An early look at the Organ Procurement and Transplantation Network explant pathology form data

General

This study used data from the OPTN. The OPTN data system includes data on all donors, wait-listed candidates, and transplant recipients in the United States, submitted by the members of the OPTN.

All analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC) and were based on OPTN data as of December 4, 2015. All P values were 2-sided. Tests for differences in the proportions of patient demographics between the HCC exception types were based on the chi-square distribution. Tests for differences in continuous variables were based on the t distribution. Logistic regression using backward elimination was used to determine factors that are significantly related to HCC recurrence.10 A time-to-event analysis was not possible because the date of recurrence was not available for all patients. In approximately 75% of cases, the recurrence was reported on the posttransplant malignancy form (PTM), whereas in the remaining 25% the recurrence was noted as the cause of death (COD) or graft loss on the posttransplant follow-up form, which does not include the date of recurrence.

Analysis of Explant Forms

OPTN explant pathology forms submitted electronically for liver transplants occurring between April 12, 2012 and December 31, 2014, were included in the analyses (n = 4550). Each transplant recipient was categorized based both on the type of HCC exception and the HCC stage (Milan) calculated from the number and size of tumors provided on the explant pathology form. There are 3 types of MELD HCC exceptions included in the analysis: stage T2 meeting policy criteria (T2), hepatocellular carcinoma not meeting policy criteria (NMC; primarily due to the tumor being beyond criteria, or being completely ablated prior to requesting MELD exception), and those listed with a non-HCC exception (“other”) where the diagnosis and narrative clearly indicated that the exception priority was being asked for HCC. Recent data presented to the Liver and Intestinal Organ Transplantation Committee suggest that approximately 70% of the “other, specify” cases are submitted following a missed extension deadline, which may be the only reason that the case did not meet criteria.11

A primary aim of this analysis was to compare the data reported on the standardized explant pathology form to the imaging data.

Posttransplant Outcomes

HCC recurrence was noted if, at any point in time following the transplant, the PTM indicated recurrent cancer, if the COD text field denoted HCC recurrence, or if the cause of graft failure (COGF) was “recurrent disease.” A time-to-recurrence analysis was not performed because the date of recurrence is only reported on the PTM form, representing only 75% of patients identified. Variables investigated in the initial model included the following: initial presenting tumor burden; age, sex, region, length of time to transplant from initial listing and from exception application; pretransplant ablative therapy; pretransplant alpha-fetoprotein (AFP); evidence of HCC on explant; Milan stage on explant; vascular invasion; satellite lesion; lymph node involvement; extrahepatic spread; and level of tumor differentiation. Because many of these variables are required only if evidence of HCC was seen on explant, the cohort for recurrence was restricted to those with evidence of HCC.

Pretransplant Exception Application Data

Thus, it is important to reiterate that “not meeting criteria” does not necessarily mean that the HCC is greater than stage T2. Nearly 95% of the applications indicated that the candidate had undergone pretransplant ablative therapy for HCC, which is why there were so many with reported tumor sizes of 0 (complete ablation). The remaining 2.9% of the patients who were not automatically approved in UNet were initially listed with an “other, specify” MELD exception. As with those patients who are designated as HCC exceptions, the OPTN has a mechanism to ensure that explant pathology forms are generated in those cases when an HCC exception ultimately is converted to an “other, specify” exception. There were significant differences in the percentage of cases not meeting policy criteria across regions, ranging from 17.5% to 31% (P < 0.001).

Demographics by final exception type prior to transplant are provided in Table 1. There were no major differences in sex, age, or diagnosis based on the type of HCC exception application at the time of transplant. The overall median age was 59.5 years. Overall, 77.3% were male. Not surprisingly, the most common diagnosis at listing was malignant neoplasms, followed by noncholestatic cirrhosis. Those with an “other, specify” application were more likely to be Asian or Hispanic. Asians were more likely to have a diagnosis of type B surface antigen positive cirrhosis than other ethnic groups, at 14.2%, versus 1% or less for other groups (data not shown).

Explant Form Data

Evidence of HCC was seen on explant in 4242 (93.2%) patients. Of the 308 recipients with no evidence of HCC on explant pathology, 87.0% had received pretransplant locoregional treatment for HCC, leaving 40 (0.9% of total) patients with no tumor found in the explant and no pretreatment indicated. Only 8.0% of the cohort had a report of a biopsy being performed; the percent with evidence of HCC on explant was only slightly lower in those with no biopsy versus those with a biopsy (5.4% versus 6.9%; P = 0.23). Of those with evidence of HCC, 23.0% reported well with tumor differentiation; 45.6% were moderate; 7.1% were poor; and treatment resulted in complete tumor necrosis in 17.6% of patients. Most recipients (75.5%) showed no vascular invasion, with 13.1% indicating microvascular invasion and 1.7% with macrovascular invasion. Two percent of pathology reports revealed lymph node involvement, <1% showed extrahepatic spread, and 6.5% reported satellite lesions.

Of the 2642 recipients that were within Milan criteria (stage T2) on imaging (as reported on the last exception application submitted prior to transplant), 601 (22.7%) exceeded criteria based on explant pathology, with 14.7% indicating stage T1, and 6.8% with no evidence of HCC, and the remaining 55.8% meeting T2 criteria (Figs. 1 and 2). Those not meeting policy criteria or with an “other, specify” MELD exception were more likely to be stage T3 or T4 on explant, at 37.5% and 34.9%, respectively. However, after stricter pretransplant imaging criteria were implemented on October 31, 2013, there was a slight decrease in the percentage of stage T4 tumors on explant when compared to exceptions submitted prior to that date (13.4% versus 11.9%; P = 0.21), with a corresponding increase in stage T3 (14.7% versus 18.6%; P = 0.003). The percentage of stage T1 and T2 cases remained approximately the same for those with a T2 or NMC exception. Those cases submitted after October 31, 2013 that did not meet radiologic criteria (but were approved by the RRB) were no more likely to have a biopsy than those who met criteria (P = 0.92).

Although most recipients had undergone pretransplant locoregional treatment for HCC, at 92.1% nationally, there were some differences by region. Regions 3, 10, and 11 fell below the average, at 86.5%, 82.4%, and 86.5%, respectively. In general, the percentage increased for those regions with typically longer waiting times to transplant, such as 1 (97.4%) and 9 (97.9%). Although 93.2% of forms indicated evidence of HCC, there were small variations by region. There were significant differences between regions in terms of downstaging (from greater than stage T2 to less than stage T2), ranging from <1% to 12% (P < 0.001).

Recurrence

Recurrence was found in 258 of the transplant recipients (5.7%), as shown in Table 2. The majority of patients with recurrence (75.1%) were identified via the PTM. Although the percentage is likely an underestimate due to the short follow-up period (the median time to recurrence for those whose recurrence was reported on the PTM was 345 days), there are some notable findings even with this small sample. As Figs. 3 and 4 show, early recurrence was much higher in stage T3 and T4 HCC on explant and for those with poor tumor differentiation, macrovascular invasion, and extrahepatic spread.

Results of a logistic regression analysis revealed several pretransplant factors as well as factors identified on explant that were predictive of HCC recurrence, as shown in Table 3. AFP (log scale) had a statistically significant impact on recurrence with an odds ratio (OR) of 1.2 per unit increase (P < 0.01). Explant findings that were significantly related to recurrence included extrahepatic spread (OR = 6.8; P < 0.01), poor tumor differentiation (OR = 2.8; P < 0.01), microvascular invasion (OR = 2.6; P < 0.01), macrovascular invasion (OR = 3.2; P < 0.01), whether the Milan stage based on the number and size of tumors on the explant form was T4 (OR = 2.4; P < 0.01) or T3 (OR = 1.9; P < 0.01), lymph node involvement (OR = 1.9; P = 0.03), and whether the HCC had been downstaged from greater than stage T2 to less than stage T2 (OR = 1.8; P = 0.01).

The length of time between initial listing or exception application and transplant was not found to be statistically significant for recurrence. Other factors examined that were not found to be significantly associated with recurrence included age, sex, pretransplant ablative therapy, and satellite lesions. The percent with recurrence for those with a biopsy was comparable to those with no biopsy (4.5% versus 4.6%). Statistically, the downstaging effect was not dependent on, or modified by, waiting time (P = 0.69). The c-statistic of the final model was 0.75, indicating good discrimination between recipients with and without recurrence in the original sample.

One other pretransplant factor, being transplanted in region 3, also increased the odds of recurrence (OR = 2.0; P < 0.01). Further analyses did not explain why region 3 would differ from other regions in terms of recurrence rates. Statistically, the effect of waiting time was independent of listing region (region 3 versus all other regions; P = 0.61). Although patients in region 3 have shorter waiting times than other regions, we could not specifically demonstrate short wait time as the etiology for recurrence. These shorter waiting times may mean that patients with poor tumor biology (and poor outcomes) are transplanted before tumor progression beyond Milan criteria becomes evident, as happens in areas with longer waiting times.12 It is also possible that this finding is due to reporting bias or other factors not collected by the OPTN. We will continue to examine this finding for possible explanations. However, because this is not a patient-level factor that would be used for policy-making or posttransplant surveillance, it was removed from the final model.