Clinical characteristics of second primary cancer in colorectal cancer patients: the impact of colorectal cancer or other second cancer occurring first

Colorectal cancer (CRC) is the second leading cause of cancer-related mortality in Europe and the USA, and there are approximately 300,000 new cases and 200,000 deaths due to CRC in these areas annually [1, 2]. It has been reported that the incidence of CRC in economically transitioning countries continues to rise and the incidence of CRC in economically developed countries has stabilized or is declining [3]. In Taiwan, CRC was the third leading cause of cancer death and the death rate was 20.2 per 100,000 in 2010 (http://​www.​mohw.​gov.​tw/​cht/​DOS/​Statistic.​aspx?​f_​list_​no=​312&​fod_​list_​no=​2622; accessed in Auguest 2012). There was a 24.69% increase in the CRC-related death rate in 2010 compared with that in 2002 and a 64.23% increase compared with that in 1996 [1]. In Taiwan, cancer is reported as the first leading cause of death and the death rate was 177.40 per 100,000 in 2010, over eight times the European and USA average. There was a 16.02% increase in the cancer death rate in 2010 compared with that in 2002 and a 36.04% increase compared with that in 1996 (http://​www.​mohw.​gov.​tw/​cht/​DOS/​Statistic.​aspx?​f_​list_​no=​312&​fod_​list_​no=​2622; accessed in Auguest 2012). Cancer patients are at high risk of developing a second cancer after the treatment of the initial cancer [1, 4, 5]. Due to improvements in early detection, supportive care, and multimodality treatments, the number of CRC survivors has increased in recent decades, and continuing careful attention should always be paid to the second primary cancer (SPC) in treating these CRC patients.

In recent years, reports have documented that the present percentage of SPC in the total number of cancers ranges from 0.73% to 11.70% [6] and the trend is increasing [5, 7]. Compared with the general population, patients with CRC are at a higher risk of developing a SPC [8]. Thus, it is important to be aware of the clinical characteristics of SPC in CRC patients for early diagnosis and effective treatment. Several analytical studies have found interesting associations between the first tumor site and the subsequent tumors; this could be useful for prevention [9‐11]. The initial characteristics of CRC as well as possible risk factors for subsequent SPC are also crucial for clinicians [12, 13]. However, no relevant information regarding the differences between the sequences of CRC or SPC occurrence is available up to the present time.

In this retrospective analysis, CRC patients with at least two primary malignancies were analyzed with regard to differences of clinicopathological features and survival (CRC first or other malignancy first). We tried to identify potential risk factors and clinical outcomes which could possibly evaluate, describe, and provide appropriate postoperative surveillance strategies for CRC patients and furthermore, identify early SPC.

The clinical and pathological data regarding CRC patients diagnosed with SPC are summarized in Table 1. There were 73 (4.3%) patients diagnosed with metachronous SPC. Sixty-eight (93.2%) patients had double cancers and five (6.8%) patients suffered from triple cancers. Of these 73 patients, 22 (30.1%) were in the group of CRCF, whereas 51 (69.9%) were in OCF. There were also no significant differences in gender, tumor size, tumor location, histological type, AJCC/UICC cancer stage, vascular invasion, perineural invasion, BMI, and comorbidity of pulmonary disease and renal disease. Meanwhile, the percentages of patients receiving adjuvant chemotherapy were not significantly different in the two groups (P = 0.206; Table 2). The mean age at the time of diagnosis of the first cancer was 64.0 years in the CRCF group and 58.0 years in OCF patients. Likewise, the mean age at the time of the second cancer between the two groups was not significantly different (67.0 ± 12.0 versus 66.9 ± 13.6). Between the first and second cancer there is a time lag of approximately 2.9 years in the CRCF group and 8.7 years in the OCF group, which was significantly different (P = 0.002; Table 2).

Table 3 shows the sites of the SPCs. In the 15 sites of SPC, gastric cancer, breast cancer, bladder cancer, and hepatocellular carcinoma (HCC) were most frequently detected, followed by cancers in the lung. The most frequent site of cancers in the group of OCF was gastric cancer, and the most frequent site of cancers in the group of CRCF were HCC and ovarian cancer. There was a significant difference in the sites of cancers in these two groups: 14 (27.5%) patients in the OCF group had gastric cancer compared to one (4.5%) patient in the CRCF group (P = 0.026). The comparison stage data of 15 gastric cancer patients and eight HCC patients between CRCF and OCF was not significantly different (P = 0.232 versus P = 0.587; Tables 4 and 5). Although there was no difference of hepatitis B virus (HBV) or hepatitis C virus (HCV) carriers between the OCF and CRCF (P = 0.165), a measure of six (27.3%) of 22 CRC patients with HCC as SPC in the CRCF group was significantly higher than two (3.9%) of 51 patients in the OCF group (P = 0.003). There was a significant difference in the sites of cancers in these two groups: no patients in the OCF group had ovarian cancer compared to two (9.1%) patients in the CRCF group (P = 0.029).

Regarding survival analysis, the cancer-specific survival of CRCF patients was significantly higher than that of the OCF patient group (P = 0.036; Figure 1A), but overall survival of CRCF patients was not significantly different between the two groups (P = 0.108; Figure 1B).

Cancer-bearing patients are assumed to have an increased incidence of developing cancer of other organs [15]. In this current investigation, we identified that 5.3% of CRC patients were diagnosed with SPC, which was consistent with previous investigations (range between 0.73% and 11.70%) [6], and the trend has been increasing in recent reports [5, 7]. The most frequent sites of cancers preceding or following CRC were as follows: gastric cancer, breast cancer, bladder cancer, and HCC. In Western studies, however, different combinations of common sites of multiple cancers are reported, most likely due to the low incidence of gastric cancer in Western populations.

Of the 51 cases in the OCF group, the stomach was the most frequent site before developing CRC as SPC. Ueno et al. indicated that when colon cancer was combined with other cancers, stomach cancer (1.4%) was the most frequently encountered neoplasm, followed by breast cancer (0.4%); and when rectal cancer was combined with other cancers, stomach cancer (0.6%) was also the most frequently encountered neoplasm, but this was followed by lung cancer (0.5%) [4]. Consequently, they suggested close follow-up to ascertain the possibility of developing secondary CRC among gastric cancer patients postoperatively. Our results are consistent with this study, since the most frequent SPC after gastric cancer was CRC. Therefore, colonoscopic examination should be arranged for the possibility of CRC in treating gastric cancer patients if they have hematochezia or ileus, in order to detect synchronous or metachronous CRC [13]. On the other hand, the liver was the most common site in CRC patients who developed SPC in the present study; however, gastric cancer remains the most frequent SPC in CRC patients from a Japanese study [4]. Despite there being no difference of HBV or HCV carriers between the OCF and CRCF, the incidence of HCC as SPC in the CRCF group was significantly higher than that in the OCF group. Patients with CRC were at a high risk of developing HCC as SPC, no matter whether the patients were HBV or HCV carriers. Follow-up strategies such as abdominal echo or computed tomography (CT) are also needed for these CRC patients to detect primary or metastatic liver lesions.

Likewise, in this study, the incidence of lung cancer together with CRC was only 0.5%, which is close to the result of the Japanese study [4]. Another study from Taiwan observed that the most common SPC in CRC patients was the liver, followed by the stomach [16]. In our study, HCC is ranked as the third leading SPC. However, only 1.26% of patients developed SPCs in another study from northern Taiwan [16]. A significantly higher proportion of Dukes’ B patients developed SPC than that of patients of other stages [16]. In our study, nine (0.53%) of TNM stage II patients developed SPC, which was not significantly different from other stages.

Birgisson et al. reveals that inclusion of SPC affects the results of cancer-specific survival rates in patients with CRC [17]. SPC generates a significant number of events during follow-up of patients with CRC, which causes worse cancer-specific survival rates when SPC is included as an event in the calculations. Almost half as many SPC cases were observed as endpoints as death from the same cancer or non-cancer-related deaths [17]. For women diagnosed with both breast cancer and CRC, the cumulative risk of death at 5 years following the second cancer diagnosis is three times more likely to be due to CRC than to breast cancer [18]. In our study, the cancer-specific survival rate of CRCF patients was significantly higher than that of the OCF patients. The study by Tsai et al. suggests that perineural invasion may be a significant factor for postoperative early relapse of colon cancer and significantly lower overall survival rates in CRC patients [19]. Despite the cancer-specific survival rate being higher in the CRCF group, the overall survival was not significantly different between the CRCF and OCF groups in our study, and this might result from the cause of death being non-cancer-related death (>60% after 8 years of follow-up). The event that occurred most often in patients treated with curative intention was non-cancer-related death [18].

We are left, finally, with several caveats: 1) there are few (if any) prospective controlled trials of screening strategies among patients who are at risk for second malignancies that can guide us with objective, evidence-based information about appropriate screening and management for patients with primary malignancies; 2) there is very little validated evidence that an increased screening frequency will improve outcomes among patients who develop second malignancies; and 3) the optimal screening modalities and strategies for patients who are at risk for second malignancies remain to be defined for most tumor sites. Additional investigative efforts in the future should strive to address these limitations in our current knowledge.

In summary, SPC in CRC remains a difficult issue for clinicians no matter whether in CRCF or OCF. In addition to a regular follow-up program, for the variety of different sites of SPC, clinicians must pay more attention to determining the potential lesions especially for the accompanying high incidence of malignancies.