In rectal cancer patients, whole-body metabolic tumor burden on a staging FDG PET/CT is an independent predictor of overall survival and progression-free survival. The use of PET/CT for primary staging may strategically help establish the best treatment options. Each unit increase of the wbTLG significantly augmented 3.6 times the risk of death (p < 0.0007).
There is a scarce amount of data evaluating specifically whole-body metabolic tumor burden for rectal cancer staging. Most of the available investigations are related to colorectal cancer, not specifically rectal cancer, and quantify the metabolic tumor burden of the primary tumor, not the whole-body metabolic tumor burden. Nevertheless, our data are consistent with a few investigations related to colorectal cancer, such as the study by Xu et al [25]. The authors performed FDG PET/CT for the primary staging of colorectal cancer and determined the metabolic tumor volume of the primary tumor. They showed that MTV has a strong relationship with prognosis; patients with low MTV had a significantly better prognosis. The authors also found that a SUVmax cutoff value of 19 separates patients with better from worst prognoses. The difficulty of studying CRC is the heterogeneity of the population and the presence of multiple variables, such as KRAS. Oner et al [26] evaluated patients that underwent FDG PET/CT for primary staging of colorectal cancer and showed that although both KRAS mutation and MTV of the primary tumor demonstrated prognostic power, MTV could not predict the presence of KRAS mutation.
In our study, both whole-body and primary tumor MTV and TLG values were significantly associated with overall survival and progression-free survival. Rectal cancer tumor volume itself is not a predictor of overall survival. More important to determine survival is the presence of local invasion and lymph node metastases, which is elegantly performed with MRI for local staging [27]. However, for the identification of synchronous tumors and distant metastases in advanced rectal cancer, FDG PET/CT has been increasingly indicated.
Interestingly, although the association of volume and metabolism had some prognostic power, still wbTLG and tuTLG played a significantly more important role in rectal cancer prognosis compared to wbMTV and tuMTV, similar to other tumor types such as lung cancer and lymphoma [28, 29]. In rectal cancer, it seems that the metabolism of the disease is a more important prognostic indicator than the volume of the disease.
For example, patients with patients with stages III and IV and low wbMTV had a better outcome (Fig. 4), whereas early-stage disease (I and II) presenting with a large metabolic volume of the tumor had a dismal outcome (Fig. 5). The patient example was classified as stage II disease but had high wbMTV on the staging FDG PET/CT; the patient died after 10 months. Therefore, whole-body metabolic tumor burden may be more powerful than tumor staging. Likewise, in patients presenting lymph node metastasis (especially pelvic and abdominal lymph nodes) with low-grade uptake (uptake below the automatic selection SUVmax threshold), there was no significant change in wbMTV compared to lower-stage patients and consequently their prognoses were similar to patients without lymph node metastases although these metastatic lesions lead to upstaging. Contrarily, patients with a highly metabolically active voluminous primary tumor, even with negative N and M in primary staging, were more likely to develop worse outcomes compared to patients with tumors that had a low metabolism and volume even with higher TNM stage.
Patients with mucinous-type tumors and localized disease, in general, present a relatively good prognosis while the mucinous type tumors with metastatic disease have dismal outcomes, especially when located in the rectum, because of their poorer response to chemotherapy and radiotherapy compared to adenocarcinomas [30–34].
Our study showed that mucinous adenocarcinomas had a similar risk of death when compared to non-mucinous adenocarcinomas (p = 1.0000), despite the low number of patients with the latter histology type. Although some studies show that mucinous adenocarcinomas tend to be less FDG-avid than adenocarcinomas [35], we did not see this pattern in our study group. The patients with mucinous tumors in our cohort presented with highly avid lesions as has been shown in the study by Anjos et al. [36]. Furthermore, a frequent problem encountered in mucinous tumors is a relative delay in diagnosis and treatment onset [37]. This outcome was seen in our population as there was an elevated risk of death in patients with mucinous tumors because 6 out of 7 were stage IIIB or IV, with a delay in diagnosis and thus in treatment onset.
The major strength of this study was demonstrating that specifically in rectal cancer, whole-body tumor burden on a staging PET/CT study is a strong independent diagnostic imaging biomarker of prognosis. There is a scarce amount of data evaluating specifically whole-body metabolic tumor burden for rectal cancer staging. Most of the available investigations are related to colorectal cancer, not specifically rectal cancer, and the majority quantify the metabolic tumor burden solely of the primary tumor, not the metabolic tumor burden of the whole body.
Although other clinical variables were also independently associated with overall survival and progression-free survival these variables do not account for the possibility of evaluating through imaging the prediction of outcome.
One major limitation of our study was that 70% of our patients were stages III or IV because our institution is a public tertiary reference hospital in a developing country and therefore treats mostly advanced rectal cancer patients as there is a delay of referral from primary and secondary institutions. Therefore, our study has major demographic and clinical differences compared to worldwide statistics and other studies [25, 38].
Even with these limitations, our investigation may allow us to take the next step in helping better stratify rectal cancer patients. Although laboratory data (hemoglobin, ANC, monocytes, and platelet levels), the presence of sarcopenia, and low BMI, and clinical staging are also associated with prognosis, these variables are incapable of independently discriminating prognosis in localized tumors with aggressive behavior or vice versa.