Studies evaluating the role of different IAPs in AM have revealed heterogeneous patient profiles while analyses of IAPs’ expression have failed to provide strong evidence that establishes IAPs as potent individual prognostic markers. In the early 2000s, two studies found a correlation between lower levels of XIAP protein with longer survival ^[1][2]. However, a third study showed no correlation between the levels of XIAP protein expression and survival was found ^[3]. In 2007, Hess et al. ^[4], analyzed pro- and anti-apoptotic gene expression in AML patients, documenting a three-gene, (including BIRC3 (c-IAP2), expression signature associated with poor overall survival (OS) ^[4]. High levels of ML-IAP/Livin protein have also been correlated with poor OS ^[5]. A later study, by Pluta et al. ^[6], evaluated correlations both between XIAP, c-IAP1, c-IAP2 and survivin with each other and with Smac/Diablo in newly diagnosed AML patients. Although XIAP showed strong correlations with c-IAP1 and c-IAP2 and the latter two with each other, no correlation with survivin was observed. Low levels of IAPs were associated with 100% complete remission (CR), while high levels of either one, two or three IAPs significantly reduced the percentage of patients attaining CR. On the other hand, a short OS was influenced by older age (>50), poor risk karyotype, low levels of Smac/DIABLO and high levels of survivin ^[6]. Along these lines, survivin expression was observed in 98% of the samples and protein and mRNA levels were shown to be higher than any other IAP measured in these patients. Taken together, results from Tamm et al. ^[1], Carter et al. ^[3], and Pluta et al. ^[6], appear to suggest an influence of survivin in OS that is stronger than that of XIAP, although further studies assessing how the presence of individual IAPs affect treatment and CR achievement are clearly required ^[1][3][6].

Importantly, different authors failed to find similar results when studying younger AML patients. Indeed, despite the classical assumption that childhood cancers are similar to adult ones, evidence against this model is becoming more obvious. Childhood AML was recently defined as a significantly different disease from adult AML at the genetic level ^[7]. However, genes encoding IAPs were not included in the genomic analysis performed on the infant cohorts as part of the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) initiative. Therefore, over time, more and more studies evaluating childhood malignancies as independent diseases have started to appear. Along these lines, some studies have assessed the role of IAPs’ expression on survival and therapeutic response in children. Over the past decade, high levels of XIAP were found to positively correlate with risk groups, worse response to chemotherapy and shorter overall and relapse-free survival ^[2]. Livin and survivin levels were recently analyzed in two AML subtypes, acute promyelocytic (APL) and non-promyelocytic (non-APL). Survivin-negative patients were observed to exhibit longer survival independently of the AML group. On the other hand, livin negative patients presented longer survivals but only in the APL group. Furthermore, risk of relapse was associated with higher expression of survivin but not livin, although both correlated with higher levels of primary WBC. Taken together, none of them were identified as independent prognostic factors, although associations with other AML poor prognostic factors were evident ^[8]. While these findings agree with those of Makhlouf et al. ^[9] and El-Mesallamy et al. ^[5], they stand in strong contrast to the results presented by Sung et al., Carter et al, Moore et al. or Ibrahim et al. ^{[10][11][12][13]}.

The overexpression of apollon was correlated with an unfavorable prognosis in pediatric patients, a finding first observed by Sung et al. ^[13] and later confirmed by Ismail et al. ^[14]. In both studies, apollon was identified as a promising prognostic factor due to its association with a higher risk of relapse, worse response to chemotherapy and decreased survival. Indeed, cutoff values for the expression of the BIRC6 gene were observed to be potent prognostic metrics for both poor clinical outcomes and worse response to therapy ^[13][14].

All in all, although the cited studies point towards associations between IAPs and AML, larger cohorts, stratification by relevant factors such as disease subtype, cytogenetic characteristics or age, and unified methodology must be analyzed in future studies. Additionally, data obtained from mRNA levels must be assessed and compared to those of protein in both adult and child populations in order to validate these IAPs as prognostic factors.

Chronic lymphocytic leukemia is the most common leukemia in adults. CLL slowly develops for extended periods of time before showing any symptoms, allowing patients to have normal lives during the early stages. However, in contrast to acute leukemias, which develop much faster, it is harder to cure. CLL is a heterogeneous disease characterized by the accumulation of monoclonal mature CD5+ B cells in lymphoid organs, bone marrow and peripheral blood. The accumulation of CLL cells has classically been suggested to be caused by the inhibition of spontaneous apoptosis, rather than by cell proliferation. In turn, given CLL cells fail to engage apoptotic programs in vivo although not in vitro, as observed by Collin et al. ^[15], it becomes apparent that the tumor microenvironment might have a role in these differences ^[15]. In fact, spontaneous apoptosis has been observed to be prohibited in patients with progressive disease compared to those with stable disease, which also correlated with higher expression of cIAP1, cIAP2, XIAP and survivin ^[16]. Lower Smac/DIABLO protein levels also correlated with progressive disease. Additionally, Grzybowska et al. ^[16] found the co-expression of survivin and cIAP1to be a negative prognostic factor, since patients presenting it showed shorter OS ^[16].

Despite the classical definition of CLL as a non-proliferative disease, some studies have paid attention to the proliferative nature of CLL cells found in ‘pseudo-follicles’ formed in lymphoid organs ^[17]. These cells are known to receive support from accessory cells, such as monocyte-derived nurse-like cells, follicular dendritic cells, BMSCs or T lymphocytes, which provide pro-survival and proliferative signals. Indeed, several pro-survival signaling pathways, including NF-kB, are known to be constitutively activated in CLL ^[18]. It is believed that such a microenvironment is accountable for resistance to chemotherapeutics. Growing evidence indicates that CLL proliferative cells might be responsible for disease relapse after treatment, so any approach targeting said microenvironment could mean a complete clear out of malignant cells ^[19]. Purroy et al. ^[20] sensitized resistant CLL cells to cytotoxic agents by inhibiting survivin expression and inducing apoptosis in an ex vivo model of this microenvironment ^[20]. Likewise, Zhu et al. ^[21] achieved sensitization of CLL cells to chemotherapy by modulating the expression of XIAP, among other antiapoptotic proteins, with microRNAs shown to be downregulated in ‘p53 wild-type’ patients from a Chinese cohort ^[21].

Colorectal cancer (CRC), also known as bowel cancer, was the third most diagnosed cancer, and the second most fatal, worldwide in 2018, according to “Las cifras del cancer” report from the SEOM (Sociedad Española de Oncología Médica). Although it is mostly a curable disease, when detected at early stages, when symptoms arise, cancer may already have advanced. Lack of activity, obesity, high-fat diet as well as alcohol and smoking are all risk factors for sporadic CRC. Altered molecular pathways documented in CRC have delineated a large number of potential biomarkers. However, given that current screening methods present specificity, sensitivity and cost limitations, make it has proven difficult to rapidly advance towards personalized and targeted therapies. Additionally, studies aiming for validation of these markers show small sample sizes, have data interpretation limitations and lack standard methodologies causing reproducibility challenges ^[22]. Despite these barrier, promising advances are being made. EGFR-targeted therapies use the KRAS gene as a predictive marker of response, and anti-VEGF is sometimes used in metastatic disease in combination with chemotherapy ^[22].

Survivin and cIAP-2 expression were distinctive in elderly groups with CRC compared to younger patients, which suggested age-related differences in CRC. Interestingly, XIAP expression was comparable in normal and cancerous tissues in both young and older patients ^[23]. The location of both cIAP-1 and cIAP-2 was assessed by Ponnelle et al. ^[24] who found that, although the proteins were present in both nucleus and cytoplasm, cIAP-1 was most frequently expressed in the nucleus (85%) and cIAP-2 in the cytoplasm (82%) in CRC samples ^[24]. However, nuclear expression of cIAP-2 was associated with lymphoid infiltrate in the stroma, suggesting a nuclear localization role of both proteins in the pathogenesis of CRC. Similarly, Karasawa et al. ^[25] observed higher cIAP-2 expression in stage II CRC with lymph node metastasis ^[25]. These patients also presented early recurrence after fluorouracil-based chemotherapy. Interestingly, lymph node metastases with low levels of cIAP-2 and TUCAN (CARD8) showed a larger percentage of five-year survival in a 2005 study of IAPs’ expression ^[26].

Correlation between XIAP expression and clinical features or prognosis in CRC have yielded conflicting results with some studies reporting no differential expression between cancerous and surrounding normal tissues, while others observed both higher protein and mRNA expression in the former ^[23][27]. Guoan et al. ^[28] showed XIAP staining with immunocytochemistry in the cytoplasm of CRC cells ^[28]. Also, a significant correlation between higher levels of XIAP and tumor differentiation, venous invasion and Duke’s staging was found. Furthermore, XIAP expression levels varied inversely with different survival indexes, such as PFS or OS, in most studies ^[27]. Takeuchi et al. ^[29] described the Akt signaling pathway as an important cascade through which XIAP appeared to be upregulated in C-Met overexpressing CRC cells ^[29].

Interestingly, livin expression significantly correlated with longer survival in a study by Lee et al. ^[27] in contrast to what was observed for XIAP, while Takeuchi et al. ^[29] did not find mRNA or protein overexpression in malignant cells ^[27][29].

BRUCE/Apollon mRNA was also found to be overexpressed in CRC tissues compared with non-neoplastic samples through cDNA microarrays ^[30] correlating with unfavorable clinical features. Smac/DIABLO levels have also been evaluated with respect to their prognostic potential in CRC and one study reported its decreased expression as an independent factor for poor prognosis ^[31].

In 2020, breast cancer was the most commonly diagnosed cancer worldwide, surpassing lung cancer. It is the most common cancer among women and the second most common cause of mortality in this population. However, the establishment of prevention programs, although still challenging, has allowed an increase in early detection and survival of patients ^[32].

In the last few decades, the discovery of hormonal alterations in breast cancer patients and the application of targeted therapies in populations with specific molecular characteristics have brought therapeutic promise. In terms of IAPs, overexpression in cancerous tissues has been generally observed in comparison to normal tissues. A trend of increasing expression in more advanced diseases is also apparent. Pluta et al., ^[33] evaluated XIAP, cIAP1, cIAP2 and survivin expression in breast cancer samples at diagnosis. The expression of XIAP and survivin was higher in more advanced cancers. Expression of cIAP-2 was higher in node-positive breast cancer, although cIAP-1 levels did not correlate with clinicopathological features of this malignancy. Higher expression of XIAP was found in most patients compared to control. Its overexpression was also associated with advanced cancer ^[33]. Yang et al. ^[34], observed an increase in XIAP expression in higher grades of ductal invasive breast carcinoma and ductal breast carcinoma in situ ^[34]. Similarly, immunohistochemical assays confirmed its expression in 84% of breast invasive ductal carcinomas with high immunoscores ^[35]. The distinction between cytoplasmic and nuclear localization has also been made. Zhang et al. ^[35], reported differential subcellular expressions in breast cancer samples when cytoplasmic XIAP was observed in all the studied cases but nuclear staining in only 43% ^[35]. Likewise, Xu et al. ^[36], found higher cytoplasmic expression compared to normal tissues, while nuclear expression showed no difference ^[36]. Cytoplasmic localization correlated to HER-2 expression status, and mutant-type (P53) status, and was considered to be a prognostic biomarker for basal-like breast cancer. Patients with higher tumor levels of XIAP showed an increased risk of relapse, as observed in database analysis. Zhang et al. ^[35], reported a significant relationship between nuclear staining of XIAP protein and shorter OS while Pluta et al. ^[33], observed associations between shorter PFS and XIAP protein expression ^[33][35]. However, Xu et al. ^[36], did not find a correlation with DFS or OS ^[36]. Tumor size, extranodal extension, triple-negative status and poorly differentiated subtypes showed direct associations with XIAP expression in the Middle Eastern population. This study showed a significant relationship with the PI3K pathway, p-AKT, Ki-67 and PARP. XIAP was proven to be an independent prognostic marker ^[37].

Similar to XIAP expression, survivin expression has been correlated with worse clinicopathological features, such as metastasis, and advanced stage, or size, of the tumor. In addition, ER and progesterone receptor-negative hormonal status was observed in to correlate with survivin level ^[38]. However, previous studies had not found any correlation with said features ^[39]. In particular, Adamkov et al. ^[40], suggested that the disagreement observed between different studies could have an origin in the subcellular localization of survivin ^[40]. Their experiments led to the determination of survivin expression as a potential prognostic factor for ductal breast carcinoma. A meta-analysis encompassing 15 studies, with a total of 2202 breast cancer patients, confirmed significant associations between positive expression of survivin and worse OS ^[41].

High expression of livin has been associated with highly-invasive breast cancer cells in opposition to a lower expression in non-invasive cells ^[42]. Li et al. ^[42], described the role of livin in cancer cell migration and invasion by means of the activation of Akt signaling and the induction of EMT in vitro and in vivo ^[42]. Furthermore, as in other malignancies, Smac/DIABLO protein was observed to be lowered in breast cancer samples compared to control ^[43].

Finally, despite the clear tendency of IAP expression to correlate with poor prognosis, Pluta et al. ^[33], who observed XIAP, cIAP1, cIAP2 and survivin expression in breast cancer samples at diagnosis, its effect on survival could not be confirmed ^[33]. IAPs’ patterns of expression in other malignancies, such as bladder, lung or renal cancers, are similar to those presented above and have been further developed by Fulda and Vucic ^[44] and Che et al. ^[45]. In glioblastomas, livin is correlated with worse PFS and OS rates. C-IAP1 and cIAP2 have been found to be overexpressed in osteosarcoma, and XIAP’s expression levels were high in esophageal, ovarian and pancreatic cancers ^[46][47]. Table 1 summarizes the IAPs and cancer.