PERSONAL OBSERVATIONS, OPINIONS, AND APPROACHES TO CANCER OF THE PANCREAS AND PERIAMPULLARY AREA

Abstract

At the end of his remarkable career, Whipple³³ could look back on a personal series of 37 pancreatoduodenectomies, a surgical procedure that has carried his name ever since,²³ although it had first been performed 26 years earlier by Kausch¹⁸ in Berlin. As for the results of this surgery, Whipple³² said that “the considerable risk (i.e., operative mortality) of 30%–35% is justified if they (the patients) can be made comfortable for even a year or two.” That was more than 50 years ago.

When the author retired from active clinical surgery in May 1998, the Mannheim Surgical Clinic could look back on 642 pancreatoduodenectomies performed over 26 years (Table 1). The overall operative and hospital mortality rate was 2.3% (including a consecutive series of 144 resections without any mortality at all), and the actual 5-year survival rate following resections for adenocarcinoma of the pancreas was 31%.

Nothing could be more foolish than attempting to compare today's surgeons with the surgical giants of the past on whose shoulders surgeons now stand. These two sets of data are separated by 50 years of progress in diagnostic imaging, perioperative supportive care of patients, and least of all, surgical technique. As a result, it is possible for any careful surgeon to reduce the operative mortality rate to below 10% and to increase the life expectancy of operable cases from 1 or 2 years to 5 or more years (in one-third of cases).

There has been progress, but the authors do not hesitate to say that ductal adenocarcinoma of the pancreas is an incurable disease. It cannot be cured by any of the available modalities. Why these sweeping and defeatist statements? The reasons are based on the analysis of personal experience.

Table 2 lists the actual (not actuarial) 5-year survival rates of 118 patients whose pancreatic cancer could be radically resected (R₀ resections) more than 5 years previously. At the end of this period, 37 patients were still alive, accounting for a 5-year survival rate of 31%. But one cannot close the books after 5 years. Life goes on, and so does the death toll from pancreatic cancer. In fact, 20, more than half, of the survivors succumbed subsequently to a late local recurrence or metastases. The ultimate cause of death could be proved by autopsy in only a few of these 20 patients.

The same picture of continued attrition is provided graphically by the survival curve of those 277 patients who underwent pancreatoduodenectomy for adenocarcinoma of the pancreas (n=174) or papilla (n=103; Fig. 1). The curves are calculated according to Kaplan-Meier, thus showing only the actuarial survival. If the resection was complete, that is, R₀ (n=122), then actuarial survival is 25.6% at 5 years; however, the curve continues to decrease thereafter.

One could argue that the definition of an R₀ resection is not objective because surgeons and pathologists may miss residual tumor macro- or microscopically; therefore, there are probably several incomplete resections (R₁ or R₂) included in this group, which may account for this attrition of long-term results.

But there is another even more cogent argument to support the earlier mentioned defeatist statements.²² The authors are referring to the long-term results following resection of early-stage pancreatic cancer. Early-stage cancer is defined by the Union Internationale Contre Cancer (UICC) classification of 1997 as a tumor of less than 2-cm in diameter, entirely confined to the pancreas, and without any evidence of lymphangiosis carcinomatosa or lymph node metastases. These early cancers are rare. They are discovered purely by chance and, in any large series, they constitute less than 5% of all operable cases. Thus, only 9 of 308 patients who underwent resection for pancreatic cancer in Mannheim fulfilled the criteria of true early cancers (Table 3).

These tumors were discovered at an early stage insofar as the duration of symptoms was relatively short—only 3 to 6 weeks. As one would expect, all but one of these patients reached 5-year survival. The last column of Table 3 shows equally clearly that (probably) all of these patients finally succumb to the disease. Six have died so far—as late as 11.5 years after the successful and apparently complete resection. If this is true of early cancers, it applies even more so to those 95% of cancers that are not discovered and resected until they have reached more advanced stages.

The earlier mentioned defeatist statements notwithstanding, a second, more positive one is: It is certainly worthwhile to do everything in our power to diagnose and treat (resect) pancreatic cancer, even if the diagnosis comes late and the resection proves to be only palliative.