only semi-quantitative measurements, falsenegative and false-positive results can occur, and the assays are technically demanding. Non-PCR based alternatives are available. An in situ-based adaption allows for the assessment of individual cells, and it can ensure that the telomerase activity is within cancerous cells and is not that from non-cancerous activated lymphocytes [
Immunohistochemistry: Immunohistochemistry-based methods, typically for hTERT, can be used on paraffinembedded tissues [
TERRA assay: The telomeric repeat-containing RNA (TERRA) assay is a relatively new assay that may prove valuable for accessing whether a tumor is telomerase positive [
Thus, TelTMM tumors can be identified using TRAP, immunohistochemistry and TERRA assays. The TRAP assay is the best characterised and the most widely validated, while developments in other telomerase activity assays, hTERT antibody design, and surrogate assays for the presence of telomerase, such as TERRA, provide additional options for adaption to the clinical setting.
The gold standard measurement for identifying ALTTMM is the presence of maintained telomere length by recombination in the absence of telomerase activity [
TRF measurement: The terminal restriction fragment (TRF) assay is most commonly used as an indication of the telomere length distribution. Based on Southern hybridization, telomere DNA is separated using electrophoresis, and a probe recognising the telomere DNA allows for visualisation of the telomere length distribution [
FISH: More recently, the increased length of the telomeres is detected by telomere DNA fluorescence in situ hybridisation (FISH) [29,30]. In this assay, ALTTMM tumors produce large fluorescent signals compared to noncancerous cells. FISH is within the capabilities of many diagnostic laboratories, and therefore this method could be applicable to the clinical setting. Again, as outlined earlier, false-positive interpretation can occur if the telomere length alone is used to assay for TMM.
APB assay: In addition to telomere DNA FISH detection combined with co-localized detection of the promyelocytic leukaemia protein (PML), is the APB assay. “ALT-associated” nuclear PML Bodies (APBs) in large form are unique to ALTTMM with few exceptions [19, 35,36]. The development of the APB detection assay by the Reddel group marked a substantial advance in identification of ALTTMM tumors because paraffin embedded sections, or even cytopathological samples, could be used [
An APB is defined as a large (≥1.4 µm2) co-localized PML protein and telomere DNA signal in the cell nucleus [35,36]. A tumor is typically considered positive if the percentage of cells containing APBs is >0.5%. Large co-localizations of PML and telomere DNA are tumor specific, whereas smaller co-localizations have been found in both neoplastic and non-neoplastic cells [67, 68]. The function of both large and small co-localizations is unknown, but upon malignant transformation to ALTTMM, large APBs have been shown to appear and disappear precisely with the gross deregulated telomere length phenotype [
C-circle assay: C-circles are ALTTMM specific and are comprised of circular telomeric C-strand (i.e., (CCCTAA)n circles) [
TERRA assay: The TERRA assay outlined earlier for TelTMM detection may also provide insight into the ALTTMM status. Cells deficient for DNMT (expected to express high levels of TERRA) were shown to possess increased heterogeneity of telomere length and increased APB frequency (hallmarks of the ALTTMM positive tumor) [
Mutation assay: Large-scale genomic analyses have discovered new markers associated with ALTTMM [73,74]. One of these is mutation in the isocitrate dehydrogenase 1 (1DH1) gene [74-76]. Mutant IDH1 is much more frequent in grade 2 - 3 astrocytomas and secondary glioblastoma compared to primary gliboblastomas [75, 76]. In ALTTMM-positive tumors, IDH1 mutations are found at a frequency of 19% - 59% and were rare in TelTMM and NDTMM glioblastomas (<5%) [37,77]. In brain tumors, the R132H variant is the most prominent IDH1 mutation and can be detected by immunohistochemistry or sequencing of tumor DNA [75,78].
Mutations in the ATRX, DAXX, and histone H3.3 (H3F3A) genes are also associated with ALTTMM [73, 79-81]. Mutations can be identified by sequencing, or in the case of ATRX and DAXX mutations, by a lack of protein expression using immunohistochemistry [79,80]. Screening for ATRX/DAXX mutations using immunohistochemistry in pancreatic neuroendocrine tumors revealed that 100% of tumors with ATRX/DAXX mutations were ALTTMM positive [
ALTTMM tumors are identified using the TRF, FISH, APB, c-circle and TERRA assays. The TRF and APB assays are the best characterized and the most widely validated, but the other methods are more readily adapted for the diagnostic laboratory. In the near future, ALTTMM tumors may be identified, or the TMM may be supported using a panel of antibodies to the various mutations that are characteristic of ALTTMM.
The prognostic significance of the TMM is dependent on the tumor type. Here we focus our attention on tumors with all four TMM subtypes. With the exception of glioblastoma, tumors with NDTMM are consistently associated with the best survival as might be expected, whilst the presence of the ALTTMM is often associated with the worst prognosis [20,39-42,82].
In glioblastoma, the TRAP, TRF and the APB assays can be used to separate tumors into TelTMM, ALTTMM, Tel/ALTTMM and NDTMM [20,36,37,83]. Patients with ALTTMM tumors have an improved prognosis compared to those with TelTMM and NDTMM [
The TRAP and TRF assays require frozen tissue, however this is not always possible when large cohorts are sought. In a cohort of 573 glioblastomas, the APB assay alone confirmed a 2-year increase in the survival rate for those with APB-positive tumors (consistent with ALTTMM) at 22% compared to 11% for those with APB negative tumors [
As suggested by the findings of Sampl et al. (2012), in an analysis of a smaller glioblastoma cohort (28 cases), the survival of individuals based on a single TMM assays did not result in a correlation with survival [
In contrast to gliomagenesis, TMM in other non-epithetlial tumors offers a quite different prognosis. Three sarcomas, (uterine sarcoma) [
In sarcoma, it is not merely the presence of a TMM, but the presence of ALTTMM that is prognostically important. The poorer prognosis for ALTTMM-positive tumors could be due to a more aggressive tumor phenoltype. The ALTTMM is more frequent in high-grade compared to low-grade tumors, and it has an increased mitotic index and tumor size compared to non-ALTTMMs [31,39,87]. The picture is less clear with disease progresssion, and metastases are frequently telomerase positive [88,89].
Although telomerase inhibition is an attractive option for treating cancer, patients with ALTTMM and NDTMM tumors would not benefit. Stratifying patients based on the TMM will be required for assigning the correct treatment, either a ALTTMM or TelTMM inhibitor, or neither. Furthermore, the prognostic data presented in this review emphasise the need to diagnostically consider and determine all four groups of tumor TMM. The methods used to detect these groups are also important, and at least two methods are currently required. The ALTTMM can be inferred from heterogeneous and long telomere lengths, high TERRA, the presence of APBs, or the c-circle assay. Telomerase activity is determined by TRAP and inferred based on low TERRA or the hTERT immunohistochemistry assay. The ability to monitor the TMM using blood samples is an attractive option in the clinical setting, and it may prove important for early tumor detection.
As yet, all combinations of TMM detection methods have not been formally validated against known ALTTMM tissues, and this evidence is urgently required to progress to clinical trials. From the above studies, a combination of telomere length and TRAP, or the APB assay and TRAP appear adequate. Newer tests such as TERRA, immunohistochemistry for multiple markers, and the c-circle analysis are available, and combinations includeing these may be better adapted to the diagnostic laboratory. Hopefully a combination can be found that will prove clinically predictive, with good sensitivity and specificity. Targeted and effective telomere maintenance pathway treatment will then become a useful addition to the oncologist’s armamentarium.